Patient Navigation for Patients Frequently Visiting the Emergency Department: A Randomized, Controlled Trial.

BACKGROUND

Emergency department (ED) superutilizers (patients with five or more visits/year) comprise only 5% of the patients seen yet comprise 25% of total ED visits. Although the reasons for this are multifactorial, the cost to the patient and the community is exceedingly high. The cost is not just monetary; care of these patients is inappropriately fragmented and their presence in the ED may contribute to overcrowding affecting the community's emergency readiness. Previous studies using staff trained to help patients navigate their care options have had conflicting results.

OBJECTIVES

The objective was to determine whether a trained patient navigator (PN) can reduce ED use and costs in superutilizers over a 1-year period.

METHODS

Superutilizers were enrolled in a prospective randomized controlled clinical trial. Patients were randomized into the treatment arm and met with a PN who reviewed their diagnosis and associated care plan and identified proper primary care services and community resources for follow-up. The remaining control group was provided standard care. Both groups were given a follow-up call and survey by the PN within 7 days of their visit who assessed primary care follow-up and patient satisfaction using a 4-point Likert scale. After 12 months, the patients' return ED visits and ED costs were compared to the year prior and primary care compliance and satisfaction were measured using Student's t-tests with Bonferroni correction or Mann-Whitney U-tests.

RESULTS

A total of 282 patients were enrolled (148 in navigation treatment group, 134 controls). Patients were similarly matched in age, race, sex, insurance, and chief complaints. Overall ED visits decreased during the 12-month study period, compared to the 12 months prior to enrollment (2,249 visits prior to 2,050 visits during study period, -8.8%). There was a greater decrease in ED visits from the preenrollment year to postenrollment year in the treatment group (1,148 visits to 996 visits, -13.2%) compared to the control group (1,101 visits to 1,054 visits, -4.3%; p < 0.05). Overall health care costs (ED and hospital) for all 282 patients decreased in the year after compared to the 12 months prior to enrollment ($3.9M to $3.1M) with a greater decrease in the navigation treatment group (-26.6%) compared to the control group (-17.5%). Patient surveys found no difference in patient satisfaction in the pre- and postenrollment periods but there was an increase in primary care physician (PCP) use over the 12-month follow-up period in the treatment group (6.42 visits/patient) compared to the control group (4.07 visits/patient; p < 0.05).

CONCLUSION

Our data showed that the overall number of return ED visits and costs did decrease for both groups, potentially inferring a placebo effect for the use of a PN; however, the decrease in ED visits and costs were greater in the treatment group. One-year follow-up noted an increase in PCP visits in the navigation group. Use of a PN may be cost-effective.