Fifty-one RCTs (30,079 participants) were included. One study had 13,100 participants and others had 30 to 4,864 participants.
Non-personal interventions: (17 trials)
Three trials assessed paper, video or faxed delivered information. In one study that used letters to participants, people who received the intervention were more likely to adhere to medication. Two trials that used letters or videos showed no statistically significant effect.
Six trials assessed electronic systems. Four of these (on beeper pill caps, automated phone systems, computer generated reports and automated home blood pressure monitoring) showed improved adherence and two (on automated phone calls and decision analysis) showed no effect.
Eight studies assessed phone calls made by either trained lay people or professionals. Three of these showed a statistically significant improvement in adherence (one did not clearly define the outcome measure) and five showed no statistically significant improvement.
Face-to-face interventions: (34 trails)
Two trials of home visiting with family member involvement resulted in improvement in adherence. Two studies that did not include families showed no statistically significant effect.
Five of six trials that assessed interventions by pharmacists in a pharmacy showed improvements in adherence. Fifteen trials were conducted in a clinic setting. Five trials undertaken by pharmacists in clinics with dispensing facilities showed an improvement in adherence; three with no dispensing showed no statistically significant difference.
Two of three trials of physicians in a clinic showed improvements, but the methods of assessing adherence were not rigorous. Four of six studies that recruited participants at the point of hospital discharge showed an improvement in adherence.