Organisation of a trauma registry in a regional Greek university hospital: the first two years experience.

OBJECTIVE: To design and implement a hospital trauma registry so as to be able to monitor the care of injured patients. SETTING: Teaching hospital, Greece. SUBJECTS: All patients admitted with trauma from January 1997. MAIN OUTCOME MEASURES: Design of a suitable form, establishment of inclusion and exclusion criteria, injury severity scoring, finding money and personnel, and getting suitable computer hardware and software for reliable collection and analysis of data. RESULTS: We experienced great difficulty in getting funding, so were unable to employ dedicated staff to collect the data, though we have had a part-time secretary to coordinate the registry whose salary has been paid by a pharmaceutical company. We have to rely on junior doctors to collect the data, which works well when they are enthusiastic (though not all are). We decided to use the data collection form used by the UK Trauma Network. We are trying to collect sufficient data to code severity by more than one system, but at present this is causing problems because busy nurses and doctors do not like filling in forms. Software has also been a problem as most of it is in English and translation is a considerable workload. The calculations are still being done manually while we work with two computer programmers to develop our own. We have submitted a research protocol to the Ministry of Health which has been accepted and this will guarantee our expenses for the next two years. CONCLUSIONS: Implementing a philosophy of continuous quality improvement is never easy, and we expect funding to be a permanent source of anxiety. Our progress so far has been good, but not as good as we hoped; however, we are optimistic that as people see the value of continuous monitoring of the system they will become more enthusiastic and supportive.

Immunohistochemical expression of p53, bcl-2, mdm2 and waf1/p21 proteins in colorectal adenocarcinomas.

The present study was undertaken to examine the distribution of p53, p21, mdm-2 and bcl-2 protein expression in human colorectal adenocarcinomas in order to obtain combined information about the immunophenotypes characterising these tumours. Formalin-fixed, paraffin-embedded tissue sections from 52 cases of colorectal adenocarcinomas were stained using immunohistochemical methods for the detection of p53, p21/waf1, mdm2 and bcl-2 proteins. P53, p21/waf1, mdm2 and bcl-2 proteins were expressed in 35/52, 45/52, 9/52 and 27/52 cases, respectively. All nine mdm2+ cases expressed p53 and p21 proteins as well. The three patterns observed in p53/p21 expression were: p53+/p21+, p53+/p21- and p53-/p21+ in 28, 7, and 17 cases, respectively. Consequently, p53+/mdm2-/p21+, p53+/mdm-/p21- and p53-/mdm2-/p21+ immunophenotypes were expressed in 19, 7, and 17 cases respectively. Four patterns of p53/bcl2 expression were identified: p53+/bcl2+, 20 cases; p53+/bcl2-, 15 cases; p53-/bcl2+, 7 cases; p53-/bcl2-, 10 cases. It was noteworthy that 9 of the 10 p53-/bcl2-tumours had negative lymph node status. The present results suggest that both p53 dependent and p53-independent induction of p21 expression may be involved in the molecular mechanisms controlling these tumours. High expression of the p53 protein in colorectal carcinomas could be due not only to p53 gene mutations but also to binding to mdm2 protein which leads to p53 protein stabilisation. In addition, tumours with p53-/bcl2- immunophenotype are frequently associated to negative lymph node status and seem to be less aggressive.