What is the best way to improve the appropriateness of outpatient referrals from primary to secondary care?
The interventions that have the best supporting evidence are the distribution of guidelines with standard referral forms and the involvement of specialist consultants in education. Disseminating guidelines without forms and providing physicians with feedback on referral patterns are not proven to be effective.
Improving the referral process is a high priority for family medicine. In 2001, the Institute of Medicine issued a report1 on the state of the health care system that included several goals for overhaul: to make the system safe, effective, patient-centered, timely, efficient, and equitable. Appropriate use of subspecialty care is a key component of these goals.1 In 2004, the Future of Family Medicine Project Leadership Committee developed a template for the transformation of the specialty and the creation of a new model of family medicine.2 A centerpiece of this document is that every patient should have a medical home.2
Grimshaw and colleagues searched for studies of interventions to change or improve outpatient referrals. They found 17 trials with 23 different comparisons. Four out of five studies reported a benefit to dissemination of guidelines with structured referral sheets (checklists to accompany referral letters). These referral sheets prompt primary care physicians to perform prereferral management or tests. In one study, use of a structured referral sheet for infertility consultation yielded absolute increases of 16 percent in the number of primary care physicians who elicited a five-point sexual history, 24 percent in the number of women who received five tests before referral, and 18 percent in the number of men who received two tests before referral.3 All of the studies evaluated referral patterns for only one condition, and only about one half of referrals were accompanied by a completed referral sheet. Overuse of referral checklists for a wider range of conditions could be counterproductive.
Two out of three studies showed involvement of consultants in educational activities to be effective. In a study assessing the impact on referrals of monthly workshops about orthopedic problems, the intervention produced an increase in the use of injections by primary care physicians (30.6 versus 11.7 percent control, P < .001), a reduction in subsequent referrals to orthopedic surgeons (35.4 versus 68.0 percent control, P < .001), and an increase in the number of patients whose symptoms resolved after one year (35.4 versus 23.7 percent control, P < .05).4
Other effective interventions included patient management with a family physician rather than an internist; attachment of a physical therapist to a primary care office; requirement of an in-house second opinion before referral; and changes in the reimbursement scheme, from capitation to a mixed capitation and fee-for-service system and from low-cost fee-for-service to high-cost fee-for-service or capitation. Strategies that were not proven effective included passive dissemination of local consensus referral guidelines, feedback on referral rates, and discussion with an independent medical advisor.