T.C.) and checked by a second (M.R., R.A., or R.W.). In an amendment to the published protocol, all articles were appraised using the Effective Public Health Practice selleckchem Project tool17 to enable assessment of all study designs with the same rubric. Appraisal considered the method of sample selection, potential for bias connected with study design, differences between groups at baseline and how these were dealt with in the analysis, assessment of outcome measures, description of the flow of patients through the study, and use of a valid and reliable primary outcome measure. Changes in medication use were reported in all included studies. However, the multitude of different formats in

which the data were provided

and the range of included study designs precluded formal pooling of the data. For example, among the randomized studies, medication use was variously reported as psychoactive drug use score, proportion of residents who had antipsychotic selleck chemicals llc medications discontinued, number of days of antipsychotic therapy per patient per month, proportion of residents taking antipsychotic medications, and dose of antipsychotic medication. Data were therefore tabulated, grouped according to study design and outcome, and discussed narratively. The electronic searches retrieved a total of 5071 unique citations. Screening of title and abstracts against the inclusion and exclusion criteria resulted Ponatinib ic50 in the retrieval of the full text of 80 articles. Fifty-nine articles were excluded because the following aspects of the article did not meet the inclusion criteria: population (n = 3), intervention (n = 14), reported outcomes (n = 1), and study design (n = 32). Six articles were published as conference abstracts only with insufficient information provided and we were unable to locate a full-text publication despite contact with authors, and 3 were duplicate publications. One additional article was located through hand searching of the bibliographies

of identified systematic review articles. The update search identified an additional 985 articles, of which 7 were retrieved in full text and 1 article met the inclusion criteria. A total of 23 articles were included, describing 22 studies. Figure 1 shows the flow of studies through the review. Table 2 shows the study characteristics of all included articles. All the included studies provided quantitative data. We did not identify any articles reporting the views and experiences of prescribers with specific interventions. Our search identified a number of qualitative articles exploring factors that influence prescribing practice in care homes; these are considered further in the discussion. Six of the studies are randomized,14, 18, 19, 20, 21 and 22 5 have a controlled design,23, 24, 25, 26, 27 and 28 and 11 are uncontrolled before and after studies.