The Influence of Physician Explanations on

Patient Preferences-about Future

Health-care States
DENNIS J. MAZUR, MD, PhD, DAVID H. HICKAM, MD, MPH

Objective. To determine the influence of alternative explanations by physicians of the

purpose of a medical intervention (intubation and ventilatory support, IVS) on three
types of patient preferences: desire for IVS, the length of time patients would find IVS
acceptable, and the minimum probability of a good medical outcome patients would
require before assenting to continued support with IVS. Methods. Structured interviews
were conducted with patients followed in a continuity care general medicine clinic at
the Department of Veterans Affairs Medical Center, Portland, Oregon. Patients were
asked to consider whether they would accept IVS at a future time. Patients were
randomly assigned to one of two explanation conditions that differed in terms of their
future medical contexts, that Is, the degrees of specification of the nature of the medical
condition patients were asked to consider. The general-explanation group was asked
to consider the future medical context of “an unspecified medical condition”; the spe-
cific-explanation group was asked to consider the future medical context of “a severe
pneumonia.” Patients were asked three questions: 1) Would you accept IVS (yes or
no)? 2) How long would you allow your physician to continue IVS?; and 3) After being
on IVS for 2-3 days, what would be the minimum chance of recovery from the con-
dition you would require before agreeing to continued IVS? “Chance of recovery” was
defined for both groups as the probability that the patient would be able to leave the
hospital and be able to take care of activities of daily living unassisted with minimal
change in his or her mental state from the pre-hospitalization status. Results. Of 186
patients (mean age = 66.5 years; mean education = 12.7 years), 97 received the
general explanation and 89 received the specific explanation about a severe pneu-
monia. Significantly fewer (p = 0.03) patients receiving the general explanation wanted
physicians to intervene with IVS (general 94% vs specific 100%). Patients receiving
the general explanation were willing to accept significantly fewer (p = 0.009) days of
intubation (general 65 days vs specific 96 days). Significantly fewer (p c 0.0001) pa-
tients receiving the general explanation wanted physicians to continue IVS when the
probability of a successful outcome was less than 50% (general 30% vs specific 64%).
Conclusion. Physician explanations-based on general (unspecified medical condi-
tion) vs specific (severe pneumonia) explanations-have a marked influence on the
duration of IVS patients would permit and the probability of a good outcome required
to continue IVS. Key words: advance directives; life support; patient preferences;
patient-physician decision making. (Med Decis Making 1997;17:56-60)

When patients cannot make decisions on their own,
it helps to have reports about their attitudes to guide
decision making. 1-3
3 Elicitation of these attitudes de-
pends on physicians’ providing information. Such
discussions are the basis for advance directives, that
is, instructions regarding patients’ preferences for
future interventions. These communications be-
Received September 26, 1995, from the Medical Service and
HSR&D Program, Department of Veterans Affairs Medical Cen-
ter, Portland, Oregon. Revision accepted for publication April 10,
1996.
Address correspondence and reprint requests to Dr. Mazur:
Medical Service (111-P), Department of Veterans Affairs Medical
Center, 3710 S.W. U.S. Veterans Hospital Road, Portland, OR
97201.
tween a physician and a patient may be achieved by
oral discussion alone or can be aided by predrafted
forms and statements that help to articulate to the
patient the multitude of issues involved in decisions
about future medical interventions.4-7
While it is important to ascertain patient prefer-
ences about specific medical interventions well be-
fore future health care states occur, there are seri-
ous problems in how these preferences are elicited.
Patient preferences regarding acceptance or rejec-
tion of a medical intervention have been shown to
be influenced by whether the information provided
to them was in a survival or mortality format.8 Mal-
loy et al.9 found that the language used in advance
directives influenced patient preferences. When the
medical intervention was described in negative

56
VOL 17/NO 1, JAN-MAR 1997
terms (emphasizing risks, side effects, and harsh
phrasings of mechanical ventilation, e.g., “machine
that controls your breathing”), preferences for life-
sustaining treatment were lower than when the in-
formation was presented in positive terms (empha-
sizing benefits of mechanical ventilation).
The previous studies examined only simple pref-
erences to accept or forego life-sustaining treat-
ment. We have conducted a study designed to gain
a more detailed understanding of how variations in
medical explanations affect the patient preferences
that could be captured in advance directives. We
studied three types of patient decisions involving a
specific technologic life-support intervention, intu-
bation and ventilatory support (IVS): 1 whether pa-
tients are willing to accept IVS, 2 ) the number of
days patients would be willing to undergo IVS, and
3) after two to three days of IVS, at what probability
of recovery from the condition would patients allow
their physicians to continue supporting them with
IVS. We also sought to expand upon the previous
research by examining how the amount of medical
detail provided in physicians’ explanations influ-
ences patients’ preferences.
Methods
This study was approved by the Subcommittee on
Human Studies of the Department of Veterans Af-
fairs Medical Center, Portland, Oregon. Study pa-
tients were being followed in a continuity care gen-
eral medicine clinic. Medical records were reviewed
prior to patients’ being asked to participate in the
study. All patients with active cognitive and psychi-
atric problems were excluded from study partici-
pation, as were all patients in moderate or severe
emotional or pain states. Informed consent was ob-
tained from all patients. Each structured interview
was conducted immediately before the patient was
seen by a general medicine clinic physician.
All structured interviews were conducted by one
of the physician investigators (DJM). Patients were
given the following description of intubation and
ventilatory support (IVS):
A hard but flexible plastic tube is inserted either in
your nose or through your mouth and then maneu-
vered into position in your windpipe. A section of the
tube will then be hooked up to a second piece of
tubing that runs from the tube in your nose (or
mouth) to the respirator or ventilator. When you are
so hooked up to the breathing machine, the respi-
rator will be adjusted to provide your lungs with the
appropriate amount of oxygen to carry out normal
lung function while you are being treated.
Patients were randomly assigned to one of two
Influence of Physician Explanations l 57
explanation conditions that differed in terms of their
medical contexts, that is, the degrees of specification
of the nature of the medical condition patients were
asked to consider. The general-explanation group
was asked to consider the future medical context of
“an unspecified medical condition”; the specific-ex-
planation group was asked to consider the future
medical context of a specific medical condition, “a
severe pneumonia.” Patients were then asked three
questions:
Question 1. Would you accept IVS? Patients
were asked to provide a yes-or-no response. Those
answering “no” did not complete the remaining two
questions, but went on to the sociodemographic and
medical history questions (described below).
Question 2. How long would you allow your
physician to continue to support you with IVS? Pa-
tients were asked whether they wanted IVS for 2-3
days, 2-3 weeks, 2-3 months, 2-3 years, or were
asked to specify the duration of days, weeks,
months, or years they found acceptable for IVS in
their own particular case. Although patients were
not asked directly whether they wanted their phy-
sician to make the decision for them and were not
asked directly whether they wanted indefinite sup-
port by IVS, these two responses given by patients
were recorded as such.
Question 3. Given that your physician could
specify your chances of survival once you are on IVS
for two to three days, at what chance of recovery
would you allow your physician to continue sup-
porting your respirations?
For Question 3, the concept of “chance of recov-
ery” was defined as the probability that the patient
would be able to leave the hospital and be able to
take care of activities of daily living unassisted with
minimal change in his or her mental state from the
pre-hospitalization status.
Patients were asked to specify the probability by a
lottery technique. All patients were asked whether
they were willing to have their physician continue
supporting their respirations with IVS if their phy-
sician estimated the “chance of recovery from IVS”
in the following fashion:
50:50 (50% chance that the patient would have a suc-
cessful recovery from IVS as compared with a 50%
chance that the patient would be dead or in a coma).
Patients who were willing to accept a 50:50 lottery
were then asked if they were willing to accept a
40:60 lottery (40% chance of a successful recovery
from IVS as compared with a 60% chance that the
58 l Mazur, Hickam
Table 1 l Duration of Time Patients Would Allow Support
for Their Medical Conditions after General vs
Specific Explanations
Number of Patients Reporting
General Specific
Explanation Explanation
(n = 97) (n = 89)
1-7 days 23% 13%
7-13 days 16% 10%
22%
14-20 days 24%
21-30 days 13% 12%
>30 davs 24% 43%
patient would be dead or in a coma). We continued
to decrease by decrements of 10 until a l0:90 lottery,
followed by a 5:95 lottery, 4:96 lottery, 3:97 lottery,
2:98 lottery, 1:99 lottery, and a less-than-l:more-
than-99 lottery. Patients who were unwilling to ac-
cept a 50:50 lottery were asked whether they would
accept a 60:40 lottery, 70:30 lottery, 80:20 lottery,
9O:lO lottery, or 100:0 lottery.
Patients were also asked to complete a sociode-
mographic questionnaire and a questionnaire de-
tailing their current or past medical problems
related to: pneumonia, heart attacks, cancer (ex-
cluding minor skin cancers), and stroke. Compari-
sons among subgroups were analyzed using Fisher’s
exact test, chi-square analyses, and t-tests for inde-
pendent samples.
Results
Structured interviews were conducted with 186
patients, most of whom were elderly men. Mean age
was 66.5 years (SD = 9.4; range = 37-84); mean level
of formal education completed was 12.7 years (SD
= 2.6; range = 6-22). Readability of the interview
text was found to be at the ninth-grade level, which
was felt to be appropriate for the educational attain-
ment of the subjects. Thirty percent of the patients
(52/186) had been hospitalized for pneumonia at
least once in their adult life; 30% (53/186) had a his-
tory of myocardial infarction; 15% (28/186) had un-
dergone coronary artery bypass-graft surgery; 10%
(18/186) had a history of cerebral vascular accident;
and 9% (17/186) had a significant cancer history.
Ninety-seven patients received the physician expla-
nation about the need for IVS in terms of a general
explanation about a medical condition; 89 patients
received the specific explanation about a severe
pneumonia.
The type of explanation and the degree to which
the explanation specified the future medical context
appeared to influence patient attitudes about IVS.
Slightly fewer (p = 0.031 patients receiving the gen-
eral explanation than patients receiving the specific
MEDICAL DECISION MAKING
explanation were willing to allow the physician to
proceed with IVS. In the general-explanation group
with the unspecified future medical context, 94% of
patients (9l/97) were willing to allow the physician
to proceed with IVS. In the specific-explanation
group with the specified future medical context (se-
vere pneumonia), 100% of the patients (89/89) were
willing to allow the physician to proceed with IVS.
The patients receiving the general explanation
with the unspecified future medical context were
willing to accept significantly fewer (x2 for trend =
6.935, df = 1, p < 0.01) days of IVS than the specific-
explanation group with the specified future medical
context (severe pneumonia1 (table 1). The general-
explanation group (unspecified future medical con-
text) was willing to accept a mean of 64.5 days (SD
= 167.8; range = O-1,095) of IVS; the specific-expla-
nation group (specified future medical context: se-
vere pneumonia1 were willing to accept a mean of
95.8 days (SD = 189.7; range = O-7301 of IVS. Also,
significantly fewer (p < 0.011 patients receiving the
general explanation were willing to accept IVS for
230 days.
Significantly fewer (x2 = ‘27.654, two-tailed p <
0.00011 patients receiving the general explanation, as
compared with the specific explanation, reported
willlngness to continue IVS when the physician’s es-
timate of the chance of reversal of the medical con-
dition was less than 50%. The general-explanation
group’s mean required probability of success was
54.8% (SD = 20.0; range = 5-100); the specific-ex-
planation group’s mean required probability of suc-
cess was 35.5% (SD = 26.7; range = l-991. In the
general-explanation group, 70% of the patients (64/
91) did not want the physicians to continue IVS un-
less the physician’s probability of reversal of the
medical condition was 250%. In the specific-expla-
nation group, 64% of the patients (54/84) were willing
Table 2 l Probabilities at Which Patients Would Allow
Physician to Continue lntubation and Ventilatory
support (IVS)
Number of Patients Repotting
General Specific
Explanation Explanation
(n=97) (n = 89)
p(success) : p(failure)
100-80 : O-20 16 4
75-60 : 25-40 17 6
50 : 50 31 20
40-25 : 60-75 19 18
20-O : 80-100 8 36
At any chance 5 3
2
Whatever physician says 1
VOL 17/NO 1, JAN-MAR 1997
to continue if the probability of reversal of the med-
ical condition was <50% (table 2).
Discussion
The stimulus for our study was an observation
about the oral communication that takes place be-
tween physicians and their patients: some physi-
cians use general descriptions; others use descrip-
tions of specific medical conditions. We focused on
the degree of specification of the future medical
context when the physician is seeking the patient’s
preferences regarding life support. Our results sug-
gest that specifying the future medical condition as
a severe pneumonia influences two types of patient
decisions: 1) how long to remain on a specific type
of life support (intubation and ventilatory support),
and 2) the likelihood of successful reversal necessary
for the patient to consent to continued life support.
In both cases, patients who were considering “an
unspecified future medical condition” desired fewer
days of support and tended to want continued sup-
port only when the probability of successful reversal
of the condition was 50% or greater.
Our findings have implications for how advance
directives should be constructed. In our patient
population of older male veterans, specific descrip-
tion of the future medical condition appeared to
have only a small effect on the desire to initiate IVS.
Nearly all patients in both groups desired IVS. How-
ever, specific description of the future medical con-
dition was associated with willingness to continue
IVS for a longer period of time and with a lower
probability of a successful outcome. These results
suggest that attitudes about IVS are complex and in-
clude preferences about both initiating and termi-
nating the procedure. Preferences about termina-
tion appear to be more context-dependent than
preferences about initiation, and the design of ad-
vance-directive-elicitation methods should account
for this phenomenon.
Past work o n advance directives has focused on
four sets of issues: physicians’ and patients’ attitudes
about the use of life-sustaining treatments, 2,10 the
abilities of physicians, nurses, and spouses to pre-
dict patients’ resuscitation preferences, 11,12 the ef-
fects of offering advance directives on medical treat-
ments and costs, 13~ and how advance directives are
actually discussed with patients3 Within the latter
category, two approaches to eliciting advance direc-
tives have been studied“,‘: nonspecific statements
(like our general explanation with nonspecific future
medical context) and scenario- and treatment-spe-
cific statements (like our specific explanation with
specified future medical context). Our study shows
that different preferences are obtained when the
Influence of Physician Explanations l 59
statement about the future medical context is
framed in general vs specific terms. Physicians must
be careful in how they interpret patient preferences
in such a situation.
In a prior study of older patients’ preferences for
mechanical ventilation, Malloy et al.9 studied how
patients’ preferences were influenced by the word-
ing used in descriptions of mechanical ventilation:
a negative description that included identification of
risks and side effects, and used harsh phrases, e.g.,
“machine that controls your breathing”; a positive
description that included benefits, e.g., “device to
help you breathe”; and a “neutral” description that
was the same description used i n an advance direc-
tive currently in use.’ They found a significant dif-
ference in preferences to initiate mechanical venti-
lation among the three groups. In our study, the
rates of patient preferences for IVS were higher than,
in any of Malloy’s three groups. However, both stud-
ies found that the wording used in descriptions of
clinical scenarios had a significant effect on the pref-
erence rate for IVS.
Murphy et al 14
4 studied the effect of disclosure of
the probability of survival on patient preferences for
cardiopulmonary resuscitation. Five percent of the
respondents stated they would undergo CPR when
given a probability of survival of 0-5%, and 22% ac-
cepted it when given a probability of survival of l0-
17%. In our study of intubation and ventilatory sup-
port, 30% of the patients given the general
explanation and 64% of the patients given the spe-
cific explanation were willing to continue IVS if the
probability of a successful outcome was less than
50%. Like this prior study, our results suggest that
patients are able to understand and use prognostic
information expressed in terms of probabilities. The
important issue for clinicians is making certain that
the probability ranges that are provided patients are
accurate in the specific contexts in which they are
presented.
Our study has several limitations. First, only one
specified medical situation was studied. The effect
might be different for other medical conditions. Sec-
ond, our description of IVS did not mention any un-
derlying illness. Third, our study was performed on
an elderly male veteran population and thus may
not be generalizable either to a younger patient pop-
ulation or to female patients. Nevertheless, the re-
sults of this study suggest that even minimal changes
in the description of a medical condition or medical
disease entity may have strong effects on the pref-
erences of patients and their willingness to accept
or refuse invasive medical interventions. Advance di-
rectives appear to be very context-dependent and in-
volve varying preferences about both initiation and
termination of life-support technologies. 7,9 These
findings challenge the validity of advance directives
60 l Mezur, Hickam MEDICAL DECISION MAKING

and highlight the need for further development of
patient-preference-elicitation methods.
References
1. Bedell SE, Delbanco TL. Choices about cardiopulmonary re-
suscitation in the hospital: when do physicians talk with pa-
tients? N Engl J Med. 1984;310:1089-93.
2. Lo B, McLeod GA, Saika G. Patient attitudes to discussing life-
sustaining treatment. Arch Intern Med. 1986;146:1613-5.
3. Shmerling RH, Bedell SE, Lilienfeld A, Delbanco TL. Dis-
cussing cardiopulmonary resuscitation: a study of elderly
outpatients. J Gen Intern Med. 1988;3:317-21.
4. Emanuel L. Advance directives: what have we learned so far?
J Clin Ethics. 1993;4:8-16.
5. Emanuel LL, Barry MJ, Stoeckle JD, Emanuel EJ. Advance
directives: do scenario based treatment choices extrapolate
to diierent decisions? Clin Res. 1992;40:61lA.
6. Emanuel LL. Does the DNR order need life-sustaining inter-
vention? Time for comprehensive advance directives. Am J
Med. 1989;86:87-90.
7. Emanuel LL, Emanuel EJ. The medical directive: a new com-
prehensive advance care document. JAMA. X%8%261:3288-
93.
8. McNeil BJ, Pauker SG, Sox HC Jr, Tversky A. On the elicita-
tion of preferences for alternative therapies. N Engl J Med.
1982;306:1259-62.
9. Malloy TR, Wigton RS, Meeske J, Tape TG. The influence of
treatment descriptions on advance medical directive deci-
sions. J Am Geriatr Soc. 1992;40:1255-60.
10. Layson RT, Adelman HM, Wallach PM, et al. Discussions
about the use of life-sustaining treatments: a literature re-
view of physicians’ and patients’ attitudes and practices. J
Clin Ethics. 1994;5:195-203.
11. Uhlmann RA, Pearlman RA, Cain KC. Understanding of el-
derly patients’ resuscitation preferences by physicians and
nurses. West J Med. 1989;150:705-7.
12. Uhlmann RF, Pearlman RA, Cain KC. Physicians’ and
spouses’ predictions of elderly patients’ resuscitation pref-
erences. J Gerontol. 1988;43:M115-M121.
13. Schneiderman LJ, Kronick R, Kaplan RM, et al. Effects of
offering advance directives on medical treatments and costs.
Ann Intern Med. 1992;117:599-606.
14. Murphy DJ, Burrows D, Santilli S, et al. The influence of the
probability of survival on patients’ preferences regarding
cardiopulmonary resuscitation. N Engl J Med. 1994;330:
545-9.

Call for Papers

MEDICAL DECISION MAKING SPECIAL ISSUE ON PHARMACOECONOMICS

Deadline for Submissions: June 1, 1997

Medical Decision Making, the official journal of the Society for Medical Decision Making, plans to publish a
special issue devoted to Pharmacoeconomics, a growing field of applied decision research, in April 1998. Manu-
scripts appropriate for this special issue are invited, provided they have not been published elsewhere nor are
currently under review by another journal. All submissions will be peer-reviewed; authors will be notified of
the editorial decision by December 1, 1997.

Magnus Johannesson (Stockholm School of Economics) and Bernie O’Brien (McMaster University) will serve as
guest editors for this special issue. They w-ill be assisted by an editorial board of leading international experts
in pharmacoeconomics.

Four copies of the manuscript should be submitted to either guest editor. Manuscripts must conform to style
guidelines for Medical Decision Making, these are published as “Information for Authors” in each issue.
l

Magnus Johannesson, PhD Bernie O’Brien, PhD

Center for Health Economics Centre for Evaluation of Medicines
Stockholm School of Economics St Joseph’s Hospital-Martha Wing, Room H-329
Box 6501 50 Charlton Avenue East
S-113 83 Stockholm Hamilton, Ontario, L8N 4A6
SWEDEN CANADA
email: hemj@hhs.se email: obrienb@fhs.csu.McMaster.CA
FAX: 46-B-302-115 FAX: l-905-521-6136