General Practitioners' Clinical Impression in the Screening for Frailty: Data From the FAP Study Pilot

The progression of frailty is marked by an increased risk of adverse health outcomes in the elderly including falls, physical and/or cognitive disability, hospitalizations, and mortality. In primary care, the general practitioner's (GP's) clinical impression about their elderly patients' frailty state seems to be a key point in identifying frail individuals in their clinical practice. The aim of this article is to examine if GPs' clinical impressions regarding frailty concurs with objective measures of the gold standard frailty phenotype as described by Fried in community-dwelling older persons.

Cross-sectional study in 14 primary care GP offices in the Toulouse area from May 1st to October 31st, 2015.

Fourteen GPs screened their patients ≥70 years old.

GPs' "frailty impression" was based on the Gérontopôle Frailty Screening Tool. "Objective measures of the five Fried frailty criteria" were obtained by a geriatric nurse through standardized testing. The capacity of the GPs' clinical impression to detect participants objectively measured as frail was examined with diagnostic values of observed sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV).

A total of 268 participants were screened by GPs and assessed by a nurse. Mean age was 81 years and 62.3% were female. According to the objective measures of Fried's criteria, frailty (three to five criteria) and pre-frailty (one to two criteria) states were identified in 31% and 45.2% of participants, respectively. The Se of the GPs' impression was good (80.39%; 95% confidence interval [CI], 74.27%-85.61%), and the Sp was moderate (64.06%; 95% CI, 5.10%-75.68%). The overall PPV of the GPs' impression was 87.70% (95% CI, 82.12%-92.04%), and the NPV was 50.51% (95% CI, 39.27%-61.91%). Although the PPV increased with age reaching 93.33% (95% CI, 85.12%-97.80%) among patients ≥ 85 years old, the NPV decreased accordingly to a minimal 21.43% (95% CI, 4.66%-50.80%) in that subgroup.

The present study highlights the importance of the GPs' clinical impression on frailty as a fair means to identify this syndrome in community-dwelling older patients in primary care. This clinical impression may not be sufficient, however, and some objective tests could be added to improve the accuracy of frailty detection in older patients in primary care.