A community-based cross- sectional study was conducted in Amroha District, previously known as Jyotiba Phule Nagar, is one of the 75 districts of Uttar Pradesh state in northern India.
Study period: From 1st August 2018 to 31st July 2019 (12 months duration)
Study Population: The primary study population comprised women of reproductive age (15 to 49 years) who had delivered a live baby within the past 12 weeks prior to the conduct of this study (RDW). The 12-weeks limit was set with the purpose of resolving recall bias by the mother.
Sample Size Estimation: The required sample size for the survey among the mothers of children aged 0–12 weeks was calculated based on the standard formula for one point sample estimation:
To ensure coverage of minimum required sample size for estimating various outcome indicators of the study the value of ‘P’ (neonatal morbidity) is taken as 50.3% .10 With the above assumption the required sample size at 95 percent level of confidence with 5 percent of permissible error in the estimates, is worked out as: n = 1.962 * 0.503 * 0.497 / 0.052 = 384. With consideration of 10% non-response rate the sample size was 427. The total sample of 427 households would be divided between the six blocks according to Probability Proportional to Size. Thus, in each block, number of villages in each direction was selected for the survey with assumption of 7 RDW per village to complete required sample size.
Study tool: A predesigned and pretested structured questionnaire in the form of multiple-choice questions was used as study tool. It was developed to determine the knowledge and practices regarding thermal care of postnatal mothers/ caregivers on newborn care. That was divided in to 2 sections.
· Section A: Socio Demographic information
· Section B: Knowledge and Practice questionnaire on newborn thermal care
The questionnaire had 12 items regarding Knowledge and Practices of mother/ caregiver about newborn care. Each correct response carried one mark; thus, the maximum possible score was 7 for knowledge and 5 for practices regarding thermal care. For each geographical area (blocks) knowledge and practices total score was calculated separately and categorized in to three grades: Good, Fair and Poor.
Sampling Procedures: The district is divided into 1133 villages, 3 tehsils, and 6 blocks. Blocks are Dhanaura, Amroha, Gajrola, Gangeshwari, Hasanpur, and Joya. From each block most populous villages were selected, one near the CHC and 2nd farthest from CHC and high priority was given to villages where ASHA and ANM were appointed. A total of 61 villages of 6 blocks were covered under the study (Fig. 1).
Procedure for Selection of Households- The number of households selected per village was fixed at 7. In each village, to select the required number of respondents systematic random sampling methodology was used. The field supervisor will move to the center of the village and select first household randomly. In the contacted households, it was verified whether the household have a child who was aged 0–12 weeks of age. If the child and the mother are present there then the household was selected and the structured questionnaire will be canvassed. If not, the investigator will move to the immediate next household for the similar enquiry.
Selection of Respondents: In each selected household where the child aged 0–12 weeks would be identified, the mother of the selected child will contact for the interview. In a selected household, if more than one child was there, the mother of the youngest child would be contacted for the interview. Training was given for data collection and the overall data collection activities were supervised by principal investigator.
Data analysis: The data was entered in MSExcel, critically analyzed and tabulated using SPSS- 20 version software. Appropriate statistical tests of significance (Logistic regression) were applied to test and validate the findings of the study. Adjusted and unadjusted odds ratios (OR) and their 95% confidence intervals (CIs) were used as indicators of the strength of association. A p-value of 0.05 or less was used as cut off level for statistical significance.