Table 3

Comparison of secondary with early latent infection: estimates and CIs for multinomial regression modelling*

Parameter	Estimate	SE (95% CI)	p Value
Intercept	175.39	142.23 (−103.63 to 454.41)	0.109
Strategic Health Authority
East Midlands (baseline^*)
East of England	13.80	306.01 (−586.52 to 614.13)	0.482
London	1.42	1.29 (−1.11 to 3.94)	0.136
North West	1.32	1.20 (−1.04 to 3.68)	0.136
South Central	11.52	215.44 (−411.13 to 434.16)	0.479
South East Coast	0.37	1.42 (−2.41 to 3.15)	0.397
South West	1.53	1.27 (−0.97 to 4.03)	0.115
West Midlands	0.37	1.37 (−2.32 to 3.06)	0.393
Yorkshire & The Humber	11.19	290.09 (−557.90 to 580.28)	0.485
Gender
Male (baseline^*)
Female	−0.74	1.60 (−3.88 to 2.39)	0.321
Whether born in the UK
UK born (baseline^*)
Non-UK born	0.61	0.36 (−0.10 to 1.32)	0.045
Ethnicity
White (baseline)
Black	0.60	0.94 (−1.24 to 2.44)	0.261
Asian	−0.77	0.66 (−2.06 to 0.52)	0.12
Other ethnicity	−0.40	0.90 (−2.17 to 1.38)	0.33
Sexual orientation
Heterosexual (baseline^*)
Men who have sex with men	0.37	0.67 (−0.95 to 1.69)	0.292
Reason for attending clinical services
Routine STI screen (baseline^*)
Symptoms	2.32	0.31 (1.71 to 2.92)	<0.001
Campaign	−11.76	342.18 (−683.04 to 659.52)	0.486
Partner notification	0.96	0.48 (0.02 to 1.90)	0.022
Antenatal screening	−12.26	717.42 (−1419.67 to 1395.16)	0.493
Other	0.33	0.39 (−0.43 to 1.10)	0.197
HIV serostatus
HIV negative (baseline^*)
HIV positive	0.97	0.29 (0.40 to 1.54)	<0.001
Infection acquired through oral sex
No (baseline^*)
Yes	−0.16	0.29 (−0.73 to 0.41)	0.29
Partner notification^†
Traceable	0.00	0.03 (−0.05 to 0.06)	0.438
Untraceable	0.00	0.01 (−0.01 to 0.01)	0.441
Location
London (baseline^*)
Brighton	−0.46	0.74 (−1.91 to −0.99)	0.268
Manchester	−0.18	0.57 (−1.31 to 0.94)	0.374
Elsewhere UK	0.85	0.57 (−0.27 to 1.96)	0.068
Overseas	−0.37	0.48 (−1.32 to 0.58)	0.223
Commercial sex worker
No (baseline^*)
Yes	−1.43	0.95 (−3.29 to 0.44)	0.067
Social network
None (baseline^*)
Sauna	−0.78	0.30 (−1.38 to −0.19)	0.005
Bar	−0.86	0.29 (−1.43 to −0.28)	0.002
Internet	−0.76	1.30 (−1.34 to −0.18)	0.005
Age
16–19 (baseline^*)
20–24	−1.30	0.27 (−3.86 to 1.25)	0.158
25–34	−0.72	1.27 (−3.22 to 1.77)	0.285
35–44	−0.76	1.29 (−3.29 to 1.77)	0.278
45+	−0.89	1.30 (−3.44 to 1.65)	0.245
Time
Year	−0.09	0.07 (−0.23 to 0.05)	0.108

Interpretation: each estimated regression coefficient provides a measure of increased (if positive) or decreased (if negative) likelihood of presenting with secondary, relative to early latent infection. For example, an HIV-positive individual would be expected to have a 1.62=exp{0.48} higher OR of presenting with primary syphilis infection (as opposed to early latent) compared with an HIV-negative patient.
↵* Baseline individual: white heterosexual, UK-born HIV-negative male, aged 16–19, who attended genitourinary medicine services in the East Midlands for a routine STI screen, likely to have acquired syphilis in London, but unlikely to have acquired infection through oral sex, not a commercial sex worker, and with no stated social/sexual networks.
↵† No baseline is shown for partner notification because it is a discrete variable.
STI, sexually transmitted infection.