Older Adults’ Preferences for Discussing Long-Term Life Expectancy: Results From a National Survey

Nancy L. Schoenborn; Ellen M. Janssen; Cynthia Boyd; John F.P. Bridges; Antonio C. Wolff; Qian-Li Xue; Craig E. Pollack

doi:10.1370/afm.2309

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Figure 1
Preferences for discussing life expectancy at various time points.

Tables

Figures
Additional Files

View popup

Table 1

Participant Characteristics

Characteristic	All (N = 878)	Preference for Discussing Life Expectancy
Characteristic	All (N = 878)	Wanted to Discuss^a (n = 363)	Did Not Want to Discuss^a (n = 515)	P Value
Age, mean (SD), y	73.4 (6.1)	73.5 (6.1)	73.3 (6.2)	.70
Female sex, No. (%)	462 (55.1)	177 (51.9)	285 (57.4)	.20
Race, No. (%)
White	575 (77.2)	261 (80.9)	314 (74.7)	.15
African American	214 (8.7)	68 (6.3)	146 (10.4)
Other	89 (14.1)	34 (12.8)	55 (14.9)
Educational level, No. (%)
<High school	61 (14.4)	13 (9.6)	48 (17.7)	<.001
High school	268 (33.2)	80 (26.2)	188 (37.9)
<4-year college	243 (24.2)	104 (25.9)	139 (23.0)
College graduate or postgraduate degrees	306 (28.3)	166 (38.2)	140 (21.5)
Health literacy,b,36 mean (SD)	13.1 (2.1)	13.2 (2.2)	13.0 (2.0)	.33
Numeracy,^c,37 mean (SD)	13.8 (3.5)	14.4 (3.4)	13.4 (3.6)	<.001
Predicted life expectancy,³¹ No. (%)
>10 years	631 (68.9)	262 (69.7)	369 (68.4)	.76
<10 years	197 (31.1)	81 (30.3)	116 (31.6)
Self-perceived life expectancy, No. (%)
≥10 years	762 (83.4)	322 (84.7)	440 (82.5)	.54
<10 years	110 (16.6)	41 (15.3)	69 (17.5)
Belief that life expectancy can be predicted,²¹ No. (%)
Disagree	446 (47.3)	155 (39.5)	291 (52.6)	<.001
Neutral	308 (38.6)	132 (39.1)	176 (38.2)
Agree	122 (14.2)	74 (21.4)	48 (9.2)
Has had life-threatening illness,³² No. (%)	252 (29.3)	120 (33.4)	132 (26.5)	.09
Has discussed life expectancy of a loved one,³² No. (%)	240 (24.8)	156 (40.7)	84 (14.0)	<.001
Complete trust in doctor,³⁵ No. (%)
Disagree	81 (8.6)	26 (7.0)	55 (9.7)	.53
Neutral	231 (27.6)	98 (27.6)	133 (27.6)
Agree	565 (63.8)	238 (65.4)	327 (62.7)
Preferred decision-making role, ³³ No. (%)
Make own decisions	533 (62.5)	220 (62.0)	313 (62.9)	.83
Shared or leave to doctor	337 (37.5)	141 (38.0)	196 (37.1)
Religion is important,³⁴ No. (%)	633 (70.8)	242 (65.5)	391 (74.5)	.02

↵a Responses to the hypothetical scenario. See Methods for description.
b Possible range 3 to 15; higher scores indicate better health literacy.
c Possible range 3 to 18; higher scores indicate better numeracy.

View popup

Table 2

Factors Associated With Preferring to Discuss Life Expectancy

Characteristic	Unadjusted Odds Ratio (95% CI)	P Value	Adjusted Odds Ratio (95% CI)^a	P Value
Age, per year	1.01 (0.98-1.03)	.62	–	–
Female sex	0.80 (0.61-1.05)	.11	–	–
Race		.05		.55
White	Ref		Ref
Black	0.56 (0.33-0.93)		0.80 (0.45-1.43)
Other	0.79 (0.53-1.17)		0.82 (0.53-1.28)
Education		<.001		.004
<High school	Ref		Ref
High school	1.27 (0.80-2.01)		1.09 (0.66-1.82)
Some college	2.06 (1.28-3.33)		1.66 (0.96-2.85)
>College	3.26 (2.05-5.19)		2.18 (1.25-3.80)
Health literacy,36 per point	1.05 (0.98-1.12)	.15	–	–
(scale 3-15)
Numeracy,³⁷ per point (scale 3-18)	1.09 (1.05-1.14)	<.001	1.04 (0.99-1.09)	.11
Predicted 10-year mortality^b,31	0.81 (0.46-1.43)	.47	–	–
Self-perceived 10-year mortality^b	1.08 (0.62-1.87)	.78	–	–
Doctors can accurately predict		<.001		<.001
life expectancy²¹
Disagree	Ref		Ref
Neutral	1.36 (1.01-1.83)		1.59 (1.14-2.21)
Agree	3.10 (2.05-4.69)		3.06 (1.93-4.86)
Has had life-threatening illness³²	1.39 (1.04-1.86)	.03	1.50 (1.07-2.09)	.02
Has discussed life expectancy	4.22 (3.05-5.85)	<.001	3.98 (2.82-5.62)	<.001
of a loved one³²
Complete trust in doctor³⁵		.36
Disagree	Ref		–	–
Neutral	1.38 (0.80-2.37)		–	–
Agree	1.44 (0.87-2.39)		–	–
Prefers shared decision making or leaving decision to doctor (vs making own decisions)³³	1.04 (0.79-1.37)	.79	–	–
Religion is important³⁴	0.65 (0.48-0.87)	.004	0.69 (0.49-0.97)	.03

Note: Preference expressed in the hypothetical patient scenario. See Methods for description.

Ref = reference group.
↵a Only variables having P ≤.05 in univariate analysis were included in multivariate analysis.
↵b Analyzed as continuous variables with range of 0 to 1 in the regression model. For example, one person’s 10-year mortality risk may be 0.02 or 2%, and another person’s 10-year mortality risk may be 0.92 or 92%. Here, the odds ratio is per 1 unit of mortality risk, ie, comparing 100% mortality risk with 0%.

View popup

Table 3

Participants’ Reasons and Preferences Around Discussing Life Expectancy

Reason/Preference	Wanted to Discuss Life Expectancy^a (n = 363)	Did Not Want to Discuss Life Expectancy^a (n = 515)
Reason for wanting to discuss life expectancy^b		–
Help patient better plan life	270 (72.3)
Important to be honest and open	151 (42.2)
Other	16 (4.6)
Reason for not wanting to discuss life expectancyb	–
Doctors cannot predict life expectancy		302 (56.7)
May worry or depress patient		272 (52.2)
Other		43 (9.7)
Acceptable for the doctor to offer to discuss life expectancy?
Yes, as long as I can say no	344 (94.8)	222 (40.1)
No, the doctor should not have brought up the topic at all	16 (5.2)	291 (59.9)
Want the doctor to discuss life expectancy with family or friends?
Yes	198 (57.9)	65 (12.3)
No	164 (42.1)	450 (87.7)
If the doctor recommends stopping cancer screening due to limited life expectancy of the patient, should doctor then discuss life expectancy with the patient?
Yes	305 (84.5)	114 (21.3)
No	58 (15.5)	401 (78.7)
How should the doctor describe life expectancy of about 5 years?
In the range of a few years	174 (49.2)	255 (48.7)
About a 50-50 chance to live another 5 years	188 (50.8)	252 (51.3)

↵a Responses to the hypothetical scenario. See Methods for description.
↵b Participants could choose more than 1 reason; therefore, percentages do not add up to 100%.

Supplemental Appendix

Supplemental Appendix

Files in this Data Supplement:

Supplemental data: Appendix - PDF file
The Article in Brief

Older Adults' Preferences for Discussing Long-Term Life Expectancy: Results From a National Survey

Nancy Li Schoenborn , and colleagues

Background Clinical practice guidelines recommend incorporating long-term life expectancy to inform a number of decisions in primary care. We aimed to examine older adults' preferences for discussing life expectancy in a national sample.

What This Study Found A majority of older adults do not wish to discuss life expectancy when presented with a hypothetical scenario on the topic. In a survey of communication around life expectancy, 878 adults age 65 years and older received a description of a hypothetical patient with limited life expectancy who is not imminently dying. Participants were asked, as the hypothetical patient, if they would like to talk with the doctor about how long they might live, if it was acceptable for the doctor to offer such discussion, whether they would want the doctor to discuss life expectancy with family or friends, and when life expectancy should be discussed. Fifty-nine percent of participants (n = 515) did not want to discuss how long they might live in the presented scenario. Among these, 291 participants did not think that the doctor should offer discussion, and 450 participants did not want the doctor to discuss life expectancy with family or friends. As estimated life expectancy increased, fewer participants felt that it should be discussed. Fifty-six percent of participants (n = 478) only wanted to discuss life expectancy if it were less than two years. Factors associated with wanting to discuss life expectancy included higher educational levels, belief that doctors can predict life expectancy, and past experiences with either a life-threatening illness or with discussing life expectancy of a loved one. Reporting that religion is important was associated with lower odds of choosing to discuss life expectancy.

Implications

Overall, this research--the first national study to examine these questions--found that long-term life expectancy can be an important factor in health care decisions for older adults, but whether, when, and how to communicate with patients about it is not clear.
The authors suggest that strategies to address this topic include assessing patient factors associated with willingness to discuss life expectancy and offering the discussion when closer to the patient's final year of life.