Another Look at Deferred Care for Nonacute Conditions

Abstract & Commentary

Source: Washington DL, et al. Next-day care for emergency department users with nonacute conditions. Ann Intern Med 2002;137:707-714.

With emergency department (ED) overcrowding on the rise, the concept of deferring care for certain ED patients is being studied again. The authors of this randomized, controlled trial sought to determine if there were differences in health status and the use of health services during a one-week follow-up between patients who received usual ED care vs. those who were referred to next-day primary care.

The investigators incorporated previously developed deferred care criteria for three symptom complexes into chief-complaint-specific data forms. These forms were used in the initial screening assessment of ED patients to identify nonacute conditions. The study took place between 7 a.m. and 3 p.m., on Mondays and Thursdays, in a level one ED that has 91,000 visits annually.

Of ambulatory adults requesting care, 1176 were screened, and 421 met deferred care criteria and were referred to a research assistant, who determined study eligibility. Patients were excluded for a number of reasons, including refusal to participate. In addition, patients were excluded who, in the nurse’s judgment, required a more detailed evaluation. Of the 421 patients who met deferred care criteria, 299 met study eligibility requirements, and of those, 143 declined enrollment, leaving 156 patients for random assignment to either usual ED care (81) or deferred care (75) at a specific time the following day at the study site’s primary care clinic.

By the end of the one-week follow-up period, 96% of the deferred care group and 95% of the usual care group had been evaluated at least once by a physician; 4% in each group had sought additional health services after their initial evaluation; and no patients were hospitalized or died. Using a previously validated tool, health status improved in both groups: 2.35 points (95% CI, 0.7-4.0) for the deferred group, and 4.20 points (95% CI, 2.2-6.2) for the usual care group, with a difference of 1.85 (95% CI, 0.69-4.39), which approached a predetermined point of clinical significance within the confidence interval. Both groups reported a reduction in number of days in bed or with disability, although the deferred care group reported less improvement in both measures and the 95% confidence intervals were sufficiently wide so that the possibility of one additional day in bed or with disability could not be excluded.

Commentary by Stephanie B. Abbuhl, MD, FACEP

This topic is not a new one to emergency physicians, who work daily amidst the crisis of ED overcrowding and who have been doing research in this area for more than 10 years. The concept of "deferring" care, which has superficial appeal, is a complex topic. Any potential program must have, at a minimum, a proven safety record in large populations, must be in compliance with the Emergency Medical Treatment and Labor Act (EMTALA), and must make sense globally for the efficiency of a health system. Unfortunately, various attempts to defer care have not held up under the scrutiny of critical review of methodology or have simply not been validated when studied in a different population.1,2 For many emergency physicians, there also is an ethical issue of turning patients away who have sought emergency care.

This study suffers from many of the same methodologic problems that others have encountered. Selection bias is a major issue in this study, where 421 patients met deferred care criteria, but only 156 (37%) were randomized. It is concerning that 34% of the patients were not included because they declined enrollment, and one wonders if these patients were sicker than those who were studied. Patients who "required a more detailed evaluation" also were excluded, suggesting that a subjective nursing evaluation was part of the fundamental screening criteria. Another methodologic issue was the small sample size, and the "negative results" may represent an insufficiently powered study and must be interpreted with caution.

From a practical point of view, the infrastructure to support next-day care simply is not in place in many health systems. In fact, outpatient appointment availability may be suffering from as much overcrowding as EDs, but there are no systems in place to measure this. Finally, the question arises of the inherent inefficiency of a system that provides a sufficiently detailed screening examination to determine if a patient meets "deferred criteria," then not treating the patient. Treatment often is only a small part of the total time spent in the evaluation of the patient.

Dr. Abbuhl, Medical Director, Department of Emergency Medicine, The Hospital of the University of Pennsylvania; Associate Professor of Emergency Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA, is on the Editorial Board of Emergency Medicine Alert.


1. Lowe RA, et al. Appropriate standards for "appropriateness" research. Ann Emerg Med 2001;37:629-632.

2. Birnbaum A, et al. Failure to validate a predictive model for refusal of care to emergency department patients. Acad Emerg Med 1994;1:213-217.