Total N=2,934a | Comprehensive Knowledgeb % (95% CI) | Adjusted Model for Comprehensive Knowledgec | ||
---|---|---|---|---|
aPR (95% CI) | P Valued | |||
Received any EVD message | ||||
No | 1,120 | 42.6 (36.8, 48.6) | Reference | |
Yes | 1,814 | 49.6 (44.4, 54.8) | 1.11 (0.99, 1.24) | .05e |
District category | ||||
Low-risk | 940 | 31.7 (25.4, 38.7) | Reference | |
High-risk | 1,994 | 47.4 (42.2, 52.6) | 1.61 (1.35, 1.93) | <.001e |
Sex | ||||
Female | 1,746 | 42.1 (34.6, 50.0) | Reference | |
Male | 1,188 | 55.6 (46.3, 64.5) | 1.18 (1.08, 1.28) | <.001e |
Age (years)f | ||||
15–24 | 639 | 47.8 (35.8, 60.0) | Reference | |
25–34 | 769 | 43.7 (35.9, 51.9) | 1.05 (0.94, 1.18) | .39 |
35–44 | 612 | 58.4 (47.4, 68.7) | 1.13 (1.02, 1.26) | .032 |
45–59 | 584 | 39.8 (24.3, 57.6) | 1.10 (0.97, 1.25) | .13 |
60 and older | 330 | 35.2 (24.2, 47.9) | 0.95 (0.81, 1.11) | .52 |
Educational level | ||||
No formal education | 489 | 41.5 (34.1, 49.3) | 0.87 (0.76, 0.98) | .027 |
Primary | 1,474 | 37.9 (26.2, 51.2) | 0.92 (0.84, 1.00) | .05e |
Secondary or higher | 971 | 55.4 (48.0, 62.5) | Referencee | |
Religion | ||||
Christian | 2,636 | 45.8 (41.5, 50.1) | Reference | |
Muslim | 244 | 50.5 (39.3, 61.6) | 1.01 (0.87, 1.17) | .92 |
Other | 54 | 39.1 (28.9, 50.2) | 1.06 (0.79, 1.41) | .71 |
Abbreviations: aPR, adjusted Prevalence Ratio estimated using Generalized Estimating Equation (GEE) modified Poisson model; CI, confidence interval; EVD, Ebola virus disease.
↵a Analysis was performed on complete cases (N=2,934) on all factors listed in the table. 551/3,485 (15.8%) respondents had missing data: on any component of comprehensive knowledge 523/3485 (15%), educational level (n=5), religion (n=2).
↵b Percent of total respondents in a specific category (weighted row percentages).
↵c Estimates from multivariable GEE modified Poisson model including priori selected factors as presented in the table. There was no evidence of multicollinearity among predictors included in multivariable model; variance inflation factor [VIF] for all factors included in multivariable model was <3.
↵d P<.05 is significant and P≥.05 is not significant (at 5% level).
↵e Adjusted model results after completing missing values using multiple imputation using chained equations (MICE): N=3,485, EVD message exposure (aPR: 1.12, 95% CI=1.01, 1.24; P=.032), district type (high risk versus low risk: aPR=1.64, 95% CI=1.41, 1.92; P<.001), gender (male versus female: aPR:1.18, 95% CI=1.08, 1.28, P<.001), and Educational level (reference=Secondary+ versus No formal education aPR: 0.87, 95% CI=0.76, 0.99, P=.036, primary aPR: 0.92, 95% CI=0.84, 1.00; P=.051).
↵f Age categorized as per the 2016 Uganda Demographic and Health Survey.25