Table 3.

Parameter estimates (95% confidence intervals) from sensitivity analyses for average HbA1c (mmol/mol) levels over the follow-up

	Baseline level at 01/2015	Pre-policy change trend in 01/2014–12/2016	Change in level at the time of the policy change in 01/2017	Post-policy change trend in 02/2017–9/2019
Primary result	52.40 (51.88 to 52.91)	0.07 (0.04 to 0.09)	0.81 (0.04 to 1.58)	0.03 (− 0.03 to 0.08)
Results from sensitivity analyses when...
...excluding 01/2017	52.52 (51.50 to 53.54)	0.06 (0.02 to 0.10)	1.00 (−0.01 to 2.01)	0.03 (−0.07 to 0.13)
...excluding 01–02/2017	52.44 (51.89 to 53.00)	0.06 (0.04 to 0.09)	0.72 (−0.13 to 1.57)	0.04 (−0.02 to 0.10)
...excluding 01–03/2017	52.81 (51.79 to 53.23)	0.06 (0.03 to 0.10)	0.76 (−0.31 to 1.83)	0.04 (−0.04 to 0.12)
...excluding 01–04/2017	52.55 (51.82 to 53.28)	0.06 (0.02 to 0.09)	0.99 (−0.10 to 2.07)	0.04 (−0.05 to 0.12)
...excluding 01–05/2017	52.59 (51.53 to 53.66)	0.05 (0.01 to 0.10)	1.28 (0.14 to 2.41)	0.03 (−0.08 to 0.15)
...excluding 01–06/2017	52.76 (51.73 to 53.80)	0.04 (0.00 to 0.09)	1.37 (−0.05 to 2.79)	0.04 (−0.09 to 0.16)
...applying 2-month time windows	53.04 (52.64 to 53.44)	0.09 (0.05 to 0.12)	0.86 (0.28 to 1.44)	0.06 (−0.01 to 0.12)
...restricting the population to those who survived until Oct 1, 2019 (n = 7669)	51.95 (51.46 to 52.45)	0.07 (0.05 to 0.10)	1.00 (0.25 to 1.75)	0.03 (−0.02 to 0.09)