Modern health care systems have placed growing emphasis on monitoring outcome of health interventions in general surgery. The aim of our study was to introduce an innovative database to assist surgeons in monitoring clinical practice outcomes in colorectal surgery. The study design involved comparing mortality rates to risk-adjusted operative scoring systems: Portsmouth-Physiological and Operative Severity Score for enUmeration of Mortality and morbidity (P-POSSUM), ColoRectal-Physiological and Operative Severity Score for enUmeration of Mortality and morbidity (CR-POSSUM) and Association of ColoProctology or Great Britain & Ireland (ACPGBI) score.

METHODS:
Physiological, operative, pathology and 30 day-mortality data of 125 consecutive patients undergoing elective and emergency surgical procedure for colorectal cancer in our Department of Surgery over a one year period from January to December 2003 were prospectively entered in the database for evaluation. Specific queries were created to analyze the information and to compare the observed with the expected mortality rates by means of the P-POSSUM, CR-POSSUM and ACPGBI scoring systems. Results were discussed at regular intervals (3 months) with the surgical staff.

RESULTS:
77 patients (61.6%) underwent emergency procedures, while 48 patients (38.4%) were operated upon electively. In 102 cases (81.6%) a colorectal resection was performed, while palliative procedures (bypass or permanent stoma) were carried out in 23 patients (18.4%) due to advanced stage of disease. Pathology showed a high number of patients with stage III and IV cancer (57 patients, 46%) as expected by the elevated number of patients operated with acute abdomen due to perforation or occlusion. The observed mortality rate was 4.8% (6/125), significantly lower than values predicted by P-POSSUM, CR-POSSUM and ACPGBI colorectal score (8.67, 8.71 and 19.38% respectively), p<0.05.

CONCLUSIONS:
Our study confirms the usefulness of a dedicated database in order to assess performance and measure outcome in a surgical department. The introduction of risk-prediction scoring systems is a useful tool to compare outcomes related to complexity of patients' condition and surgical procedures. The value of this model may be limited as a result of over-prediction of mortality, especially in the emergency setting, and by the use of the ACPGBI score that may be inadequate to accurately predict the risk of 30-day post-operative mortality.