Table 2.
Intervention fidelity.
| IRC | Guatemala | India | Peru | Rwanda | Total |
|---|---|---|---|---|---|
| Households receiving intervention 1 | 400 | 398 | 394 | 393 | 1585 |
| Stove and Initial LPG Cylinder Delivery | |||||
| Days between randomization and stove/cylinder delivery: median (Q1, Q3) | 9.0 (5.0, 15.0) |
14.0 (9.0, 20.0) |
5.0 (2.0 7.0) |
11.0 (7.0, 20.0) |
8.0 (5.0, 15.0) |
| Intervention delivery within 14 days of randomization: N (%) | 286 (71.5%) | 218 (54.8%) | 393 (99.7%) | 259 (65.9%) | 1156 (72.9%) |
| Gestational age at start of intervention (weeks): median (Q1, Q3) | 17.5 (15.4, 20.6) | 18.7 (16.4, 21.7) | 17.4 (14.6, 20.1) | 18.0 (15.7, 20.4) | 17.9 (15.4, 20.6) |
| Days under intervention during pregnancy (installation date to the end of gestation): median (Q1, Q3) | 150.0 (131.0, 167.0) | 139.0 (120.0, 157.0) | 153.5 (134.0, 176.0) | 153.0 (135.0, 170.0) | 149.0 (130.0, 168.0) |
| LPG Refill Delivery | |||||
| Time between request and delivery (days) 2: median (Q1, Q3) | 2.2 (1.9, 2.6) |
5.0 (3.0, 7.5) |
0.0 (0.0, 0.0) |
1.3 (0.5, 3.3) |
1.9 (0.0, 3.3) |
| Stove Use Reinforcement | |||||
| Participants who agreed to the stove use pledge: N (%) | 400 (100%) | 398 (100%) | 394 (100%) | 393 (100%) | 1585 (100%) |
| Participants who received LPG stove training: N (%) | 399 (99.8%) | 388 (97.5%) | 392 (99.5%) | 392 (99.8%) | 1571 (99.1%) |
| Participants with traditional stove use (TSU) who received a behavioral reinforcement visit: N (% of those with TSU) | 52 (57.1%) | 3 (7.7%) | 175 (79.9%) | 121 (71.6%) | 351 (67.8%) |
1 A total of n = 5 households randomized to intervention (2 in India, 2 in Peru, and 1 in Rwanda) exited the study after randomization and before LPG stove delivery.2 Data on delays in timely LPG delivery assume that participants requested LPG by phone prior to fieldworker visits. In some cases, participants did not call for LPG deliveries in advance of bi-weekly fieldworker visits, yielding an interval of 0 days between request and refill (most common in Peru). Given the low reporting rate of running out of LPG (Figure 2), any lack of prior refill requests does not appear to have substantially affected the continuity of LPG supply.