Skip to main content

Table 4 Final model of the association between covariates and logged HIV incidence ratea

From: Incidence of HIV in Sub-Saharan Africa, 2000–2015: The Interplay Between Social Determinants and Behavioral Risk Factors

  Estimatea 95% confidence interval, t-value Sum of squaresc F value (df = 1)c
(Intercept)  − 4.4* (− 5.8; − 2.99) t = − 6.14 (p < 0.01)  
ART Coverage  − 0.81* (− 1.34; − 0.28) t = − 3.01 (p < 0.01) 1.1 71
Years of Education Completed  − 0.39* (− 0.57; − 0.22) t = − 4.48 (p < 0.01) 17.43 1136
  1. *Statistically significant at the 5% level
  2. aMean effect of a 1-unit increase in the covariate on the percent change in HIV incidence rate
  3. bAll coefficient are based on a linear mixed effects model with random intercepts for country, random slopes for LDI, and log incidence rate for the outcome
  4. cInterpretation: reduction in residual sum of squares when the covariate was added. P-values are not provided as there is no hypothesis test. Larger F-values represent greater significance for the fixed effect’s ability to explain variation in the outcome