Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Compassionate care decreased emergency department use by homeless persons

ACP J Club. 1995 Sept-Oct;123:36. doi:10.7326/ACPJC-1995-123-2-036

Source Citation

Redelmeier DA, Molin JP, Tibshirani RJ. A randomized trial of compassionate care for the homeless in an emergency department. Lancet. 1995 May 6;345:1131-4.



To determine the effect of compassionate care of homeless adults by volunteers in an emergency department (ED) on repeat visits to the ED.


Randomized controlled trial with mean follow-up of 6.5 months.


ED of an inner-city Canadian hospital.


133 homeless persons (mean age 37 y, 83% men) who sought medical attention in the ED between May and August 1993. Patients were excluded if they were acutely psychotic with violent activity, intoxicated and exhibiting threatening behavior, unable to speak English, or medically unstable. All patients were tracked through each of 3 hospital computer systems for 1 year.


Patients were allocated to attentive contact from a trained volunteer (modified care) (n = 65) or to usual care (no contact with a volunteer) (n = 68). Volunteers were interested university students with high levels of maturity and communication skills who were trained to establish an empathetic rapport with homeless patients by listening attentively and conversing about common interests. Contact occurred throughout the patient's stay in the ED.

Main outcome measures

Number of repeat ED visits to either the study hospital or 2 regional hospitals. Patient satisfaction with the ED visit was assessed by questionnaire.

Main results

Analysis was by intention to treat. At baseline, the groups were similar for age, sex, and clinical characteristics. Patients receiving modified care made fewer mean repeat visits/mo than patients receiving usual care (0.43 vs 0.65 visits, P = 0.018). The relative reduction in the total number of visits made by the modified care group, adjusted for previous use, was 28% (95% CI 14% to 40%, P = 0.001). Time to the first repeat visit differed only for patients returning within the first month after the initial index visit. Twice as many patients in the modified care group as those in the usual care group rated their overall quality of care as excellent or very good (71% vs 36%, P < 0.005).


Compassionate care provided by volunteers in the emergency department decreased the number of repeat visits by homeless adults and improved their perceived quality of care.

Source of funding: Not stated.

For article reprint: Dr. D.A. Redelmeier, Sunnybrook Hospital, Institute for Clinical Evaluative Sciences, 2075 Bayview Avenue, North York, Ontario M4N 3M5, Canada. FAX 416-480-6048.


The study by Redelmeier and colleagues suggests that the role of the ED as a "safety net" for indigent patients may be as prevalent a problem in other health care systems as in the United States (1). The high frequency of ED visits and the low rate of medical intervention remind us that homeless persons may use emergency health care services inappropriately.

The surprising conclusion that compassion actually reduced the number of subsequent ED visits warrants a closer look at the study's methods. The number of repeat ED visits was measured using 3 hospital computer systems to match repeat visits on any 2 of 3 criteria (first name, last name, date of birth). Although inaccuracies caused by crude matching criteria would have been present in both the intervention and control groups, the small effect size may have changed if some imbalance occurred. A second concern is that this newly compassionate ED may have increased ED visits from other homeless persons who were not in the study, but this was not measured. Despite these cautions, little doubt exists that the intervention raised patient satisfaction scores, and the plausibility of the reduced demand for repeat visits is supported by an effect that dissipated after the first month of the intervention visit.

The generalizability of the intervention may be limited because premedical student volunteers may show a level of enthusiasm unattainable by other personnel, either volunteer or employed. Further, the authors recognized that paying someone to carry out the intervention would have been cost-prohibitive. The technical details of the intervention, however, may not be important. This is more of a lesson for personnel in the ED showing that compassion does not necessarily foster dependence on the ED. In addition to being inherently good, compassionate care is not economically detrimental and may even be helpful.

James L. Wofford, MD
Earl Schwartz, MDBowman Gray School of MedicineWinston-Salem, North Carolina, USA


1. Black ME, Scheuer MA, Victor C, et al. Utilisation by homeless people of acute hospital services in London. BMJ. 1991;303:958-61.