In 1973 Wennberg first described large variations in the utilization of health services between neighboring communities. Since that time there has been increasing interest in clinical standards of care that help physicians select the most efficacious treatments for their patients. Government agencies and medical specially societies are currently expending much energy developing such standards. To our knowledge, however, there has been little research aimed at developing ways to make physicians aware of the standards of care at the time when they are seeing their patients. This is particularly important in the emergency department where there is little time for reflection and research with regard to diagnostic and therapeutic decisions. We propose the development of the Emergency Department Expert Charting System (EDECS) to improve the quality and reduce the cost of emergency care while providing education for physicians-in-training. Clinical standards of care will be developed for common emergency department (ED) complaints. Physicians will use EDECS to generate the medical record, and while doing so will encounter the standards of care which are presented in the form of checklists and reminders. The computer system will simultaneously create the medical record, offer the physician advice regarding treatment and cost-containment measures, perform concurrent quality assurance, and produce patient-specific aftercare instructions. The main hypotheses to be tested are that EDECS will reduce the cost of care and improve the quality of care. In doing so EDECS will teach physicians-in-training clinical standards that define a cost-effective approach to common ambulatory complaints. The quality of care will be judged through process measures obtained from the medical record and outcome measures obtained from telephone follow-up interviews with patients. Cost of care will be calculated using the visit specific charge document and cost to charge ratios for each hospital cost center. The cost of an episode of illness will be estimated from the number of health care visits required; these data will be obtained from the medical record and the follow-up telephone interviews. The effect of EDECS on the cost and quality of care will be studied in a three stage time series off-on-off design in which four intervention complaints run parallel with three non-intervention (off-off-off) control complaints. Within each chief complaint the first """"""""off"""""""" period serves as a control for the intervention period, while the second """"""""off"""""""" period will be used to determine whether EDECS works through education (care in this stage similar to that during the intervention) or by helping the physician at the moment that care is delivered (care in this period reverts to that found in the preintervention stage). The non-intervention complaints serve to control for factors that would make the baseline care in any one period different from that in another period. If this expert charting system decreases cost while improving quality of care, it could serve as a model system which, if used nationwide, could reduce the cost of emergency care by 120 million dollars per year. The EDECS system could be easily adapted for use in the delivery of ambulatory care.
