Table 3.

The proportion of people who had discontinued PrEP 2 years after the first supply, and crude and adjusted hazards ratios for discontinuation (Cox regression analysis)

PrEP users a		Discontinued at 2 years b			Discontinuation (Cox regression) c
		%	n	p‐value d	Unadjusted HR (CI)	p‐value		Adjusted HR (CI)	p‐value
PrEP user characteristics
Sex
Male	29,169	37.7%	(10,853)	<0.001	ref			ref
Female	380	78.6%	(297)		3.65	(3.25–4.09)	<0.001	2.99	(2.65–3.38)	<0.001
Age group
18–29	8302	48.9%	(4054)	<0.001	1.76	(1.68–1.84)	<0.001	1.62	(1.55–1.70)	<0.001
30–39	10,124	35.3%	(3.566)		1.15	(1.09–1.20)	<0.001	1.15	(1.10–1.21)	<0.001
40+	11,123	31.7%	(3530)		ref			ref
Gay prevalence area e
Low	8529	44.3%	(3770)	<0.001	1.54	(1.48–1.62)	<0.001	1.02	(0.96–1.07)	0.59
Med	10,745	38.7%	(4167)		1.32	(1.26–1.39)	<0.001	1.03	(0.98–1.08)	0.29
High	9754	31.1%	(3.026)		ref			ref
Subsidy type f
General	24,368	38.2%	(9302)	<0.001	ref			ref
Concessional	5181	35.7%	(1848)		0.90	(0.86–0.95)	<0.001	0.75	(0.72–0.79)	<0.001
Year first dispensed
2018	22,482	34.3%	(7705)	<0.001	ref			ref
2019	7057	48.2%	(3445)		1.60	(1.54–1.67)	<0.001	1.38	(1.32–1.44)	<0.001
Final dispensed supply
30 tablets or less	17,188	41.5%	(7124)	<0.001	ref			ref
More than 30 tablets	12,361	32.6%	(4025)		0.82	(0.79–0.85)	<0.001	1.09	(1.04–1.13)	<0.001
Prescriber characteristics
Gay prevalence area e , g
Low	2585	64.1%	(1657)	<0.001	2.80	(2.65–2.96)	<0.001	1.98	(1.85–2.12)	<0.001
Med	7228	45.0%	(3252)		1.60	(1.53–1.67)	<0.001	1.37	(1.31–1.44)	<0.001
High	19,409	36.7%	(5956)		ref			ref
Prescriber caseload g , h
1–10	4022	61.2%	(2463)	<0.001	2.51	(2.40–2.64)	<0.001	1.78	(1.69–1.89)	<0.001
11–100	4785	46.7%	(2234)		1.65	(1.57–1.73)	<0.001	1.48	(1.41–1.56)	<0.001
>100	20,742	31.1%	(6453)		ref			ref

Abbreviations: CI, confidence interval; HR, hazards ratio.

^a People with first PrEP supply between April 2018 and September 2019 with all data censored at 2 years from the date of first supply.

^b Discontinuation is defined as more than 180 days without PrEP after the final dispensed supply at 2 years, equating one dose of dispensed PrEP to 1 day supply.

^c Time from when the definition of discontinuation could be first met until the definition was met or 2 years (censor).

^d Pearson's chi‐squared test.

^e Estimated prevalence of gay men in area of patient residence (at the time of first prescription) or prescriber practice (low <1.0%, medium 1–5.0% and high >5%).

^f Concessional subsidy is a higher level of government subsidy resulting in a lower patient co‐payment based on income, employment, health, health expenditure or disability.

^g Where one individual has been dispensed prescriptions from more than one prescriber, the higher caseload and the higher gay prevalence practice location is assigned.

^h Number of individuals for whom that patient's prescriber has prescribed PrEP.