Menstrual Hygiene Knowledge and Practice
A School-Based Cross-Sectional Study Among Schoolgirls in Kirtipur, Nepal
Overview
A school-based cross-sectional study assessing knowledge and practice of menstrual hygiene among 312 schoolgirls in Grades 7 through 9 attending two public and two private schools in Kirtipur, Kathmandu Valley, Nepal. The study identified sociodemographic predictors of good knowledge and good practice using binary logistic regression.
This study was conducted in June and July 2017 as part of graduate research in public health, and the full reproducible analysis is published as a Quarto report.
Research Context
Menstrual hygiene management remains a significant public health challenge in low- and middle-income countries. Adolescent girls face gaps in knowledge, restricted access to appropriate materials, and cultural practices that affect both their health and school attendance. In Nepal, these issues are compounded by taboos and restrictions that shape girls’ experiences of menstruation from the time of menarche.
This study was designed to quantify the prevalence of adequate knowledge and practice, and to identify which sociodemographic factors predict better outcomes — providing an evidence base for school-based interventions.
Study Design
| Parameter | Detail |
|---|---|
| Design | School-based cross-sectional study |
| Setting | Kirtipur, Kathmandu Valley, Nepal |
| Period | June – July 2017 |
| Participants | 312 schoolgirls, Grades 7–9 |
| Schools | 2 public, 2 private |
| Primary outcomes | Good knowledge (≥60% correct) and good practice (≥60% score) |
Methods
Outcome Measures
Knowledge was assessed across seven items covering the physiology, causes, and hygiene implications of menstruation. Practice was assessed across ten items covering absorbent use, disposal, genital hygiene, and school attendance behavior. Binary outcomes were defined at the 60% threshold for both.
Statistical Approach
- Descriptive statistics — frequencies and proportions for categorical variables; means and standard deviations for continuous
- Bivariate analysis — Pearson’s chi-squared tests examining associations between sociodemographic variables and knowledge or practice outcomes
- Binary logistic regression — simultaneous entry of all predictors to identify independent determinants of good knowledge and good practice, reported as adjusted odds ratios (AOR) with 95% confidence intervals
Key Findings
| Finding | Result |
|---|---|
| Good knowledge prevalence | 49.7% |
| Good practice prevalence | 43.9% |
| School absenteeism due to menstruation | 50.3% |
| Dominant predictor (both outcomes) | Education grade |
| Grade 7 vs Grade 9 — knowledge | AOR ≈ 0.10, p < .001 |
| Public vs private school — practice | AOR ≈ 0.34, p < .001 |
| School absenteeism primary reason | Dysmenorrhea (pain) |
Education grade was the strongest independent predictor of both outcomes. Girls in Grade 7 had approximately 90% lower odds of good knowledge compared to Grade 9 girls, after adjusting for all other factors. Private school attendance was strongly associated with better practice, likely reflecting differences in school infrastructure and facilities rather than knowledge alone.
Half of participants reported ever missing school due to menstruation, with dysmenorrhea as the most commonly cited reason — a finding with direct implications for school health programming.
Visualizations
The full analysis report includes:
- Forest plot — adjusted odds ratios for all predictors of good practice with 95% confidence intervals
- Score distribution — practice score distributions by school type showing the gap between public and private school girls
- Regression tables — fully formatted with
gtsummary, showing both bivariate and adjusted results
Tools and Reproducibility
| Component | Technology |
|---|---|
| Analysis | R 4.4.1 |
| Data wrangling | tidyverse (dplyr, tidyr, readr) |
| Table generation | gtsummary |
| Visualization | ggplot2 |
| Regression | Base R glm(), broom |
| Reporting | Quarto |
| Hosting | GitHub Pages |
The full analysis is reproducible from the Quarto source file. Individual participant data are not publicly released; all analyses run from anonymized survey data held securely by the author.
Limitations
- Cross-sectional design prevents causal inference
- Self-reported data on practice may be subject to social desirability bias
- Limited to four schools in one urban municipality — findings may not generalize to rural Nepal or other settings
- Study period (2017) predates significant policy changes in Nepal’s menstrual health landscape