100. Towards Harmonized Measurement of Adolescent Health: an Assessment of Global-level Indicator Alignment

Holly Newby,Andrew D. Marsh,Mariame Gueye Ba,Liliana Carvajal,Saeed Dastgiri,Howard S. Friedman,Anna Kagesten,Sarah Keogh,Ann-Beth Moller,Kun Tang,Regina Guthold
DOI: https://doi.org/10.1016/j.jadohealth.2022.11.121
IF: 7.898
2023-01-01
Journal of Adolescent Health
Abstract:The Global Action for Measurement of Adolescent health (GAMA) Advisory Group (AG) was established by WHO in collaboration with UN partners in 2018. In line with GAMA’s overarching aim to focus and harmonize global adolescent health measurement, the purpose of this study was to identify specific discrepancies in indicator recommendations promoted by different initiatives and measured through multi-country data collection instruments. First, we identified indicators included in 16 previously identified global initiatives measuring the same concepts as one of the 52 GAMA AG-proposed indicators. Second, we decomposed all indicators into four elements: 1) who: the population of interest, 2) what: the construct or behaviour measured, 3) how much: the threshold, and 4) when: the reference or recall period. Third, we compared the four elements of the GAMA AG-proposed indicators to those of other initiatives including an indicator measuring the same concept. For each initiative indicator element, we coded whether it was 1) an exact match, 2) a partial match, 3) clearly discrepant, or 4) unspecified or ambiguous. Fourth, we reviewed multi-country data collection instruments, using the same four elements to identify whether measurement and tabulation standards aligned with GAMA AG-proposed indicator elements. A total of 179 initiative-recommended indicators were identified that measured the same concept as one of the 52 GAMA AG priority indicators. Five of the GAMA AG-proposed indicators were not covered by any other initiative. Only 16 of the 179 initiative indicators reviewed were an exact match with the corresponding GAMA AG indicator across all elements. Population of interest was the most commonly discrepant element. An exact match to the population of interest was only found in 33 of the 179 reviewed initiative indicators, mainly because the majority of initiatives recommended measurement covering only part of the adolescent age range of 10-19 years. Indicators pertaining to health behaviours and risks were the most likely to be discrepant in terms of the construct, the threshold, and/or the reference period. Indicators with inconsistencies across all these elements included, for example, heavy episodic drinking and tobacco use, and dietary indicators such as fruit and vegetable consumption. Reviewed data collection instruments varied in their methods of measuring and tabulating the indicators. Definitional differences in tobacco use, for example, occurred in the construct (smoking versus any tobacco use), threshold (at least once versus daily), and the reference period (the past 30 days versus an unspecified recall period). The current adolescent measurement landscape is inconsistent, with differing recommendations on what and how to measure. This study highlights specific discrepancies in indicator recommendations across different initiatives promoting what are seemingly the same indicators and links the discrepancies to multi-country data collection instruments. The results provide an empirical basis for improved alignment and harmonization, which can in turn support countries in data collection, interpretation, and use.
What problem does this paper attempt to address?