Intended for healthcare professionals

Rapid response to:


If social determinants of health are so important, shouldn’t we ask patients about them?

BMJ 2020; 371 doi: (Published 24 November 2020) Cite this as: BMJ 2020;371:m4150

Rapid Response:

Re: If social determinants of health are so important, shouldn’t we ask patients about them?

Dear Editor,
We are really grateful for the interest shown in our paper and for the responses on the BMJ website and on social media.

Prof Buttery refers to a US medical school curriculum and its inclusion of social determinants in clinical history taking. In the UK too social determinants have been routinely included in clinical teaching. For decades, Macleod’s Clinical Examination (a standard UK text) has advised students to consider and enquire about patients’ employment, occupation, and housing, along with other social factors (the current 14th edition of Macleod’s suggests asking about schooling and financial circumstances). Yet despite this, or perhaps because of the lack of a clear rationale for those enquiries, there is an important disconnect between undergraduate precepts and actual clinical practice. We hope that our article, and the rationales that it puts forward for asking patients about social determinants, might prompt actual changes in practice.

Dr Soljak highlights the importance of occupation, both as marker of social class and as a determinant of occupational hazard. He points to those workers at higher risk in the current context of covid. Indeed, during covid, the lack of socioeconomic data for individual patients has limited our understanding of the pandemic. In the UK, thanks to the inclusion of ‘occupation’ on death certificates, we know that security guards, taxi and bus drivers have had an especially high covid death rate. But we have no combined person-level health and socioeconomic data to help us understand whether this was because people in those jobs had a higher chance of encountering covid, or whether they had more pre-existing clinical conditions that put them at risk of poor outcomes, or whether being relatively low-wage earners was enough to threaten their health.

John Robson et al provide important evidence of the acceptability in Britain of the Canadian poverty screening question ‘Do you ever have difficulty making ends meet at the end of the month?’ Of over 1,000 Hackney residents included in their study, only 2% declining to answer that question.
They also draw the apt analogy with ethnicity screening in healthcare settings, highlighting the initial scepticism, the changes that facilitated implementation (education, data entry templates, and initial financial incentives), and the eventual wholesale uptake and acknowledgement of importance.

Prof Kelly reminds us of the significance of social risk factors compared with clinical ones, but also reminds us that estimates of area deprivation (based on postcodes) and estimates of individual deprivation (based on the sort of socioeconomic enquiries that we propose) are both important and independent health risk factors. ‘Both where you live, and your personal characteristics, matter’, she says.

Dr Tomlinson draws attention to the inadequacy of current electronic medical records for collecting person-level socioeconomic data. But this technical challenge does not appear to be an insurmountable obstacle. A GP and Twitter user based in Tower Hamlets described how templates for the collection of patient-level social determinant data have been incorporated into their electronic medical records.

We were especially pleased to see our work prompting exchanges on Twitter among members of the public, researchers, healthcare providers of various disciplines, think tank employees, and NHS strategists.
We have been heartened by the amount of support for our proposals.

Andrew Moscrop, Sue Ziebland, Gary Bloch, Janet Rodriguez Iraola,

Competing interests: No competing interests

04 January 2021
andrew moscrop
Sue Ziebland, Gary Bloch, Janet Rodriguez Iraola,