Error rates among clinical pharmacists in calculating the APACHE II score

The Food and Drug Administration recently advocated the use of acuity scoring to determine those patients whose mortality benefits outweigh risks of adverse effects from drotrecogin alfa (activated). Many institutions have adopted an Acute Physiology and Chronic Health Evaluation (APACHE) II cutoff score of 25 (i.e., if > or = 25, administer the agent) as a component in determining eligibility for treatment with this agent. Concern is increasing that errors in the acquisition of APACHE scores can lead to prescribing errors with drotrecogin alfa (activated).

To determine the ability of clinical pharmacists in ascertaining APACHE II scores.

Fifteen clinical pharmacists calculated APACHE II scores for the clinical scenarios of three patients. End points were the pharmacists' APACHE II score calculation, decision to start drotrecogin alfa (activated) treatment, and evaluation of individual components of the APACHE II scores.

Of the 15 pharmacists, correct APACHE II scores were derived by nine (60%) pharmacists for patient no. 1 (APACHE score 44), by two (13%) pharmacists for patient no. 2 (APACHE score 22), and by two (13%) pharmacists for patient no. 3 (APACHE score 19). Scoring errors occurred in 32 of the 45 (71%) derivations. The pharmacists were significantly more likely to correctly prescribe drotrecogin alfa for the two patients with true APACHE scores of 19 and 44 than for the patient whose score was 22, which was closer to the prescribing cutoff score of 25 (both p<0.001, patient no. 1 vs no. 2, and patient no. 2 vs no. 3, Fisher exact test) All individual components of the score were correctly evaluated at least 80% of the time, except for mean arterial pressure, respiratory rate, and Glasgow Coma Scale score.

Caution must be used in applying results of APACHE II scores estimated by only one trained individual when deciding to administer or withhold drotrecogin alfa (activated).