gms | German Medical Science

Background: When treating older adults, one of the main factors to consider is physical frailty, which is reflected by deterioration of functional status. Deviation from gold standard treatment mostly concerns functionally impaired and disabled older adults. Because information about functional status in clinical trials is still lacking, a critical appraisal of treatment evidence is challenging.Our aim was, therefore, to identify and categorize information on functional status used in clinical trials.

Method: As part of a systematic review of randomised and non-randomised controlled clinical trials in older people with hypertension, all functional assessments used in such trials were analysed. An expert group including geriatricians, clinical pharmacologists, pharmacists and epidemiologists reached a consensus on categorizing functional status across four different levels. These levels span from functionally not impaired, slightly impaired, significantly impaired to very impaired/disabled older people. In addition, typical assessments that are often used in trial with older adults were added from the group. If feasible based on available literature, cut-off points of the functional assessments for the four different functioning levels were extracted. For each instrument without established cut-offs, cut-off points were defined by determining “which best and worst possible score would a patient in a given functional category possibly obtain?” taking into account (i) the number and weight of items and (ii) the used scoring system and its clinical interpretation.

Results: We identified 35 instruments that included measures of functional status. Frailty assessments were also included if they focused on physical frailty. While some of the assessments had established cut-offs for functional status across our predefined categories, many others had not. In most cases, no cut-offs existed for slightly impaired/pre-frail or highly impaired/very frail older adults, so that they were determined as described.

Conclusion: For the retrospective characterization of frail older people in RCTs and observational studies, we derived cut-off points for four functional status levels across different assessments. The here referred list is a starting line for the comparison of patients that were functionally assessed with different instruments in different studies. Future studies should validate allocated categories by evaluating the same patients with different scoring systems. Moreover, as general standard, future studies involving older patients should include and explicitly report functional impairment as a baseline characteristic of the enrolled participants. Expanding the CONSORT reporting guideline with this item would help to support the search, analysis, synthesis and interpretation of evidence in older people not just based on the chronological age but according to their functional status.