Our MICU monitors aggregate outcomes by monitoring APACHE II scores and in-hospital mortality of all patients (to compare observed with predicted mortality over time). APACHE cannot be used to give a prognosis for any one patient. It is used to compare observed outcomes to those expected in large cohorts of patients (e.g. over months or years). Although APACHE III may be slightly superior, APACHE II is simpler and it is free; it only requires someone to enter the data. Other indices are also available (SAPS etc.), but none as widely as the APACHE system. For the past decade interns in our MICU have computed admission APACHE II scores on all patients. All interns' patients APACHE II scores are filled in the APACHE log-book and the supervisory resident, at the end of each month, creates a monthly report of average APACHE II and observed in-hospital mortality, complications etc.

How to fill out the APACHE II

The single most important thing for you to remember when you're filling out the data - you use the WORST value recorded in the chart (including emergency department and ICU record) FOR THE FIRST 24 HOURS OF ADMISSION.

The most common errors are to assign chronic health points that should not be. For example, if there is a suspicion of cirrhosis but no biopsy or other convincing historic/exam/laboratory evidence, then the patient should not get points for cirrhosis. Also the neurologic points are commonly miscomputed. The Glasgow must be computed BEFORE the patient gets sedatives or narcotics. If you never got to see the patient before and/or there is no definitive exam in the chart, then look at the admitting nurses, ER staff and EMS notes. They almost always have a Glasgow that can be used. If you simply have no data (which should be very, very rare) the patient should get a Glasgow of 15 and no neuro points added to their APACHE score.