[Effectiveness of a program for improving drug prescriptions and hospitalization reporting at a university hospital]

To evaluate the effectiveness of a method for improving the quality of care applied to the formulation of medical prescriptions and to the release of hospitalisation reports. The actions comprised the diffusion of results of audits and recommendations of good practice.

For each of the 41 services, 30 visits from 1996 were selected at random. 3289 prescriptions and 1067 correspondence files were analyzed. The results were compared to those obtained from the previous two years.

Patient identification was complete in 44% of prescriptions, the identification of the prescribing doctor and his signature were present in 62% and 19% of cases respectively. 37% of medicines included all information. 7 indicators out of 12 for the quality of prescriptions improved (p < 0.0001). Files were found for 83% of hospital visits and 56% were sent (released) within a week. The practitioner was identified in 79% of cases, the main diagnosis in 96% and the treatment in 65% of cases. Five out of nine indicators of the quality of correspondence improved (p < 0.01). For each theme, the number of indicators improving was similar (p > 0.05). The services that improved for one theme didn't necessarily improve for the second (p > 0.05).

The evaluation of the programme, based on a strategy of quality assurance, shows modest progress. Given that the improvement of two themes for a given service are not correlated, the programme appears to sensitise professionals at an individual level rather than collectively. However, this programme is an important step for introducing a mode of continued improvement of quality.