Impact of Multi-Morbidity on Clinician Estimates of Life Expectancy among Men with Urological Cancers: A Pilot Assessment
Cynthia P. Leung, MD; Alexander S. Chiu, MD; Michael S. Leapman, MD
Yale University, New Haven, CT
BACKGROUND: Estimates of life expectancy and competing causes of death are critical in decision making for patients with urologic cancers. Tools that weigh multiple comorbidities appear to improve the accuracy of life expectancy estimates relative to estimates based on a patient's age. However, less is known about the perception of individual competing risks, or how multiple risks are integrated. Therefore, we aimed to assess the accuracy of clinicians' estimations of life expectancy in the setting of multi-morbidity.
METHODS: Under IRB approval we prospectively collected data from urology clinicians using a third party web survey (Survey Monkey). Participants were asked to estimate the life expectancy of each patient in 10 vignettes that varied by patient age and presence of comorbidity. We invited all attending physician urologists, fellows, residents, physician assistants and advanced practice registered nurses to participate by email. Responses were compared to a comorbidity-adjusted life expectancy calculator. Differences between life expectancy estimations were compared using two sided t-tests.
RESULTS: Of 83 participants who were invited to participate, we received 31 responses. Overall, the difference in clinician estimation of life expectancy from the standard life expectancy calculator ranged from -10.14 years to +2.06 years (median -1.17 years, IQR 3.18 years). Providers underestimated life expectancy in cases of low or medium comorbidity status (mean -5.67 years, p < 0.0001). Providers also underestimated life expectancy in both patients under 75 years old (mean -3.31 years, p < 0.0001) and over 75 years old (mean -1.12 years, p = 0.0053). When stratifying data based on provider level (attending vs other), there was no statistically significant difference in estimation of life expectancy with regards to age or level of comorbidity.
CONCLUSIONS: In a pilot study from a single institution, providers tended to underestimate life expectancy for patients with urologic cancers in the setting of multiple low or medium risk comorbidities. There were no significant differences in risk estimation based on provider level or degree. Further study is warranted to explore how individual medical comorbidities are perceived during decision-making.
Back to 2018 Program