Dr. David Walton (Western University) has shared an update regarding his recent polling of Network researchers and the collection of sociodemographic data in their studies:
Earlier this year I invited members of the MSK RNN to share examples of recent data collection forms on which questions related to participant characteristics (e.g., demographics, social data) were posed. I received 18 such forms from across the country, and thank those who took the time to share. I extracted all questions verbatim, including response options/structures, and thematically grouped them to better understand a) what types of participant characteristics were being asked and b) how we were asking them. I was able to identify 26 different constructs collected across the 18 forms. Importantly, I did not collect names of who submitted so the findings are anonymous. Amongst the findings were:
1. Only one construct was collected across all 18 forms: age. And where it was collected, only twice was it collected using the exact same wording (i.e.,: “Age: ____”). Others were various forms of “What is your age in years?” or “What is your birthdate” to varying levels of precision (year, month/year, day/month/year). While the differences may be subtle, they are not likely completely benign. Sex was the next most frequent question, showing up on 75% of forms, and 6 constructs (ages, sex, gender, race/ethnicity, employment, education) were collected on at least half. Across forms, it was very rare that any question was asked the exact same way twice;
2. There appears to be confusion around some terms, most commonly related to sex v gender and race v ethnicity, revealed through response options that did not always align with the question being asked. We also seem to remain committed to use of ‘other’ in our response options for many intersectional identity questions, despite a growing number of EDI offices encouraging a move away from non-inclusive language;
3. Interesting was the rarity with which questions or phrasing were aligned with those used in existing large-scale databases (e.g., StatsCan, CLSA) or guidance documents from groups like CIHI or the Federal Government of Canada. Importantly, this is not some kind of admonishment – I very much include myself in each of these findings. Rather it begs a larger question in my mind – do many of us in the MSK health research field not align with existing guidance because we’re unaware of them, or because we do not believe they adequately align with important participant characteristics or not at the level needed in our field?
Armed with this (and much more from other sources) information, I am pleased that a number of RNN members were adequately interested in this topic to join me as co-Investigators or trainees on a proposal submitted to the CIHR Fall 2023 Project Grant round titled “Rigorous Partner-Informed Creation of a ‘Core Set’ of Social Variables for Musculoskeletal Health Research in Adults to Enable Data Linkage and Optimize Exploration of Social Determinants of Health”. Co-Investigators are J. MacDermid, A. Rushton, A. Zajacova, J-S. Roy, and trainees E. Ekediegwu, M. Farzad, and A. Lawan. This is a very different type of study than I have submitted previously so will be interesting to see how it is reviewed, but success in that competition or no I still believe this is a worthy project and will seek ways to move towards standardizing a core set of sociodemographic questions for MSK health research and invite all interested to join me. Watch for future updates.