Inequality in provider and patient-initiated healthcare cancellations during Covid-19
利用英国家庭纵向调查数据，研究了2020年4-7月间由提供者和患者发起的医疗预约取消情况，发现87%的取消由提供者发起，且年龄、种族、居住地等因素导致不同群体面临双重风险，加剧了医疗不平等。
The Covid-19 pandemic adversely affected access to healthcare raising concerns about worsening health, unmet need and subsequent 'displaced' demand. Yet little is known about how this displaced demand was distributed or whether it reflected patient's decisions to cancel versus provider's decisions to ration care. Using survey data for England from the UK Household Longitudinal Study, we examine whether planned care continued (or alternative provided), was cancelled by the provider or cancelled by the patient and how these outcomes vary across socio-demographic, clinical, regional and treatment-type characteristics. We estimate weighted multinomial logit models for April-July 2020 (pooled and wave specific), including region and month effects and a region-month Covid-mortality proxy for local NHS strain. Cancellations were overwhelmingly provider-initiated (87% vs 13% patient-initiated). A clear age gradient emerges: provider-initiated cancellations rise with age while patient-initiated cancellations fall; the provider-to-patient cancellation ratio is much higher for those aged 65+ than for younger adults, consistent with providers 'moving first' for older adults under capacity constraints. Several groups experienced 'double jeopardy' with elevated risks of both provider and patient cancellation: ethnic minority respondents, people in smaller households, urban residents, and those in the North East and Yorkshire and the Humber regions. Because provider cancellations predominated, providers largely determined which treatments continued. Without safeguards, such rationing risks amplified existing inequalities, particularly for double-jeopardy groups. Backlog recovery should protect elective capacity, especially for procedures, and prioritise proactive outreach and flexible scheduling for these groups, whilst reducing patient-side barriers.