Table 3.

Hazard ratios (95% confidence interval) for BNT162b2 effectiveness against PCR-confirmed COVID-19 infection in health care workers, relative to unvaccinated.

				Unadjusted*		Adjusted**
Dose	Week	Person-years	Events†	HR	95% CI	HR	95% CI
Unvaccinated		12,748	4,680	1.00		1.00
First	0–2	4,028	1,080	0.94	(0.88, 1.01)	0.97	(0.90 to 1.04)
	3–6	4,982	310	0.47	(0.41, 0.53)	0.48	(0.42 to 0.55)
	7+	2,123	140	0.60	(0.49, 0.74)	0.61	(0.50 to 0.76)
Second	0–1	2,711	60	0.31	(0.23, 0.42)	0.33	(0.24 to 0.44)
	2–5	5,372	50	0.13	(0.09, 0.20)	0.14	(0.09 to 0.21)
	6–13	10,414	230	0.21	(0.16, 0.28)	0.23	(0.17 to 0.31)
	14–17	4,849	180	0.36	(0.28, 0.46)	0.40	(0.31 to 0.52)
	18–21	4,554	360	0.38	(0.32, 0.45)	0.43	(0.36 to 0.50)
	22–25	4,341	670	0.42	(0.37, 0.48)	0.47	(0.41 to 0.53)
	26+	4,997	1,570	0.48	(0.43, 0.54)	0.55	(0.49 to 0.61)

^*Only stratification by categories of prior number of PCR tests.

^**Main effects included staff group, sex, age, ethnicity, SES, urban/rural classification, QCovid co-morbidity score, previous number of hospitalisations, GP attendances and prescriptions over the last five years, with stratification by 'prior PCR test history'.

^†For disclosure purposes, number of events have been round to nearest 10.