Inter-observervariabiliteit bij het gebruik van de APACHE-II-score: Effect van strikte richtlijnen en training

Introduction: In spite of its general acceptance and widespread use little is known about the reliability and interobserver variability of the APACHE II scoring system. We previously reported that assessment of APACHE II scores in individual patients by a group of intensivists and residents varied widely. We subsequently identified areas were problems in data interpretation had occurred; a training program focussed on these areas was implemented, and uniform guidelines for APACHE II scoring were adopted. We then repeated our assessment of interobserver variability in attributing APACHE II scores. Methods: APACHE II scores were attributed in 11 patients admitted to our surgical ICU by 7 experienced intensivists and 9 residents. Variability in APACHE II scores was assessed; differences in scoring between experienced intensivists and residents were determined, and the results were compared to our findings before implementation of our training program and before the adoption of uniform guidelines. Results: Results are shown in the Table. Variability had decreased markedly when compared to our previous study. However, a significant degree of variability persisted, as demonstrated by the still quite wide range of scores in individual patients. Conclusion: Implementation of uniform guidelines for interpretation of data included in the APACHE II score, and regular training in its use, decreases inter-observer variability and can improve the overall reliability of the score. However, a certain degree of interobserver variability appears to be inherent in APACHE II scoring even when strict guidelines are implemented and regular training programs applied.