Linear Estimates of the Impact of the Solar Suitcase Intervention in Maternal Health Care Facilities in Uganda on Objective Measures of Light Using Sensor Dataa
Outcome | Average Before Intervention | Average After Intervention | Difference (95% CI) |
---|---|---|---|
Minutes of light per 24 hours | 856 | 997 | 141.0 (7.6, 274.3) |
Level of brightness during daytime (7:00 am to 6:00 pm) | 47.9 | 54.2 | 6.3 (−2.2, 14.8) |
Level of brightness during nighttime (6:00 pm to 7:00 am) | 14.5 | 24.7 | 10.1 (0.71, 19.6) |
Abbreviation: 95% CI, 95% confidence interval.
↵a Results show point estimates and 95% confidence intervals. Standard errors are clustered at the facility level. Regression controls for month of year and facility fixed effects. One facility (in Sequence 2) had a broken sensor and was not included. The number of minutes of light per day is calculated as the number of minutes over a threshold of 20% of the maximum seen in that facility. Results robust to using any threshold between 1% and 35% (results not shown). Level of brightness is on a 0–100 scale, as the percentage of the maximum light the sensor in each facility could read.