Fifteen clinical trials were included (n= 5,742): 14 randomised controlled trials (RCTs, n=4,349) and one quasi-experimental study (n=1,393). Sample size ranged from fewer than 100 to more than 1,000; only three studies had more than 200 in the intervention group. The outcome assessor was blinded in nine studies. Physicians were blinded in one study. Most studies showed retention rates in the mid-80s per cent or higher.
Hospital re-admissions (15 studies): Seven studies reported no statistically significant difference in unplanned re-admissions between treatment groups. Eight studies reported that the intervention was associated with a significant reduction compared to control (p<0.05 to <0.001).
Hospital days (nine studies): Seven studies reported that the intervention was associated with a statistically significant reduction in the number of hospital re-admission or length of stay (p<0.05 to <0.001).
Emergency department visits (11 studies): Three studies reported that the intervention was associated with a statistically significant reduction in presentations to the emergency department compared to control (p<0.05 to <0.001).
Mortality: The two studies that explicitly reported mortality found no difference between interventions in mortality. Most of the other studies appeared to have similar death rates between intervention groups.