Forty-eight studies were included. Study quality and reporting were poor in most of the studies with a median score of three (range two to four) out of seven. Only ten studies were RCTs.
Accessibility to primary care (25 studies): Ten studies focused on interventions to increase access to medical doctors in primary care or primary care centres. Evidence from five studies (one RCT) showed that increased numbers of primary care doctors or centres were associated with reductions in emergency department visits. Two studies reported significant reductions in the number of hospital visits with the intervention and one found no significant difference. Two of the nine studies that assessed out-of-hours services reported decreases in emergency department use after increasing hours of primary care access; however, two studies reported increases in emergency department use. Six studies (four RCTs) assessed telephone triage and consultation and none found any significant between-group differences in the number of emergency department attendances.
Demand for services
Educational interventions (six studies): One RCT found that monthly educational group meetings resulted in fewer emergency department visits than with the control group. Two quasi-experimental studies that evaluated one of three different educational/counselling interventions and another of a "care facilitator" found significant decreases in emergency department visits. The only high quality RCT did not find any significant between-group differences in emergency department utilisation.
Barrier interventions (17 studies): One randomised study found that a lack of cost-sharing (any kind of out-of-pocket payment for health care services) resulted in significantly greater emergency department use than insurance with cost-sharing. Cost-sharing reduced both urgent and non-urgent visits. Five out of six quasi-experimental studies found that emergency department cost-sharing reduced emergency department use. Five studies in USA evaluated the effects of gatekeeping (no direct access to secondary care and emergency department access by referral) and one cross-sectional study found that gatekeeping plans reduced emergency department use more than with a control group. The only RCT where the usual primary care doctor carried out the gatekeeping role found no significant differences in hospital emergencies or mortality rates.