Table 3.
Moderator Analyses of Intervention Characteristics
Moderator | k | Effect size | Standard error | Qbetween |
p (Qbetween) |
---|---|---|---|---|---|
Behavior target | 0.087 | .768 | |||
Intervention exclusively targeted adherence behavior | 41 | 0.318 | 0.073 | ||
Intervention targeted multiple health behaviors including adherence | 67 | 0.292 | 0.046 | ||
Intervention delivered to health care providers vs. patients | 4.722 | .030 | |||
Health care providers received intervention | 7 | 0.107 | 0.086 | ||
Patients received intervention | 101 | 0.316 | 0.043 | ||
Pharmacist interventionist | 0.960 | .327 | |||
Present | 28 | 0.356 | 0.063 | ||
Absent | 80 | 0.277 | 0.050 | ||
Physician interventionist | 0.361 | .548 | |||
Present | 14 | 0.369 | 0.166 | ||
Absent | 94 | 0.267 | 0.032 | ||
Ambulatory care delivery site | 0.006 | .938 | |||
Present | 25 | 0.272 | 0.118 | ||
Absent | 83 | 0.282 | 0.035 | ||
Intervention delivered in pharmacy | 0.692 | .405 | |||
Present | 9 | 0.432 | 0.166 | ||
Absent | 99 | 0.290 | 0.042 | ||
Intervention delivery | 0.709 | .400 | |||
Face-to-face delivery | 81 | 0.319 | 0.053 | ||
Mediated delivery (e.g., telephone, mail) | 27 | 0.259 | 0.048 | ||
Theory used to design intervention | 0.350 | .554 | |||
Any theory used to design intervention | 29 | 0.335 | 0.072 | ||
No theory used to design intervention | 79 | 0.284 | 0.048 |
k denotes number of comparisons, effect size is standardized mean difference, Q is a conventional heterogeneity statistic.