Multi-level evidence for the impact of pain on workplace attendance

Chronic pain is a global obstacle to well-being and workplace productivity, yet its full extent and impacts are extremely difficult to measure and monitor. Utilising two novel sets of digital footprints data, we develop a new approach to model the prevalence of pain through self-medication behaviours, with the purchases of pain relief products leveraged as a proxy indicator. In Study 1, we obtain anonymous shopping records from a pharmacy chain en masse (20,500,952 customers) and compute pain prevalence for each local authority district (LAD) in the UK, showing regional-level evidence that the prevalence of pain is associated with shorter working hours and more individuals working part-time as indicated by official statistics. In Study 2, we gather shopping history and survey responses directly from 828 participants, demonstrating individual-level evidence that those who purchase more painkillers experience more pain, are less likely to work full-time, and also more likely to be restricted in their workplace attendance. This paper introduces a novel approach to measure prevalence of pain and illustrates the detrimental effects of pain on work-related outcomes beyond absenteeism, offering key insights into how self-medication patterns can be extracted from shopping data and linked to health conditions. • Prevalence of pain is difficult to measure, but can be captured from shopping data. • We developed approached to link massive digital footprints to labour statistics. • We utilised data donation to gather ground truth about pain and workplace attendance. • Geographically, pain predicts shorter working hours and people working part-time. • Individually, pain poses a limitation on people attending workplace as usual.