Capture-recapture analysis is a statistical analysis method used to estimate populations, more traditionally animal populations, where a total population estimate can BIBF-1120 be made from the number of a species captured, tagged, and recaptured in a geographical area. This review aimed to identify all systematic reviews published from 2006 onwards that contained randomised controlled trials of balance exercise interventions, assuming that each systematic review intended to be exhaustive in its search of the scientific literature. We have worked on the assumption that each

systematic review in isolation is a ‘capture’ of trials from the total population of trials of balance exercise intervention and when a trial appeared in more than one systematic review, this trial was considered ‘recaptured’. The results of the search strategy for relevant systematic reviews

and the trials subsequently identified from those reviews are illustrated in Figure 1. This Rucaparib search strategy yielded 23 systematic reviews, which are listed in Appendix 1 (see eAddenda for Appendix 1). From these 23 systematic reviews, 145 trials were extracted and an additional 3 trials were found by scanning the reference lists of eligible trials. These 148 trials are listed in Appendix 2 (see eAddenda for Appendix 2). Analysis of the 23 systematic reviews identified in the first phase of the search using a capture-recapture analysis tool (Thompson 2007) confirmed 145 unique randomised controlled trials were identified, and gave an estimate of 17 trials missing, equating to a group review yield of 90%. Three additional trials were found by scanning reference lists of the original 145 eligible trials, leaving an estimated 14 of 162 trials theoretically missed from this analysis. Of the 148 trials identified for inclusion in this review, just over one-third (n = 60) originated from North and South America, with the remainder originating in Europe (n = 47), the Asia-Pacific region (n = 42), and the Middle East (n = 1). Most trials were set in the community

(n = 105) with others set in residential aged unless care (n = 31), hospital settings (n = 6), combined community and residential aged care (n = 5), and combined community and hospital (n = 1). The number of participants in trials ranged from 13 to 3999 (mean = 204), with a range of mean ages from 59 to 88 years (mean = 77). The majority of trials (n = 135) were trials of exercise interventions only, with the remainder (n = 13) multifactorial falls prevention interventions that included a balance exercise component. Exercise programs were primarily of mixed type of which balance exercise was one component (n = 137), while 11 trials investigated balance exercise only interventions. Some trials (n = 27) used published exercise programs such as the Otago program (Accident Compensation Corporation 2003) or the High Intensity Functional Exercise (HIFE) program (Littbrand et al 2006a).