You are viewing the site in preview mode

Skip to main content

Advertisement

Table 5 Postnatal health problems and responses as reported during day 10 survey

From: Can a community health worker administered postnatal checklist increase health-seeking behaviors and knowledge?: evidence from a randomized trial with a private maternity facility in Kiambu County, Kenya

Panel A. Maternal health problem reporting and responses
  (1) (2) (3) (4)
  Any maternal health problem reported Any maternal health problem reported Maternal problem reported and action taken Maternal problem reported and action taken
  Unadjusted Adjusted Unadjusted Adjusted
Estimation Type Logistic Logistic Logistic Logistic
CHW Call 0.47 0.23 0.73 0.50
  (0.071 - 3.08) (0.021 - 2.58) (0.095 - 5.69) (0.047 - 5.39)
CHW Home Visit 3.50* 5.58* 1.38 1.69
  (0.81 - 15.1) (0.88 - 35.4) (0.21 - 9.12) (0.18 - 15.5)
Mean of standard of care arm 0.13 0.13 0.83 0.83
p-value for test of call = visit 0.017 0.004 0.512 0.371
p-value for test of joint significance of call and visit 0.032 0.009 0.805 0.670
Controls? N Y N Y
Observations 83 79 83 53
Panel B. Newborn health problem reporting and responses
  (1) (2) (3) (4)
  Any newborn health problem reported Any newborn health problem reported Newborn problem reported and action taken Newborn problem reported and action taken
  Unadjusted Adjusted Unadjusted Adjusted
Estimation Type Logistic Logistic Logistic Logistic
CHW Call 1.27 1.21 1.30 0.95
  (0.35 - 4.54) (0.33 - 4.46) (0.28 - 6.11) (0.20 - 4.57)
CHW Home Visit 3.53** 3.78* 2.95 3.94*
  (1.01 - 12.3) (0.95 - 15.1) (0.68 - 12.9) (0.80 - 19.4)
Mean of standard of care arm 0.21 0.21 0.13 0.13
p-value for test of call = visit 0.070 0.087 0.205 0.053
p-value for test of joint significance of call and visit 0.078 0.120 0.260 0.104
Controls? N Y N Y
Observations 83 79 83 79
  1. *** p < 0.01, ** p < 0.05, * p < 0.1
  2. Notes: Odds ratios generated via logistic regression (models 1–4) are presented with 95 % confidence intervals constructed with robust standard errors. “Any maternal problem reported” is a binary variable indicating whether or not the participant reported any maternal-related health problems at 10 days postnatal. “Maternal problem reported and action taken” is a binary variable taking the value 1 if the respondent reported both any maternal health problem or concern and taking any action to address the problem, including calling a health facility or pharmacy, visiting a facility or pharmacy, or another action. “Any infant problem reported” is a binary variable indicating whether or not the participant reported any infant-related health problems at 10 days postnatal. “Newborn problem reported and action taken” is a binary variable taking the value 1 if the respondent reported both any newborn health problem or concern and taking any action to address the problem, including calling a health facility or pharmacy, visiting a facility or pharmacy, or another action. For all adjusted regressions (columns 2 and 4), individual level covariates include the female participant’s age (coded as an ordinal variable with values 18–25, 26–30, 31–35, or 36–40 years old), marital status (defined as a binary single or married), a binary variable indicating whether the female respondent was employed at any time in the past 12 months, a binary variable indicating whether the participant was enrolled in a concurrent randomized trial and received a voucher (cost subsidy) for postnatal family planning services, and a binary variable indicating whether the participant attended one or more antenatal care visits at a Jacaranda Health facility