[Assessment of targeted clinical audit in suicide attempt]

Suicide attempt is a serious condition that is frequent in France. Picardie ranks fifth in France for suicide (418 deaths in 2005 for 1,890,000 inhabitants). Suicide attempt is one of the priorities of the regional public health program. The National Agency for Accreditation and Evaluation in Health (Anaes) has designed targeted clinical audits (TCA) on various conditions to promote this method as the basic tool for quality improvement.

We investigated the contribution of TCA for improving the quality of care of suicide attempt within a regional framework in Picardie.

TCA were conducted in 12 state hospitals (eight Surgical Medicine and Obstetrics, three specialized in psychiatry, one local) between 2004 and 2006. The standards from the Anaes had 16 criteria in three fields: care on admission (n=10); assessment of family and social environment (n=2); management for after hospital care (n=4). A project manager and a MD certified in health care quality supported the medical (MD certified in acute care and in psychiatry) and nursing staff of the emergency wards. All the wards analyzed 30 patients' files for the first cycle, set up and implemented improvement actions and then performed the second cycle of data collection.

All wards fully satisfied the protocol with 30 patients' files per cycle and two cycles. In all wards the teams consisted of physicians (both certified for emergency or psychiatry) and others care providers (nurses, psychologists, social workers, secretary). For the first cycle, three criteria (patient assessment, somatic examination and coordination) met the 100% target for more than half of the wards while three criteria (sociofamily and environmental evaluation, management for after hospital care, monitoring of follow-up) did not conform by more than 50% in more than half of the wards. All wards implemented changes after the first cycle with a total of 29 interventions, each one specifically devoted to improving a particular criterion. Intervention included better coordination and communication, protocol design and reminders, and information tools. The second cycle showed modest and mixed changes. After the interventions only one criteria reached the 100% target in one ward; the degree of conformity decreased in nine cases (with a mean of -23%) and increased in 16 cases (+19%). Globally, three criteria improved by less than 10% while three slightly decreased.

G. Shaw introduced clinical audits in 1989 to boost a poorly performing system within the "clinical governance" framework, a condition quite different from the French healthcare system in 2005. Therefore, the validation of clinical audit in a different context appeared necessary. Anaes has not yet published the evaluation of this method in a peer reviewed journal. Observed changes are modest and mixed. Moreover, the true impact on care delivery appears limited and one cannot rule out that the observed improvements are in fact related to an improvement in traceability or due to Hawthorne's effect. Quality improvement methods must be evaluated and validated by scientific methods such as for new treatments with clinical research.

The feasibility of the method was excellent, due to the methodological and technical support, however the method did not significantly improve the quality of care.