Stata version 15 for windows

We grouped the 10 PA levels into three categories at baseline: ‘inactive’, (PA level 1–4), ‘moderately active’ (PA level 5–6), and ‘active’ (PA level 7–10). We then used the follow-up data on PA level to categorize participants as ‘consistently active’ (PA level 7–10 at baseline and follow-up), ‘consistently moderately active’ (PA level 5–6 at baseline and follow-up), ‘consistently inactive’ (PA level 1–4 at baseline and follow-up), ‘increased PA’ (increased PA level between baseline and follow-up), and ‘decreased PA’ (decreased PA level between baseline and follow-up).
We then used this change in PA level as the exposure variable and adjusted for the time period between the two measurements. Thus, we considered participants to be at risk from the date of the follow-up measurement until emigration, death, CRC diagnosis, or the end of the study period (31 December 2015), whichever came first.
All statistical tests were two-sided, and all statistical analyses were conducted using Stata for Windows version 15.0 (StataCorp, College Station, Texas, USA). All p values were considered statistically significant at a level of < 0.05.

Analyses of continuous outcomes will be undertaken on an intent-to-treat basis, including all participants randomised, regardless of treatment actually received or withdrawal from the study. Mixed-model repeated measures (MMRM) analyses will be used because of the ability of this approach to include participants with missing data without using correction biased techniques, such as last observation carried forward [26 ]. This approach can also accommodate and assess the strength and significance of clustering effects, enabling us to model cohort effects (including training time and location) using random effects within the MMRM models. For binary outcomes (prevalence of suicidal ideation) and skewed outcomes (severity of suicidal thoughts and behaviours), a comparable modelling approach will be used that accounts for binary and ordinal outcomes. Process evaluation will be analysed using descriptive statistics; for example, the average amount of time spent using the app during each session will be assessed. Responses to the process survey, sent to participants following completion of the trial, will be examined. The mean and frequency of quantitative responses will be calculated, and content analysis of qualitative responses will occur. Quantitative analyses will be undertaken in Stata for Windows, version 15 [27 ].

Data were analyzed using Stata for Windows, version 15 (College Station, Texas, USA). The significance level was set at 5%. Descriptive statistics were used to summarize the variables. Frequency and percentage were used for categorical variables, while median and quartile were used for the continuous variable “age,” which does not fit the normal contribution. Inferential statistics, including Wilcoxon rank-sum test and Chi-square (χ²) tests, were used to compare differences between groups with and without concerns about resuming work. Pearson's χ², likelihood ratio χ², Cramér's V, Goodman and Kruskal's γ, and Kendall's τb were used to measure the association between knowledge and practice of COVID-19 prevention (9 –12 (link)). Logistic regression was used to analyze the factors affecting concerns about resuming work.

Survival analysis was used for various events. For each analysis, the survival time was defined as the time from birth to the occurrence of each event, children who did not experience the event were censored at the earliest of the following: date of death, the last observed date in the linked data or 31 December 2017. The time-scale for the survival analysis was the age of children in years (continuous variable). The Kaplan–Meier estimator was used to calculate the cumulative probability of each outcome at each age, by the four levels of HI. A Cox proportional hazard model was used in the multivariable analysis to examine the association between HI and the first record of child maltreatment. To account for the intra-group (community) correlation standard errors were clustered at the community level. Due to the substantial increase of child protection notifications/substantiations each year, separate baseline hazard rates were estimated for annual birth cohorts in the multivariable analysis, and the cumulative probability of each outcome was reported separately for the 1999–2003 and 2004–2008 birth cohorts. In additional analysis, multivariable analyses were conducted separately for the two cohorts. All statistical analyses were conducted using Stata for Windows, Version 15 (StataCorp 2015).