Table 2. Intent-to-treat results.
| N | Controls | Any SMS | p | SMS1 | SMS2 | p1 | p2 | p3 | |
|---|---|---|---|---|---|---|---|---|---|
| Returned (as % cases analyzed) | |||||||||
| Full sample | 506 | 78/101 (77.2%) | 331/405 (81.7%) | 0.304 | 164/204 (80.4%) | 167/201 (83.1%) | 0.521 | 0.220 | 0.483 |
| HIV positive | 184 | 25/35 (71.4%) | 124/149 (83.2%) | 0.110 | 62/79 (78.5%) | 62/70 (88.6%) | 0.414 | 0.028 | 0.100 |
| Time-to-return (mean days) | |||||||||
| Full sample | 506 | 12.9 | 10.6 | 0.260 | 11.5 | 9.7 | 0.552 | 0.140 | 0.307 |
| HIV positive | 184 | 16.4 | 10.5 | 0.115 | 12.4 | 8.4 | 0.355 | 0.044 | 0.203 |
| Returned within 2 business days (as % recruits) | |||||||||
| Full sample | 506 | 52/101 (51.5%) | 251/405 (62%) | 0.054 | 125/204 (61.3%) | 126/201 (62.7%) | 0.103 | 0.062 | 0.770 |
| HIV positive | 184 | 10/35 (28.6%) | 97/149 (65.1%) | 0.000 | 50/79 (63.3%) | 47/70 (67.1%) | 0.001 | 0.000 | 0.623 |
| Returned within 5 business days (as % recruits) | |||||||||
| Full sample | 506 | 72/101 (71.3%) | 311/405 (76.8%) | 0.249 | 155/204 (76%) | 156/201 (77.6%) | 0.377 | 0.228 | 0.697 |
| HIV positive | 184 | 21/35 (60%) | 118/149 (79.2%) | 0.017 | 59/79 (74.7%) | 59/70 (84.3%) | 0.114 | 0.006 | 0.150 |
| Returned within 10 business days (as % recruits) | |||||||||
| Full sample | 506 | 75/101 (74.3%) | 326/405 (80.5%) | 0.167 | 160/204 (78.4%) | 166/201 (82.6%) | 0.415 | 0.089 | 0.291 |
| HIV positive | 184 | 24/35 (68.6%) | 123/149 (82.6%) | 0.063 | 61/79 (77.2%) | 62/70 (88.6%) | 0.328 | 0.012 | 0.068 |
| Returned within 20 business days (as % recruits) | |||||||||
| Full sample | 506 | 76/101 (75.2%) | 327/405 (81.5%) | 0.155 | 159/204 (79.1%) | 168/201 (84%) | 0.447 | 0.067 | 0.206 |
| HIV positive | 184 | 25/35 (71.4%) | 120/149 (82.2%) | 0.152 | 58/79 (76.3%) | 62/70 (88.6%) | 0.582 | 0.028 | 0.053 |
Notes: p for returned (as % recruits) and returned within x business days (as % recruits) is Pearson’s χ2 test of equal proportions (of returners / returners-within-x-days) across intervention and control groups. p for time-to-return (mean days) is a test of equality of means. p1 is test for SMS1 vs. control, p2 is test for SMS2 vs. control, p3 is test of equality across the two SMS groups.