Structuring the Issues in Informed Consent
Margaret A. Somerville*
Introduction
I. Battery or negligence?
II. Scope of disclosure
A. Determining the scope of disclosure “normally” required
by law
1. Therapeutic interventions
a. “Pure” therapy
b. Therapeutic research
2. Non-therapeutic interventions
a. Non-therapeutic research
b. Non-therapeutic non-research interventions
B. Restricting the scope of disclosure: waiver and
“therapeutic privilege”
1. Therapeutic interventions
2. Non-therapeutic
interventions
C. Extending the scope of disclosure: the patient’s
questions
III. The patient’s reception of the disclosure
A. Understanding
B. Rationality
IV. Who must disclose
V. The consent form
VI. Scope of the consent
VII. Proof
VIII. Causation
Conclusion
* Of the Faculty of Law and the Faculty of Medicine, McGill University.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
741
Introduction
Canadian courts, at all levels, have recently’ been called upon to
deal with the issues raised by consent to medical care. These deci-
sions have clarified many issues but some problems remain, prin-
cipally due to two factors. First, the courts have only been able
to deal with the issues raised by the facts of the particular cases
before them; thus, although a judgment such as the Supreme Court
of Canada’s decision in Reibl v. Hughes’ laid down relatively clear
guidelines, it did so only with respect to certain issues. Secondly,
and with all respect, some judgments are confused and confusing.
The cause of this confusion has been failure to structure adequately
the doctrine of consent before applying it to the facts of a particular
case. Such failure can result in inaccurate or vague decisions that
are difficult for both the legal and medical professions to use as a
guide to future conduct.
It is the aim of this enquiry to establish a decision-making
structure which will organize and assist in determining the legal
issues relating to the consent of competent adults in the medical
relationship. One hopes that such a structure will prove useful in
ascertaining the conduct required and in reaching decisions in
subsequent cases. To establish this structure it is first necessary
to trace the development of the law of informed consent in Canada.
This will include consideration of recent cases on consent in the
Supreme Court of Canada, how these cases have altered the law,
and the issues left unsettled.
There are many other aspects of consent which could be ex-
plored, from the purposes served by requiring it, or the values and
philosophical principles underlying it, to its role and function when
the subject of a medical intervention is either incompetent or a
minor.’ One reason for limiting this discussion to the competent
adult model is that close analysis of the issues raised by consent in
that context is a prerequisite to dealing with the more cpmplicated
problems that arise with incompetents or minors.3
1(1980) 114 D.L.R. (3d) 1, rev’g (1978) 21 O.R. (2d) 14 (C.A.).
2 For a general discussion of medical consent, see Somerville, Consent to
Medical Care (1979).
3 See, e.g., Re Eve (1980, 1981) 115 D.L.R. (3d) 283 (P.E.I.S.C. in banco),
rev’g (1979) 10 R.F.L. (2d) 317 (P.E.I.S.C.). This case is now on appeal before
the Supreme Court of Canada.
McGILL LAW JOURNAL
[Vol. 26
I. Battery or negligence?
The proper cause of action when a defective consent is alleged
may be either battery or negligence. The difference is significant
because:
It will have important bearing on such matters ‘as the incidence of the
onus of proof, causation, the importance of expert medical evidence, the
significance of medical judgment, proof of damage and, most important,
of course, the substantive basis upon which liability may be found 4
While each of these factors will not be discussed in detail here, it is
necessary to be aware of them.5
Common law courts in Canada have taken the traditional ap-
proach that the consent that is both necessary and sufficient for
avoiding a cause of action in battery in medical cases is consent to
“the basic nature and character of the operation or the procedure”.,
The difficulty in applying this rule is determining which factors
form part of the basic nature and character of an act and which do
not.7 Such determinations have been made by judges on a case-by-
case basis, with no more definite guidelines than the rule itself. But
4 Kelly v. Hazlett (1976)
15 O.R. (2d) 290, 310 (C.A.). The Supreme Court
of Canada disapproved of Kelly v. Hazlett in part in Reibl v. Hughes, supra,
note 1.
5 It should be noted that this problem of the proper cause of action and
the differences between one cause of aotion and another does not present
itself in the same way in the civil law of Quebec. Liability based on delict
or quasi-delict is governed by Art. 1053 of the Civil Code, whether or not the
offending conduct would found an action
in battery or negligence at
common law. Thus any differences in the way in which -two claims, both
falling under Art. 1053 C.C., are handled, does not depend on a difference in
the basis of liability (cause of action), as it does at common law, but on
the different facts or circumstances of the two cases. However, a claim
based on a breach of duty in the physician-patient relationship is more
likely to sound in contract in the civil law than it would at common law,
and contrary to the civilian doctrine of cumul, which would allow a plaintiff
to claim in both delict or quasi-delict and contract in relation to the same
facts which it is alleged give rise to liability, there is a trend to exclude
delictual or quasi-delictual liability to the extent that the obligation
in
issue has a contractual basis. Support for this trend depends on also
excluding a doctrine of option, as argued by Professor P.-A. Cr6peau (La
responsabilitg mddicale et hospitali&re dans la jurisprudence qudbdcoise rd,
cente (1960) R. du B. 434, 470-2), but the doctrine of option was recently en-
dorsed by a unanimous bench of the Supreme Court of Canada in Wabasso
Ltd. v. The National Drying Co. (dated 22 June 1981).
6 Kelly v. Hazlett, supra, note 4, 313.
7There may be liability in negligence even though there may be no
liability in battery for failing to disclose consequences which are not part
of the basic nature and consequences of an act: see infra.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
743
some judges have tried to formulate a clearer, more objective rule,
which would help determine when non-disclosure of information
or failure to obtain consent should give rise to a cause of action in
battery.”
The situation facing
the judges can be represented diagramma-
tically:
–
collateral features
of the act
basic
nature
of the
act
The outer square represents all the con-
sequences or risks to which the patient
must consent if liability in tort (battery
or negligence) for failure to obtain con-
sent is to be avoided. The inner square
represents the factors that make up the
basic nature and character of the act
of touching. Failure to obtain consent
to these factors will give rise to a cause
of action in battery. Thus, whether or
not battery lies depends on where the
inner line is drawn.
That judges will vary in drawing this line, even with respect
to the same facts, can be seen by comparing the decision of the
majority of the Supreme Court of Canada in R. v. Bolduc & Bird9
with that of the dissent and that of the Court of Appeal of British
Columbia in the same case. Likewise, the judgment of the majority
of the Court of Appeal of Ontario in R. v. Maurantonio’ can be
compared with that of the dissent. These are criminal assault
(battery) cases, but the rules governing consent in criminal law
and in the tort of battery not only have common origins but are
directly comparable. Such cases demonstrate that because criminal
assault (or battery) will not lie if there is consent as to the basic
nature and character of the act, liability will depend on whether or
not the feature to which consent has not been obtained forms part
of the act’s basic nature and character. Thus, to the extent that
there is discretion involved in determining whether or not a particu-
lar feature forms part of the basic nature and character of the act
(“basic features”),
there is discretion as to the imposition of
liability. The two possible alternative analyses of any given fact
situation that gives rise to this discretion may be represented as
follows:
8 See, e.g., Kelly v. Hazlett, supra, note 4.
9 [1967] S.C.R. 677, rev’g (1967) 61 D.L.R. (2d) 494 (B.C.C.A.).
10 [1968] O.R. 145.
Analysis I:
CONSENT
toI
McGILL LAW JOURNAL
[Vol. 26
NO CONSENT
= NO BATTERY
ACT
“COLLATERAL FEATURES”
which consists of
“BASIC FEATURES”
A.B.C.
Analysis II:
CONSENT
to
X.Y.Z.
NO CONSENT
to1I
=-
BATTERY
ACT
which consists of
“BASIC FEATURES”
A.B.C.
X
“COLLATERAL FEATURES”
Y.Z.
From this diagram it can be seen that, depending on whether or
not X n1 is held to be a “collateral feature” or a “basic feature” of
the act alleged to constitute criminal assault or the tort of battery,
the necessary consent will or will not be present, respectively, and
liability will be determined accordingly.
There is another way in which a holding as to whether or not
battery-avoiding consent is present can be varied. This does not
require altering the characterization of a feature of the act from
“basic” to “collateral” or vice versa, but rather makes consent to
the act conditional upon the collateral features being as represented.
Using the same model this can be represented as follows:
n In R. v. Bolduc & Bird (supra, note 9) and R. v. Maurantonio (supra,
note 10), the feature which X would represent was as to the holding of
professional medical qualifications by the person observing or conducting,
respectively, a medical examination. The question in both cases was whether
having such qualifications was part of the basic nature and character of
a medical examination. If it were, a person without such qualifications could
not conduct a true medical examination and, if he attempted to do so, the
act consented to would not be the act carried out and assault would be
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
745
NO CONSENT
to
t
X.Y.Z.
“COLLATERAL FEATURES”
THEREFORE
NO CONSENT
to1
ACT
which consists of
“BASIC FEATURES”
A.B.C.
(to which there was consent)
established. If, on the other hand, such qualifications are only a collateral
feature, there may be consent to the act of examination and hence no
assault, although fraud as to the qualifications of the person conducting it.
In R. v. Maurantonio the majority of the Ontario Court of Appeal held that
the patients seen by the accused had only consented to being examined
by a medically qualified person, that such qualiflications related to the
basic nature and quality of the .act of examination and, consequently, fraud
as to the holding of medical qualifications vitiated consent to the examination.
In contrast, the dissent held that the holding of professional qualifications
did not relate to the nature and quality of the act of examination. In
R. v. Bolduc & Bird the majority of the Supreme Court held that “[t]he
fraud that was practised on [the patient] was not as to the nature and
quality of what was to be done but as to Bird’s identity as a medical
intern” (supra, note 9, 680). In this case the person who actually examined
and carried out a medical procedure on the woman patient was medically
qualified, but his “voyeur” friend, Bird, who did not touch her, was
not. The majority of the Court stated that if Bird had touched the woman
there would have been an “unlawful assault”, which indicates
that they
treated fraud as to qualifications as going to the basic nature and character
of an act and as capable of vitiating assault-avoiding consent. But here
there was no actus reus of an assault, no touching by the person without
qualifications and hence the presence or absence of consent to such an
act was irrelevant. In comparison, Mr Justice Lord of the British Columbia
Court of Appeal held that the patient’s consent was vitiated because she
only consented to being examined in the presence of medically qualified
persons and when this was not the case there had been “a drastic change
in ‘the nature and quality of the act and …
the consent she
gave cannot be considered as a consent to the thing done to her and cannot
be relied upon by the appellants” (supra, note 9, 495). Lord J.A. thus
characterized what constituted the act of assault more broadly than the
Supreme Court and hence found that consent to the act had been vitiated
by fraud as to one of the accused’s qualifications.
[therefore]
McGILL LAW JOURNAL
[Vol. 26
Pursuant to this analysis, it is irrelevant whether X is characterized
as a “basic” or “collateral” feature, as even if X is a “collateral”
feature, if X is not as represented, the consent will fail to “flow
through” as a valid consent to the act. R. v. Williams,12 in which
the accused, a choir-master, persuaded a young woman that sexual
intercourse was therapy for her voice, is probably an example of
a court taking such an approach. Depending on the circumstances,
whether an act is therapeutic for the patient could be regarded as
a “collateral” feature of the act or could relate to the basic nature
of the act. However, accepting that in the particular circumstances
characterization of an act as therapy is a collateral feature, battery
could still lie when there is fraud or misrepresentation in this
respect. For instance, if, as in the Williams case, a collateral feature
of an act was not as represented (that is the act of sexual inter-
course was not voice therapy),
then despite the consent to the
act itself (having sexual intercourse), that consent would be invalid,
as the Court held, because consent to the act was conditional on
the collateral feature (that the act was therapy) being as represent-
ed. Although this example may seem very far removed from a
normal medical context, it may have important applications. For
instance, if a patient were misled to the effect that a particular
procedure was therapeutic when in fact it was performed for the
purposes of non-therapeutic research, battery-avoiding consent could
similarly be vitiated.
How has the law outlined above been applied in the medical
relationship? The requirement that there be consent to the basic
nature and character of the operation or procedure means that a
physician must disclose all inevitable consequences of a proposed
procedure in order to obtain battery-avoiding consent. Consequently,
all courts faced with the issue have held that there will be a cause
of action in battery when a physician does something to which the
patient has not consented at all, or which the patient has expressly
requested not be done 3 or has refused. Battery could also be estab-
lished where the physician’s act was essentially different in nature
from that to which the patient consented. For instance, if the patient
was told that the purpose of the operation was to relieve pain, but
not told that the consequences would include sterility, any consent
given would be invalidated and a cause of action in battery would
be available.
12 [1923] 1 K.B. 340 (C.C.A.).
13See Allan v. New Mount Sinai Hospital (1980) 28 O.R. 356 (H.C.); see
also Beausoleil v. La Communautg des Soeurs de la Charitd de la Providence
[19651 B.R. 37 for a similar situation in Quebec law.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
747
Some courts have also held that knowledge of certain risks could
be so material to understanding of the basic nature and character
of an operation that failure to disclose them would vitiate battery-
avoiding consent. 14 In other words, it has been held that not only
non-disclosure of inevitable results of a procedure can vitiate
battery-avoiding consent, but also non-disclosure of risks of which
knowledge was “essential to an informed decision to undergo the
operation”. 15 It is with respect to failures to disclose a risk, as
compared with an inevitable consequence, that the Supreme Court
of Canada has probably restricted the availability of an action in
battery.
It is not easy to decide, from a policy point of view, whether
or not a cause of action in battery should be allowed for non-
disclosure of certain risks. The argument that it should be allowed
is that some risks are so serious that they necessarily relate to
the basic nature and character of an operation and, therefore, their
non-disclosure should give rise to a battery action. The difficulty
is that as only some risks have this effect, how is the line to be
drawn between those that do and those that do not? The alternative
solution, which may be the position adopted by the Supreme Court
of Canada in Reibl v. Hughes, is that any liability for non-disclosure
of a risk can only lie in negligence and not in battery. 6
[Aictions of battery in respect of surgical or other medical treatment
should be confined to cases where surgery or treatment has been per-
formed or given to which there has been no consent at all or where,
emergency situations aside, surgery or treatment has been performed
or given beyond that to which there was consent ….
[U]nless there
has been misrepresentation or fraud to secure consent to the treatment,
a failure to disclose the attendant risks, however serious, should go to
negligence rather than to battery. Although such a failure relates to an
informed choice of submitting to or refusing recommended and appro-
priate treatment, it arises as the breach of an anterior duty of due care,
comparable in legal obligation to the duty of due care in carrying out
the particular treatment to which the patient has consented. It is not
a test of the validity of the consent.’ 7
That is, non-disclosure of a risk will not give rise to a cause of
action in battery except when there has been a “misrepresentation
or fraud to secure consent to … treatment”. But what can constitute
misrepresentation within
this rule, and why does such mis-
representation allow non-disclosure of a risk to give rise to an action
in battery, whereas without such misrepresentation it does not?
14 E.g., Kelly v. Hazlett, supra, note 4, 313.
15 Ibid.
16 Supra, note 1, 7-10.
17.Ibid., 10-1.
McGILL LAW JOURNAL
[Vol. 26
In exploring these issues it is necessary to determine the basis
of the holding of the Supreme Court referred to above. Is it, first,
that non-disclosure of a risk does not amount to misrepresentation
(although that of an inevitable consequence can); or, second, that
information about risks can never relate to the basic nature and
character of an operation and thus cannot vitiate battery-avoiding
consent; or both propositions; or neither? Presumably, the Court
has established the second proposition rather than the first, as it
is difficult to draw a distinction between one type of non-disclosure
(non-disclosure of inevitable consequences) constituting misrepre-
sentation, and the other (non-disclosure of risks) not doing so.
It may be argued that because there is a pre-existing duty to disclose
inevitable consequences, but not risks, non-disclosure of inevitable
consequences constitutes a misrepresentation, while that of risks
does not. But this pre-empts the question to which an answer is
sought, that is, just what information is there a duty to disclose
so as to avoid an action in battery? Moreover, reliance on a rule
that total non-disclosure of a risk does not give rise to a cause of
action in battery, because a total non-disclosure cannot constitute
misrepresentation, would not exclude a partial disclosure. But an
approach that recognizes partial but not total non-disclosure of
risks as misrepresentation would be artificial and could give rise
to fortuitous results. Further, the distinction between nonfeasance
and misfeasance has no place when there is a pre-existing duty
relationship, as there is between physician and patient.
If the Court has not relied on a rule that a non-disclosure is
unable to constitute misrepresentation, has it held that the risks
of a procedure cannot relate to its basic nature and character? This
question can be explored by asking what the situation would be
where a significant and serious risk was grossly misrepresented,
rather than undisclosed. This could, arguably, give rise to a cause
of action in battery within the Court’s ruling. However, the fact
that misrepresentation of a risk could give rise to a cause of
action in battery means, by definition, that the misrepresentation
must relate to the basic nature and character of the act. If this
is true it shows that the action in battery is not excluded because
of the nature of the misrepresentation, that is because the mis-
representation related to a risk and a risk cannot relate to the basic
nature and character of a procedure, but that battery is excluded
on some other basis.
If the above analysis is accepted the basis of the Supreme
Court’s holding is not that total non-disclosure of risks cannot
constitute misrepresentation, nor that risks cannot relate to the
19811
STRUCTURING THE ISSUES IN INFORMED CONSENT
749
basic nature and character of a medical procedure. This leaves the
question which reveals the key to the basis of the Supreme Court’s
ruling still unanswered. Why does “misrepresentation or fraud to
secure consent to the treatment” cause non-disclosure of a risk to
give rise to a cause of action in battery, where it would not do
so if the elements of misrepresentation or fraud were not present?
The true test of whether or not a cause of action in battery
will lie for non-disclosure of a risk, provided the risk is serious
and sufficiently likely of occurrence to relate to the basic nature and
character of the act carried out, depends on the nature of the
physician’s conduct with respect to the non-disclosure. Not only
the nature of the undisclosed information is significant, but also the
nature of the failure to disclose. It is proposed that, if the physician
negligently (i.e., unintentionally) fails to disclose or misrepresents
a risk, he will be liable in negligence. If he intentionally does
either of these things the action will also lie in battery, provided
that the risk which is not disclosed or is misrepresented is fun-
damental enough to relate to the basic nature and character of
the procedure.’8 Thus the presence or absence of intention with
respect to the non-disclosure of a risk which relates to the basic
nature and character of an intervention will determine the causes
of action available for failure to obtain consent to that risk. By
contrast, when the non-disclosure relates to an inevitable conse-
quence of an intervention, intention or lack of it in relation to
the non-disclosure is irrelevant to establishing a cause of action in
battery. This is true because the intention necessary to support a
cause of action for battery arising from an intentional, non-con-
sensual touching of the kind which occurs is present in carrying
out the act which has those inevitable consequences, regardless of
the presence or absence of intention with respect to the non-
disclbsure.
Hence, what is being suggested is that the intention necessary
to support an intentional tort will be found in relation to a different
element of the tortious act (that is, either the touching or the non-
disclosure) depending on whether the failure in obtaining consent
‘sIt may be argued that it is inappropriate to have both battery and
negligence available on the same set of facts. Historically, this was not
unusual (see Prichard, Trespass, Case and the Rule in Williams v. Holland
[19641 C.L.J. 234), and the reason for which this was true is equally applic-
able here. It occurs because the plaintiff pleads in a lesser cause of action
in terms of the culpability of the defendant than he might have done; i.e.,
he pleads in negligence (or, formerly, in an action on the case) when he
could have pleaded in battery.
McGILL LAW JOURNAL
[Vol. 26
relates to failure to inform of risks or of inevitable consequences,
but in both cases, the necessary intention may be present.
In relation to determining whether the touching itself was
intentional, there is a key concept: the question which must be
asked is not simply whether there was consent to a touching,
which in most cases there will be, but whether there was consent
to touching of that kind or in that manner. Likewise, it is relevant
to ask not only whether there was intention to touch, but whether
there was intention to touch in that manner. The concept of in-
tention “to touch in that manner” is broader and more precise
than the concept of just touching. It includes the inevitable conse-
quences and purposes of an intervention, as well as the touching
itself. Because risks, by definition, may not occur, it is not possible
to find the required intention to touch in the manner which results
from risks occurring simply by demonstrating their crystallization.
Any proof of intention relates, rather, to the act of non-disclosure
of these risks. By contrast, when what occurs is an inevitable con-
sequence of an intervention it can be presumed that this was
intended, as in tort a reasonable person is presumed to intend the
inevitable consequences of his acts. Consequently, in the latter case,
the necessary intention to support intentional touching in that
manner and a prima facie tort of battery is established by proving
the touching, and the only question is whether or not there was
sufficient consent.
An objection could be raised here that two different entities
are being compared: in one case the question asked is whether or
not there is intentional non-disclosure of information to which the
patient is entitled: that is, is there intentional failure to obtain
consent? In the other case the question is whether there is inten-
tional touching in a situation where the failure to obtain consent
to that touching may have been intentional or unintentional. It is
submitted that there is no contradiction between these two ap-
proaches. Battery is an intentional tort, which may be established
through the intention “to touch in that manner” or the intentional
failure to obtain the necessary consent. It is just that demonstrating
the latter is superfluous when it can be shown that there was an
intention to touch in a certain manner to which there was no
consent.
There is one further problem with the Supreme Court’s approach
to actions in battery for failure to obtain adequate consent. Some
risks are so important that most people would regard them as
an essential part of any description of the basic nature and
character of a procedure. For instance, the fact that an operation
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
751
carries a substantial risk of death would cause most persons to
characterize that operation as being of a serious nature. Further,
can any real distinction be drawn between failure to disclose, for
instance, that as a result of an operation a person will certainly be
rendered sterile, and failure to disclose that there is a substantial
risk of this occurring? It is submitted that the law should not try
to draw distinctions that do not accord with generally held views
as to what factors constitute the basic nature and character of an
act, no matter how conceptually pleasing and easy of application
the resulting rule may be.
Thus, with all respect, it is submitted that to the extent that
the Supreme Court has limited the availability of an action in
battery by stating the law to be that risks do not relate to the
basic nature and character of an act and, consequently, their non-
disclosure cannot vitiate battery-avoiding consent, the ruling may
not be desirable. However, as shown above, the effect of the
Supreme Court’s ruling on the availability of a battery action will
vary, depending on how it is analyzed. The analysis suggested ac-
cepts the Supreme Court’s ruling but minimizes
its effect of
making unavailable a battery action that otherwise would have
been available under Canadian common law.
The approach suggested may be summarized as follows: in all
cases where lack of consent is alleged, one is arguing either that
there was no consent at all to the touching or to touching in that
manner. The first question is whether the touching itself and in
that manner was intentional. It is highly unlikely that the touching
itself will be unintentional, but this is not true of the manner of
the touching. The manner of the touching includes two types of
consequences: inevitable consequences and risks which eventuate.
When the feature of the touching to which it is alleged there was
no consent is an inevitable consequence, then there will necessarily
have been an intention to touch in that manner. When, on the
other hand, the touching is of that manner because of the crystal-
lization of a risk, touching in that manner is unintentional (unless,
possibly, the risk which eventuates was of very high probability,
but this case will not be considered here). In the second case, it
may initially seem that battery should not lie for non-disclosure
of a risk, as the act of which the plaintiff complains –
that he was
touched in a manner to which he did not consent – was un-
intentional. However, a second question is relevant: whether the
failure to obtain consent was intentional. It
is suggested that
where it is intentional, and provided the non-disclosure is of a
sufficiently serious and probable risk that the risk can be said to
McGILL LAW JOURNAL
[Vol. 26
form part of the basic nature and character of the intervention, the
necessary intention for a cause of action in battery will exist. Such
an approach would allow non-disclosure of certain risks because
of “misrepresentation or fraud” to give rise to a cause of action
in battery, as the Supreme Court suggests. It would include within
the notion of misrepresentation some total non-disclosures; that is,
intentional concealment of certain risks would suffice. Thus, only
the unintentional non-disclosure of a risk that relates to the basic
nature and character of an act would not be actionable in battery.
The suggested approach and the correlation of the variables it in-
cludes can be demonstrated in the following way:
Non-disclosure:
of inevitable consequence
of “sufficiently serious
and probable” risk
INTENTIONAL
UNINTENTIONAL
Battery
Battery
Battery
Negligence
to obtaining
Defects in consent which do not give rise to a cause of action
in battery will be either actionable in negligence (or possibly in
contract 9) or will not be actionable at all. The dividing line
between those that are actionable in negligence and those that are
not actionable at all is determined by whether or not the physician
has breached the standard of care required of him by the law
relating to negligence with respect
the patient’s
consent. The substantive content of this duty is discussed below.
Finally, it is appropriate here to note that some confusion may
be caused because two doctrines “consent” and “informed consent”
are not distinguished. It is suggested that the word “consent” be
reserved to refer to the substantive entity which must be present
to avoid liability in battery and similarly for the term “informed
consent” in relation to negligence. It is proposed that with respect
to the substantive content of consent the traditional notion should
be retained. This means that consent will be present when there
is consent to the basic nature and character of the act. Informed
consent is a more extensive concept that also comprehends consent
to certain consequences or risks of consequences. But, as further
discussion will show, there has not always been consensus as to
its requirements. It necessarily includes all elements of the consent
doctrine, but the reverse is not true. Thus a physician may have
obtained sufficient consent to avoid liability in battery, but not in
negligence.
‘9 See, e.g., the discussion in Dendaas (Tylor) v. Yackel (1980) 109 D.LR.
(3d) 455 (B.C.S.C.).
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
753
II. Scope of disclosure
What must physicians disclose in order to fulfill their legal
duty to obtain informed consent?20 This has been a vexed issue for
the courts and some problems remain. However, the Supreme Court
of Canada has now settled one matter on which there was uncer-
tainty, that is, the standard for determining the scope of disclosure.
A discussion of the scope of disclosure of information required
in any given circumstances in order to obtain informed consent,
like any decision by a court concerning it, involves three determi-
nations, which must be undertaken in a particular order. First, the
scope of disclosure “normally” required by the law must be de-
termined. Within this context, the effect of a proposed intervention
being therapeutic or non-therapeutic and “practice” (“pure” thera-
py) or research,21 must be considered. Secondly, it must be asked
whether this “normal” scope of disclosure may be restricted in
some circumstances, which necessitates consideration of waiver
and “therapeutic privilege”. Thirdly, does the law recognize any
mechanism for extending the “normal” scope of disclosure? Here
one must consider the effect of the patient’s questions.
A. Determining the scope of disclosure “normally” required by law
1. Therapeutic interventions
a. “Pure” therapy
The physician-patient relationship is a fiduciary one 2 which
means that unlike “arm’s length” transactions (where the law only
imposes a duty to be honest and, in certain circumstances, not
to be negligent with respect to one’s statements), there is a positive
duty on the part of the physician to inform the patient. This type
of duty is sometimes described as one of trust, confidence and
conscience. It arises when the law recognizes an imbalance in power
or status or position between the parties to a relationship, such
that the person in the position of authority must prove that he did
not abuse this authority. In the medical relationship the physician
would abuse his authority if he failed to accord sufficient respect
to the patient’s rights to autonomy and inviolability.
20Because the scope of disclosure necessary for informed consent is wider
than that for consent, the more extensive concept is surveyed here except
where the contrary is stated.
21 For definition of the terms “practice” and “research”, see Somerville,
Clarifying the Concepts of Research Ethics: A Second Filtration [1981]
Clinical Research [forthcoming].
22 See Kenny v. Lockwood [1932] O.R. 141 (C.A.).
McGILL LAW JOURNAL
[Vol. 26
The rights of the patient require, if it is at all possible, that the
patient be placed in a situation where he can make his own deci-
sions regarding medical treatment. In order to make such decisions
the patient needs information. In deciding what information must
be supplied to the patient, two situations must be distinguished:
where the patient asks questions and where he does not.
Where the patient has not asked any questions about a purely
therapeutic (i.e., non-research) treatment which he is to receive,
what must the physician disclose to fulfill his legal duty to inform?
The court could adopt several possible responses: what the physi-
cian thinks the patient ought to be told (the subjective-physician
standard); what the reasonable physician of that specialty would
tell the patient in the same circumstances (the reasonable physician
or the “professional disclosure” 23 standard); what the reasonable
patient would want to know in those circumstances (the reasonable
patient or the “full disclosure”24 standard); or what that patient
wants to know (the subjective-patient standard). These four possi-
ble standards require progressively more information to be dis-
closed. This can be represented diagrammatically as follows:
SCOPE OF DISCLOSURE
0%
1) Subjective physician standard
2) Objective reasonable physician or
“professional disclosure” standard
3) Objective reasonable patient or
“full disclosure” standard2
4) Subjective patient standard
100%
23 See Picard, “The Tempest of Informed Consent” in Klar, Studies in
Canadian Tort Law (1977), 129.
24 Ibid.
25 It may be unfortunate that the term “full disclosure” has been adopted
to describe the standard of disclosure based on informing the patient of
all those consequences and risks which would be material to the reasonable
1981)
STRUCTURING THE ISSUES IN INFORMED CONSENT
755
It is possible to fiied examples of acceptance of each of these
standards in various Canadian and American cases. However,
almost without exception, courts now reject the first standard as
being too paternalistic and as aot conforming to the usual legal
standard of care exacted of physicians in relation to other duties
apart from consent: that is, a physician may not, solely accord-
ing to his own discretion, establish what is a legally acceptable
standard of conduct in relation to a particular matter, but must,
at least, conform to the norms of the profession in that respect.
Likewise, the fourth stdndard is rejected as affording the patient
too much the advantage of hindsight, as the patient can always
claim that he wanted to know a certain fact that W~as undisclos-
ed. Consequently, the choice is between the “reasonable physician”
and the “reasonable patient” standards and, until very recently,
most Canadian courts had adopted the “reasonable physician”
standard as the norm 6 In doing so it was not uncommon to
find a court distinguishing those American judgments27 which had
accepted the “reasonable patient” standard. 8 However, the Supreme
Court of Canada, in Hopp v. Lepp, and even more clearly in Reibl v.
Hughes, changed the required legal standard of disclosure from
standard 2 to standard 3. The latter standard requires a physician
to disclose all those consequences and risks which would be
material to a reasonable patient in the same circumstances. And
“[a] risk is … material when a reasonable person, in what the
physician knows or should know to be the patient’s position, would
be likely to attach significance to the risk or cluster of risks in
deciding whether or not to undergo the proposed therapy. ’29
This change in the law does not mean that medical evidence is
irrelevant to setting the required scope of disclosure. As the Su-
preme Court explains,30 it is just that such evidence is no longer
patient in the same circumstances. This would be true, if the term “full
disclosure” were to cause confusion, as it might. The denotation of the term
“full” tends to indicate that all possible information must be disclosed, but,
as can be seen from the diagram this is not necessarily the case.
2 See, e.g., Kelly v. Hazlett, supra, note 4; Reibl v. Hughes, supra, note 1
(C.A.); Lepp v. Hopp (1979) 98 D.L.R. (3d) 464 (Alta C.A.), rev’d sub nom.
Hopp v. Lepp (1980) 112 D.L.R. (3d) 67 (S.C.C.).
27 See, e.g., Canterbury v. Spence 464 F. 2d 772 (D.C. 1972).
28 See, e.g., Kelly v. Hazlett, supra, note 4, 319.
2OWaltz & Scheuneman, Informed Consent to Therapy (1970) 64 Nw. U. L.
Rev. 628, 640, adopted by the Court in Canterbury v. Spence, supra, note 27
and quoted by Laskin C.J.C. in Hopp v. Lepp, supra, note 26, 80.
30 Reibl v. Hughes, supra, note 1, 12-4.
McGILL LAW JOURNAL
[Vol. 26
sufficient to determine the full scope of disclosure that the law
requires: ” In particular medical evidence is still necessary
is needed
to show the existence of
to establish (1) risks inherent in a given procedure or treatment, (2) the
consequences of leaving the ailment untreated, (3) alternative means of
treatment and their risks, and (4) the cause of the injury suffered by
the plaintiff-patient. Finally, if the defendant-physician claims a pri-
vilege, expert testimony
(1). an
emergency which would eliminate the need for obtaining consent, and
(2) the impact upon the patient of risk disolosure where a full disclosure
appears medically unwarrantedV2
What would the reasonable physician disclose and how is this
to be determined? The court will hear expert evidence on what
physicians of that specialty tell their patients, but this evidence does
not automatically become part of the accepted legal standard, even
as far as disclosure of medical risks is concerned. It is taken into
account when the court determines the legal standard 33 These
two standards (the standard of disclosure of the profession and the
one determined by the court) will usually be identical, but need
not necessarily be so. This allows the court a discretion to reject
or alter the accepted professional standard if it considers it inade-
quateY4
Factors that will invariably require disclosure because they
will always be material to any reasonable patient include the
nature of the procedure, its seriousness, its purpose and whether
it is necessary or elective. With respect to other effects of a given
medical intervention that must be disclosed, the distinction between
inevitable consequences and risks is again relevant.
Inevitable consequences of the proposed treatment, and of its
alternatives, including foregoing all treatment, must be disclosed.
The only exception to this would be if the consequences
in
question were matters of such common knowledge that the physi-
cian could assume the patient knows of them and, consequently,
that there is no need to make an express disclosure in order to
inform the patient. Of course, if the physician had actual knowledge
that the patient was not so informed any presumption that the
31 See Somerville, Legal Investigation of a Medical Investigation [1981]
Alberta L. Rev. [forthcoming].
3 2 See Reibl v. Hughes, supra, note 1, 13-4: the passage quoted is from a
discussion of Canterbury v. Spence, supra, note 27, in New Trends in In-
formed Consent (1975) 54 Neb. L. Rev. 66, 90 (emphasis added).
3 See Reibl v. Hughes, supra, note 1; Hopp v. Lepp, supra, note 26, 80
(S.C.C.).
34 For an example of a court exercising such a discretion, see Helling v.
Carey 519 P. 2d 981 (Wash. 1974).
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
757
patient was informed arising from “common knowledge” would
be rebutted, and the physician would have an obligation to inform
the patient expressly of these consequences.
Risks include chances of benefit and chances of harm. A physi-
cian must describe the nature of the benefit hoped for and the
likelihood of achieving it in relation to each possible alternative
treatment, including no treatment.
The initial factors to be taken into account in deciding which
risks of harm must be disclosed are the probability of the risk
occurring and its seriousness. The more serious the consequences
and the higher the probability, the more likely it is that the patient
should be informed. Not only the “material risks”‘ 5 of accepting
treatment must be outlined, but also those of alternative treatments
and of refusing any treatment, as a reasonable patient with a con-
dition needing treatment would consider disclosure of the “material
risks” of alternative treatments or refusing treatment necessary
to his decision.38 This latter duty may relate not only to the risks
of refusal of treatment, but also to the risks of refusal of diagnostic
tests, if a recent American case is to be followed. In that case a
physician was held liable for failing to warn a woman patient of
the dangers of not having a “Pap” smear test for cervical cancer,
when the woman refused the test because she said she could not
afford to pay the cost. The woman’s children subsequently recover-
ed damages from the physician when their mother died of this
disease. 37
It is difficult to specify the degrees of probability or the types
of risks of harm which necessitate disclosure, although some Amer-
ican statutes have attempted to delineate both of these factors,
but it is suggested that any risk with greater than one per cent
probability of causing irreversible morbidity should be disclosed.
Risks of death with a lesser probability of occurring should also
be disclosed, but for this most serious risk, it is difficult to state
a minimum level.
In Reibl v. Hughes the majority of the Court of Appeal of Ontario
seemed to suggest that a physician may “set-off” these risks against
each other in determining whether a sufficient probability of risk
was present to require disclosure.3 8 The Supreme Court of Canada
clearly rejected such an approach.3 9 In fact, in Hopp v. Lepp the
35 Reibl v. Hughes, supra, note 1, 5 (S.C.C.).
36 Ibid., 13.
37 Truman v. Tho.nas 611 P. 2d 902 (Cal. 1980).
38 Reibl v. Hughes, supra, note 1, 23 (C.A.).
3 Supra, note 1, 12.
McGILL LAW JOURNAL
[Vol. 26
Supreme Court seemed to indicate that telling the patient the
risks of not having treatment could reassure him about having it.4″
It may be argued that such a “set-off” approach is justified if the
risks of having or foregoing a procedure are of the same kind, even
if it cannot be accepted where the risks of having or foregoing the
procedure are of different natures. But to accept such a “set-off”
rule even when the risks are identical would mean, for example,
that a ten per cent risk of death in having a procedure could be
set off against a ten per cent risk of death in foregoing it, and like-
wise, a similar 0.01 per cent risk of death could be set-off, with the
result that in neither case need the risks be disclosed. Clearly,
this would be an unsatisfactory result as the two cases are not
comparable in terms of whether disclosure should be required.
The aim of requiring that the patient be informed is to allow him
to measure the risks and benefits of various alternatives when
making a decision as to which course of conduct he will pursue.
Thus, it is not, and should not be, the overall “net” risk which is
the probability determining whether or not disclosure is required,
but rather the individual “gross” risks of the alternative courses of
conduct which are open to the patient.
Another reason for not allowing such a “set-off” is that in riskier
procedures the patient may want to be put on notice to seek a
second opinion or, perhaps, the most expert operator or the most
up-to-date facilities for undertaking the procedure. Moreover, in
many cases the nature of the risks involved in having or foregoing
a procedure may not be of the same kind, or even if they are, the
times at which they are likely to occur may differ.”
It could be
that the patient’s value-preference in any of these respects will
differ from the physician’s. For instance, the patient may prefer a
risk of death to a risk of a life as a quadriplegic, whereas the physi-
cian may not. Likewise, another person, who is the parent of a
young family, may prefer a risk which is likely to eventuate in the
future rather than one which is immediate. As it is the patient who
must live with the results of the decision, the patient’s preference
must prevail.
The final point to be made about assessing the probability and
seriousness of medical risks in order to determine whether they
must be disclosed is that the assessment must be carried out in
relation to the particular patient involved.4 This involves two en-
40 Hopp v. Lepp, supra, note 26, 79-80 (S.C.C.).
41 Reibl v. Hughes, supra, note 1, 32-5 (S.C.C.).
42See Kelly v. Hazlett, supra, note 4, 305-7; Reibl v. Hughes, supra, note 1;
Hopp v. Lepp, supra, note 26, 80 (S.C.C.).
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
759
quiries: first, is the patient particularly susceptible to any given
risk in terms of its probability of occurrence or seriousness?
Secondly, do the patient’s circumstances make an otherwise im-
material risk material to him.
Although a subjective or objective assessment with respect to
the first enquiry will very often reach the same result, this is not
necessarily the case. For instance, if the physician takes into ac-
count any special vulnerability of the patient to certain risks of
which the physician is aware or, as a reasonable physician, should
be aware, the assessment of the seriousness and probability of the
risks in relation to the particular patient may vary from what it
would be for the hypothetical, “reasonable” patient.
The second enquiry, which determines material risks according
to those risks which would be material to a reasonable patient in
the same circumstances, introduces a subjective factor, but one
which is assessed objectively, as all tests which take into account
the “same circumstances” invariably do. The circumstances which
will be taken into account are wider than simply medical considera-
tions and will be established by evidence from persons other than
just physicians. They can include, for instance, the fact that a
patient may wish to delay an operation for financial reasons 3 The
presence of such relevant extra-medical circumstances will be
established by evidence of the patient and other witnesses 4 4 Pre-
sumably such circumstances would be limited to those that were
known to the physician, or those that ought to have been known, or
would have been known to him if he had made the enquiries a:
reasonable physician ought to have made in the circumstances.
Courts also use classifications other than the probability and
seriousness of any given risk in determining whether or not that
risk must be disclosed. Some risks that are indicated as needing
disclosure within these other classifications would also require
disclosure under a probability/seriousness classification, and, there-
fore, the additional test may be regarded as superfluous. Alterna-
tively, some risks needing disclosure under one of these tests may
be exempted from disclosure on other grounds, because, for in-
stance, they are matters of “common knowledge” and in such cases
the additional test may be regarded as irrelevant. But despite such
overlap or exemption from disclosure, these classifications are
worth examining for the added information they give on how
the courts are applying the scope of disclosure standard.
43Reibt v. Hughes, supra, note 1, 6 (S.C.C.).
44 Ibid., 13.
McGILL LAW JOURNAL
[Vol. 26
Risks have been classified as “definite”, “special”, and “usual”.45
“Definite risk” is assessed on the basis of the likely effect of the
proposed procedure on the particular patient, taking into account
his disabilities. “Special risk” consists of the risks of operations in,
general, such as the risk involved with general anaesthesia. Risks
have also been divided into two groups, “definitive” and “col-
lateral”.46 “Definitive risks” are those which are so important that
their omission would give rise to a misdescription of the nature and
character of the procedure. In consequence, a consent given in
ignorance of these risks “is no consent at all”.47 In comparison,
“collateral risks” are those risks that do not fit the description of
“definitive risks” and do not fall within the category of risks not
requiring disclosure. The current validity or usefulness of the
distinction between “definitive” and “collateral” risks would need
to be questioned, however, if it were accepted that the Supreme
Court has held that the risks attached to an act cannot relate to
its basic nature and character.48
It has been held that the physician’s duty of disclosure “does not
require warning the patient of dangers incident to or possible in
any surgical procedure, such as the dangers of anaesthesia or the
risk of infection, matters which men of ordinary knowledge are
presumed to appreciate”. 49 This is the “common knowledge” ex-
ception to requiring disclosure, but there are other risks so common
that expert evidence is not needed to assess them, but which must
be disclosed in order to avoid liability in battery. 0
In summary, one can derive from the cases a classification of
five possible classes of risks:
45 Kelly v. Hazlett, supra, note 4, 319.
46 Reibl v. Hughes (1977) 16 O.R. (2d) 306, 312 (H.C.).
47Ibid.
48See text between notes 14 and 19, supra.
4 9 Reibl v. Hughes, supra, note 46, 312 (HC.).
50 In Reibl v. Hughes the trial judge, Haines J., states that when the cause
of action is in battery “[tjhere is no issue for which medical evidence is
required” (ibid., 311). (Presumably such evidence is not needed because the
tribunal of fact can itself assess the required scope of disclosure, rather
than the evidence not being -needed because disclosure is not required due
to the matter being one of “common knowledge”.) But-Mr Justice Haines
also contemplates that non-disclosure of some risks could give rise to a cause
of action in battery (ibid., 312). In conjunction, these two statements en-
visage that there are risks which must be disclosed in order to obtain battery-
avoiding consent, but which do not need medical evidence in order to
establish this. It is this group of risks which is referred to here and in the
second category listed in the paragraph next following.
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
761
1.
2.
3.
4.
5.
those risks so common that disclosure is not necessary and consent
to run them may be presumed;
those risks so common that expert evidence is not needed to assess
them, but which must be disclosed and consent to which may not
be presumed;
those risks which require expert evidence for their definition and
which must be disclosed as it is competent medical practice to do so;
those risks which must be disclosed because they would be material
to the reasonable patient in the same circumstances; and, finally,
risks falling within none of the previous four categories.
Thus at both ends of this classification there is a category of risks
not requiring disclosure and, in the middle, three categories, where
what has to be disclosed is determined by a different test according
to the categorization of the risk”
This scheme may seem cumbersome but it is necessary, because
it is very unclear in some cases why the court has held that a
particular risk need not be disclosed. It
is important that the
justification of any non-disclosure be clear because, while non-
disclosure of a certain risk may be perfectly acceptable on one
ground, it may be unacceptable (and set a bad precedent) on
another. For instance, not to require express disclosure of a risk
because it is “usual” (i.e., a matter of “common knowledge”) is a
very different justification from not requiring its disclosure on the
basis that the probability of the consequences occurring, and their
seriousness should they do so, do not warrant requiring disclosure.
Confusion as to exactly which justification is being relied upon
leads, in turn, to confusion in interpreting the resulting judgments.
It is submitted, with respect, that the Supreme Court of Canada’s
decision in Hopp v. Lepp displays some of this confusion.52 For
instance, the Court holds that “[p]robable risks, which must be
disclosed, have been contrasted with mere possibilities (as, for
example, risks involved in any operation), but this dichotomy
cannot be absolute because it ought to take note of whether a risk
51 It could be asked whether there are any risks in category 3 which would
not be included in category 4 and therefore whether category 3 is super-
fluous. However, even if there are not any additional risks added by category
3, it may be important to assess which risks should be disclosed from the
standpoints of both the reasonable physician and the reasonable patient,
as this may give a more complete analysis. In other words, in determining
which risks are material to a reasonable patient, it
is relevant to ask
which risks should be disclosed according to competent medical practice
in the same circumstances. Moreover, such an approach reflects
the
historical development of standard-setting by the courts in relation to the
physician’s duty of disclosure.
52 Supra, note 26, 71-2, 75, 77, 79, 801.
McGILL LAW JOURNAL
[Vol. 26
is or is not quite remote, and here the gravity of the consequences,
if a risk should materialize, must be brought into account”. 3 Is
the Court speaking here of justifying non-disclosure of a risk on
the basis of a low frequency, taking into account the seriousness
of the consequences if the risk eventuates, or is the justification
one of common knowledge that the particular risk is involved in all
operations? The two justifications are not of the same nature and
content, although in this passage the Court could be interpreted
as equating them.
b. Therapeutic research
There is one further therapeutic situation in which the scope of
disclosure should be considered, that in which therapeutic research
is undertaken. The definition of medical research is a difficult and
controversial matter.5 4 However, therapeutic research may be dis-
tinguished from “pure” therapy; although in both there is an
intention to benefit therapeutically the particular patient on whom
it is carried out (which is the feature distinguishing therapeutic
research from non-therapeutic research),
in therapeutic research
the procedure has research aspects such as novelty, or it cannot
be said in advance that it has a reasonable chance of success, or
it is part of a formal research endeavour such as a controlled trial.
Cryderman v. Ringrose55 and Zimmer v. Ringrose”, are two of
the few cases in which a court has considered therapeutic research.
In both cases the procedure involved was sexual sterilization using
an experimental method. With the exception of the trial judge in
the Zimmer case, all the courts treated the procedure undertaken
as therapeutic, and one query is whether they were correct in
doing so, as purely contraceptive sterilization has usually been
classified as non-therapeutic. Moreover, in the Zimmer case, the
Court of Appeal of Alberta held that the procedure used “was
experimental only in the sense that it represented an innovation
in sterilization techniques which were relatively untried. 5 7 In other
words, the Court seems to have held that the procedure was therapy
and was not research, both of which findings, it is respectfully
submitted, are open to question.
5 Ibid., 80.
54 See Somerville, supra, note 21.
55 (1978) 89 D.L.R. (3d) 32 (Alta C.A.). The decision affirms the judgment
of Stevenson D.C.J. and the report includes the reasons of the trial judge.
56 (1981) 28 A.R. 69 (C.A.), rev’g in part (1978) 13 A.R. 181 (S.C., T.D.): leave
to appeal to the Supreme Court of Canada denied (1981) 28 A.R. 92 (S.C.C.).
57Supra, note 56, 77.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
763
The above approach was consistent with that taken in Cryder-
man v. Ringrose, where the Court of Appeal of Alberta affirmed
the trial judgment which rejected the application of Halushka v.
University of Saskatchewan-8 as being “of very limited assistance
because it involves a case of pure medical experimentation and
different considerations must there apply.” 59 It was held that
“[w]hen an experimental procedure is employed the common law
requires a high degree of … disclosure to the patient of the fact
that the treatment is new and risky.”60 The standard of care re-
quired, presumably in relation to performance of the procedure
and disclosure of information, is “that of a reasonable medical man
considering all the circumstances …
. [However,] [t]he disclosure
must be critically analysed when a new procedure is prescribed
by its very inventor.”61 These holdings must now be interpreted in
conformity with the recent Supreme Court judgment on the
standard of disclosure in the “pure” therapy context, as the Court
of Appeal of Alberta does in the Zimmer case: it is clear that the
appropriate standard of disclosure would not be that of a “reason-
able medical man”, but would be what the reasonable patient-
subject would want to know in the circumstances. Further, it is
suggested, “the circumstances” should be identified as being at
least those of therapeutic research. But how full is “full disclosure”
when therapeutic research rather than “pure” therapy is involved,
and does the content of the disclosure differ depending on whether
the research is therapeutic or non-therapeutic?
Clearly, full disclosure in the research context will require dis-
closure that research is being undertaken and, in this respect, even
if no other factors differed, the disclosure required may be regard-
ed as “fuller” than that required in relation to “pure” therapy. It
is also likely that whatever the theoretical similarities of the re-
quired scopes of disclosure for therapeutic research and “pure”
therapy, in practice courts would be more stringent with respect
to what is required to be disclosed in the research context and less
willing to allow justification of any non-disclosures. Further, a
patient-subject must be advised whether any given research pro-
cedure is therapeutic or non-therapeutic. This is also a relevant
circumstance which could alter what would be material information
to a research subject in all the circumstances in which he finds
58 (1965) 53 D.L.R. (2d) 436 (Sask. C.A.).
59Cryderman v. Ringrose, supra, note 55, 41 per Stevenson D.C.J.
0 Ibid.
01 Ibid., 41-2.
McGILL LAW JOURNAL
[Vol. 26
himself.62 Despite the holdings in Cryderman v. Ringrose and
Zimmer v. Ringrose, it is suggested that the following approach
should be adopted: any non-disclosure of information in the thera-
peutic research situation which would be required to be disclosed
in the non-therapeutic context, must be justified, if any such
justification exists,6 on some basis other than simple reliance” on
the fact that in one case the research is therapeutic and in the
other non-therapeutic.
Because, in the past, courts have varied the required scope of
disclosure according to whether or not an intervention was thera-
peutic, and because this distinction may still be of some relevance
in this respect, it is worth examining what courts have said about
the scope of disclosure in cases of non-therapeutic interventions.
2. Non-therapeutic interventions
a. Non-therapeutic research
The scope of disclosure required for non-therapeutic research
interventions is well stated in the judgment of Hall J.A., speaking
for the Court of Appeal of Saskatchewan, in Halushka v. University
of Saskatchewan:64
[Tihe duty imposed upon those engaged in medical research … to those
who offer themselves as subjects for experimentation … is at least as
great as, if not greater than, the duty owed by the ordinary physician or
surgeon to his patient … The subject of medical experimentation
is
entitled to a full and frank disclosure of all the facts, probabilities and
opinions which a reasonable man might be expected to consider before
giving his consent.95
The standard of disclosure just cited applies to all non-thera-
peutic medical research interventions whether these are carried out
on patients or on healthy volunteer subjects, as was the case in
Halushka. In other words, it is important to distinguish between
identifying the situation as being one in which therapy will be
undertaken, and identifying whether or not a particular inter-
vention carried out in that situation is intended to be therapeutic
62 To the contrary, the Medical Research Council of Canada, in Ethics in
Human Experimentation [Report No. 6]
rejected any distinction
between
therapeutic and non-therapeutic research with respect to ethical
requirements, including the obtaining of consent. It would be open to a court
to adopt these guidelines as expositive of generally accepted practice as
to the appropriate standard of disclosure required in relation to aill medical
research.
(1978)
63See the discussion of therapeutic privilege, infra.
64Supra, note 58.
65 Ibid., 443-4.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
765
for the patient on whom it is performed. One may have a healthy
or a sick volunteer subject, but in either case, if the intervention
is non-therapeutic, it is the most exacting standard of disclosure
which applies. There may be an understandable tendency to classify
a situation by its overriding characteristic, and when a sick person
is involved this is almost invariably therapy. In such circumstances
there is a danger that certain non-therapeutic interventions may not
be indentified as such and that therapeutic research may be regard-
ed as “pure” therapy, resulting in failure to apply the more stringent
standard or standards of disclosure that these require.66
interventions
b. Non-therapeutic non-research interventions
One of the most common non-therapeutic
is
aesthetic or cosmetic surgery. Although courts have characterized
such interventions as therapeutic, on the basis that they are psy-
chologically beneficial, the tendency to adopt this approach has
lessened. In many cases the aesthetic surgery involved was clearly
not within the usual meaning of “therapy”, but courts felt com-
pelled to identify it as such as it was thought that non-therapeutic
surgical interventions were illegal.67 Whether or not this belief was,
or still is, correct, it is certainly not upheld in practice, as aesthetic
surgery is manifestly commonplace. It is arguable that the law
has changed to allow non-therapeutic interventions if certain con-
ditions are fulfilled. The most important of these conditions are
that the subject of the intervention give his fully informed consent,
that the intervention not be contrary to public policy or “public
order and good morals” and, possibly, that the subject is adult.68
In a recent British Columbia case, Petty v. MacKay,”9 the patient,
whose occupation was that of “an exotic dancer”, alleged negligence
on the part of the defendant doctor in failing to inform her of the
risks of plastic surgery in the form of a “modified abdomino-
plasty”. Although the judge paid very little direct attention to the
required scope of disclosure in relation to such interventions, he
referred to it in passing when speaking on the issue of causation.
It was held that the plaintiff had not proved causation, as
a reasonable and prudent person, in the circumstances of the plaintiff,
including the knowledge and experience of the plaintiff and her desire
G For a discussion of the disclosure standards required in therapeutic
research see text at note 54 et seq.
67 See Somerville, Medical Interventions and the Criminal Law: Lawful
or Excusable Wounding? (1980) 26 McGill L.J. 82.
68 Ibid.
69 (1979) 14 B.C.L.R. 382 (S.C.) (emphasis added).
McGILL LAW JOURNAL
[Vol. 26
to attain a state of cosmetic perfection or near-perfection, would have
gone ahead with this operation even if the risks involved had been fully
discussed with that reasonable and prudent person prior to the opera-
tion.70
There is no indication what the judge intended by “fully discussed”
in this statement, but it has been suggested that, as the judge in-
corporated the report of a specialist in plastic surgery into his
judgment, he may have been accepting the “professional disclosure
standard”:” If this is true, the standard of disclosure now required
would be the more exacting “reasonable patient” or “full disclosure”
standard adopted by the Supreme Court. As this would take into
account that undertaking the plastic surgery involved here was
clearly a non-therapeutic intervention, it may be that a “fuller” or
more demanding standard of disclosure would apply than for a
therapeutic intervention. 2
One of the difficulties which used to exist when a “professional
disclosure” standard applied to therapeutic interventions and a
“full disclosure” standard to non-therapeutic ones, was that the
patient would sometimes request that an aesthetic, non-therapeutic
procedure be performed in the course of what was otherwise a
therapeutic intervention. This tended to cause confusion in the
standard to be applied, especially in determining whether risks of
procedures, such as anaesthetics, which were necessary for both the
therapeutic and non-therapeutic intervention, were then subject to
a “professional” or a “full disclosure” standard. In Kelly v. Hazlett
just this situation existed,73 and the Court applied the “professional
disclosure” standard to all aspects of the procedure which was
undertaken.74 This difficulty has now been solved with the uniform
adoption of the “full disclosure” standard.
70 Ibid., 388.
71 Picard, [Comment: Petty v. MacKay] (1979) 10 C.C.L.T. 99.
72 See the recent case of Hankins v. Papillon [1980] C.S. 879, 881 per Roth-
man J.: “In cases of plastic surgery … where the decision to be made by
the patient is more subjective and personal than therapeutic, I believe the
doctor has a duty to be especially careful to disclose completely all the
risks and, certainly, any special risks, as well as the consequences for the
patient should such risks materialize.”
73Note that the Court in Kelly v. Hazlett, supra, note 4, 315, held that
“cosmetic values have to be weighed carefully in the balance. There is more
than mere improved appearance involved … [a cosmetic procedure may be]
done for the ‘functioning of the individual in society’.” The Court seems
to accept that the latter purpose may be therapeutic, but it did not find
that this was the surgeon’s aim in undertaking the cosmetic procedure in
this case.
74Ibid., 319-20.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
767
The discussion throughout Part A relates to the “normal” stand-
ard of disclosure required by the law in order to obtain “informed”
consent from competent adults for medical interventions
to be
performed on themselves. It is now necessary to ask under what
conditions this “normal” scope of disclosure may be validly restrict-
ed in relation to this same class of persons.
B. Restricting the scope of disclosure: waiver and “therapeutic
privilege”
1. Therapeutic interventions
Risk that should be disclosed under the “normal” scope of
disclosure requirements, as outlined above, may still be exempted
from disclosure either because of a valid waiver by the patient
of the right to be informed, or pursuant to the doctrine of thera-
peutic privilege.
Allowing a patient the right to waive his right to be informed
is part of the full recognition of a patient’s right to autonomy or
self-determination. However, it should be clear that the patient
has exercised his right of choice in deciding not to be informed
and, consequently, waiver should usually be express rather than
implied, and should occur in a situation where the physician has
made it clear that he is willing to inform the patient.75
The doctrine of therapeutic privilege has been described as
having such a broad operation that it is thought to apply when
informing the patient will cause him to “become anxious and
apprehensive and so nervous that he might be reluctant …
to go
forward with a procedure that is necessary.” 76 It is submitted, how-
ever, that a preferable approach is that the physician may rely on
therapeutic privilege to justify a non-disclosure of risks where the
reasonable physician in the same circumstances would have be-
lieved that the disclosure, in itself, would physically or mentally
harm the patient to some significant degree. The doctrine should
not apply if the only reason for the non-disclosure is that it may
cause the patient to refuse treatment that the physician regards as
necessary.
As has been noted elsewhere, such an approach does not seem
inconsistent with recent statements by the Supreme Court of
75 See the discussion of the scope of consent, sections IV and V, infra,
particularly in relation to Kelly v. Hazlett and the extension of the scope of
consent, or waiver of information, clause, referred to at p. 309 of that case.
76 McLean v., Weir, Goff & Royal Inland Hospital [1977] 5 W.W.R. 609, 621
(B.C.S.C.) (testimony of Dr Goff, accepted by the Court), aff’d [1980] 4
W.W.R. 330 (B.C.C.A.). See Picard, [Comment] (1977) 3 C.C.L.T. 87.
McGILL LAW JOURNAL
[Vol. 26
Canada.77 In Hopp v. Lepp, the Court had the following to say
about the scope of the doctrine of therapeutic privilege:
No doubt, a surgeon has some leeway in assessing the emotional condi-
tion of the patient and how the prospect of an operation weighs upon
him; the apprehension, if any, of the patient, which may require placat-
ing; his reluctance, if any, to submit to an operation which, if the
surgeon honestly believes that the operation is necessary for the preserva-
tion of the patient’s life or health, may demand a detailed explanation
of why it is necessary.78
Further, Chief Justice Laskin, speaking for the Court, was
far from persuaded that the surgeon should decide on his own not to
warn of the probable risk … if the course of treatment contemplated
is administered. A surgeon is better advised to give the warning, which
may be coupled with a warning of the likely consequence if the treat-
ment is rejected. The patient may wish for a second opinion, whatever be
the eminence of his attending physician. It should not be for that physi-
cian to decide that the patient will be unable to make a choice and, in
consequence, omit to warn him of risks.79
In Reibl v. Hughes, the Supreme Court acknowledged that “it may
be the case that a particular patient may, because of emotional
factors, be unable to cope with facts relevant to recommended
surgery or treatment and the doctor may, in such a case, be
justified in withholding or generalizing information as to which
he would otherwise be required to be more specific.”80 However,
after considering the facts of the case, the Court held that the
normal disclosure requirements had not been modified as “there
was no evidence that the plaintiff was emotionally taut or unable
to accept disclosure of the grave risk to which he would be exposed
by submitting to surgery.’
Thus, while recognizing the existence of the doctrine of thera-
peutic privilege, the Supreme Court seems to have taken a rather
restrictive approach to its application. The physician may be re-
quired to disclose to the patient and then placate and explain to
avoid adverse effects of the disclosure, rather than avoid a dialogue
that may cause anxiety to the patient, and possibly the physician.
There is a suggestion that therapeutic privilege may only operate
when someone else, presumably another physician, agrees that its
application is warranted. Moreover, it would seem that before the
privilege will apply, account must be taken of whether any adverse
effects of disclosing the risks of the proposed procedure to the
77 See Somerville, supra, note 31.
78 Hopp v. Lepp, supra, note 26, 77.
79Ibid., 79-80.
8OReibl v. Hughes, supra, note 1, 13.
81 Ibid., 34.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
769
patient can be off-set by disclosing to him the risks of not having
the procedure. Despite the somewhat mandatory language used
by the Supreme Court in this regard, if such a set-off did not seem
likely to occur with respect to a particular patient, or if disclosing
the risks of not having treatment could additionally harm the
patient’s physical or mental health, then, presumably, therapeutic
privilege would apply. Finally, the Supreme Court indicates that
the privilege is not always complete; that is, when the patient can
cope with the information in a more generalized form this must be
given, and in such cases total non-disclosure will not be justified.
Any reliance on the doctrine of therapeutic privilege should be
clearly distinguished from other justifications for non-disclosure.
For instance, in a British Columbia case, McLean v. Weir,82 an expert
witness, a physician-radiologist, in giving evidence for the purpose
of determining the required scope of disclosure of risks, stated that
certain “complications are known to have occurred, but they are
rare and we do not tell the patient that those things are possible
because we do not expect them to happen.” 3 This evidence was
referred to in the judgment at first instance, but in such a way that
the distinction between a non-disclosure justified by therapeutic
privilege and that justified by low probability of the risk occurring,
was not clearly drawn. Further, the Court of Appeal of British
Columbia may have continued this confusion by approving the
physician-radiologist’s approach which “avoided rare and unexpect-
ed complications which would serve only to make the patient
anxious and apprehensive”. 4 Although it is most unlikely to have
been intended, this statement of the Court of Appeal could be in-
terpreted as formulating a justification for non-disclosure that re-
quires the presence of both the criteria mentioned. That is, non-
disclosure of certain risks would only be justified if they were both
“rare and unexpected” and would “make the patient anxious and
apprehensive”. 85
It is submitted that what falls within therapeutic privilege and
therefore constitutes an exception to the “normally” required scope
of disclosure, should be determined according to both the “reason-
able patient” and the “subjective patient” standards. More precisely,
estimates of the effect of the disclosure must be made in relation
to the reasonable patient and if non-disclosure is indicated it should
s2 Supra, note 76.
83 Ibid., 622.
84 Supra, note 76, 336.
85 This section of the discussion is taken from Somerville, supra, note 31,
where McLean v. Weir, Goff & Royal Inland Hospital is analyzed in detail.
McGILL LAW JOURNAL
[Vol. 26
then be considered whether this is also justified in relation to the
particular patient, taking into account subjective factors of which
the physician is aware or ought to be aware. Conversely, if disclosure
is indicated according to an “effect of the disclosure on a reasonable
patient” test, non-disclosure may still be justified by subjective con-
siderations. The subjective test thus determines whether or not
therapeutic privilege justifies a particular non-disclosure. But the
suggested two-stage approach, because it also requires an objective
assessment, is important in determining as precisely as possible
whether any given non-disclosure falls within this doctrine.8s
The doctrine of therapeutic privilege is usually regarded as
being a particular example of the defence of necessity. Even if this
is correct, therapeutic privilege differs from most examples of the
necessity defence in that it is the patient’s interests that are both
invaded and purportedly promoted by the otherwise legally action-
able conduct, whereas, in general, a defence of necessity operates
when one person’s interests are invaded to promote
those of
another. However, the principle underlying the defence of necessity
is that the conduct in question is justified as it is undertaken to
avoid a greater harm, despite the fact that it is otherwise prohibited
by the law and may cause harm. This is also true of a justification
of therapeutic privilege and, consequently, even if the latter is not
an example of the defence of necessity, it does have in common with
it this underlying principle.87
It is interesting to note that in Kelly v. Hazlett the doctrine of
therapeutic privilege is stated as being relevant to a cause of action
in negligence and not one in battery. The judge only refers to
negligence when he says that there may be no cause of action if
the non-disclosure was “justified by reasonable medical considera-
tions”.88 It is therefore arguable that a doctor could not withhold
all information, as compared with some information, under a
justification of therapeutic privilege. If this line of reasoning were
8GAn alternative way of reaching the same result without using a two-
stage approach is to hold that the information which may be withheld from
the patient is to be judged according to the effect of the information on a
reasonable patient in the same circumstances as the patient, and that such
“circumstances” include subjective factors related to the patient of which the
physician knew, or ought to have known.
87Although it may seem pedantic to be concerned about whether or not
therapeutic privilege is an example of the defence of necessity, this could
instance, if it is not of this nature, it will not be subject
be important. For
to the same criteria for application or of limitation as the defence of
necessity.
88 Kelly v. Hazlett, supra, note 4, 313.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
771
followed it would mean that a physician must at least disclose the
basic nature and character of the procedure and its inevitable con-
sequences, even if there were a privilege, in the circumstances, with
respect to non-disclosure of risks. If therapeutic privilege is ex-
cluded in the battery situation this demonstrates another way in
which it differs from the defence of necessity, as the latter can
apply to excuse battery. This is shown in the medical context by the
way in which courts and legal commentators use the defence of
necessity to justify emergency medical interventions where the
patient is unable to consent.8 9 This difference may be just another
indication that the doctrines of necessity and therapeutic privilege
are not of the same nature. But even if this is true, the possible
reasons for excluding the application of the doctrine of therapeutic
privilege in relation to a cause of action in battery merit further
consideration.
One reason for adopting such an approach would be to prevent
physicians from, in effect, imposing treatment on patients by
withholding certain basic information. Courts and commentators
in many jurisdictions have vacillated when facing the problem of
what to do about the competent adult patient who refuses life-
saving treatment. Some juristsP have proposed that when a patient
refuses consent in such circumstances a physician may be able to
override his will without incurring legal liability because a defence
such as necessity applies. But if, as suggested above, the defence
of therapeutic privilege is inapplicable, the physician may be liable
for acting without giving the patient certain basic information,
where this is possible. In other words, there could be circumstances
in which, because the defence of necessity applies, a physician is
justified in overriding a patient’s decision, but the physician is not
justified in failing to inform the patient about the intervention
which constitutes that overriding, because “therapeutic privilege”
does not apply. This distinction could be important in protecting
rights of the patient, such as autonomy, because if he is informed
9 AIthough it should be queried, in light of the previous discussion,
whether or not the use of the defence of necessity is appropriate to justify
emergency medical interventions as, similarly to the situation when the
doctrine of therapeutic privilege applies, it is the patient’s interests which
are both invaded and promoted in intervening in an emergency. The use
of necessity to justify emergency interventions may represent an extension
by the modem law of the defence of necessity and may explain why tradi-
tionally the justification relied upon in relation to such interventions was
implied consent.
0OSee, e.g., Mayrand, L’inviolabilitd de la personne humaine (1975), Nos
38-40, pp. 47-9.
McGILL LAW JOURNAL
[Vol. 26
the patient may at least challenge the justification for overriding his
wishes.
The previous discussion demonstrates the precision with which
any justification for contravening a patient’s rights must be identi-
fied, if those rights are to be protected to the full extent required
by the law. In particular, even if therapeutic privilege is a species
of necessity, it does not apply in the same conditions as other forms
of the necessity defence. Hence, it is inadequate and could set
dangerous precedents to allow blanket defences simply on the basis
of a wide and undifferentiated doctrine of necessity. Rather, each
situation must be examined to see whether the particular necessity
claimed fulfills the conditions required to make it sufficient legal
justification.
The approach taken above contemplates that in certain circum-
stances a patient’s right to make a decision concerning himself may
not be respected, but that failure to honour this right cannot take
the form of obtaining the patient’s compliance by withholding in-
formation from him or not telling him that his decision is not
being respected.
Alternatively, excluding the application of therapeutic privilege
in a battery action may be a recognition of a patient’s right to
exercise his right to refuse treatment. More explicitly, to the extent
of the information required to be given to avoid a cause of action
in battery, the law may be requiring that the patient’s autonomy
be respected as the predominant value and, to this degree at least,
that the patient be informed in order to exercise his right to refuse
treatment.
No matter which of these two possible explanations is accepted,
they both demonstrate that there are limits beyond which the law
will not allow the individual’s right of self-determination to be
ignored, at least by way of the justification of therapeutic privilege.
A further enquiry is whether a defence of “therapeutic privilege”
can ever apply in a medical research situation. It is submitted that
the answer is negative. If a physician is unable to obtain informed
consent to a research intervention, because to inform the patient
adequately may physically or mentally harm the patient, then the
physician may not rely on the doctrine of therapeutic privilege and
carry out the research. In such a situation, the patient must be
given some form of accepted treatment, in which case the doctrine
of therapeutic privilege is potentially available.
The only extremely rare and exceptional case in which the
doctrine of therapeutic privilege may apply to therapeutic research
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
773
would be where the patient’s life was in danger, or his health seri-
ously threatened, and the research intervention was the only hope of
helping the patient, and the conditions for the application of
therapeutic privilege were otherwise fulfilled. However, considera-
tion is needed before allowing such an exception: it could too easily
make dying persons the subjects of research interventions without
their fully informed consent.
Finally, it should be noted that as therapeutic privilege operates
as a legal justification or defence, the burden of proof of its
applicability will be borne by the person seeking to rely upon it –
the physician.
2. Non-therapeutic interventions
Neither waiver nor therapeutic privilege ever apply to non-
therapeutic medical interventions. In Halushka v. University of
Saskatchewan the Court had the following to say in relation to a
non-therapeutic medical research context:
There can be no exceptions to the ordinary requirements of disclosure
in the case of research as there may well be in ordinary medical practice.
The researcher does not have to balance the probable effect of lack of
treatment against the risks involved in the treatment itself. The example
of risks being properly hidden from a patient when it is important that
he should not worry can have no application in the field of research.91
This rule will be applied most easily in non-therapeutic research
and in relation to interventions such as cosmetic surgery. More
difficulty may arise when, for instance, a cosmetic procedure is
combined with a therapeutic one, as was the case in Kelly v. Hazlett.
The Court in that case expressly stated that therapeutic privilege
was inapplicable. The basis for this holding was not only that there
was “no danger in alarming the patient”,92 but, that “since the
operation was for cosmetic purposes only, it would have been
prudent to attempt to alarm the patient”.” Thus the Court relied on
a rule that the doctrine of “therapeutic privilege” did not apply to
non-therapeutic interventions and “non-therapeutic” was defined in
the traditional, strict sense of the term.
C. Extending the scope of disclosure: the patient’s questions
The first case on informed consent to reach the Supreme Court
of Canada, Hopp v. Lepp,94 involved the effect of a patient’s question
on the required scope of disclosure. Unfortunately, the full content
O1 Supra, note 58, 444.
92 Supra, note 4, 319.
93 Ibid.
94 Supra, note 26.
McGILL LAW JOURNAL
[Vol. 26
of the question that the plaintiff-patient alleged he had asked was
not explicit, but needed to be implied from evidence of a conversa-
tion between the patient and the surgeon, which was open to more
than one interpretation. The Supreme Court of Canada disagreed
with the majority of the Court of Appeal of AlbertaP5 as to what
implications could be drawn from this. As a result, the Supreme
Court reversed the Court of Appeal’s holding that the normal scope
of the duty of disclosure had been extended by the patient’s en-
quiries. However, all judges agreed, and it has never been question-
ed, that a physician must answer any specific questions posed by
the patient, even when they relate to matters that would otherwise
not require disclosure. Consequently, the patient has the power to
extend
the normal disclosure
standard.
the scope of disclosure beyond
It is worth noting that the majority of the Court of Appeal of
Alberta were relatively willing to find a specific question about the
risks and seriousness of a proposed operation in the plaintiff’s
query as to whether or not he would be as well to have the proposed
operation in Lethbridge as in Calgary. Despite the fact that the
Supreme Court rejected this implication as an unwarranted extra-
polation from the evidence, it does indicate an approach a court
may take in another case. In other words, the question which must
be answered by the physician may not always be explicit; it may be
implied from a statement or question of the patient. Presumably, a
question could also be inferred from the patient’s conduct in that
it either would have put the reasonable physician on notice, or did
provide the physician with subjective knowledge, that the patient
required additional information. Although it would be undesirable
and unfair to physicians to take this extension of the normal scope
of disclosure by implied questions too far, it is not unreasonable
to expect a physician to give the patient the additional information
that he knew, or a reasonable physician in those circumstances
would have known, that the patient was seeking.
Finally, it must be asked what would be the effect of a patient’s
claiming that the scope of disclosure was extended, or even that the
normal scope was maintained by his questions, and the physi-
cian’s arguing that the scope was restricted, or that the specific
question need not have been answered, on the basis of therapeutic
privilege. As deception should, almost certainly, never be allowed
in the physician-patient relationship and as, in these circumstances,
even a failure to reply may be deceptive or give rise to mis-
representation, it is submitted that the doctrine of therapeutic
95 Ibid.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
775
privilege would not apply. In such circumstances the physician could
tell the patient that he considers it better for the patient not to
discuss the matter, and he could seek the patient’s agreement to
this. Such an approach would amount to waiver by the patient of
his right to be informed, rather than reliance by the physician on
the doctrine of therapeutic privilege as a defence to not informing.
However, great care would have to be exercised to avoid coercing
the patient in this situation, and it would seem that the patient
could prevent the application of therapeutic privilege where it
would otherwise apply by his express request for information.
Therefore, to the extent that the patient by his questions extends
or maintains the normal scope of disclosure, this would override
any application of the doctrine of therapeutic privilege restricting
that scope.
It is possible to modify the diagram used previously to re-
present the normally required scope of disclosure, to include ex-
tensions or restrictions of that scope:
SCOPE OF DISCLOSURE
0%
1) Subjective physician standard
–
–
1(a) Therapeutic privilege or waiver
2) Objective physician standard
3) Normal (reasonable patient) standard
— —–
-.
3(a) Patient’s questions
–
4) Subjective physician standard
100%
It can be seen that the normal, reasonable patient standard for
determining the scope of disclosure, Standard 3, can be restricted
or enlarged, theoretically to requiring that anywhere between 0%
and 100% information be given. 0
9 Positions 1(a) and 3(a) are arbitrarily chosen here simply for purposes
of exposition.
McGILL LAW JOURNAL
[Vol. 26
III. The patient’s reception of the disclosure
Leaving aside questions of involuntariness or coercion, as Cana-
dian courts have not yet been called upon to deal with these matters
in relation to consent to medical care,97 the issue now raised is
whether the physician has fulfilled his legal duty to a competent
adult patient with respect to obtaining that patient’s consent simply
by disclosing all the information that the law requires to be disclos-
ed. Does the law require the patient to display certain characteristics
upon receiving information before the consent based upon that
information is considered valid? The two characteristics most often
considered in this respect are understanding and rationality. The
fomer relates to input criteria, the latter to outcome ones.
A. Understanding
Must the patient understand the information given by the physi-
cian in order to give consent or informed consent? This question
was raised in Kelly v. Hazlett and was dealt with separately inr
relation to battery and negligence.
The plaintiff-patient, in that case, had been given 100 mg of
pethidine just before she purportedly gave consent. In relation to
battery-avoiding consent the judge did not believe
that it could be suggested otherwise than that the giving of the consent
under such circumstances, at the very least, leaves the validity of the
consent open to question … and that it would be incumbent on the
defendant to prove affirmatively that the effect of this sedation probably
did not adversely affect the patient’s understanding of the basic nature
of the contemplated operation.98
In other words, the surgeon must show
that the combination of sedation, and … [the patient’s] labile condition,
had not blotted the information from her mind, respecting the basic
nature and character of the operation when she made her demand. In
such circumstances he has shown a sufficient consent to avoid liability
on the basis of battery. 9
But, if the plaintiff
did not know the basic nature of the operation … and … all she was
asking for was a result not a procedure, -and she manifested this lack
of knowledge to the defendant, then her apparent consent to the opera-
tion, notwithstanding her clear desire for the result, would be inef-
fective.’ 00
97For discussion of these questions, see Somerville, supra, note 2.
98 Supra, note 4, 317 (emphasis added).
99 Ibid., 318.
100 Ibid., 317.
19811
STRUCTURING THE ISSUES IN INFORMED CONSENT
777
In other words, if the doctor knows or if he ought to know that the
plaintiff does not understand the basic nature of the operation, the
doctor is, in the absence of other justification, liable in battery if
he performs the operation. Moreover, proof of battery-avoiding con-
sent by the physician will require proof of the necessary degree of
understanding of the information by the patient.101
With respect to the patient’s understanding of any collateral
risks that the physician must disclose in order to avoid liability
in negligence, the defendant surgeon in Kelly v. Hazlett admitted
that he “could sense that … [the plaintiff] was not understand-
ing what … [he] was attempting to communicate with her by
words.”‘1 2 That is, in this particular case, the surgeon had subjective
knowledge of the patient’s lack of understanding. The Court held
that it is the doctor’s “duty to be satisfied that … [the risk] had
been brought home to the patient before he could reasonably regard
her apparent consent as being valid.’ 1 3 The consent needed to avoid
liability in negligence, that is the consent required in relation to
collateral risks, “involves both awareness and assent”.1 However,
“it would be quite unreasonable, and the law does not call for it, to
expect the doctor to see into the mind of the patient to satisfy him-
self that the patient not only understands the risks but also puts the
degree of emphasis on them which the doctor considers to be
reasonable”.0 5
Although the remarks of the Court in Kelly v. Hazlett in both the
battery and negligence contexts could be read as requiring actual
subjective understanding by the patient of the disclosed informa-
tion, it is clear that this is not the case, as the physician may rely
on the patient’s consent if he could reasonably have thought at that
time that she was aware of the basic nature and character of the
special risks of the operation. 10 6 Hence the physician may rely qn
101 Ibid.
102 Ibid., (emphasis added).
103DIbid., 319. Note that “apparent consent”
is to be contrasted with
“apparent understanding” (see discussion of the latter term, infra). “Apparent
consent” seems to mean that consent
there is
apparent subjective understanding of the information on which it is based.
is only apparent unless
104 Ibid., 318 (emphasis added).
105 Ibid.
106 Ibid., 317-8. There is a statement in ‘the majority judgment of the
Ontario Court of Appeal in Reibl v. Hughes (supra, note 1, 24) which can
be compared with this approach. The Court held that although the trial
judge “rejected the defendant’s explanation that the plaintiff was aware
of a risk of a stroke as a risk of the surgery”, he made “no specific finding
of credibility and indeed does not disbelieve the defendant’s evidence that
McGILL LAW JOURNAL
[Vol. 26
the patient’s consent if it is given pursuant to apparent subjective
understanding by the patient of the information disclosed. It
is
submitted that this is the most satisfactory approach. A fully sub-
jective standard requiring actual understanding by the patient is
too onerous for physicians and may not be in the best interests of
patients, as the patient may not want to understand the information,
or to make the intellectual effort to understand it, or have the
physician bothering him to ensure that he does understand. Thus
a physician may rely on the consent of a patient as a defence to an
action in battery if the physician shows that a reasonable physician
in those circumstances would have thought that this patient ap-
parently understood the basic nature and character of the operation,
provided, always, that the particular physician had no subjective
knowledge that this patient did not understand. When a patient
sues in negligence for breach of a physician’s duty to inform him
and alleges that the breach consists in the physician’s failure to
ensure that he, the patient, understood the information, the plaintiff-
patient must prove on the balance of probabilities that a reason-
able patient in the same circumstances as those in which the
plaintiff found himself would not have understood the information
communicated to him, or that he did not understand the informa-
tion, and the physician knew this.
This approach can be examined in view of Reibl v. Hughes. At
the trial level the test of when the requirement of understanding
of information by the patient will be fulfilled was formulated in
different terms with respect to battery and negligence. It was held
that “the law of battery in effect places on a physician a strict duty
to explain to his patient, in language which the patient can un-
derstand, the essential nature and quality of the treatment he is to
undergo.’ 1 7 To avoid negligence liability, on the other hand, the
doctor must have “take[n] sufficient care to convey to the plaintiff
and assure that the plaintiff understood the gravity, nature and
extent of risks specifically attendant on the [procedure] “.108
It is
he thought the plainhiff understood the risk.” There is no discussion of
whether the physician’s belief would be tested against an objective or sub-
jective standard to determine whether it would be sufficient to enable him
to rely on the consent of the patient as a legal defence, but it appears that
the majority of the Court allowed a subjective standard in this regard. That
is, if this physician thought
the required
disclosure (rather than the test being whether the reasonable physician
would have thought this in the circumstances), the patient’s consent will
be valid as far as requiring the patient to understand the information is
concerned.
the patient understood
that
0 7 Supra, note 46, 311 (emphasis added).
1
108 Ibid., 314 (emphasis added).
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
779
not exactly clear what standard of care is being required of the
physician here in order to avoid negligence liability for non-disclo-
sure, but it seems to vacillate between one of taking all reasonable
means that a reasonable physician would take to ensure under-
standing and one of actually requiring that this result be achieved.
In comparison, it is much clearer that actual understanding by the
patient of the information is necessary to obtain battery-avoiding
consent.
This approach should be contrasted with that of the majority
of the Court of Appeal in the same case. In discussing whether the
patient understood the purpose for. which the surgery was being
undertaken, a matter which would relate to battery-avoiding con-
sent, it was held that “[i]f as the patient said, the doctor did his
best to tell him about the surgery, and the patient had some dif-
ficulty in understanding it, there was some obligation to have told
the doctor what troubled him.”‘1 9 This approach is acceptable to
the extent that it maintains that the test for validity of the patient’s
consent, as far as non-understanding of information is concerned,
should be determined in the absence of subjective knowledge on
the part of the physician that the patient does not understand, on
the basis of whether the reasonable physician, taking all the cir-
cumstances into account, would have thought the patient un-
derstood. But, to the extent that it establishes only an obligation
on the physician to ensure understanding by the patient if the
patient expressly indicates that he does not understand, it should
not be accepted. To make the content of the physician’s duty depend
on the patient fulfilling some obligation, such as asking questions,
may overlook the power and status differential
in the doctor-patient
relationship. From a practical point of view, such a power imbalance both
makes the patient less likely to ask questions or to understand what he
is told and makes him reluctant to disclose this to the doctor. Further,
the patient may not even know enough to ask appropriate questions,
or to know he does not understand the answers, or he m~y be too
emotionally upset to realize this.110
The proposed test requiring “apparent understanding” of informa-
tion by the patient avoids such difficulties, as it would require the
physician to assess at least, as a reasonable doctor would, whether
the patient apparently understands the information he has been
given, and to act accordingly.
It should also be asked what the trial judge in Reibl v. Hughes
meant by qualifying the doctor’s duty as “strict” as far as the
100 Supra, note 1, 21.
110 Somerville, supra, note 31.
McGILL LAW JOURNAL
[Vol. 26
patient’s understanding of the information necessary to give battery-
avoiding consent was concerned. This characterization can be
compared with the way in which the Court described the require-
ment of understanding in relation to the negligence-avoiding disclo-
sure, that the physician took “sufficient care” in giving information
to the patient.
It is suggested that the word “strict” may have two connotations
in this context. First, the physician may be said to have what, in
civil law terminology, is referred to as an “obligation of result”.’
This would mean that the physician does not fulfill his legal duty
merely by taking all steps that a reasonable physician would take
in the circumstances to ensure that a patient understands the basic
nature and character of the proposed procedure (which would
amount to an obligation of “means”) but, rather, the physician
must ensure the result that the patient does understand. The
Supreme Court of Canada reviewed the evidence given in this case
by the plaintiff-patient as to his understanding of the information
disclosed to him by the physician and held that “it must have been
obvious to the defendant [physician] that the plaintiff had some
difficulty with the English language and that he should, therefore,
have made certain that he was understood”.” 2 The first point to
note about this statement is that any duty of the physician to
ensure actual understanding of information by the patient is not
limited to issues of battery. In fact, the statement was made only in
relation to a cause of action in negligence for non-disclosure, as
the Court had already decided that no cause of action in battery lay
on the facts. Read literally, the Supreme Court’s statement means
that the physician must ensure subjective understanding of the
information by the patient. This is more stringent as regards the
obligation placed on the physician than the test of “apparent un-
derstanding”. With all respect, the Supreme Court’s approach
places too onerous an obligation upon the physician and may not
be in the best interests of patients, if not in relation to the infor-
mation which needs to be disclosed to obtain battery-avoiding con-
sent, at least in relation to that required to avoid negligent non-
disclosure. As explained above, it amounts to an obligation of
result –
that is, an obligation to ensure a particular eventuality,
in this case understanding, rather than an obligation of means
which is an obligation to take all the steps that a reasonable physi-
11 See Crepeau, Le contenu obligationnel d’un contrat (1965) 43 Can. Bar
Rev. 1.
“12Reibl v. Hughes, supra, note 1, 34 (emphasis added).
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
781
cian in the circumstances would take in attempting to ensure un-
derstanding. For this reason, it is suggested that the Supreme Court
statement should be read only in relation to the case then before it.
Where a physician has subjective knowledge that a person is not
completely familiar with the language in which he is being ad-
dressed, then the physician, as a reasonable physician, must take
greater care than would otherwise be required in attempting to
ensure understanding. Rather than requiring subjective understand-
ing by the patient, this approach maintains the objective standard
of requiring the physician to do what a reasonable physician Would
do in the circumstances, and modifies those circumstances to take
into account the patient’s difficulties with the language.
The second possible interpretation of the word “strict” in Reibl
v. Hughes”3 may be that the physician is strictly liable for all
untoward results if he acts without the patient’s informed and
understanding consent as to the basic nature and character of
the intervention. In other words, the physician will be subject to
a risk of strict liability rather than fault liability with respect to
such results. For recovery of damages, this would mean that even
where the untoward results were not caused by any fault of the
physician, but were simply the crystallization of one of the un-
avoidable risks of the procedure, the physician would be liable to
compensate for these if he acted without consent. Although the
battery can itself be regarded as constituting fault on the part of
the physician, the point is that there is no need to establish the
presence of fault in the sense of negligence in undertaking or
carrying out the procedure or in failing to obtain the patient’s
consent to the running of risks, in order to recover compensation
for risks which eventuate. 114 Moreover, once liability in battery
113 See Reibl v. Hughes, supra, note 46, 311 (H.C.).
114This situation is interesting from the point of view of legal analysis.
Liability is often characterized as either fault, or strict or no-fault. The
former usually refers just to negligence liability. Battery was originally
a tort of strict liability (in that the necessary and sufficient criterion for
liability was simply directness of the physical contact). Later, and possibly
in connection with the element of wilfulness or intention becoming relevant
to battery, a defence was recognized in that the defendant was not liable if
he could prove that the tort occurred “utterly without his fault” (Weaver
v. Ward (1616) Hob. 134, 80 E.R. 284 (K.B.)). The point is, therefore, that
fault is relevant to whether a person will be held liable in battery, but
it is not the same degree or nature of fault as in negligence. Thus, in a
sense, battery may be described as a stricter liability than negligence, but
it is not strict liability in the sense that presence or absence of fault is
totally irrelevant.
McGILL LAW JOURNAL
[Vol. 26
is established, the damages recoverable will often be more extensive
than those recoverable in negligence, as the test of causation is
one of directness, rather than reasonable foreseeability, which may
be more limiting. A recent case on point is Allan v. New Mount Sinai
Hospital.”5 This case did not involve a defect of understanding by
the patient, but a direction by the patient that her left arm was
not to be used for intravenous administration of an anaesthetic
agent, which the anaesthetist ignored. The Court found that there
was no negligence on the part of the anaesthetist and yet awarded
damages to the plaintiff for all the untoward results of the pro-
cedure, whether they were reasonably foreseeable or not. The basis
for this holding was that as the plaintiff had specifically requested
that the procedure not be carried out in the way in which it was,
to contravene her wishes constituted battery and, consequently, all
directly resulting damages were recoverable.
Finally, the relationship between understanding of information
by the patient and coercion or duress should be considered. Some-
times a patient fails to understand information that he must be
given, because that information is given after the patient has re-
ceived medication and the medication affects his ability to com-
prehend what he is being told. Alternatively, the patient may un-
derstand such information but the medication causes him to act
upon it in a way in which he would not have acted, had he not re-
ceived the medication. Some cases fail to distinguish these two
types of defect and in either situation treat the patient’s consent
as defective because of the coercive effects of the medication.
Coercion or duress means that the patient has consented against
his will, and his consent is voidable. Such consent is to be contrasted
with decisions taken in the absence of will (e.g., in a state akin to
automatism), in which case the purported consent is void ab initio.
In a sense, lack of understanding is more akin to a partial absence
of will than coercion, unless the lack of understanding was deli-
berately induced in order to coerce the person’s decision. In law,
this distinction could be important not only because the nature of
the defect in the consent varies in the two cases, but also because
conduct that is contrary to someone’s will is often more repre-
hensible than that done in the absence of his will, and this could
affect such matters as the award of punitive or aggravated damages.
Thus, is the effect of the medication on the patient of such a nature
and degree that a purported consent given after receiving it is really
given in the absence of the patient’s will and therefore the consent
115 Supra, note 13.
19813
STRUCTURING THE ISSUES IN INFORMED CONSENT
783
is void?”” Or has the effect of the medication been to limit the
patient’s ability to understand what he is being told to the extent
that his consent is defective because of lack of understanding? Or
has the medication made the patient unduly susceptible to being
influenced in his decision? Showing that- the patient reached a
decision which he would not normally have taken may be evidence
of any of the above and it may be important to identify exactly
which situation occurred. The point is that the same circumstances
may raise questions both of lack of understanding and of coercion
or duress, and both should be considered when determining the
adequacy of the patient’s consent.
B. Rationality
Does the law require rationality”17 of the patient’s decision as a
substantive element of a valid consent?
The first consideration is the relationship between understand-
ing and rationality. If, as argued above, the law requires that the
patient apparently understand the required disclosure of informa-
tion in order to give a valid consent, does this mean that the law
is seeking to promote rationality of the patient’s decision and,
further, if his decision is adjudged irrational, may it be ignored or
overridden on this basis?
Even if it is accepted that understanding of the required disclos-
ed information is being mandated in order to promote rationality,
it must be asked whether this means rationality of the decision-
making process or rationality of the decision itself or both. Although
understanding may promote rationality in both of these respects,
it is usually only the rationality of the decision outcome that is
relevant to the law, and understanding may not be an essential
116 See Kelly v. Hazlett, supra, note 4; Beausoleil v. La Communautd des
Soeurs de la Charitg de la Providence, supra, note 13.
“7It is being presumed here that there are certain objective criteria
according to which a decision can be classified as rational or irrational, but
what these criteria are is open to debate. For instance, in The Framing of
Decisions and the Psychology of Choice (1981) 211 Science 453, 458, Tversky
and Kahneman state that “the modern theory of rational choice has adopted
the coherence of specific preferences as the sole criterion of rationality.
…
[But] [c]onsistency is only one aspect of the lay notion of rational
behavior. … [Tihe common conception of rationality also requires that
preferences or utilities for particular outcomes should be predictive of the
experiences of satisfaction or displeasure associated with their occurrence.
Thus, a man could be judged irrational either because his preferences are
contradictory or because his desires and aversions do not reflect his
pleasures and pains.”
McGILL LAW JOURNAL
[Vol. 26
condition for this. It is quite possible for a decision to be judged
rational by an objective bystander, when the reasons on which it
was based were quite irrational. Moreover, the law’s requirement
of understanding of information by the patient may be seen as
promoting autonomy rather than rationality. In this case, to re-
quire rationality either of the patient’s decision-making process
or its outcome would be to contradict directly the value of self-
determination which is being promoted by requiring understanding,
because self-determination requires recognition of the competent
patient’s right, for no matter what reason or on what basis, to
determine what shall be done to himself.
to
limits
It is submitted that the preferable approach is to view un-
derstanding as promoting autonomy, rather than rationality. Any
irrationality of a decision-making process or
legal
decision outcome should then be set by declaring the person
factually or legally incompetent. Thus, the right to autonomy would
mean that the competent patient could make irrational decisions
concerning himself without the law overriding such decisions.
Further, in order to give proper scope to such a rule, it is necessary
to recognize that irrationality of the decision-making process or
of decision outcomes does not of itself indicate incompetence,
although in some circumstances it may be evidence of this.
The issue of irrationality of the patient’s decision was considered
in Kelly v. Hazlett. The defendant surgeon gave evidence that he
considered the plaintiff “irrational”, that is, “irrational from the
point of view of not being able to think the way that [he, the
doctor] was thinking, which [he] thought was more rational”.” 8
The actual irrationality referred to was the patient’s decision to
undergo the operation only if the cosmetic procedure were in-
cluded. In the result the Court held that the patient’s “apparent
consent” was insufficient to protect the surgeon from liability in
negligence on the basis of failure to obtain informed consent. It is
not exactly clear how much the irrationality of the patient’s deci-
sion influenced his holding, but it seems that such irrationality
should at least put the physician on notice that the patient’s
“decision was not based upon any knowledge or appreciation of the
risk”,” 9 in which case the physician may not rely on the consent as
being valid. This case is probably a demonstration of a court looking
to the rationality of the patient’s decision to indicate both whether
the patient had the required understanding of the risks that must
11 Supra, note 4, 316.
“1 Ibid., 318.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
785
be understood in order to give informed consent, and whether the
physician had, or ought to have had, knowledge of any lack of
understanding on the part of the patient.
In summary, the question of rationality in matters of consent
is difficult. To allow (or even more so, require) second-guessing of
the patient’s decision according to whether that decision is rational
may seriously detract from autonomy (which it is the purpose of
consent to protect). The more a person’s decision deviates from
what the person assessing that decision would decide in the same
circumstances, and the more serious the consequences of that de-
cision, the more likely it is that the person making the decision
will be labelled incompetent and his decision irrational.120 Rather
than judging the rationality of a patient’s decision and validating
or invalidating consent on that basis, the better solution (which
was probably the approach of the Court in Kelly v. Hazlett) is to
adopt understanding by the patient of the information required to
be disclosed as the necessary safeguard. Pursuant to such an ap-
proach, provided the patient is otherwise judged to be competent,
if the physician has no subjective knowledge that the patient lacks
understanding and the reasonable physician would have believed
that the patient apparently understood the information disclosed,
the resulting consent may be relied upon as valid whether or not the
patient’s decision is considered rational.
In a paternalistic physician-patient relationship, consent was
often the imprimatur of the doctor’s rational decision-making on
the patient’s behalf. Under a doctrine of informed consent the aim
is to enable the patient to make a decision on his own behalf. A
remaining question is how far the physician is justified in carrying
out a patient’s informed, irrational decisions. The physician is far
less likely to be acting within legal or ethical limits when imple-
menting such a decision requires a positive intervention on his
part, than when it is a situation in which he must desist from
violating the patient’s physical or mental integrity against his will.
More explicitly, a patient’s irrational refusal of treatment should
be accorded greater respect than his irrational demand for it.
IV. Who must disclose
There are two questions involved here. Does a referring doctor
have an obligation to disclose to a patient any risks of a procedure
to be carried out by the physician to whom the patient is referred?
120 See Roth, Meisel & Lidz, Tests of Competency to Consent to Treatment
(1977) 134 Am. J. of Psychiatry 279.
McGILL LAW JOURNAL
[Vol. 26
To what extent may a physician rely on the consent obtained on
his behalf by someone else?
It has been held in two British Columbia cases 12 1 that the re-
ferring physician does not have any duty to disclose the risks of
a procedure to be carried out by another physician to whom he has
referred the patient. In both cases the referral was to a more
specialized medical practitioner who, presumably, would be more
familiar than the referring doctor with the risks to be disclosed to
the patient. It has been argued elsewhere12 2 that such an approach
may be justified in these circumstances, but that there could, in
some instances, be a danger in extending this reasoning to situa-
tions where a more junior colleague carries out the procedure,
which is commonly the case in instances of medical treatment by
an alternative doctor.
It has also been suggested that referral to a specialist can be
regarded as part of the treatment undertaken by the first physician
and that he has some general duty of disclosure in this situation,
such as to warn the patient to ask about risks of the other pro-
cedure. 2′ – However, a statement by the Court in Strachan v. Simpson
rejects such an approach. It was held that there was “no claim
… for failure to warn, because it
against [the referring doctor]
… about a pro-
was not his duty to warn [the plaintiff-patient]
cedure to be carried out by [the surgeon to whom he was re4
ferred] “24
In this case the referral was to a specialist surgeon
and not to a junior colleague, and the ruling can therefore be res-
tricted to such a situation. Moreover, a distinction can be drawn
between a duty to warn about risks of a procedure and a duty to
warn the patient to ask about such risks.
The appropriateness of delegating the duty to obtain informed
consent, which is a common practice in Canada, does not appear
to have been in issue as yet in any Canadian case. Except in cir-
cumstances where one physician has referred the patient to another
and, on this basis, has argued that he did not have a duty either
of disclosure or to obtain informed consent, no physician has
pleaded that his duty was fulfilled by someone else, or that liability
for failure to obtain the required consent did not lie with him
because he had non-negligently delegated the task of obtaining it.
Such an argument would be rejected because it is clearly the duty
121 McLean v. Weir, Goff & Royal Inland Hospital, supra, note 76; Strachan
v. Simpson (1980) 10 C.C.L.T. 145 (B.C.S.C.).
122 Somerville, supra, note 31; Picard, supra, note 76, 90.
123 See note 122, supra.
124 Strachan v. Simpson, supra, note 121, 173.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
787
of the intervenor either to obtain consent himself or to ensure that
adequate consent has been obtained. Consequently, to the extent
that a delegee failed to obtain consent, the intervenor would be
liable for an intervention performed without adequate consent.
One legal analysis of the approach outlined above would be that,
even if performance of the physician’s duty to inform the patient
is delegable, liability for non-performance of that duty is non-deleg-
able. In other words, the physician has a duty to ensure that the
patient has been adequately informed, even if in law he need not
carry out the informing process himself. Expressed in another way,
the physician is directly liable for breach of his own duty of ensur-
ing that consent is obtained, rather than vicariously liable for his
delegee’s failure to obtain informed consent. Consequently, for the
physician to be liable, it is not necessary to establish a bond of
authority, such as a master-servant relationship, between the physi-
cian and his delegee. Under such an analysis the breach of duty
by the delegee quantifies the damage caused by the physician’s
breach of duty, and does not establish the liability of the physician
by the mechanism of vicarious liability.
It is also worth noting that disclosure by a physician other than
the one carrying out the procedure may protect the latter from
liability even though he did not know of the disclosure. This will
be true if the court holds, as it did in Davidson v. Connaught Labo-
ratories”5 that despite the physician’s non-disclosure of the risks
of undergoing rabies vaccination, the patient knew of them because
he had been so informed by another physician. Hence, even if the
non-disclosure were a breach of duty by the physician, it had not
caused the plaintiff’s damage, as the disclosure would not have
altered the decision which the patient made. 2 6
V. The consent form
Canadian courts have taken the view that a patient’s signature
to a consent form does not of itself mean that the required consent
has been obtained: 27 as one judge has described it, such a procedure
may be just “the ritual of the signing of the consent form”.128
Although written consent may not be sufficient, it may still be
necessary because, for instance, there is a legal requirement that
125 (1980) 14 C.C.L.T. 251 (Ont. H.C.).
126 See text following note 138, infra.
127 See, e.g., Kelly v. Hazlett, supra, note 4; Lepp v. Hopp, supra, note
26 (C.A.).
128 Kelly v. Hlazlett, supra, note 4, 318.
McGILL LAW JOURNAL
[Vol. 26
consent to surgical operations or general anaesthesia be in writing. 29
Whether or not a written consent is sufficient depends on the
substance of the information, the communication of information
and the reality of the agreement that is symbolized by the patient’s
written consent, and not just on the form or formal expression of
that consent in writing. This distinction is sometimes made in the
law by saying that the written document does not constitute the
consent, but is evidence which must be taken together with all
other evidence in deciding whether or not legally sufficient consent
was present. Consequently, a signed consent form is not, as some
physicians have tended to believe, an effective shield against
liability. Moreover, when a health professional undertakes an inter-
vention on a person from whom he has not personally obtained
consent, he cannot simply rely on a signed consent form as de-
monstrating that the required consent has been obtained. Assuming
that the intervenor may validly delegate the obtaining of the ne-
cessary consent, he must at least ascertain its adequacy or take the
risk of legal liability if it is inadequate.
What are the responsibilities of a witness to a consent form?
The possible purposes of such a signature could be either to witness
that the legally required consent was obtained, or to witness that
the signature on the form is that of the patient. In the absence of
indications to the contrary (such as a statement included in the
form itself, or legislation requiring that the witness certifies as to
the substantive presence of legally valid consent), the witness will
normally only attest that the signature is the patient’s.
VI. Scope of the consent
It is clear that when a patient expressly consents to a certain
medical procedure he does not consent to any other procedure
unless this can be implied from the express consent, his conduct,
or the surrounding circumstances. Any intervention not covered
by express or implied consent would be unlawful unless justified
on some other basis, such as necessity.
Difficult problems can arise in relation to whether or not
there was consent to particular aspects of an intervention, rather
than to the intervention in general. The most common of these
problems occurs when the patient alleges that he consented to a
129 See, e.g., O.-in-C. 3322-72 (8 November 1972), made pursuant to Loi
sur les services de santg et les services sociaux, L.Q. 1971, c. 48 [now L.R.Q.,
c. S-5].
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
789
procedure but only if it were carried out in a particular manner, 3 ‘
or by a particular person. In each case it will be a question of fact
as to what the consent covered. It is suggested that, like the pro-
posed test of apparent understanding of information by the patient,
the extent of the patient’s consent should be determined according
to an apparent scope of consent test: that is, what would the reason-
able physician in these circumstances have thought, or what did
this physician subjectively know, were the limits to the scope of
the patient’s consent? The alternative approach would be to deter-
mine the scope of the patient’s consent by asking what the reason-
able patient in these circumstances would have thought he was con-
senting to? Whether the reasonable physician or the reasonable
patient standard is used, the final determination of the apparent
scope of consent must be reasonable. In other words, the apparent
scope of consent must not be unduly restricted so as to cause un-
easiness among physicians as to what they are entitled to do pursuant
to the patient’s consent; but neither should it be extended in such
a way that it is the patient’s implied, rather than express, consent
that is being relied upon in circumstances where this was not
necessary. Certain matters need express consent. In normal cir-
cumstances a surgeon could not argue, for example, that allowing
another surgeon to carry out a procedure on the patient, when the
patient had not expressly consented to this, was justified, as it fell
within the apparent scope of the patient’s consent. Rather, the
patient must be clearly aware of such substitution, and consent to
it. In the latter situation, the scope of the patient’s consent, with
respect to substitution of the surgeon, is explicit, and there is
apparent consent present. That is, the distinction between the sug-
gested apparent scope of consent rule and apparent consent should
be kept in mind as they are not the same entity.
It should also be kept in mind that this apparent scope of con-
sent test cannot be viewed in isolation but must be seen as operating
in conjunction with other requirements such as the required scope
of disclosure of information to the patient and his apparent un-
derstanding of this information. Hence, before a procedure can
fall within the apparent scope of the patient’s consent, the physician
must have met all disclosure requirements and there must be ap-
parent understanding of this information by the patient.
Does the fact that the patient has acknowledged something in
writing extend the apparent scope of consent where this would not
otherwise be the case? It was noted in Kelly v. Hazlett
130 See Allan v. New Mount Sinai Hospital, supra, note 13.
McGILL LAW JOURNAL
(Vol. 26
that the defendant raises no argument based on the language of the
consent form. Specifically, he places no reliance on “such additional
operations or procedures as are considered necessary or desirable in
the judgment of the surgeon” or on the patient certifying that she is
“aware of the contents and significance of this statement”. 3
Thus the fact of signing a consent form does not, of itself, extend
the apparent scope of consent any more than it constitutes apparent
understanding. It may be that, as a matter of public policy, the
apparent scope of the patient’s consent cannot be extended by
having him agree to clauses as vague and general as the one referred
to in Kelly v. Hazlett. That is,
the patient must give reasonably
specific consent and any justification of an intervention to which
this has not been obtained must be on some basis other than
32
consent.1
The above discussion may be thought to indicate that clauses
such as that in the consent form signed by the patient in Kelly v.
Hazlett are of no legal effect. But this may not be correct as such
a clause may constitute valid waiver of the right to be informed
and thus an alternative justification to informed consent for carry-
ing out the intervention.”‘ For such a clause to constitute a waiver,
the patient must understand that *that is its nature and freely and
voluntarily consent to it
in circumstances where he knows the
physician is willing to inform him. In this case the apparent scope
of the patient’s consent would be extended by such a clause to
include all procedures which fell within its terms although the
131 Supra, note 4, 309.
132 Such a basis would be necessity. For instance, if during a surgical pro-
cedure a surgeon discovers a condition which is immediately dangerous to
the patient’s life or health, or a condition which could be dealt with in a
separate and later operation but a second operation would “unduly” endanger
the patient’s life or health, the surgeon may act without consent under a
doctrine of necessity: see, e.g., Tabor v. Scobee 254 S.W. 2d 474 (Ky 1952).
However, if the necessity should have been anticipated before the operation
being commenced, at which time the patient’s consent could have been
obtained, the surgeon may be liable for failing to obtain consent: see Marshall
v. Curry [1933] 3 D.L.R. 260 (N.S.S.C.). It should be noted that if a clause
such as that in the consent form discussed in Kelly v. Hazlett were only
meant to include situations such as necessity, it would be superfluous as a
defence exists in such circumstances regardless of consent. In fact the clause
cited in Kelly v. Hazlett is much wider than this as it covers “procedures …
considered … desirable in the judgment of the surgeon” (supra, note 4, 309).
133Waiver involves consent to the extent that there must be consent to
the waiver of the right to be informed. But such consent is to be distinguish-
ed from consent to run the risks (or, even, consent to the consequences)
of the intervention itself, which is the usual justification for undertaking an
intervention.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
791
patient had not been informed of the risks of these procedures, or
even consented to the procedures themselves. Thus valid waiver by
a patient of the right to be informed can affect the apparent scope
of his consent.
VII. Proof
The fault alleged in cases involving failure to obtain consent
or informed consent is an omission. This omission either consists
of a non-disclosure of information, or, less frequently, a failure to
obtain consent itself, that is, failure to obtain a legally adequate
expression of the patient’s will indicating that he agrees with the
proposed course of conduct.
In battery the burden of proof of consent is on the physician,3 4
which means that the physician must prove positively the presence
of the necessary and sufficient consent on the balance of proba-
bilities. When the cause of action is negligence the plaintiff patient
must prove absence of the necessary and sufficient consent, also
on the balance of probabilities. In practice the patient may not be
able to do much more with respect to proof than to allege that the
omission said to constitute the fault took place. It is then a matter
of whether or not the court believes the plaintiff. This means that
the physician facing such an allegation will usually have an eviden-
tiary burden of proving the omission to disclose required informa-
tion or obtain consent did not take place. The evidence brought
in such circumstances will include the physician’s own denial of
the alleged omission. Not infrequently courts must decide this
issue simply according to which of the two parties, plaintiff or
defendant, they believe. In negligence actions where the evidence
leaves the court in equal doubt as to whom to believe, the plaintiff
will lose, as he carries the legal burden of proof of the absence of
the necessary and sufficient consent: he necessarily loses his claim
in negligence if he cannot establish failure to disclose the required
information or lack of the necessary consent on the balance of pro-
babilities. Conversely, when the cause of action is in battery, the
defendant physician will lose in cases of equal doubt, as he has the
legal burden of proving the presence of the necessary disclosure of
information and consent.
It has been claimed that “[t]he greatest problem facing a phy-
sician is to prove what he did disclose”.’35 Almost without exception
134 See, e.g., Kelly v. Hazlett, supra, note 4, 318.
135Ferguson, Informed Consent: What the Law Requires (1980) 1 Health
Law in Canada 56, 57.
McGILL LAW JOURNAL
[Vol. 26
a trial takes place years after the event and, as one author has
commented, this may result in the courts’ relying on a formula
such as the following:
[the patient] recalled with precision the exact words she spoke, as well
as the exact words used by the doctor. I accept her evidence. This opera-
tion was a very important event for her, and, consequently, it is one that
she remembers distinctly.
Doctor
did not expressly deny [the patient’s] version
of the conversation with him. For the doctor, of course, it was only one
instance of many thousands of such procedures. Thus, it is understand-
ably very difficult for him to recollect exactly what was said. His
evidence, since he did not, in fact, recall anything that was actually said,
was largely speculative; he merely surmises what he may have said
and what she may have said.13
Clearly, combinations of circumstances such as those raised here
could be a cause of injustice to physicians. The best course may be
for physicians to keep full and careful records regarding their
disclosures of information to patients.
A signed or written consent form is prima facie evidence of con-
sent. Likewise, if a patient had been given reading material, or
audio-visual aids had been used to inform him about a medical
procedure, these could be produced in evidence as proof that the
required disclosure of information was made. It would, of course,
be a further question whether there was apparent understanding of
this information by the patient and, consequently, a legally valid
consent. Where such evidence is produced, one of the factors that
will be assessed in order to determine whether the patient un-
derstood what he was being told will be the complexity of the
information material, including the nature and structure of the
language used. It is worth noting in this respect that a recent
analysis of’consent forms found that they gave information in
language too complex to be readily understood by the average
person. 3 7
It may be thought that a physician could prove that he made
the necessary disclosure to a particular patient by bringing other
patients as witnesses to give evidence that in similar circumstances
the physician made the required disclosure to them. However, the,
problem with such an approach is that in all likelihood it is
barred by the rules on similar fact evidence. These prohibit
proving a fact by proving other similar facts and then drawing an
136 Ibid., citing Allan v. New Mount Sinai Hospital, supra, note 13.
137Grundner, On the Readability of Surgical Consent Forms (1980) 302
New England J. Medicine 900.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
793
inference that the fact to be proved is established because in other
similar cases the like fact has been proven to be present. 138
VIII. Causation
Canadian courts have faced many of the same problems
in
developing a test for determining when damage will be deemed in
law to have been caused by failure to disclose required information
to a patient as they have establishing a test for the required scope
of that disclosure. The two problems are related.
Causation is the necessary link in law between the fault or
breach of duty and the damage which occurs. The usual test of
causation-in-fact, when a single cause is alleged, is the “but for”
test: but for the alleged wrongdoing, would the damage have oc-
curred? If the answer to this question is no, causation-in-fact is
present. The test of causation-in-law (except where the only damage
suffered is “pure” economic loss) is, whether damage of that kind
occurring in that way was reasonably foreseeable as a result of the
breach of duty? If it was, and the damage is not too remote,’3
it
will be recoverable.140
Proof of damage is not necessary to establish a cause of action
in battery and, consequently, neither is proof of causation. But when
damage results from an intentional tort, the test of its recoverability
is the directness of the damage.14′ This test can be regarded simply
as requiring that causation-in-fact be shown. It should be noted
that this means showing only that the touching to which no consent
138 See Cross, Evidence, 5th ed. (1979), 355 et seq. It should be noted here
that such evidence would be admissible if it fell outside the limitations on
introducing similar fact evidence, for instance, if it constituted proof of
custom by habitual conduct on the part of that physician.
130 Damage is not too remote if recovery is not barred by any of the legal
rules on the remoteness of damage. Such a rule is that certain damage is not
of a nature that the law regards as non-compensable.
140 See generally Somerville, A Diagrammatic Approach to Causation (1978)
24 McGill L..
442, 442-3. Causation-in-law may be relevant in two ways: an
essential to establishing a cause of action and to identify which items of
damage will be compensated. These two aspects of causation-in-law may
be separated conceptually, for purposes of analysis, by referring to rules
relating to the former as rules on causation of damage, and those relating
to the latter as rules on remoteness of damage, although it should be
noted that these terms, causation and remoteness, are often used inter-
changeably to refer to either or both those aspects of the causation-in-law
doctrine.
141 See Turner v. Thorne & Thorne (1959) 21 D.L.R. (2d) 29 (Ont. H.C.); see
also Somerville, supra, note 140.
McGILL LAW JOURNAL
[Vol. 26
was given directly caused the damage, and but for the touching
the damage would not have occurred. It is not required to show
also, as in negligence, that but for the failure to obtain consent
the damage would have been avoided.142 By contrast, proof of
damage, and hence proof of causation-in-fact and causation-in-law,
is essential to establishing a cause of action in negligence. Thus
the amount of damages recoverable may differ depending on
whether a cause of action lies in battery or negligence, because
direct but unforeseeable damages, or non-negligently inflicted ones,
are recoverable in battery but not in negligence. 4 ‘ The recent Ontario
case of Allan v. New Mount Sinai Hospital14 4 demonstrates such a
situation. The Court held that all direct damages were recoverable
in battery, which lay because a procedure had been conducted
without consent, despite the fact that there was no negligence in-
volved in undertaking or carrying out the procedure which gave
rise to those damages. The result is that if the plaintiff patient can
prove battery, as in the Allan case, she can recover compensation
for damages resulting from the crystallization of an unavoidable or
unforeseeable risk of the procedure. Moreover, establishing battery
may alter the amount of compensation payable even where the
patient also has a cause of action in negligence for breach of duty
in undertaking or carrying out a procedure, as pursuant to the
latter claim some damages may not be recoverable on the basis
that they were not reasonably foreseeable. 145
Assuming that the tribunal of fact is convinced that a poten-
tially actionable non-disclosure of information has occurred in a
medical context, how does it judge for the purposes of the law
of negligence whether or not the non-disclosure is the cause-in-fact
and the cause-in-law of the damage?
In non-disclosure of information cases, most discussion of causa-
tion problems by the courts has related to establishing causation-
in-fact. Causation-in-law has not usually raised difficulties because
in all cases the patient has alleged that the occurrence of the non-
disclosed risk that should have been disclosed constitutes his
damage. In such cases it is not necessary to be more precise as to
142 It should be queried, however, whether the same test of causation
of damage applies as in negligence, if battery lies because of an intentional
non-disclosure of information relating to the basic nature and character
of an act.
143 Note that some unforeseeable damages are recoverable in negligence
under rules such as the “take your victim as you find him” or “egg-shell
skull plaintiff” rule.
144 Supra, note 13.
145 See generally Somerville, supra, note 140.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
795
whether the patient’s damage consists of having made a different
decision from the one that would have been made if the required
disclosure had been made, or whether it consists of the ill-effects
suffered as a result of the undisclosed risk eventuating. In either
case once it is shown that the non-disclosure is the cause-in-fact of
the patient’s damage, that damage is ipso facto reasonably foresee-
able and hence proof of causation-in-fact also establishes causation-
in-law.
46
The test of whether the causation-in-fact is present is as follows:
if the physician had made the required disclosure, would the patient
have made a different decision about having or foregoing the treat-
ment, and thus have avoided the damage? If the patient’s decision
would have been different, and that different decision would have
avoided the damage, then the non-disclosure is the cause-in-fact
of the damage for the purposes of the law. If proper disclosure
would not have given rise to a different decision, or if a different
decision would not have avoided the damage (provided, in the latter
case, if multiple causes are present that the decision taken was not
a contributing cause to the damage and provided that if no other
sufficient causes had been present the decision taken would not
have been sufficient to cause the damage), then causation-in-fact is
not established. But this seemingly simple formula hides many
difficulties.
The first difficulty has been in deciding whether the test of
causation-in-fact should be subjective or objective. Canadian courts
have vacillated between the two’ 47 and have even suggested a
combination.148 The difference between the two tests is captured
in the different formulations of the basic question for determining
,whether or not causation-in-fact is present. Under a subjective
approach the test is: if the required disclosure of information had
been made to this patient would he, on the balance of probabilities,
have made a different decision regarding treatment? Under an
objective approach the question is: if the reasonable patient in the
patient’s circumstances had been given the required information
would that reasonable patient, on the balance of probabilities,
have made a different decision?
146For discussion of
the difficulties which may arise
in relation to
establishing causation-in-law, see text following note 156, infra.
147 For instance, a subjective test was adopted in Kelly v. Hazlett, supra,
note 4, 320, Strachan v. Simpson, supra, note 121, 179, and an objective test
in Petty v. MacKay, supra, note 69, 392. See elso Picard, supra, note 71.
14BReibl v. Hughes, supra, note 1 (C.A.). For critical comment on this
approach, see Petty v. MacKay, supra, note 69.
McGILL LAW JOURNAL
[Vol. 26
The decision as to which test of causation will be adopted is
probably related to the test that the law will use to set the required
scope of disclosure of information to the patient. When that scope
was determined according to what the reasonable doctor would tell
a patient in those circumstances, a subjective patient test of causa-
tion was generally adopted. This was understandable because a
subjective test, in favouring the patient, balanced any prejudice
that was caused to the patient by the reasonable physician standard
having been applied to determine what he should have been told.
Thus, although the subjective test of causation gave the patient
the benefit of hindsight (and quite clearly the patient would be
relying on such hindsight, as he would not be suing unless he felt
that he would have decided differently if he had been given the
the
information that he thought he should have been given),
court could limit liability of the physician, and in fact the medical
profession could limit its own liability to some extent 149 by not
requiring that the particular matter be disclosed. 50 In such a case
the patient would have no cause of action because there was no
breach of the duty of disclosure, rather than his action in negligence
failing because the defendant’s breach of duty was not the cause-
in-fact of the patient’s damage. But the Supreme Court of Canada
has now adopted a more patient-favouring standard for scope of
disclosure of information, which is what the reasonable patient in
those circumstances would want to know. This change may alter
the decision as to which test of causation it is more appropriate to
employ.
In Reibl v. Hughes’51 the Supreme Court rejected the subjective
test of causation as being too favourable to plaintiff patients as
they would always have the benefit of hindsight and would claim
that their decision would have been different if the proper disclo-
sure had been made. In fact the Court had before it the plaintiff’s
evidence that he would not have accepted the operation if he had
known that the particular undisclosed risk which eventuated could
occur. 1 52 On the other hand, the Court recognized that the problem
140 This is true because when a reasonable-physician standard of disclosure
is used custom in the medical profession and expert medical evidence are
influential in determining that standard.
150 Although a court can still limit liability of a physician by finding that
there was no duty to disclose a certain matter, there is more opportunity
to do this under the narrower reasonable-physician scope of disclosure
standard than under the wider reasonable-patient one.
151 Supra, note 1.
1 2 Ibid., 17 et seq.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
797
with adopting a purely objective test of causation is that one would
assume that the reasonable person would not refuse an operation
which had been recommended by a physician.153 Thus, under a
purely objective approach, the fact that a physician recommended
treatment would automatically establish causation-in-fact. The
Court rejected this. It held that the risks of having surgery and
foregoing surgery must be taken into account to see what the reason-
able patient would decide. Where the benefits of having a procedure
clearly outweigh any risks of doing so and the risks of foregoing it
are serious, then a reasonable patient would be likely to follow his
physician’s recommendation to undergo a certain procedure. In
other words, in reaching a decision, the reasonable patient would
take into account the risks and benefits of the alternatives available
to him and set them off against each other. But the results of this
assessment do not yield a final answer as to whether causation-in-
fact is present. The Supreme Court held that special considerations
affecting the particular plaintiff-patient must also be taken into
account in determining what the reasonable patient would decide.
This is achieved by characterizing such considerations as “circum-
stances” and placing the “reasonable patient” in the same cir-
cumstances as the plaintiff. Such circumstances include matters
that the plaintiff-patient thought important, provided the defendant
physician knew, or ought to have known, of them. The proviso
suggested here is necessary to introduce some measure of objectivity
into what may be taken into account as material circumstances.
As can be seen, the process of developing a rule to test for
the presence or absence of causation-in-fact in medical non-dis-
closure cases is one of continuous adjustment between purely
subjective and purely objective criteria. This can be demonstrated
by a more particular example: disclosure that there was an im-
mediate risk of paralysis from an operation, as compared with a
more delayed risk of the same nature from postponing the opera-
tion, as was the case in Reibl v. Hughes, would be material to the
reasonable patient in reaching a decision whether or not to undergo
the operation immediately rather than later in circumstances where
the patient had approximately a year and a half before being eligible
for a full disability pension from his employer. The physician ought
to know that he should enquire about the patient’s position in such
respects, or should make a disclosure of risks which takes such
possibilities into account, and, if the physician fails to do so, it
appears that he does so at the risk of being held liable to com-
153 Ibid., 16.
McGILL LAW JOURNAL
[Vol. 26
pensate. Consequently, if, as the Supreme Court held, a reasonable
patient in the position of the plaintiff in Reibl v. Hughes would,
on the balance of probabilities, refuse surgery which is recom-
mended treatment from the purely medical point of view, damage,
such as the plaintiff’s loss of his pension, arising from having the
surgery will have been caused in fact (and, as it was reasonably
foreseeable, caused in law) by the non-disclosure of information.
In summary, the test of causation-in-fact can be formulated as
what the reasonable patient in the patient’s particular position
would, on the balance of probabilities, have done if “all material and
special risks of going ahead with the surgery or foregoing it were
made known to him”.”-, And, it bears repeating, “[a] risk is …
material when a reasonable person, in what the physician knows or
should know to be the patient’s position, would be likely to attach
significance to the risk or cluster of risks in deciding whether or
not to undergo the proposed therapy”.155
There are some other points to be emphasized here. First, it
has already been seen that the patient’s questions will affect the
scope of disclosure requirements. They will also affect the deter-
mination of the presence or absence of causation. The patient’s
questions can show factors important to the patient in reaching
his decision about treatment, which may not be important to the
reasonable patient. These factors can be taken into account in
testing causation by making them part of the circumstances in
which the reasonable patient finds himself and which will in-
fluence the latter’s hypothetical decision. Thus to the extent that
the patient informs the physician of subjective factors which would
not be material to the reasonable man, but which are important
to him, the latter can enter into and, therefore, affect the out-
come of the test of causation.
Second, the Supreme Court’s test of causation will tend to
mean that if the medical risks of having or foregoing a procedure
are equally balanced it is for the patient to make up his mind,.
as in this case it is not obvious what a reasonable patient would
decide, and the defendant physician has deprived the patient of
the chance of making a decision on the basis of adequate informa-
tion. In this respect, however, it must be remembered that as the
patient has the burden of proving causation on the balance of
154Reibl v. Hughes, supra, note 1, 16 (S.C.C.).
’55 Hopp v. Lepp, supra, note 26, 80 (S.C.C.), quoting Waltz & Scheuneman,
supra, note 29, 640. This passage was adopted by the Court in Canterbury
v. Spence, supra, note 27.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
799
probabilities, he will be unsuccessful in his claim if the court finds
the evidence to be equally in favour of a reasonable patient, to
whom the required disclosure of information had been made,
making or not making the same decision as the plaintiff. By com-
parison, if, for instance, the risks are strongly in favour of having
surgery, it will be more difficult for the patient to prove causation
–
that is, to prove that he would have refused the operation if
the full disclosure had been made –
unless he can show some
circumstances apart from the medical risks which would affect the
decision of a reasonable patient, and which, one assumes, in fact
affected him, such as was the case in Reibf v. Hughes with respect
to the disability pension. Consequently, as the Supreme Court takes
care to note, the objective test of causation does not mean that the
issue of causation is completely in the physician’s hands. This
correlates with the approach which has been taken to setting the
standard for scope of disclosure of information.
Thus, although the Supreme Court adopts an objective test of
causation, they mean to include within it certain subjective factors
by placing the reasonable patient in the same circumstances as the
particular plaintiff patient. This can be compared with the alterna-
tive objective-subjective approach of the Court of Appeal of On-
tario in Reibl v. Hughes, where it was suggested that first a purely
objective test of causation be applied, followed by a subjective
test of what that particular patient would have decided if the proper
disclosure had been made 56 The subjective test was meant to test
the validity of the decision outcome reached under the objective
test, but, as was pointed out in Petty v. MacKay, if such an ap-
proach is adopted, “[w]hat is the court to do where the plaintiff’s
evidence [as to what his decision would have been if the required
disclosure had been made] is in conflict with the objective test (the
standard employed by a reasonable and prudent person) ? The judg-
ment of Brooke J.A. [speaking for the majority of the Court of
Appeal of Ontario in Reibl v. Hughes] does not, with respect, pro-
vide an answer to this question.’1 5 7 The answer, it is respectfully
submitted, depends on the way in which the patient’s subjective
decision varies from what a reasonable patient would have decided.
If the reasonable patient would have decided in the same way as
the patient decided, then it is irrelevant that the particular patient
would have decided differently if the required disclosure had been
made. A more interesting question is what a court would find with
156 Supra, note 1.
157Supra, note 69, 392.
McGILL LAW JOURNAL
[Vol. 26
respect to causation where it could be proved that, although the
undisclosed information would have affected the decision of a
reasonable patient, it would not have affected the decision of this
patient. It would seem unjust to find causation present and yet,
in tort liability subjective considerations are only
in general,
allowed to override objective ones to the extent that this imposes
greater and not lesser liability on the defendant. However, the latter
statement may need to be restricted to cases where the defendant
seeks to avoid liability imposed according to an objective standard
by arguing subjective factors. It may not apply where the plaintiff
has no claim on a subjective basis but does on an objective one.
In such circumstances the defendant is not avoiding liability by
relying on subjective factors relating to his own conduct; rather,
the plaintiff, in a sense, does not have “clean hands”. To the extent
that this discussion represents a real issue there may be merit in
having a dual objective-subjective test of causation in which sub-
jective factors may sometimes bar a recovery which would have
been allowed under a purely objective standard.
The approach of the Supreme Court in introducing subjective
factors within an objective test of causation is more unitary in
concept than the dualized objective-subjective test outlined above
and is probably more limited as to which subjective factors may be
introduced. The Chief Justice said:
the patient’s particular concerns must … be reasonably based; otherwise,
there would be more subjectivity than would be warranted under an
objective test. Thus, for example, fears which are not related to the
material risks which should have been but were not disclosed would
not be causative factors. However, economic considerations could reason-
ably go to causation where, for example, the loss of an eye as a result of
non-disclosure of a material risk brings about the loss of a job for which
good eyesight is required. In short, although account must be taken of
a patient’s particular position, a position which will vary with
the
patient, it must be objectively assessed in terms of reasonableness. 158
The passage just cited raises some difficulties. What is meant
by “fears which are not related to the material risks which should
have been but were not disclosed”? Does the Court mean to exclude
as causative factors just fears related to risks which did not re-
quire disclosure, or fears related to disclosed material risks, or
both? The former interpretation presents no difficulty. But if the
Court is saying that fears related to disclosed risks can play no
role as causal factors in relation to the patient’s decision, there is
a problem. The reasonable patient could, and almost certainly
15SReibl v. Hughes, supra, note 1, 17 (emphasis added).
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
801
would, take such fears into account as part of the “cluster of
risks”‘ 59 relevant to his reaching a decision about treatment. Con-
sequently, such fears do play a r6le in assessing causation with
respect to an undisclosed risk. It may be that the Court is saying
that disclosed risks can never give rise to a cause of action even
when there has been a non-disclosure. This is a complex causation
problem which will be discussed presently.
The second question raised in relation to the Supreme Court’s
statement is what is meant by saying that the “patient’s particular
concerns must … be reasonably based”. This should only apply to
concerns of the patient of which the physician was not subjectively
aware and of which it would be unreasonable to deem the physician
to have knowledge because such concerns would not be concerns
is submitted that, no matter how un-
of a reasonable patient. It
reasonable a particular patient’s concerns, if the physician knows of
them and if there is a risk relating to these concerns, then if that
risk is not disclosed and the patient undergoes treatment which he
would have refused had that particular risk been disclosed, non-
disclosure of that risk will be the cause-in-fact of its realization.
The discussion so far may be summarized: non-disclosure of
information which there was a duty to disclose will be the cause-
in-fact of damage resulting from a patient’s decision to accept or
forego treatment if, on the balance of probabilities, disclosure
would have caused a reasonable patient in the plaintiff’s position
to alter his decision concerning treatment. The circumstances that
will be taken into account in assessing the plaintiff’s position in-
clude all those relevant circumstances in which the plaintiff patient
finds himself and of which the physician knew or ought to have
known, and will also include any additional subjective factors which
the plaintiff patient made known to the physician as being im-
portant to him in reaching his decision. Referring to the facts in
Reibl v. Hughes, the Supreme Court sums it up this way:
Relevant in this case to the issue whether a reasonable person in the
plaintiff’s position would have declined surgery at the particular time is
the fact that he was within about one and one-half years of earning
pension benefits if he continued at his job; that there was no neurological
deficit then apparent; that there was no immediate emergency making
the surgery imperative; that there was a grave risk of a stroke or
worse during or as a result of the operation, while the risk of a stroke
without it was in the future, with no precise time fixed or which could
be fixed except as a guess of three or more years ahead. Since, on the
trial Judge’s finding, the plaintiff was under the mistaken impression, as
159 See Waltz & Scheuneman, supra, note 27, 640, quoted by Laskin C.J.C. in
Hopp v. Lepp, supra, note 26, 80.
McGILL LAW JOURNAL
[Vol. 26
a result of the defendant’s breach of the duty of disclosure, that the
surgery would relieve his continuing headaches,
in the
opinion of a reasonable person in the plaintiff’s position, also weigh
against submitting to the surgery at the particular time.
… [A] reasonable person in the plaintiff’s position would, on a balance
of probabilities, have opted against the surgery rather than undergoing
it at the particular time. 00
this would
Thus the plaintiff had proved causation-in-fact for the purposes of
a cause of action in negligence by showing, on the balance of proba-
bilities, not that he would never have undertaken this surgery and,
consequently, never have run the risk which eventuated, but that
he would not have undertaken the procedure at the time at which
he did if it were not for the defendant’s negligent non-disclosure.
Further, the plaintiff proved that this variation in the time of the
performance of the procedure and, consequently, of the crystalliza-
tion of the risk which occurred, constituted a damage to him.1′ 1
This last point is important because it demonstrates that simply
showing that at some time in the future the same treatment may
be needed, or the same situation could result without treatment,
will not necessarily operate as a defence against a claim for dama-
ges for negligent non-disclosure of information. This is true because
proof of such facts does not destroy the causal link between the
defendant’s breach of duty, the non-disclosure, and the plaintiff’s
damage, when the nature of the damage alleged is acceleration of
the condition of which the plaintiff complains rather than a claim
that the condition would never have occurred at all except for the
defendant’s breach of duty. Obviously, there are limits with respect
160 Supra, note 1, 35.
161 In Reibl v. Hughes the acceleration of the running of the risk, and
hence of its crystallization, amounted to damage in two ways: first, it
should be regarded as causing economic loss of an aggravated kind as far
as the damage represented by loss of the pension is concerned (as all
persons unable to work usually suffer some economic loss and, consequently,
unless aggravated loss were
required, any acceleration of damage could
be regarded as causing economic loss). Secondly, the patient has the damage
of suffering the disability sooner rather than later. This second loss could
be present in many cases and there is no reason why it should not be
compensable when standing alone. This
in Strachan v.
Simpson, supra, note 121. Exploratory spinal surgery was performed on the
patient. It involved a risk of paraplegia which was not disclosed and which
resulted, but without the operation it was likely that the plaintiff-patient
would have become totally paraplegic within a year. The Court held that
the plaintiff would not have consented to the surgery had the risk been
explained and awarded general damages of $22,000.00 because of acceleration
of the condition of paraplegia, which acceleration
(not the condition of
paraplegia itself) would have been avoided by not having the operation.
is demopstrated
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
803
to defining acceleration of a condition as a head of damage. For
instance, if an operation were in the nature of an emergency, the
plaintiff could not prove causation by characterizing the damage
as acceleration of a certain condition, because to delay the proce-
dure would not be an option open to the reasonable plaintiff.
Further, there may be situations of de minimis acceleration, or
difficulty in proving that any acceleration of a condition resulted
from the intervention and not other causes. In these circumstances
the plaintiff has not proven causation, or even, in some instances,
has not proven damage itself. The latter is true because when the
nature of the alleged damage is acceleration of the occurrence of
a condition, it is necessary to prove damage, that is acceleration,
before looking for a causal link. If the condition would have oc-
curred naturally at the same time, there is no acceleration, and
hence no damage, rather than no causation.16 2
Depending on the nature of the alleged damage, its actual
pathogenesis may be irrelevant to causation as no matter what
this was the physician will be liable. For instance, in Hankins v.
Papillon’ 16 3 the plaintiff-patient underwent a dermabrasion pro-
cedure to remove brown marks from her face, which had been
caused by taking oral contraceptives. These marks reappeared after
the treatment. The Court held that the plaintiff had failed to
prove that the physician had not made an adequate disclosure of
the risks. It noted, however, that even had this been achieved
“the causal link is not free from doubt”. 64 The Court did not
elaborate on this statement, but it raises an important point.
Depending on whether the plaintiff alleges her damage consists
of undergoing the procedure or suffering the reappearance of the
brown marks, there may or may not be causation-in-fact. In the
former case it would be irrelevant to proving causation arising
from a non-disclosure whether the reappearance of the marks was
caused by the procedure which was performed to remove them, or
was simply a continued effect of having taken the oral contracep-
tives, as in either situation the physician’s fault would lie in failing
162 If the cause of the condition occurring when it did was something
other than natural causes, and if this constitutes an acceleration of the
occurrence, it is a further question whether any breach of duty by the
physician contributed to such acceleration. If the physician’s breach of
duty did so it may be a cause, for the purposes of the law, of the damage
constituted by acceleration of the condition, despite the fact that other
sufficient causes were also present.
163Supra, note 72.
164 Ibid., 883.
McGILL LAW JOURNAL
[Vol. 26
to warn of the possibility of such reappearance, and this omission
could cause the plaintiff to agree to a procedure which she would
not otherwise have undergone. In other words, whatever the source
of the reappearance of the marks, the necessary causal link between
the physician’s fault (the failure to inform) and the patient’s
damage (deciding to undergo and undergoing the procedure) will
be present. If, by comparison, the damage consists of the reap-
pearance of the brown marks, then the plaintiff must prove that
this was caused by having the procedure as only then will the phy-
sician’s breach be the cause-in-fact of this damage. A finding as
to causation-in-law can also depend on exactly what is identified
as constituting the plaintiff’s damage. This will be discussed present-
ly.
A further matter relating to causation also merits discussion
here, although it has been canvassed elsewhere. 165 This is the
question of whether it must be the undisclosed risk which eventuates
in order to give rise to a positive finding of causation. More pre-
cisely, suppose that a procedure carries two risks which require
disclosure, risk A and risk B. The physician discloses risk A but
not risk B. Risk A eventuates. But if risk B had been disclosed the
reasonable patient in the patient’s circumstances would have refused
the procedure and avoided risk A. Has causation-in-fact been proven
for the purposes of an action in negligence for non-disclosure? Two
approaches are possible: one can argue that each risk is to be
treated separately, that risk A was disclosed to the patient, that he
agreed to run risk A, that risk A eventuated as the plaintiff knew
it could when he consented to run it, and therefore, no cause of
action based on non-disclosure arises. This may be designated a
separatist approach.16 6 Alternatively, one can argue that the patient
agrees to run a bundle of risks and that he would have avoided
running risk A if risk B had been disclosed to him. Therefore, the
failure to disclose risk B caused him to run risk A and, consequently,
he may recover in a cause of action based on negligent non-disclosure
when risk A eventuates. This can be regarded as a cumulative
approach. The Canadian courts have not been called upon to decide
which approach to adopt. However, it is worth noting that the
Supreme Court of Canada in Hopp v. Lepp, at least without disap.
proval and possibly with approval, cites Halushka v. University of
Saskatchewan to the following effect: “[i]n the view of the Court
165 Somerville, supra, note 31.
160 In Scott v. Bradford 606 P. 2d 554, 559 (Okla 1979) the Court held that
“[a]bsent any occurrence of the undisclosed risk, a physician’s failure to
reveal its possibility is not actionable.”
1981J
STRUCTURING THE ISSUES IN INFORMED CONSENT
805
of Appeal [of Saskatchewan],
there were undisclosed or mis-
represented facts and, it added ‘[they] need not concern matters
which directly cause the ultimate damage if they are of a nature
which might influence the judgment upon which the consent is
based’.’ 16 7 This statement could support the view that causation
can be established in some circumstances when a disclosed risk
eventuates, provided that there was an undisclosed risk present
which should have been disclosed and which would have affected
the decision of the reasonable patient whether or not to undergo
the procedure. On the other hand, the statement of the Supreme
Court already discussed,6 8 that “fears which are not related to the
material risks which should have been but were not disclosed would
not be causative factors”, 16 9 may indicate the contrary.
However, if the cumulative test for establishing causation-in-
fact is accepted, there is a further problem involving causation-in-
law. Assume that there has been a non-disclosure of risk A that
was required to be disclosed, and if it had been disclosed the
reasonable patient in the same circumstances would have refused
the procedure, and risk X, which did not require disclosure,
eventuates. Under the cumulative approach to causation-in-fact,
non-disclosure of risk A is the cause-in-fact of risk X eventuating.
But is it the cause-in-law? As stated previously, the test of causation-
in-law is whether that type of damage occurring in that way was
reasonably foreseeable. Hence, the answer to this question in any
given circumstances depends first on what a risk being reasonably
foreseeable means and, second, on how the damage and the damag-
ing event is described.
The concept of reasonable foreseeability is wide. A reasonably
foreseeable risk is one which “would occur to the mind of a reason-
able man in the position of the [defendant] and which he would not
brush aside as far-fetched”.170 A risk may be reasonably foreseeable
although it is extremely unlikely and may be described as only a
remote possibility..7 “A risk of injury which is quite unlikely to
occur … may nevertheless be plainly foreseeable.”‘172
17 Supra, note 29, 78, quoting Hall J.A. in Halushka v. University of Sask-
atchewan, supra, note 58, 445.
168 See text at note 151, supra.
169 Reibl v. Hughes, supra, note 1, 17.
170Overseas Tankship (U.K.) Ltd v. The Miller Steamship Co. Pty [The
Wagon Mound (No. 2)] [1967] 1 A.C. 617, 643-4 (P.C.) per Lord Reid.
171 Wyong Shire Council v. Shirt (1980) 29 A.L.R. 217 (Aust. H.C.).
172 Ibid., 221.
McGILL LAW JOURNAL
[Vol. 26
The problem of describing the damage and the damaging event
in order to ask whether the nature of X and the way in which it
occurred were reasonably foreseeable results of the defendant’s
conduct must now be considered. In this respect the more general
the terms in which a court, in the exercise of its discretion, frames
the description of what needs to have been reasonably foreseen,
the more likely it is that reasonable foreseeability of the damage,
and causation-in-law, will be found.
Reasonable foreseeability of the way in which the damage oc-
curred will not usually be in issue in cases concerning non-dis-
closure of medical information, because if the damage itself is
foreseeable, in general, the way in which it occurred will also be
foreseeable. The exception will be when a known type of risk
occurs, but arises from an unknown source.’ 7
3 In such cases, de-
pending on how generally or precisely the court describes the
damaging event, for exemple, either as paralysis arising in the
course of an angiogram or as paralysis arising from an unknown
source of the risk of paralysis during an angiogram, the way in
which the damage occurred will or will not be foreseeable, res-
pectively. 7 4
The fact that there may be an issue as to what constitutes the
patient’s damage in relation to non-disclosure of information has
already been considered. Does the damage consist of the patient’s
decision to run and then running known or unknown risks (includ-
ing X), which he would not have run if the required disclosure had
been made? Or is the patient’s damage constituted by occurrence of
X? If the damage is of the former nature, then proof of causation-
in-fact will automatically establish causation-in-law, because the
damage is necessarily foreseeable as a result of the non-disclosure.
Under this approach, proof of X will simply quantify the amount
of damages the patient can recover. It may be that the reason why
there is no discussion of causation-in-law, as distinct from causation-
in-fact, in any of the Canadian cases on informed consent, is that
the courts are adopting such an approach. It is impossible to be
categorical about this, however, because the fact situations so far
presented have not made the distinction critical.
By contrast, if the damage is constituted by the occurrence of
X, the question for establishing causation-in-law is whether or not
X was a reasonably foreseeable risk of the procedure which the
plaintiff underwent as a result of the defendant’s breach of duty.
173 McLean v. Weir, Goff & Royal Inland Hospital, supra, note 76.
174 See Somerville, supra, note 31.
1981]
STRUCTURING THE ISSUES IN INFORMED CONSENT
807
The answer may not always be clear, though in this respect it is
important to note that a distinction must be made between risks
which require disclosure and risks which are reasonably foresee-
able: the latter is a much broader class.
to characterizing
Consequently, under either approach
the
patient’s damage, causation-in-law’ may be present respecting the
occurrence of a risk that did not require disclosure when there has
been a breach of duty in failing to disclose another risk. It bears
repeating that this argument depends on accepting that the pre-
sence of causation-in-fact does not depend on the undisclosed risk
eventuating. If this approach is not adopted, then a discussion of
causation-in-law is irrelevant except with respect to undisclosed
risks which eventuate and form the basis of a claim for negligent
non-disclosure, as only in such cases will causation-in-fact be present.
Finally, one approach to avoiding some of the more complicated
causation problems arising from non-disclosure of information
would be to adopt a concept similar to the civil law doctrine of
“loss of a chance” (“la perte d’une chance”).175 This approach means
that the damage complained of would not be the eventual injury
(which merely quantifies- the amount of damage caused), but the
loss of the chance of acting with full knowledge of the options
and alternatives available. 176 In effect this theory makes the same
event –
both the wrongful act or breach of
duty and the damage itself. Consequently, the necessity to find the
causal link between the breach of duty and the resulting damage is
eliminated. The doctrine, as developed in French law, postulates that
the physician owes a duty to the patient to give him all chances of
being cured and the physician will be liable if he deprives the
patient of such chances. The extension suggested here is that liability
could also lie for failing to give the patient a chance to choose,
when he was entitled in law to have a choice. To some extent, this
approach introduces a discretion that allows a court to award
damages in cases where it feels that it is just to do so but where
there are difficulties in finding proof of causation of damage, as
traditionally required. Whether it is a good idea to introduce such
notions into the law will remain a matter of debate.
the non-disclosure –
’75See Boyer Chammard & Monzein, La responsabilitg m~dicale (1974),
92-105.
176 This doctrine could be regarded as an overt expression and formaliza-
tion of the decision which must be taken as to what constitutes the patient’s
damage. The doctrine is equivalent to holding that the patient’s damage
consists in having made and followed a decision to have or forego treatment
in the absence of a proper disclosure of information, rather than in the
occurrence of a risk of that procedure.
McGILL LAW JOURNAL
[Vol. 26
Conclusion
There has recently been much development in the law relating
to informed consent in the medical relationship. But many un-
resolved issues remain, even in the fact situations that have so far
been presented before Canadian courts, and the full range of
potential fact situations involving informed consent and the issues
which they can present are very extensive.
It is most important, at this early stage, to ensure that decisions
involving informed consent are not made on an ad hoc basis.
Rather, fundamental principles must be carefully worked out and
their necessary and intricate variations studied. This can be done
best by structuring a framework of questions and subquestions
that must be considered in a certain order. The approach outlined
in this paper represents one such framework.
The question of informed consent has raised complex and in-
teresting legal, ethical and policy issues. This development is far
from finished. The Supreme Court has now firmly established the
foundations of the house that Canadian law will build in this area
and, possibly, has completed the external structure. But there is
still much left to be done. We should strive for a structure that is
unified, harmonious and livable in the view of patients, the com-
munity in general, physicians and lawyers. We must also keep in
mind that the structure itself needs a certain amount of built-in
flexibility to accommodate changes or new additions that will in-
evitably be needed in this rapidly changing area. Unless such
flexibility is incorporated damage to the structure can result from
changes both in technology and in the mores of a given society.
The aim must be to develop a basic structure which can accom-
modate and incorporate such changes without itself being destroyed.
