Differences in preoperative frailty assessment of surgical candidates by sex, age, and race
Date
Language
Embargo Lift Date
Department
Committee Members
Degree
Degree Year
Department
Grantor
Journal Title
Journal ISSN
Volume Title
Found At
Abstract
Introduction: Surgical decision-making often relies on a surgeon's subjective assessment of a patient's frailty status to undergo surgery. Certain patient demographics can influence subjective judgment when compared to validated objective assessments. In this study, we explore the relationship between subjective and objective frailty assessments according to patient age, sex, and race.
Methods: Patients were prospectively enrolled in urology, general surgery, and surgical oncology clinics. Using a visual analog scale (0-100), operating surgeons independently rated the patient's frailty status. Objective frailty was classified using the Fried Frailty Criteria ranging from 0 to 5. Multivariable proportional odds models were conducted to examine the potential association of factors with objective frailty, according to surgeon frailty rating. Subgroup analysis according to patient sex, race, and age was also performed.
Results: Seven male surgeons assessed 203 patients preoperatively with a median age of 65. A majority of patients were male (61 %), white (67 %), and 60 % and 40 % underwent urologic and general surgery/surgical oncology procedures respectively. Increased subjective surgeon rating (OR 1.69; p < 0.001) was significantly associated with the presence of objective frailty. On subgroup analysis, a higher magnitude of such association was observed more in females (OR 1.86; p = 0.0007), non-white (OR 1.84; p = 0.0019), and older (>60, OR 1.75; p = 0.0001) patients, compared to male (OR 1.45; p = 0.0243), non-white (OR 1.48; p = 0.0109) and patients under 60 (OR 1.47; p = 0.0823).
Conclusion: The surgeon's subjective assessment of frailty demonstrated tendencies to rate older, female, and non-white patients as frail; however, differences in patient sex, age, and race were not statistically significant.