Hospital quality indicators: a systematic review

Abstract

PurposeThe purpose of this paper is to identify and describe hospital quality indicators, classifying them according to Donabedian’s structure, process and outcome model and in specific domains (quality, safety, infection and mortality) in two care divisions: inpatient and emergency services.Design/methodology/approachA systematic review identified hospital clinical indicators. Two independent investigators evaluated 70 articles/documents located in electronic databases and nine documents from the grey literature, 35 were included in the systematic review.FindingsIn total, 248 hospital-based indicators were classified as infection, safety, quality and mortality domains. Only 10.2 percent were identified in more than one article/document and 47 percent showed how they were calculated/obtained. Although there are scientific papers on developing, validating and hospital indicator assessment, most indicators were obtained from technical reports, government publications or health professional associations.Research limitations/implicationsThis review identified several hospital structure, process and outcome quality indicators, which are used by different national and international groups in both research and clinical practice. Comparing performance between healthcare organizations was difficult. Common clinical care standard indicators used by different networks, programs and institutions are essential to hospital quality benchmarking.Originality/valueTo the authors’ knowledge, this is the first systematic review to identify and describe hospital quality indicators after a comprehensive search in MEDLINE/PubMed, etc., and the grey literature, aiming to identify as many indicators as possible. Few studies evaluate the indicators, and most are found only in the grey literature, and have been published mostly by government agencies. Documents published in scientific journals usually refer to a specific indicator or to constructing an indicator. However, indicators most commonly found are not supported by reliability or validity studies.