Disability screening but not in regression model | In regression model | p Value | Number of missing values/4219 (%) | |
Allocation to intervention in original trial | 316/901 (35%) | 1786/3318 (54%) | <0.001 | |
Household appliance score: none of assets on list | 475/901 (53%) | 1754/3318 (53%) | 0.801 | |
Household appliance score: most expensive assets | 58/901 (6%) | 187/3318 (6%) | ||
Positive disability screen | 73/901 (8%) | 241/3318 (7%) | 0.395 | |
Included in initial regression model but dropped from final model | ||||
Four skilled antenatal care visits | 87/901 (10%) | 296/3315 (9%) | 0.501 | 3 (<1%) |
Possible symptoms of preeclampsia | 499/873 (57%) | 1697/3210 (53%) | 0.024 | 136 (3%) |
Delivery room heated | 681/901 (76%) | 2651/3318 (80%) | 0.005 | 0 |
Infant had fever in neonatal period | 156/901 (17%) | 489/3318 (15%) | 0.057 | 0 |
Retained in final logistic regression model | ||||
Maternal underweight (BMI<18∙5) | 89/675 (13%) | 382/3318 (12%) | 0.220 | 226 (5%) |
Mother cohabited with husband when <16 years old | 178/716 (25%) | 812/3318 (24%) | 0.827 | 185 (4%) |
Skilled birth attendant | 18/733 (3%) | 91/3318 (3%) | 0.664 | 168 (4%) |
Infant size ‘very small’ | 12/901 (1%) | 108/3318 (3%) | 0.002 | 0 |
Infant illness: diarrhoea and repeated vomiting | 46/901 (5%) | 159/3318 (5%) | 0.698 | 0 |
Mean weight in grams at 1 month (SD) | 4395(963) n=335 | 4271(898) n=3318 | 0.017 | 566 (13%) |
BMI, body mass index.