Leverhulme Centre for Demographic Science, University of Oxford, Oxford, UK.
Leverhulme Centre for Demographic Science, University of Oxford, Oxford, UK
BMJ Open. 2021 Oct 25;11(10):e054200. doi: 10.1136/bmjopen-2021-054200.
Non-pharmaceutical interventions (NPIs), including wearing face covering/masks, social distancing and working from home, have been introduced to control SARS-CoV-2 infections. We provide individual-level empirical evidence of whether adherence reduces infections.
The COVID-19 Infection Study (CIS) was used from 10 May 2020 to 2 February 2021 with 409 009 COVID-19 nose and throat swab tests nested in 72 866 households for 100 138 individuals in the labour force aged 18-64.
ORs for a positive COVID-19 test were calculated using multilevel logistic regression models, stratified by sex and time, by an index of autonomy to abide by NPIs, adjusted for various socioeconomic and behavioural covariates.
Inability to comply with NPIs predicted higher infections when individuals reported not wearing a face covering outside. The main effect for inability to comply was OR 0.79 (95% CI 0.67 to 0.92), for wearing face covering/masks was OR 0.29 (95% CI 0.15 to 0.56) and the interaction term being OR 1.25 (95% CI 1.07 to 1.46). The youngest age groups had a significantly higher risk of infection (OR 1.52, 95% CI 1.28 to 1.82) as did women in larger households (OR 1.04, 95% CI 1.02 to 1.06). Effects varied over time with autonomy to follow NPIs only significant in the pre-second lockdown May-November 2020 period. Wearing a face covering outside was a significant predictor of a lower chance of infection before mid-December 2020 when a stricter second lockdown was implemented (OR 0.44, 95% CI 0.27 to 0.73).
The protective effect of wearing a face covering/mask was the strongest for those who were the most unable to comply with NPIs. Higher infection rates were in younger groups and women in large households. Wearing a face covering or mask outside the home consistently and significantly predicted lower infection before the 2020 Christmas period and among women.
非药物干预措施（NPIs），包括佩戴口罩、保持社交距离和居家办公，已被用于控制 SARS-CoV-2 感染。我们提供了个体层面的证据，证明遵守 NPIs 可以降低感染风险。
COVID-19 感染研究（CIS）于 2020 年 5 月 10 日至 2021 年 2 月 2 日使用，在 72866 户家庭中嵌套了 409009 次 COVID-19 鼻喉拭子检测，共有 100138 名 18-64 岁的在职人员。
使用多水平逻辑回归模型，按性别和时间分层，以遵守 NPIs 的自主指数为指标，调整了各种社会经济和行为协变量，计算了 COVID-19 检测阳性的比值比（OR）。
当个体报告外出时不戴口罩时，无法遵守 NPIs 预测感染率更高。无法遵守 NPIs 的主要作用是 OR 0.79（95%CI 0.67 至 0.92），戴口罩/面罩的作用是 OR 0.29（95%CI 0.15 至 0.56），交互项为 OR 1.25（95%CI 1.07 至 1.46）。年龄最小的组感染风险显著较高（OR 1.52，95%CI 1.28 至 1.82），大家庭的女性也是如此（OR 1.04，95%CI 1.02 至 1.06）。随着时间的推移，自主遵守 NPIs 的效果也发生了变化，仅在 2020 年 5 月至 11 月第二次封锁前的时期具有统计学意义。外出时戴口罩是感染几率较低的重要预测因素，在 2020 年 12 月中旬实施更严格的第二次封锁之前（OR 0.44，95%CI 0.27 至 0.73）。
对于最无法遵守 NPIs 的人来说，戴口罩/面罩的保护作用最强。感染率较高的是年轻群体和大家庭中的女性。在 2020 年圣诞节前和女性中，始终如一地且显著地预测外出时戴口罩或面罩可降低感染风险。
