INTRODUCTION

A discussion of life expectancy allows patients to make choices that fit with their health context.1,2 Our recently published systematic review on patient preferences for discussing life expectancy found that, in 24 out of 31 of the included studies, the majority of patients (≥50%) reported a positive attitude towards discussing life expectancy.3 Previous studies have focused mostly on patients with cancer, while few have focused on older adults with frailty.3 We examined Danish geriatric outpatients’ preferences for discussing life expectancy with a physician.

METHODS

We conducted our survey in the geriatric outpatient clinic at Odense University Hospital, Denmark, from September to November 2020. Patients were excluded if their visit concerned a dementia assessment, if they did not speak Danish, or if the nurses deemed them of low cognitive ability.

We developed a survey based on previous literature3 and with input from Danish experts on the topic. We piloted the survey among four patients, using cognitive interviews, and revised it into a final version. The survey had two sections, one exploring patient characteristics and one exploring patient preferences for discussing life expectancy. One author (EB) interviewed patients in a private room in the geriatric outpatient clinic. Patients were given as much time as needed to provide their answers. The interviews lasted on average 20 minutes. The outcome of interest was patients’ willingness to discuss life expectancy with a physician, defined as: not willing at all, to a low degree, to a medium degree, or to a high degree. We reported results using descriptive statistics.

The study was registered in the Region of Southern Denmark’s repository (approval 20/35470). The Regional Committees on Health Research Ethics waived registration due to the study design (case number 20202000-146).

RESULTS

Ninety-three patients were invited to participate in the study of which 70 completed the survey (response rate: 75%) (Table 1).

Table 1. Patient Characteristics for the Study Population and Stratified for Patients Who Wanted to Discuss Life Expectancy and Patients Who Did Not Want to Discuss Life Expectancy*

The most prominent degree of willingness to discuss life expectancy was to a high degree (51%, n=36) followed by to a medium degree (26%, n=18), to a low degree (11%, n=8), and not willing at all (11%, n=8) (Table 2). Most patients (87%, n=61) had not previously been offered a discussion about their estimated life expectancy by a physician. However, 81% (n=57) of patients deemed it appropriate for a physician to initiate such a discussion as long as the patients had the opportunity to decline the offer. Over half (63%, n=44) of patients had little or no trust in their physician’s ability to predict life expectancy correctly. This was a prominent reason for not wanting to discuss life expectancy (75%, n=12) among those unwilling to discuss (Table 2).

Table 2. Patients’ Willingness to Discuss Life Expectancy with a Physician As Well As Reasons for Wanting/Not Wanting to Discuss Life Expectancy with a Physician for the Study Population and Stratified for Patients Who Wanted to Discuss Life Expectancy (Willing to a Medium Degree or to a High Degree) and Patients Who Did Not Want to Discuss Life Expectancy (Willing to a Low Degree or Not Willing At All).* Patients Indicating a Positive Preference Towards Discussing Life Expectancy Were Asked to Disagree/Agree to Statements About Wanting to Discuss Life Expectancy, While Patients Indicating a Negative Preference Were Asked Statements About Not Wanting to Discuss. Both Groups Had the Option to Give Other Reasons†

DISCUSSION

Most patients in our study reported having not been previously offered a discussion about life expectancy by a physician. However, four out of five patients deemed it appropriate for a physician to offer the discussion. This is consistent with previous studies among similar patient groups.4,5,6 However, a survey conducted among 878 older persons, using an online hypothetical patient scenario,6 found that most people unwilling to discuss their life expectancy did also not find it appropriate for a patient to even be offered a discussion on the topic. This may suggest that patients’ preferences are affected by the context in which the topic is posed.

Our study population might have been in a different mindset compared to people considering a hypothetical patient scenario, as we approached patients directly following their clinic appointment and phrased the question directly about them. Given that our setting is closer to where such a discussion might normally take place during a real-life consultation, it is likely more reflective of preferences and attitudes in clinical practice.

Our survey was susceptible to recall bias and social desirability bias, possibly leading to an overrepresentation of positive preferences and underestimation of previous discussions. Further, our results may not be representative of patients of other nationalities or different healthcare settings.

Our findings suggest that physicians should generally offer to discuss life expectancy with most geriatric outpatients without fear of patients’ reactions as long as the patients have the option to decline the offer. However, this needs to be confirmed in other care contexts for older people living with frailty.