An important concern of health systems facing scarce resources is to cover the care needs with lower costs. As population based surveys are less and less affordable, to use hospital data currently gathered for financial purposes and other statistics already available could result in update information and important savings for planning health care. Methods By using the national hospital database (hospitalization days, inpatient cases), statistical reports about incidence, hospital morbidity and

more »

... populations, relied on multivariate statistics, correlation and multi-linear regression, patterns of use were identified, defining either gaps or overuse in 42 districts and 8 regions. Results The data included 5763652, 24157793 hospitalization days in chronic, acute care wards spent by 262300 and 3876170 inpatients. Correlation coefficient showed lower values (0,1-0,4) for acute and chronic wards of psychiatry, palliative care, higher (0,7-0,8) for laparoscopic surgery, oncology and gynecologic oncology and extremely high (0,91-0,97) for other acute care wards while a low or medium correlation was found for the majority of chronic wards. After a matching between type of disorders and the wards where most likely these could be assisted, rates by 1000 persons of target population (children, adults, fertile women, new born) were computed. Disparities between reported inpatients days and those expected were established so the services were qualified as appropriately, over or under used. Conclusions Using the hospital data on number of cases and hospitalization days plus resident population, incidence and hospital morbidity is a way of providing update and useful assessment of needs and their coverage. The main uses of multilinear regressioncausal analysis, forecasting an effect and trend forecasting -are discussed. Key messages: Hospital data collected for other purposes than needs assessment could be used in order to measure the care needs and care use patterns When such exploitation (of dabases already available, gathered for other purposes) is performed, efforts for corroboration of information and cautions for biases identification become compulsory Clinical audit as a quality improvement tool in emergency care. A systematic literature review