Table 4. Differences between participants’ characteristics and likelihood of PrEP use.

	City of residence				Experience taking regular medication				HIV testing				Transactional sex at present
	MD	MD			MD	MD			MD	MD			MD	MD
	B	C	U	r	Yes	No	U	r	Yes	No	U	R	Yes	No	U	r
Willingness to take PrEP	3	3	7815	0.07	3	3	693	0.16*	3	3	7511	0.08	4	3	3103	0.08
As soon as it becomes available	3	3	5639	0.14*	3.5	3	463	0.09	3	3	6051	0.06	3	3	2482	0.09
Despite side effects	3	3	7879	0.07	4	3	1194	0.02	3	3	7643	0.06	3	3	3061	0.09
Despite having to pay	4	4	7382	0.13*	4	4	1114	0.04	4	4	8098	0.01	4	4	3091	0.08
Despite having to use condoms	4	4	8227	0.03	4	3.5	1042	0.06	4	4	8075	0.01	4	4	2926	0.11
Despite being tested regularly for HIV	3	3	8231	0.02	3.5	3	908	0.10	3	3	7605	0.07	4	3	2527	0.18 **
Would you share PrEP	3	3	8376	0.01	3	3	1063	0.05	3	3	7979	0.02	3	3	3309	0.04
Would you sell PrEP	2	1	7149	0.14*	1	2	903	0.03	2	2	7411	0.09	3	2	2508	0.18 **
Would want partner(s) to know	4	4	8019	0.05	4.5	4	1035	0.15*	4	4	7263	0.10	4	4	2917	0.11
Would find taking PrEP embarrassing	1	1	8385	0.01	1	1	1059	0.07	1	1	7751	0.06	1	1	3491	0.01
Would you feel anxious about taking PrEP	2	2	8417	0.00	2.5	2	1091	0.04	2	2	7421	0.08	2	2	3102	0.07
Would taking PrEP give you hope	3	3	8360	0.01	4	3	1017	0.07	3	3	8179	0.00	4	3	2986	0.10

^**Difference is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed).

MD: median. U: Mann Whitney’s U-statistic. r: effect size estimate (r = Z/√N; N = number of observations). B: Bangkok; C: Chiang Mai.