Assessing patients’ social risks: what can England learn from emerging evidence in the US?

INTRODUCTION

Recognition that social factors, such as income and housing, influence health outcomes has led healthcare systems to invest in improving people’s social conditions. In England, a range of national policies encourage NHS organisations to act on social factors that shape health and inequalities.¹ Since 2020, GP contracts have included funding for ‘social prescribing link workers’ in local practices — staff to connect patients with community activities and services that could help address unmet social needs, such as food assistance or benefits advice.²

Identifying patients’ social needs is an important part of social prescribing, but guidance on how to assess for social needs in general practice is limited, and there are no national guidelines for practices to follow. Variations in practice and lack of robust evaluation of social prescribing schemes in England mean we are not certain that existing interventions are reaching patients with the greatest social needs, or effectively helping them.^3,4

One option to help target social prescribing interventions is to develop more systematic approaches to assessing patients’ social needs — for instance, by standardising social risk assessments and screening all patients in a practice. This would mean introducing some kind of ‘social risk assessment’ in which patients are asked questions about select social issues, such as social isolation or food insecurity, and responses are recorded in medical notes. This is an approach already being taken in the US, where multiple major healthcare organisations have developed standardised social risk assessments and several federal policies incentivise using those instruments for more universal screening.^5,6

The idea is that more standardised and widespread screening for social risks reduces potential bias in who gets asked, as well as stigma for patients, and enables more information to be collected to support person-centred care. But there are also …