Skip to main content

Table 4 Model-based estimates and effect sizes for primary child development outcomes

From: Promoting parent-child relationships and preventing violence via home-visiting: a pre-post cluster randomised trial among Rwandan families linked to social protection programmes

OutcomesMixed model difference in difference estimatesa
Estimated coefficient (95% CI)Effect Size
Cohen’s d (95% CI)b or
Binary: OR (95% CI)
CHILD DEVELOPMENT (N = 1084)
ECD stimulation in the home
 HOME [0–43] (continuous)3.85 (3.20, 4.50)d = 0.87 (0.74, 0.99)
 OMCI [0–57] (continuous)3.06 (1.57, 4.56)d = 0.29 (0.17, 0.41)
 FCI (ECD activities) (continuous)1.25 (1.01, 1.48)d = 0.73 (0.60, 0.86)
Child nutrition, health and safety
 Dietary Diversity [0–7 food groups] (continuous)0.45 (0.26, 0.64)d = 0.35 (0.22, 0.47)
 Diarrhoea prevalence (%)−0.28 (−0.67, 0.11)0.76 (0.51,1.11)
 Diarrhoea care seeking (%) c1.49 (0.66, 2.31)4.43 (1.95, 10.10)
 Fever and cough prevalence (%)− 0.18 (− 0.56, − 0.19)0.83 (0.57, 1.21)
 Fever and cough care seeking (%) d1.19 (0.60,1.77)3.28 (1.82, 5.89)
Child caretaking practices and child safety
 Use of any harsh discipline (%)−1.22 (−1.67, − 0.76)0.30 (0.19, 0.47)
 Exclusive nonviolent discipline (%)0.92 (0.16, 1.68)2.50 (1.17, 5.34)
CAREGIVER OUTCOMES (N = 1498)
Caregiver mental health
 Screens for internalising problems (%) e− 0.54 (− 0.96, − 0.13)0.58 (0.38, 0.88)
Shared decision-makingf
 Action when child sick (%)0.72 (0.27, 1.18)2.06 (1.31, 3.26)
 What child eats (%)0.35 (−0.18, − 0.88)1.43 (0.84, 2.43)
Intimate partner violence
 Perpetration, male caregivers (%) g− 0.11 (− 0.97, 0.75)0.90 (0.38, 2.12)
 Victimisation, female caregivers (%) h−0.72 (−1.43, − 0.01)0.49 (0.24, 1.00)
HOUSEHOLD OUTCOMES (N = 1049)
Water, hygiene and sanitation
 Place with soap to wash hands (%)0.86 (0.42, 1.31)2.37 (1.52, 3.69)
 Water treatment (%)1.22 (0.77, 1.67)3.39 (2.16, 5.30)
 Accessing clean water (%)0.65 (0.01, 1.29)1.91 (1.01, 3.62)
  1. HOME Home Observation for Measurement of the Environment, OMCI The Observation of Mother-Child Interaction, FCI Family Care Indicators
  2. a Coefficients and effect sizes represent the “difference-in-difference” or “time-by-treatment” interaction between the two groups
  3. b Cohen’s d estimated from the regression coefficient for continuous outcomes using the pooled standard deviation of the outcome at baseline
  4. c Among those with prevalent diarrhoea (N = 376 at baseline and N = 394 at post-intervention)
  5. d Among those with prevalent fever or cough (N = 595 at baseline and N = 707 at post-intervention)
  6. e Scored ≥1.75 on the Hopkins Symptom Checklist-25 Questionnaire
  7. f Among married or cohabitating mothers and fathers (N = 913)
  8. g Among male caregivers reporting a current intimate partner at baseline (N = 450)
  9. h Among female caregivers reporting a current intimate partner at baseline (N = 523)
\