Table 3 Illustrated checklist of practical steps of sex/gender sensitivity in the stages of the research process

 1.1. Review of existing sex/gender-based knowledge - Are there differences/similarities between and within sex/gender groups? - What are the biological and social causes? - Are there different results across time, space or cultures?

- Differences were found in the literature, e.g. between boys and girls in neonatal mortality in high-income countries (boys are at greater risk) and South Asia (sometimes girls experience more neonatal mortality), and in the early (days 1–7) and late (days 8–28) neonatal period - Biological explanations favouring survival of girls (height/weight, maturity of the lungs, sex steroid influences of the immune system), more relevant in high-income countries - Sociocultural explanations for girls’ risk of neonatal mortality in South Asia: gender preference, differential care-seeking behaviours, birth order and family composition, perceptions of illnesses

 1.2. Evaluation of the knowledge base. What is the sex/gender-related gap?

“Unanswered questions remain regarding the impact that biological (immutable factors specific to the newborn or his/her mother) and environmental factors (mutable external factors) have on sex specific trends in neonatal mortality” [81]

 1.3. Formulation of sex/gender-related study aim and research question to address the knowledge gap

“… biological and environmental factors that might explain sex differences in neonatal mortality…” [81]

 2.1 Definition of sex/gender-related biological and social factors based on a theoretical model

- Biological factors typically indicating a higher risk for neonatal mortality in males: birth outcomes such as weight, gestational age, respiratory depression, malformations - Social/environmental factors which may indicate a gender preference: peri- and postnatal care such as feeding practices, hygiene and skin care practices, warming practices and care-seeking behaviours

 2.2 Selection of sex/gender sensitive outcome and exposure measures

- Sex/gender-based justification of the outcome measure early/late neonatal mortality

 2.3 Sample size calculation is justified with respect to sex/gender-related study aims, e.g. to detect differences between or within sex/gender groups

- Secondary analysis of a population-based randomised trial, 23,662 newborns were included in the analysis

 3.1 Analytic strategy, statistical modelling is justified with respect to the sex/gender-related aims of the study

- Stratified analysis by sex/gender and ethnicity, explorative examination of sociodemographic, newborn and maternal characteristics; model building strategy reflected the four conditions: biological vs. social/environmental factors, early vs. late neonatal period

 3.2 The analysis is conducted stratified by sex/gender (if appropriate) but avoids overemphasis of sex/gender

- Differentiation by ethnic groups (Pahadi and Madeshi)

 3.3 Sex/gender stratified presentation of sample characteristics

- Sociodemographic characteristics are reported to not be meaningfully different between boys and girls

 3.4 Sex/gender differences and similarities are reported

- Biological factor, care practices and crude mortality rates were presented by sex/gender and differed significantly - Multivariate models analysing biological and social/environmental factors in the early and the late neonatal period showed no influence of care related factors - Further exploration showed social factors in one ethnic group to be related with excess mortality in the late neonatal period

 4.1 Findings are discussed in the context of existing literature; unexpected results, strength and weaknesses of the study with regard to sex/gender aspects are interpreted

- Main results are discussed with regard to: • Expectations concerning early vs. late neonatal period • Seasonal influences on food availability for pregnant women • Newborn care services favoured boys, providing evidence of gender preference • Differences within the group of girls depending on ethnic group (Pahadi, Madeshi) and prior sex composition of siblings - Missing values on birth weight are discussed as a limitation, but did not affect sex/gender-related factors

 4.2 Implications for research and practice of the main sex/gender-related findings are discussed

- Important issues are highlighted: (1) neonatal analysis must be stratified by early and late period, (2) biology has a greater impact on early, environmental factors on late neonatal mortality, (3) the explanation model ‘gender preferences’ is oversimplified as it applies only to a certain group