Skip to main content

Table 4 Multivariate Logit regression model results on the determinants of parental recall bias by vaccine types

From: Validity of parental recalls to estimate vaccination coverage: evidence from Tanzania

Characteristics/determinants Vaccine types
BCG OPV DPT Measles
OR (p-value) OR (p-value) OR (p-value) OR (p-value)
Individual-level
 Child age (months) 0.93 (0.172) 0.95 (0.307) 0.88 (0.006)*** 1.28 (0.247)
 Maternal age (years) 1.02 (0.491) 1.02 (0.423) 1.06 (0.049)** 1.05 (0.233)
 Married 1.25 (0.518) 0.80 (0.466) 1.51 (0.240) 0.61 (0.205)
 Insured 0.67 (0.520) 1.14 (0.763) 2.09 (0.047)** 1.84 (0.256)
 Muslim 1.84 (0.160) 1.87 (0.120) 1.48 (0.313) 1.67 (0.300)
 No education 0.51 (0.137) 0.86 (0.664) 0.96 (0.904) 2.76 (0.019)**
 Farmer 1.30 (0.468) 1.33 (0.392) 1.16 (0.647) 0.75 (0.486)
 Parity (# of births) 1.02 (0.874) 1.19 (0.104) 0.94 (0.563) 0.81 (0.219)
Household-level
 Rural households 0.66 (0.300) 0.69 (0.304) 0.91 (0.807) 1.25 (0.655)
 Pwani households 0.76 (0.394) 1.12 (0.692) 1.50 (0.176) 2.48 (0.018)**
 Quintile 1-poorest 1.49 (0.447) 0.75 (0.559) 0.74 (0.556) 0.56 (0.362)
 Quintile 2-poor 0.51 (0.285) 0.85 (0.738) 1.09 (0.853) 0.36 (0.115)
 Quintile 3-middle 1.01 (0.993) 0.51 (0.183) 0.37 (0.077)* 0.96 (0.940)
 Quintile 4-least poor 0.60 (0.361) 0.56 (0.239) 0.89 (0.808) 0.69 (0.516)
 Quintile 5-least poorest (reference)     
 Observations (N) 2324 2324 2324 634
  1. Notes: OR Odds Ratio; The estimation for measles vaccines was restricted to children with at least 9 months of age; OPV and DPT vaccination status considered at least one dose; The dependent variable (recall bias) took the value of 1 if the two data sources disagree and 0 otherwise; These results remain unchanged when adjusted for clustering at the facility level in the analysis; *** denotes significance at 1%, ** at 5%, and * at 10% level