Research Square · Aug 6, 2026
Frailty trajectories and the prognostic value of two-stage PRISMA-7 followed by Clinical Frailty Scale screening for mortality in community- dwelling older adults: a longitudinal general-practice study in South Tyrol, Italy
By Mahlknecht A, Wiedermann CJ, Barbieri V et al.
Background. Frailty is dynamic and predicts mortality, but longitudinal, real-world evidence on the patient-completed PRISMA-7 and on the two-stage pathway of a positive PRISMA-7 followed by the Clinical Frailty Scale (CFS) in primary care is scarce. We describe frailty transitions and quantify how baseline status predicts death, accounting for the competing risk of mortality. Methods. In a two-wave observational study in general practices across South Tyrol, Italy, community-dwelling adults aged ≥ 75 were screened with the self-completed PRISMA-7 and, when positive (≥ 3), graded by the clinician on the CFS. Baseline status was classified as Class A (PRISMA-7 negative), B (positive, CFS < 5), C (positive and CFS ≥ 5), D (positive, CFS not graded), or unclassifiable. The primary outcome was all-cause mortality. Analyses used Kaplan–Meier and log-rank tests, Cox proportional-hazards models, and competing-risk methods, emphasising effect sizes and 95% confidence intervals (CI) over p-values. Results. The longitudinal cohort comprised 12,340 patients (median age 81 years, 57.3% women; assessments a median of 18.7 months apart; median mortality follow-up 24.4 months); 4,966 were newly enrolled at follow-up. Baseline Classes A, B and C accounted for 51.9%, 20.6% and 19.6% (4.0% ungraded, 3.9% unclassifiable). Among patients re-screened at follow-up (by protocol, those not already frail on both instruments), PRISMA-7 status worsened more often than it improved (1,265 vs 603 discordant transitions; McNemar p < 0.001), and 20.9% of screen-positives reverted (95% CI 19.5–22.4). Mortality rose steeply across classes (A 2.1%, B 6.9%, C 23.5%; log-rank p < 0.001): versus Class A, the hazard of death was 8.01 (95% CI 6.34–10.12) for Class C and rose 1.69-fold per CFS point (95% CI 1.60–1.78); Harrell's C was 0.80. Accounting for competing mortality lowered estimated improvement in Class C from 7.2% (survivors) to 4.9%. Conclusions. Frailty reversibility is real but limited and is overestimated in studies when the competing risk of death is ignored. Because worsening predominated, repeated screening is most valuable in patients not yet frail, in whom worsening can be caught early, whereas the already-frail are identified at baseline. Baseline PRISMA-7 and CFS status was strongly prognostic for mortality, supporting systematic two-stage screening led by the patient-completed PRISMA-7, in line with Italy's national chronic-disease policy.