Understanding the role of peer group membership in reducing HIV-related risk and vulnerability among female sex workers in Karnataka, India

Figures & data

Table 1. Multivariate analysis of membership in a peer group by socio-demographic and sex work characteristics: BTS, 2010.

	Membership in group/collective
Socio-demographic and sex work factors	N	(%)	Adjusted OR (95% CI)	p-Value
Age in completed years
<=24	245	66.5	Ref.
25–29	379	73.3	1.074 (0.712, 1.618)	0.618
30–34	341	76.4	1.240 (0.805, 1.910)	0.252
35+	774	81.6	1.505 (0.987, 2.293)	<0.05
Educational status
Literate	478	75.6	Ref.
Illiterate	1261	77.1	1.075 (0.840, 1.375)	0.273
Marital status
Never married	224	75.4	Ref.
Currently married	782	76.9	1.039 (0.734, 1.473)	0.431
Divorced/separated	733	76.8	1.036 (0.729, 1.471)	0.805
Migration
Migrant	123	59.2	Ref.
Non-migrant	1615	77.9	1.600 (1.006, 2.545)	<0.05
Having source of income other than sex work
No	551	72.2	Ref.
Yes	1186	78.7	1.118 (0.843, 1.483)	0.480
Place of solicitation
Brothel/lodge/dhaba	77	53.2	Ref.
Home/rented room	181	73.6	1.591 (0.858, 2.951)	0.112
Public place (“street”)	1016	78.6	1.999 (1.191, 3.355)	<0.01
Other	465	76.9	1.796 (1.038, 3.110)	<0.05
Duration in sex work
<2 years	189	65.4	Ref.
2–4 years	641	74.6	0.984 (0.634, 1.526)	0.206
5–9 years	476	78.4	1.080 (0.665, 1.753)	0.172
10+years	433	82.7	1.269 (0.740, 2.177)	<0.05
Weekly client volume
10+	263	65.6	Ref.
5–9	452	78.6	1.931 (1.365, 2.729)	<0.01
<5	1024	78.6	1.927 (1.433, 2.591)	<0.01
Duration of exposure to program
<=2 years	716	69.6	Ref.
>2 years	1006	82.5	2.042 (1.620, 2.572)	<0.01
District
Belgaum	382	85.3	Ref.
Gulbarga	374	77.4	0.563 (0.375, 0.845)	<0.01
Gadag	287	86.6	1.105 (0.678, 1.800)	0.714
Dharwad	341	71.6	0.483 (0.317, 0.736)	<0.01
Solapur	355	63.3	0.354 (0.222, 0.566)	<0.001
Totals	1739	76.7

Notes: Analysis excluded 11 cases for which information on group membership was missing. Ref: Categories with the lowest collective membership were taken as reference category for the logistic regression model. The model was adjusted for all the socio-demographic and sex work characteristics mentioned in the table.

Table 2. Behavioral outcomes by peer group or collective membership status: BTS, 2010.

Predictor	Condom use with regular partner at last sex	Consistent condom use with all partner/clients	Experienced violence in the past six months	Beaten/forced to have sex in past one year	Did not give bribe to police to avoid trouble	Obtained any form of identification document in past five years
Membership in peer group or collective
No (N=409)	49.8	90.5	28.2	19.5	82.3	61.6
Yes (N=1330)	54.3	91.3	19.7	15.8	88.5	67.7
Significance (χ^2)	0.163	0.626	<0.001	0.090	<0.01	<0.05
Adjusted OR (95% CI)^a^,^b	1.250 (0.928, 1.684)	1.071 (0.707, 1.623)	0.700 (0.531, 0.922)	0.837 (0.615, 1.138)	1.460 (1.035, 2.059)	1.227 (0.955, 1.577)
p-Value	0.14	0.75	0.01	0.26	<0.05	0.090

Notes: The analysis excluded 11 cases for which information on membership was missing.

^a The reference category for the logistic regression model was no membership in a peer group or collective.

^b The model was adjusted for all of the socioeconomic and sex work characteristics mentioned in Table 1.

Table 3. Biological outcomes by peer group or collective membership status: IBBA, 2005–2008.

Predictor	Gonorrhoea and/or Chlamydia	Syphilis	HIV infection
Membership in peer group or collective
No (N=2937)	9.6	10.3	18.6
Yes (N=1762)	5.2	8.2	16.9
Significance (χ^2)	<0.001	0.05	0.17
Adjusted OR (95% CI)^a^,^b	0.603 (0.468, 0.777)	0.741 (0.583, 0.940)	0.890 (0.741, 1.069)
p-Value	<0.001	0.01	0.11

Notes: ^aThe reference category for the logistic regression model was no membership in a peer group or collective.

^b The Model was adjusted for all of the socioeconomic and sex work characteristics mentioned in Table 1.

Table 4. Propensity score matching analysis: Effect of peer group or collective membership on various outcome measures, comparing FSW peer group or collective members with matched controls: BTS, 2010 and IBBA, 2005–2008.

	Membership in peer group/collective
Outcomes	Treated (%)	Matched controls (%)	Average treatment effect ATT (adjusted effect of membership)	Bootstrap standard error	p-Value	Data source
Condom use with regular partner	54.7	49.7	5.0	0.0091	<0.10	BTS, 2010
Consistent condom use with all clients/partner	89.2	86.7	2.5	0.0105	–
Experienced violence in past six months	22.2	31.7	−9.5	0.0055	<0.001
Beaten or forced to have sex in the past one year	17.6	23.8	−6.2	0.0978	–
Given nothing to police to avoid trouble in past six months	87.5	82.3	5.2	0.0051	<0.001
Obtained any form of identification document in last five years	70.5	65.1	5.4	0.0065	<0.05
Gonorrhoea and/or Chlamydia	5.7	10.6	−4.9	0.0077	<0.001	IBBA, 2005–2008
Syphilis	8.4	10.3	−1.9	0.0078	<0.05
HIV infection	16.2	17.3	−1.1	0.0111	–