Table 2 Descriptive characteristics of model variables
Variable | Mean (95% CI) | Median (min., max.) | Data transformation |
|---|---|---|---|
Output variable | |||
 DALYS | 0.23 (0.22–0.25) | 0.23 (0.15–0.42) | No transformation |
Outputs | |||
 County public health expenditure | 2114.40 (1795.23–2433.58) | 2018 (326–6018) | No transformation |
 County private health expenditure | 2310.83 (2079.90–2541.76) | 2,156 (814–4242) | No transformation |
 Number of healthcare facilities | 2.42 (2.17–2.67) | 2.3 (1–4.9) | No transformation |
Exogenous variables | |||
 Level of corruption | 1.67 (1.44–1.90) | 1.52 (1–5.53) | Square root |
 HIV burden | 4.45% (3.14–5.76%) | 3.50% (0.10–21.00%) | Logit |
 Development budget absorption | 98.09% (83.74–112.45%) | 97.70% (38.80–353.60%) | Logit |
 Recurrent budget absorption | 95.21% (88.95–101.47%) | 98.60% (20.50–166.30%) | Logit |
 Alcohol consumption | 20.83% (17.98–23.68%) | 20% (0–40%) | No transformation |
 Level of literacy | 91.38% (88.11–94.66%) | 96.40% (53.80–99.70%) | Logit |
 Quality of care | 75.70% (73.91–77.48%) | 74.70% (62.00–91.20%) | No transformation |
 Population density | 509.15 (164.03–854.27) | 221 (6–6247) | Log |
 County economic performance | 144,146.6 (121,684.4–166,608.9) | 135,135.50 (40,464.25–384,156.7) | Log |
 Private | 41.55% (36.56–46.55%) | 37% (2–80%) | No transformation |
 Absenteeism | 51.51% (48.95–54.08%) | 50.30% (24.90–67.60%) | No transformation |
 Water | 65.33% (60.06–70.59%) | 67.20% (27.80–97.10%) | No transformation |
 Smoking | 8.53% (6.93–10.13%) | 9% (0–40%) | No transformation |
 Autonomy |  |  |  |
  Full autonomy—10 counties | |||
  Partial autonomy—9 counties | |||
  No autonomy—28 counties | |||