JMIR Publications

ABSTRACT

Background:

Frailty is associated with postoperative morbidity and mortality. Preoperative screening and management of frail persons improves postoperative outcomes. The Clinical Risk Analysis Index (RAI-C) is a validated provider-based screening tool for assessing frailty in surgical populations. Patient self-screening for frailty may provide an alternative to provider-based screening if resources are limited; however, the agreement between these two methods has not been previously explored.

Objective:

The objective this quality improvement project was to examine the agreement between provider-completed vs. patient-completed RAI-C assessments to inform best practices for implementation of a preoperative frailty screening program.

Methods:

Orthopedic physicians and physician assistants completed the RAI-C assessment on Veterans 65 years of age and older undergoing elective total joint arthroplasty and documented scores into the electronic health record during their preoperative clinic evaluation. Participants were then mailed the same RAI-C form after preoperative evaluation and returned responses to study coordinators. Agreement between provider-completed and patient-completed RAI-C assessments and differences within individual domains were compared.

Results:

A total of 49 participants 65 years of age and older presenting for total hip or knee arthroplasty underwent RAI-C assessment between November 2022 and August 2023. Thirty participants (61%) completed and returned a post-visit independent RAI-C assessment. There was moderate but statistically significant correlation between patient-completed and provider-completed RAI-C assessment (r =.60; P <.001). Although 67% of participants were classified as robust, frail, or very frail by both methods, 30% of participants were re-classified from robust to frail or frail to very frail based on patient-completed RAI-C assessment. Agreement was the lowest between provider and patient-completed screening questions regarding memory and weight loss. Providers were more likely than participants to assign higher levels of functioning on questions assessing activities of daily living.

Conclusions:

RAI-C had moderate agreement when completed by providers vs. the participants themselves with a third of patient-completed screens resulting in a higher frailty classification. Future studies will need to explore the differences between and accuracy of RAI-C screening approaches to inform best practices for preoperative RAI-C assessment implementation.