Computers may enhance the clinician's diagnostic and therapeutic acumen if proper data and well-defined algorithms exist. Computers also can manage patient data rapidly and efficiently, suggesting and even implementing therapy. Why then aren't computers in widespread use today? First, unquantifiable observations not amenable to computer analysis (e.g., pain, skin perfusion) are inherent to the clinician's evaluation of diagnosis, therapy and response to therapy. Second, computers do not reassess ineffective or detrimental therapeutic maneuvers and modify therapy as do clinicians. Third, much disagreement among skilled intensivists exists concerning algorithms because the pathophysiology of many diseases is poorly understood. Fourth, most computerized data are still indirect inferences of specific organ function and must be combined with non-quantifiable observation to diagnose and treat critical illness. Finally, the high cost and difficulty of demonstrating efficacy have restricted computer utilization in intensive care units (ICU's). Either costs must decrease or quality of care improve, neither of which has been documented adequately. Perhaps those ICU's which are understaffed but have skilled medical and nursing personnel can benefit from computer technology to improve quality of care but this remains to be proven. © 1979 IEEE.
