Dear all, The next seminar in our series of statistics talks for people interested in quantitative methods in biology and medicine will take place on Thursday 20 March 2014 at 11h00 in the Salle Delachaux (Biopôle 2, premier étage, Route de la Corniche 10, Lausanne, M2: Vennes). The speaker will be Marie-Annick Le Pogam (IUMSP) who will speak about Statistical concerns about quality and safety indicators based on hospital inpatient administrative data Abstract: Reflective of an increased demand for transparency and accountability, it has become common practice world while to release quality report-cards comparing hospital-specific outcomes. Health authorities are interested in identifying any providers whose outcome deviates substantially from some standard or target (which can be the national average), after accounting for confounding factors beyond their control, primarily patient case-mix. Furthermore, in order to control health expenditures in the context of both the global financial and economic crisis and the increase of health spending due to global ageing, some governments tried to improve hospital efficiency using pay-for-performance incentive programs that link quality to payment. Within this context, providing valid and reliable outcome measures of hospital care at a national level has become challenging for researchers. Because older patients are subject to frequent and numerous comorbidities, polypharmacy, impaired hepatic and renal function and both functional and cognitive decline, their admission to acute care hospitals are at high risk or adverse events (AE). As in many countries, the Swiss Federal council has endorsed the WHO Health 2020 policy framework for health and well-being in order to align Swiss health system optimally with future challenges. The main objectives that shape the action plan include 1) the production of quality and safety indicators – focusing on outcome indicators - to inform health authorities’ decisions, 2) the public reporting of healthcare information to inform patients’ choice, 3) health services’ benchmarking to improve efficiency and quality of care. The AHRQ patient safety indicators (PSIs) are now broadly used to estimate the occurrence of potential adverse events (AE) and in-hospital complications following surgeries, medical procedures and childbirth. PSIs provide hospital AE or complication adjusted-rates at a national level using routinely collected hospital discharge abstract administrative data. In my PhD project, we hypothesize that AHRQ PSIs based on Swiss inpatient administrative data could provide valid and reliable screening and assessment tools for elderly patient safety in acute care hospitals. They should be “refined” in order to consistently identify potentially “outlying” hospitals according to their complication rates and therefore of interest for further investigation. Empirical statistical analysis will have to face several methodological issues that are still discussed in the international scientific literature: • The choice between fixed and random effects models and between frequentist and Bayesian approach to improve the validity and reliability of safety measures for hospital profiling; • The effect of unmeasured differences in case-mix (due to the quality of administrative data and data availability) on their validity and reliability; • The interpretation of unexplained variability between hospitals after adjusting for some hospital characteristics (and for patient case-mix); • The ways of dealing with both “overdispersion” and “excess zeros” of hospital raw data or rates; • And methods to account simultaneously for various hospital sizes (and more especially for small hospitals). Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) ---------------------------------------------------- Unité de Statistique Institut Universitaire de Médecine Sociale et Préventive Centre Hospitalier Universitaire et Université de Lausanne Biopôle 2 Route de la Corniche 10 CH-1010 Lausanne Tél : ++41 (021) 314 73 28 Fax : ++41 (021) 314 73 73 Email: Valentin.Rousson@chuv.chmailto:Valentin.Rousson@chuv.ch