Background The NHS long term plan commits to ‘digital first primary care’ by 2024. Increasing reliance on digital access may disadvantage those who do not use the internet. We aimed to assess changes in internet use in adults over 50 years of age before and during the coronavirus pandemic.

Methods Participants in the English Longitudinal Study for Ageing were asked how often they used the internet or email in Wave 9 (W9) from June 2018 to June 2019 and COVID Wave 1 (CW1) from June to July 2020. Response options were daily, weekly, monthly, every 3 months or never. Multivariate logistic regression on weighted data was performed to assess variation by sex, age group (50 to state pension age (SPA), SPA to 74, 75 and over), and wealth quintile.

Results 5,142 core participants responded to both W9 and CW1. Of these, 553 (10.75%; 95% confidence interval (CI) 9.71 to 11.89) reported never using the internet in W9 and 733 (14.26%; 13.05 to 15.57) in CW1. Of those aged 75 and older, 320 (30.64%; 26.87 to 34.87) were ‘never users’ in W9 and 419 (40.03%; 35.51 to 45.03) in CW1. Univariate analysis found that the odds of reporting ‘never use’ were higher for women than men (W9 odds ratio (OR) 1.39; (CI) 1.11 to 1.73, CW1 1.35; 1.11 to 1.66), older age groups (W9 4.21; 3.36 to 5.27, CW1 4.24; 3.50 to 5.14), and less wealthy quintiles (W9 1.18; 1.10 to 1.26, CW1 1.19; 1.11 to 1.27). Multivariate analysis found that age was the most important predictor of never using the internet. The odds for older age groups were 4.73; 3.81 to 5.89 (W9) and 4.93; 4.09 to 5.93 (CW1). The differences between women and men, and between wealth quintiles, were no longer statistically significant.

Conclusion The proportion of participants reporting that they never used the internet increased slightly during the pandemic and included 4 in every 10 of those aged 75 and older. A limitation is that W9 data were collected using a paper survey delivered by an interviewer, and CW1 were administered over the phone or internet. Overall, there is a substantial risk that a ‘digital first primary care’ policy will create barriers for those aged over 75 years. It will be important to maintain alternative access routes to avoid increasing barriers to health care access and subsequent inequalities in the care provided to older people in England.