BACKGROUND

Administrative database, used previously for control of cost, patient flow and invoicing, offer to researchers a large sample of patients representative of population providing interesting informations in the field of descriptive and analytic epidemiology with less cost.

AIM

To assess the usefulness of administrative database for quality of care and research.

METHODS

It was a prospective study concerning 4690 hospitalisations in Department B of General Surgery of hôpital Charles Nicolle during a period of 18 months, between June 1st, 2008 and December 31st, 2009. A descriptive analysis followed by a pronostic study with a univariate and multivariate analysis was performed. RÉSULTATS : Our study showed the usefulness of an administrative database in assessing the quality of care, it allowed us to determine postoperative mortality rate (2.7%), deep morbidity (2.5%), parietal morbidity (1.2%), medical complications (6%), nosocomial infections (3.6%) and re intervention (2.7%), with independent predictive factors of these events. To reduce the incidence of these events we should reduce length of pre-operative stay, prevent intra operative accidents, avoid intra operative bleeding in order to reduce intra operative transfusions and avoid as far as possible the stay in ICU Independent predictors of post trauma death are multiple trauma [OR: 6.14, 95% (from 1.68 to 16.94), p = 0.002], a traumatized patient in distress on arrival [OR: 8.74, 95% (3.59 -27.77), p = 0.000] and overall medical complications [OR: 13.18, 95% (from 4.01 to 31.25), p=0.000]. The ISS is a good discriminative indice to assess the severity and life-threatening risk.

CONCLUSION

Administrative databases provide information on the efficiency of care, it helps to realise observational studies on large samples representative of the population at low cost. They are very useful in the research, despite the lack of clinical data.