Abstract

Human immunodeficiency virus (HIV) prevention delivered in gay venues in US cities has been found to be effective in reducing HIV transmission in the 1990s but effects might not be generalizable to different times and settings. Doubts have been raised about: outreach's ability to address skills and explore personal behaviour; big-city commercial gay venues being appropriate sites for outreach because of gossip and social surveillance; and acceptability of outreach by professionals rather than ‘popular opinion formers’. We evaluated coverage, feasibility, acceptability and perceived impact of venue-based HIV prevention outreach by professionals in London, employing observation, surveys and interviews with venue-users, and focus groups/semi-structured interviews with workers. We found high coverage especially among target groups. Addressing negotiation skills and personal behaviour was feasible but required worker motivation and skill. Social surveillance rarely impeded work. Gay men generally found outreach acceptable and useful, and professionals were not regarded negatively. Impact on knowledge was commonly reported; impacts on negotiation skills and reflection on personal behaviour were more common among men experiencing longer contacts. In conclusion, professional HIV prevention outreach in gay venues in large cities is a feasible and acceptable intervention with significant potential impacts. Workers need to be well briefed and trained to maximize impact.

Introduction

Human immunodeficiency virus (HIV) infection among gay men remains an issue of great public health concern [1]; recent syphilis outbreaks underline the importance of sexual health interventions with this group [2]. HIV prevention delivered by outreach volunteers in gay venues in small US cities in the 1990s was found to be effective in reducing rates of sexually transmitted infections (STIs), a proxy for HIV transmission [3–5]. However, it cannot be assumed that the effectiveness of such work is generalizable to other times and settings [6–8] and there is a paucity of more recent intervention studies from other settings to assess this [9]. A number of factors might impede the effectiveness of such interventions a decade on and in larger gay communities.

Firstly, in the third decade of the HIV epidemic and with the advent of effective antiretroviral therapies, HIV prevention needs to engage with and support diverse, personal strategies for avoiding HIV transmission, which might, for example, involve using condoms with casual and/or known HIV sero-discordant partners and the negotiated non-use of condoms for partners of known HIV sero-concordancy. Whereas needs for awareness of HIV and knowledge of condom use were critical in the 1980s and 1990s; need for negotiation skills and for reflection about one's own HIV status and sexual behaviour might now be more important [10–13]. There is some evidence that outreach may be less effective in addressing these needs because educators find it unfeasible to initiate such discussion [14] and instead continue to impart factual information that gay men already know.

Secondly, whereas HIV prevention outreach has had demonstrable impact in the small, cohesive gay communities studied by Kelly et al., this might not be the case in larger communities characterized by diversity and lower levels of mutual support among gay men. In their evaluation of outreach in gay venues in Glasgow, Flowers and Hart [15] report that many big-city venues were characterized by high levels of gossip and mutually unsupportive cliques [15] which inhibited frank communication. Research by Hartley et al. [16] even casts doubt on whether outreach can reach a sufficient proportion of gay men in big cities to enable community-wide impacts, a suggestion also raised by Hart et al. [13].

Finally, whereas volunteers regarded as ‘popular opinion formers’ within gay communities demonstrably could reach out to educate and influence their peers in Kelly's studies, this might not be so with others undertaking such work. Today's HIV prevention workforce is increasingly professional [17] and while gay men may regard community self-education positively, they might view professional outreach as invasive and interfering. It has been suggested that outreach using paid workers is less acceptable and effective than that involving popular opinion formers [13, 18].

In the face of such doubts about the generalizable effectiveness of HIV prevention outreach in gay venues, this evaluation aimed to examine such work in London to inform local prevention planning and also as a case study to assess the wider potential of professional HIV prevention outreach within large gay communities in an era of new HIV prevention needs. The evaluation set out to answer the following specific questions:

  • (i) What is the coverage of the intervention and is this likely to be sufficient to enable community-wide impact?

  • (ii) Is it feasible to address negotiation skills and personal behaviour in outreach work?

  • (iii) Is HIV prevention outreach in gay venues acceptable to gay men?

  • (iv) What is the likely impact of outreach?

Methods

The intervention was provided by workers from two inter-collaborating teams, one voluntary sector and the other statutory sector. The former employed a manager, three full-time workers plus eight sessional workers, all gay men. The latter employed eight full-time workers, five being gay men and three straight women. Outreach was done in gay pubs, clubs and bars, and aimed to increase (i) men's desire not to be involved in HIV exposure, (ii) negotiation skills, (iii) knowledge about HIV and HIV testing, (iv) awareness of potential for HIV exposure in their own behaviour and (v) knowledge of other HIV and sexual health services. Outreach comprised one element of a multi-intervention programme, based on empowerment and ecological models of health promotion [19, 20] addressing biological, psychological and social ‘unmet needs’ with the aim of enabling men to reduce their vulnerability to HIV infection [21]. No formal psychological model or social learning theory informed the intervention or wider programme. Homosexually active men living in London were targeted, particularly sub-groups identified as at particularly great risk of infection [12]: those who had tested HIV sero-positive; those having >10 sexual partners per year; and those aged 25 years and under. The work had a number of delivery objectives: identifying venues and liaising with managers; achieving targets for number of visits; staffing a stand displaying leaflets; engaging in conversations with men at the stand and around the venue; providing information, advice and referral and distributing leaflets, condoms and lubricant to men.

Our evaluation was set out to explore ‘process’ in terms of the coverage of target groups, the feasibility of delivering intervention objectives and the acceptability of the intervention's aims and objectives with the target group. It also sought to explore the intervention's potential ‘impact’ in terms of whether it addressed gay men's needs, both as these were conceptualized within the programme (using quantitative data) and as the men themselves perceived these (using qualitative data). The evaluation employed a multi-method design involving non-participant observation of the work, a cross-sectional survey of men using venues, semi-structured interviews with a sub-sample of these men and focus groups and semi-structured interviews with workers. The post hoc funding of the evaluation prevented any experimental or interrupted time-series assessment of outcomes.

We purposively selected 29 venues using criteria of diversity of location, type of venue (bar/pub/club) and use by target groups (workers' perceptions), informed by previous profiling of London's commercial gay scene [22, 23]. Our venues also varied in the ethnicity of users and whether sex occurred on the premises. Evaluation fieldworkers (male and female) accompanied workers on visits to venues and afterwards recorded observations on structured schedules (type of venue; numbers, ages and ethnicity of men; planning of the session; engagement with men in the venue; topics addressed; factors influencing delivery and responses).

At these visits, fieldworkers also distributed self-completion, anonymous questionnaires to men. These asked about socio-demographic characteristics, use of venues and sexual behaviour in the last 12 months, HIV sero-status and views on the usefulness and impact of outreach work. Questions on socio-demographics, behaviour and sero-status were taken from previous instruments [24]. Questionnaires were piloted with gay men recruited as evaluation fieldworkers prior to their training. Fieldworkers explained the aims of the research and asked men for oral consent to participate. It proved impractical for fieldworkers to record refusal rates but a debriefing session established that these were of the order of 10–20%. Information about interviews was also provided on the front of questionnaires and men interested in being interviewed were asked to supply their first name and contact e-mail/telephone number. About half did so. These were then contacted to confirm their willingness to participate and offered a choice of a male or female interviewer. Interviews were carried out with a convenience sample of 45 men either face-to-face at our premises or by telephone. The semi-structured interview guide covered use of venues and views on HIV prevention outreach. Interview data were confidentialized and all contact details destroyed after interview.

The manager of each team was interviewed and focus groups were carried out with workers. Interviews and focus group guides examined the planning of outreach, aims and objectives, factors influencing feasibility and perceptions of the impact of the work. All interviews and focus groups were audiotaped and transcribed. The ethical acceptability of the study was approved by a committee of the university hosting the research.

We analysed survey data from men reporting homosexual activity ever and use of gay venues in the previous 12 months. Men were categorized according to socio-demographic and behavioural factors and HIV sero-status. Men were defined as potentially having experienced sero-discordant unprotected anal intercourse (sdUAI) if they reported a partner of different or unknown HIV sero-status. An ethnic category ‘Black and minority ethnic’ (BME) was created from our original ‘Asian/Asian British’, ‘Black/Black British’ and ‘other ethnic group’ categories since analysis of each group separately would have produced too few cases for statistical analysis. Quantitative analysis was facilitated by the use of Stata 7.0 software. Chi-square analysis produced P-values reported in Tables I and II.

Table I

Survey participants, and their contact and satisfaction with intervention components

  n (%) Workers staffing stand with leaflets
 
Leaflet from worker
 
Condom/lubricant from worker
 
≤5-min conversation with worker
 
>5-min conversation with worker
 
   Seen (%) Leaflet useful (%) Received (%) Useful (%) Received (%) Useful (%) Occurred (%) Useful (%) Occurred (%) Useful (%) 
Age <25 years 160 (14.3) 98 (69.5) 71 (51.8) 82 (58.2) 73 (53.3) 89 (63.1) 77 (56.2) 56 (40.3) 41 (29.9) 32 (23.0) 30 (21.9) 
 25–39 years 665 (59.5) 404 (65.5) 297 (52.7) 354 (57.4) 281 (49.8) 384 (62.2) 335 (59.4) 183 (30.1) 116 (20.6) 100 (16.53) 79 (14.0) 
 40+ years 293 (26.2) 180 (68.5) 132 (55.0) 157 (59.7) 130 (54.2) 154 (58.3) 130 (54.2) 78 (29.7) 48 (20.0) 49 (18.6) 36 (15.0) 
Ethnic group White 920 (87.6) 553 (65.2) 415 (52.9) 476 (56.1) 382 (48.7) 509 (60.0) 441 (56.3) 257 (30.5) 161 (20.5) 136 (16.2) 111 (14.2) 
 BME 130 (12.4) 96 (80.7)** 65 (59.6) 86 (72.3)** 75 (68.8)** 81 (68.6) 73 (67.0)* 48 (41.7)* 36 (33.0)** 33 (28.7)** 26 (23.9)** 
Educational qualifications None 47 (4.4) 21 (53.9) 12 (30.8) 22 (56.4) 14 (35.9) 24 (63.2) 15 (38.5) 12 (31.6) 5 (12.8) 9 (24.3) 3 (7.7) 
 School level 412 (38.7) 260 (67.2) 202 (56.3) 223 (57.6) 190 (52.9) 248 (64.1) 218 (60.7) 132 (34.2) 93 (25.9) 73 (18.9) 66 (18.4) 
 Degree or higher 605 (56.9) 378 (68.2) 272 (53.7)* 324 (58.5) 259 (51.1) 328 (59.2) 288 (56.8)* 158 (29.0) 98 (19.33)* 89 (16.5) 69 (13.6) 
Sex partners Male only 1026 (95.0) 634 (67.3) 469 (53.7) 548 (58.1) 448 (51.3) 548 (58.1) 509 (58.2) 298 (31.9) 193 (22.1) 170 (18.3) 137 (15.7) 
 Male and female 54 (5.0) 28 (59.6) 21 (46.7) 27 (58.7) 22 (48.9) 27 (58.7) 24 (53.3) 12 (26.1) 8 (17.8) 6 (13.0) 6 (13.3) 
Number of male partners <11 571 (67.5) 336 (63.2) 265 (54.4) 296 (55.6) 232 (47.6) 316 (59.6) 268 (55.0) 152 (28.9) 99 (20.3) 89 (17.0) 72 (14.8) 
 11+ 275 (32.5) 174 (68.0) 117 (49.0) 150 (58.6) 120 (50.2) 163 (62.9) 146 (61.1) 82 (32.0) 52 (21.8) 39 (15.3) 31 (13.0) 
HIV status Untested 277 (26.5) 161 (64.4) 101 (44.1) 127 (50.8) 102 (44.5) 143 (57.2) 125 (54.6) 58 (23.5) 35 (15.3) 30 (12.2) 24 (10.5) 
 Negative 662 (63.4) 412 (67.7) 311 (55.0) 367 (60.4) 300 (53.0) 385 (63.3) 336 (59.4) 202 (33.7) 132 (23.3) 111 (18.6) 93 (16.4) 
 Positive 105 (10.1) 75 (75.0)* 59 (62.1)** 68 (68.0)** 57 (60.0)* 64 (63.4) 55 (57.9) 43 (43.0)** 29 (30.5)** 30 (29.7)** 23 (24.2)** 
sdUAI Yes/do not know 130 (17.4) 83 (70.34) 58 (53.2) 73 (61.9) 61 (56.0) 84 (70.6) 73 (67.0) 40 (33.6) 27 (24.8) 21 (17.8) 19 (17.4) 
 No 619 (82.6) 372 (64.8) 274 (52.1) 316 (55.1) 257 (48.9) 339 (59.1)* 295 (56.1) 171 (30.1) 106 (20.2) 89 (15.6) 71 (13.5) 
Use of scene <1 per week 406 (39.6) 235 (63.3) 168 (49.6) 205 (55.3) 168 (49.6) 213 (57.4) 185 (54.6) 105 (28.3) 75 (22.1) 71 (19.2) 58 (17.1) 
 ≥1 per week 619 (60.4) 414 (69.7)* 304 (55.7) 355 (59.8) 287 (52.6) 383 (64.3)* 331 (60.6) 196 (33.5) 116 (21.3) 92 (15.7) 75 (13.4) 
All  1119 683 (66.8) 501 (53.2) 593 (58.0) 484 (51.4) 628 (61.4) 543 (57.6) 317 (31.4) 205 (21.8) 181 (18.0) 145 (15.4) 
  n (%) Workers staffing stand with leaflets
 
Leaflet from worker
 
Condom/lubricant from worker
 
≤5-min conversation with worker
 
>5-min conversation with worker
 
   Seen (%) Leaflet useful (%) Received (%) Useful (%) Received (%) Useful (%) Occurred (%) Useful (%) Occurred (%) Useful (%) 
Age <25 years 160 (14.3) 98 (69.5) 71 (51.8) 82 (58.2) 73 (53.3) 89 (63.1) 77 (56.2) 56 (40.3) 41 (29.9) 32 (23.0) 30 (21.9) 
 25–39 years 665 (59.5) 404 (65.5) 297 (52.7) 354 (57.4) 281 (49.8) 384 (62.2) 335 (59.4) 183 (30.1) 116 (20.6) 100 (16.53) 79 (14.0) 
 40+ years 293 (26.2) 180 (68.5) 132 (55.0) 157 (59.7) 130 (54.2) 154 (58.3) 130 (54.2) 78 (29.7) 48 (20.0) 49 (18.6) 36 (15.0) 
Ethnic group White 920 (87.6) 553 (65.2) 415 (52.9) 476 (56.1) 382 (48.7) 509 (60.0) 441 (56.3) 257 (30.5) 161 (20.5) 136 (16.2) 111 (14.2) 
 BME 130 (12.4) 96 (80.7)** 65 (59.6) 86 (72.3)** 75 (68.8)** 81 (68.6) 73 (67.0)* 48 (41.7)* 36 (33.0)** 33 (28.7)** 26 (23.9)** 
Educational qualifications None 47 (4.4) 21 (53.9) 12 (30.8) 22 (56.4) 14 (35.9) 24 (63.2) 15 (38.5) 12 (31.6) 5 (12.8) 9 (24.3) 3 (7.7) 
 School level 412 (38.7) 260 (67.2) 202 (56.3) 223 (57.6) 190 (52.9) 248 (64.1) 218 (60.7) 132 (34.2) 93 (25.9) 73 (18.9) 66 (18.4) 
 Degree or higher 605 (56.9) 378 (68.2) 272 (53.7)* 324 (58.5) 259 (51.1) 328 (59.2) 288 (56.8)* 158 (29.0) 98 (19.33)* 89 (16.5) 69 (13.6) 
Sex partners Male only 1026 (95.0) 634 (67.3) 469 (53.7) 548 (58.1) 448 (51.3) 548 (58.1) 509 (58.2) 298 (31.9) 193 (22.1) 170 (18.3) 137 (15.7) 
 Male and female 54 (5.0) 28 (59.6) 21 (46.7) 27 (58.7) 22 (48.9) 27 (58.7) 24 (53.3) 12 (26.1) 8 (17.8) 6 (13.0) 6 (13.3) 
Number of male partners <11 571 (67.5) 336 (63.2) 265 (54.4) 296 (55.6) 232 (47.6) 316 (59.6) 268 (55.0) 152 (28.9) 99 (20.3) 89 (17.0) 72 (14.8) 
 11+ 275 (32.5) 174 (68.0) 117 (49.0) 150 (58.6) 120 (50.2) 163 (62.9) 146 (61.1) 82 (32.0) 52 (21.8) 39 (15.3) 31 (13.0) 
HIV status Untested 277 (26.5) 161 (64.4) 101 (44.1) 127 (50.8) 102 (44.5) 143 (57.2) 125 (54.6) 58 (23.5) 35 (15.3) 30 (12.2) 24 (10.5) 
 Negative 662 (63.4) 412 (67.7) 311 (55.0) 367 (60.4) 300 (53.0) 385 (63.3) 336 (59.4) 202 (33.7) 132 (23.3) 111 (18.6) 93 (16.4) 
 Positive 105 (10.1) 75 (75.0)* 59 (62.1)** 68 (68.0)** 57 (60.0)* 64 (63.4) 55 (57.9) 43 (43.0)** 29 (30.5)** 30 (29.7)** 23 (24.2)** 
sdUAI Yes/do not know 130 (17.4) 83 (70.34) 58 (53.2) 73 (61.9) 61 (56.0) 84 (70.6) 73 (67.0) 40 (33.6) 27 (24.8) 21 (17.8) 19 (17.4) 
 No 619 (82.6) 372 (64.8) 274 (52.1) 316 (55.1) 257 (48.9) 339 (59.1)* 295 (56.1) 171 (30.1) 106 (20.2) 89 (15.6) 71 (13.5) 
Use of scene <1 per week 406 (39.6) 235 (63.3) 168 (49.6) 205 (55.3) 168 (49.6) 213 (57.4) 185 (54.6) 105 (28.3) 75 (22.1) 71 (19.2) 58 (17.1) 
 ≥1 per week 619 (60.4) 414 (69.7)* 304 (55.7) 355 (59.8) 287 (52.6) 383 (64.3)* 331 (60.6) 196 (33.5) 116 (21.3) 92 (15.7) 75 (13.4) 
All  1119 683 (66.8) 501 (53.2) 593 (58.0) 484 (51.4) 628 (61.4) 543 (57.6) 317 (31.4) 205 (21.8) 181 (18.0) 145 (15.4) 
*

P < 0.05,

**

P < 0.01.

Table II

Self-reported impact of intervention among those reporting any contact with health promotion and long conversation (LC)

  Impact on HIV knowledge
 
Impact on STI knowledge
 
Impact on knowledge of clinics
 
Impact on HIV test knowledge
 
Impact on knowledge of other services
 
Impact on negotiation skills
 
Impact on reflection about own sexual behaviour
 
Impact on hepatitis A/B vaccination
 
  Any LC Any LC Any LC Any LC Any LC Any LC Any LC Any LC 
Age <25 years 63 (50.4) 20 (74.1) 56 (44.8) 15 (55.6) 38 (30.4) 15 (55.6) 45 (36.0) 14 (51.9) 43 (34.4) 13 (48.2) 31 (24.8) 15 (55.6) 68 (54.4) 22 (81.5) 36 (28.8) 14 (51.9) 
 25–39 years 289 (51.2) 71 (75.5) 288 (51.0) 69 (73.4) 199 (35.2) 60 (63.8) 222 (39.3) 63 (67.0) 220 (38.9) 64 (68.1) 135 (23.9) 44 (46.8) 309 (54.7) 67 (71.3) 146 (25.8) 49 (52.1) 
 40+ years 109 (46.8) 24 (54.6)* 113 (48.5) 24 (54.6)* 67 (28.8) 17 (38.6)* 84 (36.1) 19 (43.2)* 82 (35.2) 24 (54.6) 60 (25.8) 12 (27.3) 141 (60.5) 33 (75.0)* 72 (30.9) 21 (47.7) 
Ethnicity White 374 (48.0) 85 (66.9) 374 (48.0) 79 (62.2) 242 (31.1) 68 (53.5) 281 (36.1) 71 (55.9) 276 (35.4) 75 (59.1) 180 (23.1) 49 (38.6) 425 (54.6) 92 (72.4) 203 (26.1) 60 (47.2) 
 BME 68 (66.7)** 24 (82.8) 62 (60.8)* 23 (79.3) 46 (45.1)** 20 (69.0) 54 (52.9)** 21 (72.4) 51 (50.0)** 21 (72.4) 33 (32.4)* 15 (51.7) 64 (62.8) 23 (79.3) 36 (35.3)* 19 (65.5) 
Education None 13 (36.1) 2 (25.0) 14 (38.9) 2 (25.0) 15 (41.7) 4 (50.0) 13 (36.1) 2 (25.0) 16 (44.4) 3 (37.5) 10 (27.8) 1 (12.5) 19 (52.8) 5 (62.5) 11 (30.6) 3 (37.5) 
 School level 189 (54.0) 48 (70.6) 180 (51.4) 44 (64.7) 119 (34.0) 40 (58.8) 137 (39.1) 38 (55.9) 144 (41.1) 45 (66.2) 93 (26.6) 27 (39.7) 214 (61.1) 51 (75.0) 104 (29.7) 34 (50.0) 
 Degree or higher 245 (48.5) 59 (72.8)* 244 (48.3) 56 (69.1)* 161 (31.9) 44 (54.3) 189 (37.4) 51 (63.0) 174 (34.5) 48 (59.3) 115 (22.8) 38 (46.9) 264 (52.3)* 58 (71.6) 127 (25.2) 41 (50.6) 
Sex of partners Male only 428 (49.8) 109 (69.4) 426 (49.6) 102 (65.0) 284 (33.1) 87 (55.4) 323 (37.6) 90 (57.3) 320 (37.3) 96 (61.2) 207 (24.1) 65 (41.4) 483 (56.2) 115 (73.3) 238 (27.7) 79 (50.3) 
 Male and female 25 (58.1) 4 (80.0) 19 (44.2) 3 (60.0) 16 (37.2) 3 (60.0) 19 (44.2) 3 (60.0) 20 (46.5) 2 (40.0) 14 (32.6) 3 (60.0) 22 (51.2) 4 (80.0) 7 (16.3) 2 (40.0) 
Number of male partners <11 237 (50.1) 65 (78.3) 233 (49.3) 55 (66.3) 148 (31.3) 51 (61.5) 177 (37.4) 54 (65.1) 170 (35.9) 54 (65.1) 109 (23.0) 36 (43.4) 265 (56.0) 64 (77.1) 126 (26.6) 45 (54.2) 
 11+ 113 (46.3) 23 (59.0)* 114 (46.7) 24 (61.5) 77 (31.6) 20 (51.3) 81 (33.2) 18 (46.2)* 91 (37.3) 24 (61.5) 49 (20.1) 10 (25.6) 126 (51.6) 28 (71.8) 64 (26.2) 20 (51.3) 
HIV status Untested 97 (44.3) 16 (61.5) 92 (42.0) 14 (53.9) 62 (28.3) 14 (53.9) 75 (34.3) 12 (46.2) 61 (27.9) 14 (53.9) 48 (21.9) 10 (38.5) 118 (53.9) 17 (65.4) 42 (19.2) 11 (42.3) 
 Negative 297 (53.5) 79 (76.0) 291 (52.4) 72 (69.2) 194 (35.0) 62 (59.6) 224 (40.4) 68 (65.4) 223 (40.2) 67 (64.4) 142 (25.6) 50 (48.1) 319 (57.5) 82 (78.9) 166 (29.9) 56 (53.9) 
 Positive 41 (43.2)* 17 (60.7) 45 (47.4)* 18 (64.3) 29 (30.5) 13 (46.4) 27 (28.4)* 12 (42.9)* 39 (41.1)** 16 (57.1) 23 (24.2) 8 (28.6) 46 (48.4) 18 (64.3) 29 (30.5)** 14 (50.0) 
sdUAI Yes/do not know 63 (54.8) 14 (73.7) 62 (53.9) 15 (79.0) 38 (33.0) 11 (57.9) 48 (41.7) 13 (68.4) 43 (37.4) 12 (63.2) 28 (24.4) 6 (31.6) 70 (60.9) 16 (84.2) 29 (25.2) 10 (52.6) 
 No 246 (48.1) 55 (65.5) 252 (49.2) 51 (60.7) 161 (31.5) 46 (54.8) 181 (35.4) 49 (58.3) 187 (36.5) 52 (61.9) 118 (23.1) 38 (45.2) 277 (54.1) 55 (65.5) 137 (26.8) 40 (47.6) 
Use of commercial scene Weekly or less 164 (48.1) 46 (68.7) 150 (44.0) 43 (64.2) 115 (33.7) 37 (55.2) 120 (35.2) 39 (58.2) 120 (35.2) 37 (55.2) 87 (25.5) 27 (40.3) 192 (56.3) 49 (73.1) 90 (26.4) 37 (55.2) 
 Weekly or more 265 (50.2) 58 (69.9) 280 (53.0) 57 (68.7) 172 (32.6) 49 (59.0) 208 (39.4) 50 (60.2) 203 (38.5) 55 (66.3) 122 (23.1) 37 (44.6) 290 (54.9) 61 (73.5) 142 (26.9) 38 (45.8) 
All  461 (49.9) 115 (69.7) 457 (49.5) 108 (65.5) 305 (33.0) 92 (55.8)*** 351 (38.0) 96 (58.2)*** 346 (37.5) 101 (61.2)*** 226 (24.5) 71 (43.0)*** 518 (56.1) 122 (73.9) 254 (27.5) 84 (50.9)*** 
  Impact on HIV knowledge
 
Impact on STI knowledge
 
Impact on knowledge of clinics
 
Impact on HIV test knowledge
 
Impact on knowledge of other services
 
Impact on negotiation skills
 
Impact on reflection about own sexual behaviour
 
Impact on hepatitis A/B vaccination
 
  Any LC Any LC Any LC Any LC Any LC Any LC Any LC Any LC 
Age <25 years 63 (50.4) 20 (74.1) 56 (44.8) 15 (55.6) 38 (30.4) 15 (55.6) 45 (36.0) 14 (51.9) 43 (34.4) 13 (48.2) 31 (24.8) 15 (55.6) 68 (54.4) 22 (81.5) 36 (28.8) 14 (51.9) 
 25–39 years 289 (51.2) 71 (75.5) 288 (51.0) 69 (73.4) 199 (35.2) 60 (63.8) 222 (39.3) 63 (67.0) 220 (38.9) 64 (68.1) 135 (23.9) 44 (46.8) 309 (54.7) 67 (71.3) 146 (25.8) 49 (52.1) 
 40+ years 109 (46.8) 24 (54.6)* 113 (48.5) 24 (54.6)* 67 (28.8) 17 (38.6)* 84 (36.1) 19 (43.2)* 82 (35.2) 24 (54.6) 60 (25.8) 12 (27.3) 141 (60.5) 33 (75.0)* 72 (30.9) 21 (47.7) 
Ethnicity White 374 (48.0) 85 (66.9) 374 (48.0) 79 (62.2) 242 (31.1) 68 (53.5) 281 (36.1) 71 (55.9) 276 (35.4) 75 (59.1) 180 (23.1) 49 (38.6) 425 (54.6) 92 (72.4) 203 (26.1) 60 (47.2) 
 BME 68 (66.7)** 24 (82.8) 62 (60.8)* 23 (79.3) 46 (45.1)** 20 (69.0) 54 (52.9)** 21 (72.4) 51 (50.0)** 21 (72.4) 33 (32.4)* 15 (51.7) 64 (62.8) 23 (79.3) 36 (35.3)* 19 (65.5) 
Education None 13 (36.1) 2 (25.0) 14 (38.9) 2 (25.0) 15 (41.7) 4 (50.0) 13 (36.1) 2 (25.0) 16 (44.4) 3 (37.5) 10 (27.8) 1 (12.5) 19 (52.8) 5 (62.5) 11 (30.6) 3 (37.5) 
 School level 189 (54.0) 48 (70.6) 180 (51.4) 44 (64.7) 119 (34.0) 40 (58.8) 137 (39.1) 38 (55.9) 144 (41.1) 45 (66.2) 93 (26.6) 27 (39.7) 214 (61.1) 51 (75.0) 104 (29.7) 34 (50.0) 
 Degree or higher 245 (48.5) 59 (72.8)* 244 (48.3) 56 (69.1)* 161 (31.9) 44 (54.3) 189 (37.4) 51 (63.0) 174 (34.5) 48 (59.3) 115 (22.8) 38 (46.9) 264 (52.3)* 58 (71.6) 127 (25.2) 41 (50.6) 
Sex of partners Male only 428 (49.8) 109 (69.4) 426 (49.6) 102 (65.0) 284 (33.1) 87 (55.4) 323 (37.6) 90 (57.3) 320 (37.3) 96 (61.2) 207 (24.1) 65 (41.4) 483 (56.2) 115 (73.3) 238 (27.7) 79 (50.3) 
 Male and female 25 (58.1) 4 (80.0) 19 (44.2) 3 (60.0) 16 (37.2) 3 (60.0) 19 (44.2) 3 (60.0) 20 (46.5) 2 (40.0) 14 (32.6) 3 (60.0) 22 (51.2) 4 (80.0) 7 (16.3) 2 (40.0) 
Number of male partners <11 237 (50.1) 65 (78.3) 233 (49.3) 55 (66.3) 148 (31.3) 51 (61.5) 177 (37.4) 54 (65.1) 170 (35.9) 54 (65.1) 109 (23.0) 36 (43.4) 265 (56.0) 64 (77.1) 126 (26.6) 45 (54.2) 
 11+ 113 (46.3) 23 (59.0)* 114 (46.7) 24 (61.5) 77 (31.6) 20 (51.3) 81 (33.2) 18 (46.2)* 91 (37.3) 24 (61.5) 49 (20.1) 10 (25.6) 126 (51.6) 28 (71.8) 64 (26.2) 20 (51.3) 
HIV status Untested 97 (44.3) 16 (61.5) 92 (42.0) 14 (53.9) 62 (28.3) 14 (53.9) 75 (34.3) 12 (46.2) 61 (27.9) 14 (53.9) 48 (21.9) 10 (38.5) 118 (53.9) 17 (65.4) 42 (19.2) 11 (42.3) 
 Negative 297 (53.5) 79 (76.0) 291 (52.4) 72 (69.2) 194 (35.0) 62 (59.6) 224 (40.4) 68 (65.4) 223 (40.2) 67 (64.4) 142 (25.6) 50 (48.1) 319 (57.5) 82 (78.9) 166 (29.9) 56 (53.9) 
 Positive 41 (43.2)* 17 (60.7) 45 (47.4)* 18 (64.3) 29 (30.5) 13 (46.4) 27 (28.4)* 12 (42.9)* 39 (41.1)** 16 (57.1) 23 (24.2) 8 (28.6) 46 (48.4) 18 (64.3) 29 (30.5)** 14 (50.0) 
sdUAI Yes/do not know 63 (54.8) 14 (73.7) 62 (53.9) 15 (79.0) 38 (33.0) 11 (57.9) 48 (41.7) 13 (68.4) 43 (37.4) 12 (63.2) 28 (24.4) 6 (31.6) 70 (60.9) 16 (84.2) 29 (25.2) 10 (52.6) 
 No 246 (48.1) 55 (65.5) 252 (49.2) 51 (60.7) 161 (31.5) 46 (54.8) 181 (35.4) 49 (58.3) 187 (36.5) 52 (61.9) 118 (23.1) 38 (45.2) 277 (54.1) 55 (65.5) 137 (26.8) 40 (47.6) 
Use of commercial scene Weekly or less 164 (48.1) 46 (68.7) 150 (44.0) 43 (64.2) 115 (33.7) 37 (55.2) 120 (35.2) 39 (58.2) 120 (35.2) 37 (55.2) 87 (25.5) 27 (40.3) 192 (56.3) 49 (73.1) 90 (26.4) 37 (55.2) 
 Weekly or more 265 (50.2) 58 (69.9) 280 (53.0) 57 (68.7) 172 (32.6) 49 (59.0) 208 (39.4) 50 (60.2) 203 (38.5) 55 (66.3) 122 (23.1) 37 (44.6) 290 (54.9) 61 (73.5) 142 (26.9) 38 (45.8) 
All  461 (49.9) 115 (69.7) 457 (49.5) 108 (65.5) 305 (33.0) 92 (55.8)*** 351 (38.0) 96 (58.2)*** 346 (37.5) 101 (61.2)*** 226 (24.5) 71 (43.0)*** 518 (56.1) 122 (73.9) 254 (27.5) 84 (50.9)*** 
*

P < 0.05,

**

P < 0.01 (association between impact and socio-demographics),

***

P < 0.05 (association between impact and intervention).

Qualitative analysis of interview transcripts identified key themes using Lofland and Lofland's [25] key-wording and memoing-based approach, with iterative testing and refinement of analyses in the light of further data to enhance validity. The research took a broadly interpretivist approach; qualitative analysis aimed to develop a research account informed by participants' own understandings of their social reality [26]. Qualitative and quantitative data were regarded as complementary. Qualitative data provided rich exploration of participants' views about acceptability and feasibility as well as perceived impact, while quantitative data enabled an assessment of the frequency and variability of simple measures of these factors.

Results

Fieldworkers observed 32 sessions of outreach work across 20 different bars, pubs and clubs. Our survey involved 1119 men, with just below 85% reporting London residence. Our sample was diverse in terms of age (median 34 years) and largely White, with high rates of post-compulsory formal education and predominantly exclusively homosexual behaviour. Around a third reported >10 partners in the last 12 months. About a tenth of men knew themselves to be HIV sero-positive with another quarter not knowing their sero-status (Table I). About a sixth of the sample reported having engaged in potentially sdUAI, as defined above. About two-thirds of men used gay venues weekly or more (Table I). Men participating in semi-structured interviews were socio-demographically similar. Median age of interviewees was 35 years; 39 were White, 2 described themselves as Asian and 3 as ‘other’; 1 man reported having no educational qualifications, 18 men reported school-level qualifications, 22 men reported a degree or higher qualification and 4 men were not asked.

Interviews were carried out with the agency managers. Three focus groups with the seven core workers (two women) and seven sessional workers drawn from both agencies were undertaken. Two statutory agency workers, sick at the time of the focus group, were interviewed separately. Most staff were aged between 26 and 35 years with just two aged between 36 and 45 years. Most described themselves as White, two as Black/Black British and one as ‘other’.

Intervention coverage

Outreach involving two workers occurred daily. Bars and pubs in central London were usually worked between noon and 9.00 p.m., whereas bars and pubs outside of central London, as well as clubs, were usually worked between 9.00 p.m. and 2.00 a.m. Workers reported that most visits to venues involved staffing a stand, talking with men and distributing resources. Each week, at least one shift (more for the statutory agency working in central London) involved workers rapidly visiting many bars primarily to distribute large numbers of leaflets, condoms and lubricant.

About two-thirds of men present in venues had on at least one occasion noticed outreach workers staffing stands (Table I) and slightly fewer had received a leaflet or a condom and/or lubricant. BME and HIV sero-positive men were more likely than others to have seen a stand or received a leaflet, men potentially engaging in sdUAI were more likely to receive a condom/lubricant and men who used gay venues weekly or more were more likely to have seen a stand and to have received a condom/lubricant.

Most workers did not know to which groups the intervention was intended to be targeted but, when told who these were, said that these were the groups that they did target. Several workers said that men from BME groups and men who used drugs were also targeted. Workers reported that decisions about when to work in a venue were mainly guided by their collective experiences as to which groups of men used a venue, and at what times and in which venues different sorts of work was feasible. More formal profiling information was also used by one agency. This consisted of information about the venue type, capacity, atmosphere, age and ethnicity of clientele, layout and ideas for the possible types of work that could be carried out.

Our observations confirmed that outreach workers handed out leaflets and condoms as well as lubricant to large numbers of men. Outreach workers put considerable thought into placing stands in positions that would maximize impact while minimizing the likelihood they would interfere with social interaction or discourage men from approaching the stand by placing it in too overlooked a position. Good relations with venue staff were crucial in negotiating the best positions for stands. Occasionally, observations suggested that outreach workers chose or were forced by circumstances to operate in areas of a venue, such as the entrance or in an area off the main part of the venue, which impeded their ability to engage men in contacts.

Several gay men interviewed complained that outreach workers remained at their stands too much and did not proactively engage men:

I think they do need to make … the initial approach because if they just stand there … people aren't likely to come up to them.

Initiation of contacts appeared from observations, and interviews with men and workers to require worker pro-activity, which varied. Workers suggested that the following promoted pro-activity: communication skills; clarity of purpose; and maintaining enthusiasm by varying the approach they took to the work, for example, in terms of what topics they focused on in each session. Both workers and gay men reported that handing out of leaflets, and particularly condoms and lubricant, facilitated the initiation of conversations.

Observations suggest that drunkenness and drug use among some venue-users sometimes impeded workers successfully making contacts. Men's desire to socialize rather than engage in contacts with outreach workers could also sometimes render outreach challenging. Workers did however minimize these problems by opting to visit venues characterized by intense socializing or drunkenness/drug use in the early evening and/or during week nights.

Feasibility of addressing negotiation skills and personal behaviour

About a third of men surveyed had engaged in a short (≤5 min) conversation with a worker, and about a fifth in a longer conversation (Table I). BME and HIV sero-positive men were more likely than others to have had a short or long conversation. Observations confirmed that most contacts were short. Where conversations occurred, these usually focused, according to observations as well as interviews with workers and venue-users, on the provision of basic factual information rather than engaging with men's specific needs and how these might be addressed. Some longer contacts were observed, some of which involved in-depth conversations addressing men's personal needs.

Some workers were much more able to undertake such conversations than others and this required both communication skills and the belief that this was a key activity. Workers were generally not aware of the formal aims of the interventions and were uncertain about whether these included engaging men in intensive one-to-one conversations. Statutory workers were more likely to consider in-depth conversations a key part of outreach and some were particularly skilled at initiating and negotiating such conversations. Most voluntary agency workers instead saw outreach as being largely about general awareness raising. One such worker said that he thought it possible to talk with men in venues, but only when a man chose to initiate such a conversation:

If [men] wanted to initiate … conversations, to talk to us, then we could. But that‘s not something that we would probably … bring up, because that’s very personal.

Some workers stressed in interviews that they took care to assess men's state of mind and attitude in making contacts and only attempted in-depth conversations with men likely to be receptive to these. A few men did suggest that the lack of privacy, and in some venues an atmosphere of cliqueness or social surveillance, sometimes impeded outreach work because men would not want to discuss personal matters where they might be overheard and subject to ridicule or embarrassment. One man commented:

[Men] will tell you they're busy and all the rest of it, but basically I think a lot of people are frightened.

Another said:

They don't particularly want to talk to someone about [personal matters] perhaps with their friends … it might be a bit kind of embarrassing.

However, such views were rare and the extent of such perceptions was said to vary between venues. The degree of ‘cliqueness’ and ‘bitchiness’ of venues was actually a factor men used to categorize venues, alongside the profile of men, in terms for example of age and dress-style, using the venue. Crowding could also sometimes impede in-depth work in some but by no means all commercial venues. This also varied between different nights of the week. Background noise from music or talking very occasionally impeded conversations where privacy might be required. One gay man commented:

All you can hear is a mutter of voices and the last thing you want to do is start saying to someone ‘what, what, what?’

Workers took both crowding and noise into account in deciding the most appropriate times to visit venues, choosing, for example, to visit venues in central London in the early evening. Workers also opted to operate in quieter and less crowded areas within venues and/or to act to reduce noise, for example, by arranging with venue staff for loudspeakers near their stand to be turned down.

Acceptability of outreach work

Most men experiencing outreach found it useful. Reported usefulness ranged from 65% in the case of short conversations, 73% for workers staffing stands, 80% for long conversations, 82% for workers handing out leaflets to 86% in the case of workers distributing condoms/lubricant (Table I). BME men were significantly more likely to rate receiving leaflets and condoms/lubricant from workers and having short or long conversations as useful. HIV sero-positive men more often rated leaflets from stands or workers and short or long conversations as useful. Those with low formal education less often reported receiving a leaflet from a stand, a condom/lubricant from a worker or a short conversation as useful. Men potentially engaging in sdUAI were more likely to rate receiving a condom/lubricant from a worker as useful.

Men who had not had contact with outreach workers were asked why. Of these, 54% reported it was because they had not noticed workers, 31% said they did not need to talk to workers, 4% said workers were not ‘their sort of people’, 8% said they would like to but felt embarrassed, 20% said they were busy doing other things and 13% said they did not want to think about health promotion when out.

Most men interviewed supported the presence of workers in venues:

I think they do a very good job and I know where to go to get information if I need it.

A few complained that the gravity of HIV prevention jarred with the fun atmosphere of the setting. One man found the presence of workers in a bar unacceptable because their ‘preachy’ style was incompatible with the venue's relaxed feel:

You know when you were at school, and you used to have prefects…. It felt like that kind of thing…. You don't necessarily want to be a part of that because it's really not cool.

There were varying views as to what characteristics affected the acceptability of a particular educator. None of those interviewed considered the professional status of the educators problematic. Around half thought that women could undertake the work in all venues except those where sex occurred throughout the premises or that were men-only. Workers from the statutory agency which employed both men and women felt that this enabled them to meet the diversity of needs and preferences of gay men, but those from the voluntary agency considered gay men as better placed to provide outreach. None of the men interviewed raised the ethnicity of workers as an important factor, although BME men were under-represented among interviewees. Workers felt that when working in settings strongly identified as ‘Black venues’, it was easier to make contact with men and probably more acceptable to venue-users if a Black worker was present. Most men interviewed thought workers should be matched to particular venues by age and dress-style, and this was also emphasized by workers from one agency. Some men interviewed thought that physically attractive workers would find it easier to engage with men while others felt that someone of more ‘ordinary’ appearance was more ‘accessible’.

The impact of outreach work

Of men experiencing any form of outreach work, about half thought it had impacted on their knowledge of HIV or STIs, and slightly fewer thought it had made them reflect on their own sexual behaviour (Table II). Around a third thought it had made an impact on their knowledge of the location of sexual health clinics, HIV-testing facilities and other HIV services. About a quarter thought it had an impact on their negotiation skills, with a similar proportion believing it had encouraged them to be vaccinated against hepatitis A/B. Those experiencing conversations of >5 min were significantly more likely to report impacts on negotiation skills, hepatitis vaccination, HIV-testing knowledge, knowledge of other services and knowledge of sexual health clinics but not on knowledge of HIV or STIs, or reflection on one's own sexual behaviour.

BME men were more likely than others to report impacts other than reflecting on their own behaviour, from any intervention (Table II). Men aged ≥40 years were less likely to report impact on knowledge of HIV, STIs or clinics, or reflection on their sexual behaviour from long conversations. Those with low formal education were less likely to report impacts on HIV and STI knowledge from long conversations, and on reflection about behaviour, from any intervention. Men with >10 partners per year were less likely to report impacts on HIV and HIV-testing knowledge from long conversations. HIV sero-positive men were, unsurprisingly, less likely to report impact on knowledge of HIV knowledge and testing as well as being less likely to report impacts on knowledge about other services and hepatitis A/B vaccination. Men who used gay venues weekly or more were more likely to report impact on STI knowledge than others.

Some gay men interviewed suggested that outreach work would be most likely to bring about significant outcomes if it engaged with men's own behaviour and personal needs. Some felt that general awareness raising and information provision could have no impact. However, others questioned this view, arguing that short contacts or even merely the sight of workers could reinforce awareness of the importance of HIV prevention or prompt reflection about one's own risk-taking:

I think it keeps it in your … mind. You actually think about things more often … as opposed to, you know if you don't get reminded of it, you tend to let things slip and forget about things.

One man suggested that in order to achieve such reinforcement, a more consistent presence of outreach workers was required in venues:

I think probably if they were more obviously around, … almost in every bar you went to or club or whatever … so it was like a constant warning, like on a cigarette packet, maybe it might make a difference.

However, another suggested that coverage must not go above saturation point as this might lead to desensitization.

Some men suggested that face-to-face distribution of safer sex messages from workers increased its impact and helped guard against information fatigue:

I think that you're immune to [posters], to a lot of what you see. This is all the same old stuff…. There is a call for face-to-face interaction.

Some men also suggested that written materials were more likely to be read if workers handed them out:

The written word or having something you can … take with you and read when you're in a more sober frame of mind or when you have a bit more privacy … it's much more helpful than … having discussions. However, I think it does need a person to be there … to dish that out … because if it's just left on the side I don't think that it will get distributed properly.

It was suggested that another strength of outreach was that it could refer men to interventions in other settings where in-depth face-to-face communication about personal matters was more feasible. Some men reported such referrals:

I … asked him some questions…. ‘Where was the local [sexual health] clinic?’ … Yeah, he told me at the top of Tottenham Court Road … that there was one up there and that there was also a clinic … in Soho itself.

Other men sought out such services after reading written information from workers:

[It was in] those leaflets … that I actually found out about a clinic … where you could go to…. If you like that … encouraged me to do that.

Discussion

Our questionnaire sample appears broadly representative of men using gay venues in London and resembles that participating in another recent survey of London gay men recruited via clinic and community settings [27], in terms of age, ethnicity, education, sexual behaviour and known HIV sero-positivity. More men in our study did not know their HIV sero-status, probably reflecting our sample not being recruited in clinics. The gay men interviewed were also broadly similar in socio-demographic profile.

Our evaluation suggests that HIV prevention outreach in venues can reach sufficiently large numbers of men to achieve community-wide impacts. Many visits were made and large numbers of men reported contact with the intervention, coverage surpassing, for example, that previously reported by Hart et al. [13]. Workers needed to be proactive in making contacts rather than merely passively greeting men approaching their stand and they varied in their ability to do so. Although our survey may overestimate coverage, given that questionnaires were distributed in venues where outreach was ongoing, this is likely to be moderated by our questions referring to contact in the past tense. Our evaluation suggests that outreach work can be targeted specifically at those in most need, the work in London successfully targeting BME, HIV sero-positive and younger men.

Our results also suggest that it was feasible for workers to initiate conversations with gay men and, in them, to address negotiation skills and men's own behaviour. A significant minority of gay men had experienced a conversation with a worker of >5 min duration, particularly among target sub-groups, and most men found this acceptable. Some workers were, as Flowers et al. [14] have previously reported, reluctant to engage in personal discussions. Such discussions required workers to see this as a priority, to be well motivated and to have good communication skills. Initiating contact and having longer conversations addressing topics such as negotiation skills and personal behaviour were acceptable to gay men and only rarely impeded by environmental factors, such as noise and crowding, or social factors, such as atmospheres of social surveillance and gossip. These findings are in contrast to those of Flowers and Hart [15] from Glasgow, perhaps reflecting important cultural differences between London and Glasgow.

A high proportion of men found the interventions acceptable and useful, particularly those from targeted sub-groups, although men with little formal education were least satisfied. Only a few regarded the intervention as invasive of their leisure space. There was no consensus among providers as to whether or not only gay men should provide outreach; this aspect of provision was under-theorized. Men did not identify the professional status of educators as problematic and many were comfortable with straight female as well as gay male workers. We found no evidence to support the suggestion that interventions not employing ‘popular opinion formers’ [13, 18] are likely to have less impact.

Without either a comparison group or an interrupted time-series design, our evaluation provides data on impact of limited rigour. We surveyed men about their experience of outreach work and the benefits they thought this had brought. Because these findings are likely to be subject to some information bias, they should be interpreted cautiously. However, the reported pattern of impact seems plausible, in that outcomes such as increased knowledge about HIV were more often reported than improved negotiation skills and the latter were more often achieved among those experiencing in-depth conversations.

Although findings from our interviews with men must be treated with caution because of the small, non-random sample, these provide some support for the suggestion that even if, as Flowers et al. [14] found, health promoters cannot easily engage with men's personal circumstances, outreach work may still play an important role in maintaining awareness of HIV/acquired immunodeficiency syndome as an important issue, encouraging men to reflect inwardly on their behaviour, delivering written HIV prevention materials and referring men to other interventions where in-depth, personal discussion is more feasible.

In conclusion, our evaluation found gay-venue-based HIV prevention outreach work to have high coverage, to have the potential to address emerging needs such as for negotiation skills and reflection about one's own behaviour (even if this was not always realized), to be acceptable and not greatly hindered by gossip or cliqueness and to have a variety of potential impacts. Our findings strongly support those of Hart et al. [13] on the importance of adequate training and supervision of outreach workers. These findings from London are likely to be pertinent to other large gay communities in cities across Europe, North America and Australasia, given the striking similarities in terms of diverse composition and commercial culture [28]. However, some cultural variations do arise between such cities, for example, in the acceptability of discussion of personal issues in London but not in Glasgow, and these need careful consideration by those tailoring outreach to particular city contexts. Our results are likely to be much less readily applicable to work in cities where non-Western notions of same-sex behaviour remain influential or which lack a large commercialized gay scene [29].

We would like to thank the London National Health Service Primary Care Trusts for funding the study and the following individuals for their help in the research: Paul Boyce, David Clover, Helen Corbin, Daron Oram, Colin Turnbull, Ricardo Vasconcelos, Ford Hickson, Peter Weatherburn, Sarah Jones, Andrew Bibby, Robert Goodwin, Tim Green, Will Nutland and Glyn Thomas.

References

1.
Centers for Disease Control and Prevention
HIV prevalence, unrecognized infection, and HIV testing among men who have sex with men—five U.S. cities, June 2004 –April 2005
MMWR Morb Mortal Wkly Rep
 , 
2005
, vol. 
52
 (pg. 
597
-
601
)
2.
Kahn
RH
Heffelfinger
JD
Berman
SM
Syphilis outbreaks among men who have sex with men: a public health trend of concern
Sex Transm Dis
 , 
2002
, vol. 
29
 (pg. 
285
-
7
)
3.
Kelly
JA
St Lawrence
JS
Diaz
YE
, et al.  . 
HIV risk behavior reduction following intervention with key opinion leaders of a population: an experimental analysis
Am J Public Health
 , 
1991
, vol. 
81
 (pg. 
168
-
71
)
4.
Kelly
JA
St Lawrence
JS
Stevenson
LY
, et al.  . 
Community AIDS/HIV risk reduction: the effects of endorsements by popular people in three cities
Am J Public Health
 , 
1992
, vol. 
82
 (pg. 
1483
-
1489
)
5.
Kelly
JA
Murphy
DA
Sikkema
KJ
, et al.  . 
Randomized controlled community level HIV prevention and intervention for sexual risk behavior among homosexual men in US cities
Lancet
 , 
1997
, vol. 
350
 (pg. 
1500
-
5
)
6.
Miller
RL
Klotz
D
Eckholdt
HM
HIV prevention with male prostitutes and patrons of hustler bars: replication of an HIV prevention intervention
Am J Community Psychol
 , 
1998
, vol. 
26
 (pg. 
97
-
141
)
7.
Hart
G
Elford
J
Stephenson
J
Imrie
J
Bonell
C
The limits of generalisability: community-based sexual health interventions among gay men
Effective Sexual Health Interventions
 , 
2003
Oxford
Oxford University Press
(pg. 
191
-
200
)
8.
Kelly
JA
Heckman
TG
Stevenson
LY
, et al.  . 
Transfer of research-based HIV prevention interventions to community service providers: fidelity and adaptation
AIDS Educ Prev
 , 
2000
, vol. 
12
 
Suppl. A
(pg. 
87
-
98
)
9.
Johnson
WD
Hedges
LV
Ramirez
G
, et al.  . 
HIV prevention research for men who have sex with men: a systematic review and meta-analysis
J Acquir Immune Defic Syndr
 , 
2002
, vol. 
30
 
Suppl. 1
(pg. 
S118
-
29
)
10.
Strathdee
SA
Hogg
RS
Martindale
SL
, et al.  . 
Determinants of sexual risk-taking among young HIV-negative gay and bisexual men
J Acquir Immune Defic Syndr
 , 
1998
, vol. 
19
 (pg. 
61
-
6
)
11.
Imrie
J
Fenton
K
Billington
A
HIV prevention and homosexual men: should we be optimistic about the new millennium?
Sex Transm Infect
 , 
1999
, vol. 
75
 (pg. 
369
-
76
)
12.
Hickson
F
Weatherburn
P
Reid
D
Stephen
M
Out and About: Findings from the United Kingdom Gay Men's Sex Survey 2002
 , 
2003
London
Sigma Research
13.
Hart
GJ
Williamson
LM
Flowers
P
Good in parts: the Gay Men's Task Force in Glasgow—a response to Kelly
AIDS Care
 , 
2004
, vol. 
16
 (pg. 
159
-
165
)
14.
Flowers
P
Hart
GJ
Williamson
LM
Frankis
JS
Der
GJ
Does bar-based peer-led sexual health promotion have a community-level effect amongst gay men in Scotland?
Int J STD AIDS
 , 
2002
, vol. 
13
 (pg. 
102
-
8
)
15.
Flowers
P
Hart
G
Aggleton
P
Hart
G
Davies
P
Everyone on the scene is so cliquey: are gay bars an appropriate context for a community-based peer-ed intervention?
Families and Communities Responding to AIDS
 , 
1999
London
UCL Press
(pg. 
83
-
98
)
16.
Hartley
M
Hickson
F
Weatherburn
P
Scene It? Done It? Who Had Contact with Gay Men's HIV Prevention in London in 1999?
 , 
1999
London
Sigma Research
17.
Wohlfeiler
D
From community to clients: the professionalisation of HIV prevention among gay men and its implications for intervention selection
Sex Transm Infect
 , 
2002
, vol. 
78
 
Suppl. 1
(pg. 
176
-
82
)
18.
Kelly
JA
Popular opinion leaders and HIV prevention peer education: resolving discrepant findings, and implications for the development of effective community programmes
AIDS Care
 , 
2004
, vol. 
16
 (pg. 
139
-
50
)
19.
Caplan
R
The importance of social theory for health promotion: from discussion to reflexivity
Health Promot Int
 , 
1993
, vol. 
8
 (pg. 
147
-
57
)
20.
Dahlgren
G
Whitehead
M
Policies and Strategies to Promote Social Equity in Health
 , 
1991
Stockholm
Institute of Futures Studies
21.
Hickson
F
Nutland
W
Weatherburn
P
, et al.  . 
Making It Count: A Collaborative Framework to Reduce the Incidence of HIV Infection During Sex Between Men
 , 
2003
London
Sigma Research
22.
Annetts
J
Eisenstadt
K
Gatter
P
Gay and Bisexual Men's HIV Prevention Service Review and Mapping Project
 , 
1996
London
Lambeth, Southwark and Lewisham Health Authority
23.
Kelley
P
Pebody
R
Scott
P
How Far Will You Go?
 , 
1996
London
Gay Men Fighting AIDS
24.
Reid
D
Weatherburn
P
Hickson
F
Stephens
M
Know the Score. Findings from the National Gay Men's Sex Survey 2001
 , 
2002
London
Sigma Research
25.
Lofland
J
Lofland
L
Analyzing Social Settings: A Guide to Qualitative Observation and Analysis
 , 
1995
Belmont, CA
Wadsworth
26.
Giddens
A
The Constitution of Society
 , 
1984
Cambridge
Polity Press
27.
Dodds
J
Mercey
D
Sexual Health Survey of Gay Men London 2003
 , 
2004
London
Royal Free and University College Medical School
28.
Bell
D
Binnie
J
Authenticating queer space: citizenship, urbanism and governance
Urban Stud
 , 
2004
, vol. 
41
 (pg. 
1807
-
20
)
29.
Bhende
AA
Jenkins
C
Sexual culture and the risk environment of HIV/AIDS
AIDS
 , 
1998
, vol. 
12
 
Suppl. B
(pg. 
S51
-
8
)

Comments

0 Comments