Health Survey for England 2020-21 Feasibility Study

In March 2020, face-to-face in-home interviewing was suspended for the Health Survey for England (HSE) and other government surveys in response to the COVID-19 pandemic and the government’s guidance. In response, the HSE 2020 survey fieldwork design was adapted to explore remote data collection through a push-to-web design. The HSE 2020-21 Feasibility Study tested the suitability of transferring some key HSE survey content from face-to-face interviewing to self-completion modes e.g. online and paper self-completion. This publication describes the key findings and response rates achieved, and considers the data quality and the accuracy of survey estimates.

Key findings

• Overall response rates achieved using remote data collection methods were far lower compared with the face-to-face survey: a household response rate of 25% compared with 60% achieved on HSE 2019.
• This was also true for acceptability rates for a future nurse visit and data linkage consents. 52% of adults agreed to a nurse visit compared with 85% in HSE 2019. Similarly, data linkage agreement rates were higher in HSE 2019: 93% agreeing to health records data linkage and 96% to civil registration mortality data linkage compared with 79% and 80% in the HSE FS.   
• The socio-demographic profile of the responding sample was compared with the face-to-face survey. Responding adults in the HSE FS were more likely to be older, from white backgrounds and living in the more affluent areas of the country. These biases persisted after non-response weighting.
• Questions, response options and format were kept identical where possible to the face-to-face survey. However, the complexity of the survey meant that this was not possible across every question or mode, potentially resulting in measurement error.
• For some health measures there were no significant differences between the survey estimates across modes. These included estimates for acute sickness; e-cigarette use; alcohol drunk in the last year and at least once a week. 
• There were a number of significant differences between the estimates for some key measures including: general health (very bad/bad), limiting long term illness, rates of current smoking, levels of physical activity; and portions of fruit and vegetables consumed. For these measures it is difficult to determine the causes of the differences as they are likely to be a combination of differences in the profile of the achieved sample, mode effects, or changes in behaviours because of the COVID-19 pandemic.

Researchers: Dhriti Mandalia, Beverley Bates and Katie Ridout.

Acknowledgement to: Rebecca Light and Mari Toomse-Smith