A systematic review of community pharmacies' staff diagnostic assessment and performance in patient consultations.

BACKGROUND: Increases in patients seeking advice at pharmacies has led to pharmacy staff engaging in diagnostic behaviours. Approaches to diagnosis include using mnemonics and clinical reasoning.

OBJECTIVES: The primary aim of this review was to assess the degree to which the criteria researchers use to evaluate diagnostic performance in pharmacy consultations, in studies that have simulated patients or vignettes, conform with a clinical reasoning and a mnemonic framework. A secondary aim of the review was to characterize staff performance in the studies, based on the authors' comments of their results.

METHODS: MEDLINE, EMBASE and Web of Science were searched between October 2016 and April 2017. Only peer-reviewed studies assessing pharmacy staff's diagnostic performance using simulated patients or vignettes were eligible for inclusion. Data were extracted about how each study's criteria conformed with clinical reasoning and mnemonic frameworks. A scoring system between 0 and 4 was devised to determine the degree to which studies aligned to these two approaches. Risk of bias was assessed using the NHI Study Quality Assessment Tools. The review was registered in PROSPERO with identification number CRD42017054827.

RESULTS: Sixty-eight studies (55 cross-sectional, 11 educational interventions and 2 RCTs) with sample sizes between 10 and 2700 were included in the review. Most studies were of poor or fair quality. Performance of pharmacy staff was overwhelmingly reported as poor by study authors. This was the case regardless of geography, scenario used, or assessment framework adopted. Scrutiny on how authors arrived at these conclusions revealed that mnemonic criteria were employed to assess pharmacy staff's diagnostic performance rather than a clinical reasoning approach.

CONCLUSIONS: Potentially important aspects of the decision-making process, such as clinical reasoning, were left unexplored. The number and geographic distribution of the included studies is a strength of this review; however, a validated tool was not employed.