Bias towards dementia: are hip fracture trials excluding too many patients? A systematic review

Abstract

Patients with hip fractures are older and often present many co-morbidities, including dementia. These patients cannot answer quality of life questionnaires and are generally excluded from trials. We hypothesized that a significant number of patients are being excluded from these studies and this may impact outcomes. This was a two part study; the first analyzing databases of two ongoing large-scale multi-centred hip fracture trials and the second being a systematic review. The FAITH and HEALTH studies were analyzed for exclusion incidence directly related to dementia. The second part consisted of a systematic search of all relevant studies within the last 20 years. In the FAITH study, a total of 1690 subjects were excluded, 375 (22.2%) of which were due to dementia or cognitive impairment. In the HEALTH study, 575 were excluded with dementia/cognitive impairment representing 207 patients (36%). Following the systematic review, 251 articles were identified 17 of which were retained. The overall prevalence of dementia was 27.9% (range 2-51%). Only two studies compared demented and non-demented groups. In these studies significant increases in both mortality and complications were found. In summary, when investigating hip fractures, choosing appropriate objective endpoints is essential to ensure results are also applicable to patients with dementia.