The impact of computer-generated messages on childhood immunization coverage

Recent evaluations of computer-generated reminder/recall messages have suggested that they are an inexpensive, labor-saving method of improving office visitation rates of childhood immunization providers. This study assesses the sustained impact of computer-generated messages on immunization coverage during the first two years of life.

Randomized, controlled trial.

County health department in the Denver metropolitan area.

Children (n = 1227) 60 to 90 days of age who had received the first dose of diphtheria-tetanus-pertussis (DTP) and/or poliovirus vaccines.

Households of children were randomized into four groups to receive: telephone messages followed by letters (Group A); telephone messages alone (Group B); letters only (Group C); or no notification (Group D). Households in the intervention groups (A, B, and C) received up to five computer-generated telephone messages and/or up to four letters each time their children became due for immunization(s).

Immunization series completion at 24 months of age.

Children whose families were randomized to receive any of the interventions were 21% more likely to have completed the immunization series by 24 months of age than were children randomized into the control group (49.2% vs 40.9%; RR [rate ratio] = .21; CI [confidence interval] = 1.01, 1.44). While not statistically significant, children in Group A were 23% more likely to complete their immunization series by 24 months of age than those in the control group (50.2% vs 40.9%; RR = 1.23; CI = 1.00, 1.52). No differences were detected among the intervention groups. The costs per additional child completing the series by 24 months of age in Group A was $226 ($79 after start-up costs were discounted).

Computer-generated contacts, either by phone or by mail (or both combined), used each time vaccines become due, are efficacious in increasing immunization coverage of children under 2 years of age.