engleski

The analysis of the highly aggregated national hospitalisations registry data as the primary source for epidemiological research of sle using stratification methods and pareto charts

As a part of "Epidemiology of SLE in Croatia"project, an analysis of existing data from the national hospitalisations registry was performed. To employ stratification methods and Pareto charts as statistical tools for analysis of the distribution of hospital resources utilisation for SLE patients in order to define further epidemiological research priorities. To estimate national and regional, crude and sex specific hospitalisation rates for SLE. Aggregated data for all patients, residents of all 21 counties in Croatia, hospitalised in 82 units of 64 Croatian hospitals in 2001 and 2002 from the national hospitalisations registry in Croatian National Institute for Public Health were analysed. All database entries of patients discharged under the diagnosis of SLE (M32 - ICD10) were taken into consideration for further analysis. Manual data check-up and consolidation were performed. Standard statistical methods, Pareto chart analysis and stratification methods were employed using Stat Soft "Statistica 6". Demographic data from the 2001 Population Census were used. In 2001, 22 of 82 hospital units reported 394 hospitalisations of 271 patients, 38 male and 233 female. In 2002, 25 of 82 hospital units reported 457 hospitalisations of 279 patients, 39 male and 240 female. Median age for males was 41/37 and for females 43/43 years in 2001 and 2002, respectively. Only 48 patients hospitalised in 2001 were re-hospitalised in 2002 (17, 7%). Crude hospitalisation rates (per 100000) in 2001 and 2002 were 6, 11 and 6, 29. Sex specific hospitalisation rates in 2001/2002 were 1, 78/1, 83 for males and 10, 12/10, 43 for females. Regional crude hospitalisation rate median for 2001/2002 was 5, 36/5, 26 (25th% 3, 50/3, 88 ; 75th% 6, 76/7, 33, min rate 1, 29/1, 86 max rate 13, 58/11, 1). Stratification analysis and Pareto chart analysis showed that there is a striking similarity of patient populations in two consecutive years, probably caused by "disease selection". Most patients were hospitalised in regional university hospital centres in specialised rheumatological units. In two counties (Osijecko-baranjska county and Zadar county) a consistently higher than average rates were found. In yet another county a significant increase in SLE hospitalisation rate was noted (Primorsko-goranska county) Pareto charts and stratification methods were successfuly used for analysis of highly aggregated and complex data on hospitalisations of patients with SLE, retrieved from the Croatian national hospitalisation registry. They can be used to stress major public health and epidemiological problems requiring further detailed investigational effort.

SLE; epidemiology; hospitalisations registry; Pareto chart; Croatia

nije evidentirano