[Application of a critical incident reporting and analysis system in an anesthesiology department]. / Utilización de un sistema de comunicación y análisis de incidentes críticos en un servicio de anestesia.

OBJECTIVES: To ascertain the changes in anesthesia-related morbidity and mortality after application of a scheme for reporting critical incidents and to assess the effect of implementing preventive measures against the detected errors. PATIENTS AND METHODS: We defined a critical incident to be any situation in which the margin of safety for the patient was reduced or might have been reduced. We analyzed data from the period between January 1999 and December 2004. RESULTS: The number of critical incidents was 547 (0.79% of 68627 anesthetic procedures). Human error was identified in 279 incidents (51%). The most frequent factors underlying errors were wrong diagnosis of the situation, communication problems, and failure to check equipment and drugs. The patient suffered no adverse effect in 81.8% of the incidents; 78.9% were considered preventable. Introducing an equipment checklist before anesthesia reduced the number of incidents from 90 events in 21809 cases in 31 months to 34 events out of 22064 cases in 29 months; chi2 test, P < 0.05; odds ratio (OR), 2.68; 95% confidence interval (CI), 1.80-3.98). Labeling syringes reduced errors in the administration of medications from 45 errors in 21 809 cases in 31 months to 27 in 22064 cases in 29 months; chi2, P < 0.05; OR, 1.68; 95% CI, 1.04-2.72. CONCLUSIONS: Corrective measures were adopted as a result of the incident reporting scheme. Some of the measures led to a statistically significant reduction in equipment and drug administration errors.