Brief training of HIV medical providers increases their frequency of delivering prevention counselling to patients at risk of transmitting HIV to others

Abstract

Objective:

The aim of this study was to examine whether brief training of human immunodeficiency virus (HIV) medical providers increased the frequency with which they routinely delivered prevention counselling to patients, and whether patient characteristics were associated with receipt of that counselling.

Design:

Longitudinal.

Setting:

Seven HIV clinics in the United States.

Method:

Medical providers received training to deliver counselling to their HIV-positive patients. The evaluation was conducted with a longitudinal cohort (N = 729) of patients at baseline (before providers received training) and at six and 12 months after the intervention was initiated. Participants reported on receipt of routine counselling (at more than half of their primary care visits) about safer sex, disclosing HIV infection to sex partners, and alcohol/drug use. Logistic regression and generalized estimating equations (GEE) methods assessed changes in receipt of counselling over time.

Results:

The percentage of participants reporting routine safer-sex counselling increased significantly across time (b = 0.12, standard error [SE] = 0.04; p = 0.004), as did receipt of routine counselling about disclosing their HIV infection to sex partners (b = 0.21, SE = 0.05; p < 0.001). The trend was not significant for counselling about alcohol or drug use. Sub-group analysis indicated that the upward trend in routine safer-sex counselling was significant for men who have sex with men (MSM), for participants who had more than one sexual partner in the past three months, and for those who had a CD4 count of > 500cells/mm³.

Conclusion:

Brief training increased the frequency with which HIV medical providers routinely delivered prevention counselling to their patients at risk of transmitting HIV to others.

Keywords

HIV medical providers prevention counselling prevention with positives

Introduction

It is estimated that 79% of human immunodeficiency virus (HIV)-infected persons in the United States are aware of their seropositive status.¹ As HIV testing efforts increase, proportionally more infected persons will become aware of their infection. As linkage-to-care efforts improve, more HIV-diagnosed persons will enter medical care soon after their diagnosis. Thus the HIV medical care setting becomes an increasingly important environment for reaching large numbers of HIV-diagnosed persons for the purpose of preventing HIV transmission. Many HIV patients are on antiretroviral therapy (ART) and have suppressed viral load,² which dramatically reduces the likelihood of transmitting HIV to others.³ But not all patients are on ART, and not all patients have suppressed virus. In addition, studies indicate that up to 25% of HIV-diagnosed persons engage in unprotected anal or vaginal intercourse with partners at risk for infection.^4–6 Thus it is extremely important that clinic-based ‘prevention with positives’ interventions have a strong sexual behaviour component that complements the ART/adherence component.

The US Centres for Disease Control (CDC) and Prevention⁷ and the Institute of Medicine⁸ recommend that HIV patients receive brief prevention counselling from their HIV medical providers during primary care visits. The extent to which HIV medical providers deliver prevention counselling to their patients and whether it can be improved through relatively brief training remains unclear. Herein, we report data on provider counselling that was part of an intervention programme providers delivered to their patients during primary care exams. The intervention was called ‘Positive Steps for Patients with HIV’.⁹ Prior to implementing the intervention, providers received training in delivering prevention counselling to their patients. We sought to examine whether that training led to an increase in the percentage of HIV patients who reported that they received routine prevention counselling from their medical providers. Further, we examined whether changes in provider counselling occurred among patients in general or whether it was focused on higher-risk patients (e.g. those who were sexually active, those who had multiple partners). Finally, independent of any changes in provider counselling following training, we examined whether patient reports of counselling varied by patient characteristics such as sexual orientation, ethnicity, age, patients’ disease status, and risky sexual activity.

Methods

Overview

The project was conducted at seven HIV clinics in public health departments or university medical centres in 2005–2006. The clinics were located in Denver, Colorado; Brooklyn, New York; Chapel Hill, North Carolina; Nashville, Tennessee; Kansas City, Missouri; and Atlanta, Georgia (two sites). Each clinic implemented the same standardized intervention; there were no concurrent control clinics.

Study staff at each clinic enrolled a measurement cohort, administered a baseline questionnaire, and collected baseline medical record data for each cohort participant. Medical providers then were trained to deliver a brief prevention counselling intervention to all of their patients (not just cohort participants) during routine medical visits on topics such as safer sex, HIV disclosure, and drug and alcohol use (see Thrun et al.¹⁰ for details about training). Specifically, a brief behavioural screening tool was administered to HIV-positive patients to assess their HIV risk behaviours (such as condom use, alcohol and drug use) during regular clinic visits to help prompt providers’ delivery of prevention counselling. This intervention ran for a 12-month CDC-funded period with no formal end. Two follow-up cohort assessments and medical chart reviews were made at approximately six and 12 months after the baseline assessment. The protocol was granted an exemption from institutional review board (IRB) review at the CDC but was reviewed and approved by the IRB at each local site.

Recruitment

Patients were eligible to enroll in the cohort if they were 18 years of age or older, able to provide informed consent, able to understand English, reported being HIV-positive for at least six months, had received care at least once before at the clinic, and anticipated receiving care at the clinic for the next 12 months. In addition, cohort patients had to be sexually active (either oral, anal, or vaginal sex) or had to have injected non-prescription drugs in the three months prior to enrollment.

Approximately 200 cohort patients were recruited at Brooklyn, Chapel Hill, Nashville, and Denver, and 100–120 were recruited at the Kansas City and Atlanta sites, which had smaller patient populations. Study recruiters attempted to approach all patients who presented at the clinic during the three- to four-month recruitment period. Recruitment flyers describing the study were posted in locations near the HIV clinics (e.g. pharmacy, laboratory). Enrollees completed a baseline questionnaire, were asked to complete two additional questionnaires in the next 12 months, were told that providers would not have access to their questionnaire responses, and that medical information would be abstracted from their medical records. Participants received a small monetary compensation after completing each questionnaire. Medical providers were not informed which patients were in the cohort.

Measures collected from the cohort

Cohort participants completed an audio-computer assisted self-interview (ACASI) at baseline (before the medical providers received training) and at two follow-up periods. The follow-up assessments were coordinated, whenever possible, with patients’ return medical visits at the clinic. Participants reported how often they received prevention counselling from their medical providers (physician, physician assistant, nurse practitioner) on the following three topics during the past six months: (1) safer sex; (2) disclosing their HIV infection to sexual partners; and (3) non-prescription drug or alcohol use. Participants responded using a five-point scale (‘never’, ‘less than half of clinic visits’, ‘half of clinic visits’, ‘more than half’, and ‘every clinic visit’). These items were used in a prior study.¹¹ For each of the three prevention topics, routine counselling was defined as having received counselling at ‘every clinic visit’ or ‘more than half of the clinic visits’ in the past six months.

Additional variables in the analysis (listed in Table 1) included age, education, race/ethnicity, sexual orientation (behaviourally defined by the gender of sex partners in the prior three months), number of sexual partners in the past three months, occurrence of unprotected anal or vaginal sex in the past three months, use of non-prescription drugs in the past three months (see note in Table 1 for the list of substances), and alcohol use in the past three months. In addition, study staff conducted medical chart reviews (CD4 cell counts/mm³ and viral load levels) of cohort participants. Laboratory results closest in time with the dates of each of the three ACASI assessments were used. Several of these variables (CD4 counts, viral load, sexual orientation based on gender of sex partners, sexual behaviour, number of sex partners, alcohol use, and drug use) were treated in the analysis as time-dependent variables because of their potential to change across the 12-month follow-up period.

Table 1.

Baseline characteristics of the analytic cohort ‘Positive Steps’, 2005–2006.

Patient characteristics at baseline	N = 729^a (%)
Age group (years)
18–34	157 (21.5)
35–45	327 (44.9)
> 45	245 (33.6)
Race/ethnicity
Black/African American	425 (58.3)
White	198 (27.2)
Hispanic	76 (10.4)
Other^b	30 (4.1)
Educational level
Less than high school	185 (25.4)
High school graduate	229 (31.4)
Some college	208 (28.5)
College graduate	107 (14.7)
HIV clinic location
Brooklyn, New York	165 (22.6)
Denver, Colorado	142 (19.4)
Kansas City, Missouri	68 (9.3)
Chapel Hill, North Carolina	107 (14.7)
Atlanta, Georgia (1)	61 (8.4)
Atlanta, Georgia (2)	58 (8.0)
Nashville, Tennessee	128 (17.6)
Sexual orientation^c
Heterosexual women	231 (31.7)
MSM	340 (46.6)
Heterosexual men	158 (21.7)
Number of sexual partners in past 3 months
1 sexual partner	467 (64.1)
>1 sexual partner	232 (31.8)
Sexual behaviour in the past 3 months
Any unprotected anal or vaginal sex	300 (41.2)
Alcohol use in the past 3 months
< 1 drink a month to 1 drink per week	288 (39.5)
2 ≥ drinks per week	149 (20.4)
None	274 (37.6)
Drug use in the past 3 months^d
Used 1 non-prescription drug	102 (14.0)
> 1 non-prescription drug	81 (11.1)
None	546 (74.9)
Time since HIV diagnosis
< 2 years	102 (14.0)
3 to 7 years	239 (32.8)
>8 years	388 (53.2)
CD4 cell count^e
< 200 cells/mm	145 (19.9)
200–500 cells/mm³	324 (44.4)
> 500 cells/mm³	255 (35.0)
Viral load^e
< 400 copies/mL	345 (47.3)
401–20,000 copies/mL	200 (27.4)
20,001–100,000 copies/mL	106 (14.5)
>100,000 copies/mL	71 (9.7)

Notes: ^aN is based on longitudinal data for the safer-sex counselling variable. Also, sample sizes for some of the variables are lower due to missing data; ^bRepresents any person who reported being either Native American or Alaskan Native, Asian, Native Hawaiian or other Pacific Islander, or some other race; ^cThis variable was defined behaviourally by the gender of sex partners in the past three months; ^dIncludes non-prescription use of any of the following: crack, cocaine, marijuana, hashish, speed, methamphetamines, ecstasy, special K/ketamine, GHB, hallucinogens (such as LSD, mushrooms, peyote, mescaline, PCP, angel dust), inhalents (such as glue, poppers), opiates, or barbiturates; ^eBased on clinic medical chart records closest temporally to the baseline assessment.

Provider counselling intervention

After all cohort participants completed the baseline ACASI, medical providers and clinic staff attended a four-hour group intervention training session at the clinic conducted by a team of prevention counselling specialists. The initial four-hour group training session focused on: (1) using a brief screening tool during primary care visits to identify prevention topics that may need attention with patients; (2) enhancing communication skills; and (3) practicing brief behavioural counselling. A group booster training session was given at the clinic one to two months into the intervention. Providers were encouraged to use the screening tool and give attention to safer sex, HIV disclosure, and/or safer drug use as needed at all primary (but not urgent) HIV medical care visits.

Statistical analyses

The analyses were conducted using data from participants who completed all three waves of ACASI assessments. We calculated the percentage of participants who reported that they received routine counselling on each of the three prevention topics at baseline (before providers received training) and at each of the two follow-up periods. Linear trends across time were examined for each outcome measure using logistic regression with generalized estimating equations (GEE) that adjusted for the correlation in patients’ responses across time. The strength of the trend (regression coefficient) reflects the change in the percentage of participants who reported routine counselling across the three time periods. These are unadjusted trends; the adjusted trends are described later. Next, we tested whether the linear trends were consistent across sub-groups of patients by modelling sub-group by time period interaction effects for each outcome measure. If an interaction effect reached p ≤ .10, we conducted a follow-up analysis to identify the pattern of the interaction by examining the linear trend within each of the sub-groups of a variable. Finally, we examined whether there were overall (i.e. pooled across time) associations between patient characteristics listed in Table 1 and receipt of routine prevention counselling. In these main effect models, variables that reached p ≤ .20 in bivariate analysis were entered into a multivariate model and analysed with logistic regressions using GEE methods. These models also included the assessment time period as a variable and, thus, afforded the opportunity to examine the adjusted linear trends after controlling for other variables. All analyses were conducted in SAS 9.2 (SAS Institute Cary, North Carolina).

Results

Enrollment and analytic sample

Of 2,451 patients approached at the seven HIV clinics, 2,082 (85%) agreed to be screened for eligibility; 1,282 patients were eligible and 800 were not. Of the 800, 94% were ineligible because they were neither sexually active nor had they injected drugs in the past three months. Of the 1,282 eligible patients, 1,109 (87%) agreed to participate. A total of 729 of the 1,109 participants (66%) completed all three waves of ACASIs and had complete data on ‘receipt of routine safer-sex counselling’. These 729 patients constituted the longitudinal cohort for this outcome measure and for description of sample characteristics at baseline. The cohort size for the other two outcomes – namely, receipt of routine counselling about HIV disclosure (n = 633) and about use of alcohol/drugs (n = 500) – were smaller because some participants responded ‘not applicable’ to those items at baseline and/or at the follow-up periods. This was most pronounced for alcohol/drug counselling because many participants who responded in the ACASI that they did not use alcohol or drugs selected ‘not applicable’ on that counselling item.

Enrollees not included in the longitudinal analyses did not differ from the analytic sample on behavioural (reports of provider counselling, unprotected vaginal or anal intercourse) or demographic (gender, income, education, sexual orientation, age) factors measured at baseline (data not presented).

Sample characteristics

The analytic sample of 729 participants at baseline was predominately male (68%) and African American (58%). Forty-five per cent of the participants were 35–45 years of age and the majority had at least a high school education. Forty-seven per cent were men who had sex with men (MSM); 22% were heterosexual men. In the past three months, 25% used one or more non-prescription drugs, 32% reported having more than one sexual partner, and 41% had unprotected anal or vaginal sex. Table 1 displays additional information about the sample.

Receipt of routine prevention counselling across time

Table 2 displays the percentages of participants who reported that providers routinely discussed prevention topics with them during the assessment periods. Participants’ reports of receipt of routine safer-sex counselling from their medical providers had a significant, upward linear trend across time (b = 0.12, SE = 0.04; p = 0.004), as did participants’ reports of receiving counselling about disclosing their HIV infection to sexual partners (b = 0.21, SE = 0.05; p < 0.001). The trend was not significant for counselling about alcohol or drug use (b = 0.01, SE = 0.05, p = 0.82).

Table 2.

Percentage of participants who reported receiving routine^a prevention counselling from their HIV medical providers at each assessment period, ‘Positive Steps’, 2005–2006.

Prevention topic	Baseline (%)	6 month (%)	12 month (%)	Unadjusted trend across survey periods
				b (SE)	P value
Safer sex (n = 729)^b	57.1	60.9	63.0	0.12 (0.04)	0.004
Disclose HIV infection to sex partners (n = 633)^b	47.1	55.1	57.8	0.21 (0.05)	< 0.001
Alcohol and drug use (n = 500)^b	58.2	59.4	58.8	0.01 (0.05)	0.82

Notes: ^aRoutine is defined as having received counselling at more than half of clinic visits in the past six months; ^bThe sample sizes vary across HIV prevention topics because ‘not applicable’ was a potential response option for each outcome. Many patients who were not alcohol or drug users had the tendency to select ‘not applicable’ for provider discussions of alcohol and drug use. Similarly, a sizable number of patients responded ‘not applicable’ for HIV disclosure discussion; b = unstandardized regression coefficient; SE = standard error.

Trends across time by patient sub-groups

We identified three significant (p < 0.05) interaction effects (time × participant characteristics) with respect to receipt of safer-sex counselling. The three participant characteristics that comprised the interactions were sexual orientation, number of sex partners in the past three months, and CD4 cell count. As seen in Table 3, there was a significant upward trend in receipt of routine safer-sex counselling for MSM (but not for heterosexual men or heterosexual women), for participants who had more than one sex partner in the past three months (but not for participants with just one sex partner), and for those who had a CD4 count > 500cells/mm³ (but not for participants with lower CD4 cell counts). The only other appreciable result was a marginal interaction effect (p < 0.10) for the alcohol/drug use counselling outcome. There was a non-significant upward trend in the receipt of alcohol/drug counselling among those who reported that they had not engaged in unprotected anal or vaginal intercourse in the prior three months, and a non-significant downward trend in alcohol/drug counselling among those who engaged in that risk behaviour (Table 3).

Table 3.

Percentage of participants who reported receiving routine^a prevention counselling from their HIV medical providers at each assessment period by participant sub-groups, ‘Positive Steps’, 2005–2006.

	Baseline (%)	6 month (%)	12 month (%)	Trends across assessment periods within sub-groups b (SE)
SAFER-SEX COUNSELLING
Sexual orientation^b,c
Heterosexual women	65.4	68.4	68.4	0.06 (0.08)
MSM	47.1	53.2	57.9	0.22 (0.06)^*
Heterosexual men	66.5	66.5	65.8	−0.02 (0.09)
Sexual partners in the past 3 months^c
1 sexual partner	60.4	62.0	63.2	0.09 (0.07)
> 1 sexual partner	48.3	51.5	60.6	0.35 (0.09)**
CD4 count (cells/ mm³)^c
< 200	60.7	58.4	54.8	−0.05 (0.11)
200–500	58.6	60.4	61.9	0.05 (0.07)
> 500	54.1	63.5	68.5	0.29 (0.07)**
ALCOHOL AND DRUG USE COUNSELLING
Sexual behaviour in the past 3 months^c
No unprotected vaginal or anal intercourse	55.5	59.8	60.6	0.11 (0.06)
Any unprotected vaginal or anal intercourse	61.5	58.8	52.6	−0.11 (0.09)

Notes: All variables were screened for interactions with assessment period using generalized estimating equations (GEE). The variables included in this table had significant (p < 0.05) or marginally significant (p < 0.10) interactions and thus sub-group trends were examined for those variables. There were no interaction effects involving the prevention topic ‘Disclosing HIV infection to sex partners’; ^aRoutine is defined as having received counselling at more than half of clinic visits in the past six months; ^bThis variable was defined behaviourally by the gender of sex partners in the past three months; ^cTime-dependent variable; b = unstandardized regression coefficient; SE = standard error.

p < 0.05; ** p < 0.01.

Association of patient characteristics and receipt of prevention counselling

Table 4 displays the results of the main-effect models of the multivariate correlates of receipt of routine counselling on the three prevention topics. The odds of receiving safer-sex counselling were higher among younger participants (18–34 and 35–45 vs. > 45) and African Americans (vs. whites). For all three prevention topics, the odds of receiving counselling were lower among MSM (vs. heterosexual men) and those who had been diagnosed with HIV infection eight or more years prior to enrollment in the study. Finally, as shown in the last row of Table 4, there was a significant upward linear trend across time in the percentage of participants who received safer-sex counselling and HIV disclosure counselling after statistically controlling for the other variables in Table 4.

Table 4.

Multivariate correlates of receipt of routine^a prevention counselling from HIV medical providers averaged across assessment periods, ‘Positive Steps’, 2005–2006.

Variable	Sub-groups	Safer-sex counselling	HIV disclosure counselling	Alcohol and drug use counselling
		Adjusted OR (95% CI)	Adjusted OR (95% CI)	Adjusted OR (95% CI)
Age group (years)	18–34	1.61* (1.08, 2.39)	1.40 (0.94, 2.07)	—
	35–45	1.51** (1.11, 2.07)	1.24 (0.89,1.72)	—
	> 45	ref	ref	ref
Race/ethnicity	Black/African American	1.45* (1.03, 2.03)	1.43 (0.99, 2.05)	1.46 (0.99, 2.13)
	Hispanic	1.26 (0.77, 2.07)	1.42 (0.85,2.37)	1.33 (0.78, 2.26)
	Other	0.97 (0.49,1.90)	0.92 (0.42, 2.01)	1.23 (0.52, 2.93)
	White	ref	ref	ref
Educational level	Less than high school	—	—	1.57 (0.94, 2.61)
	High school	—	—	1.44 (0.91,2.28)
	Some college	—	—	1.11 (0.69-1.78)
	College graduate	ref	ref	ref
HIV clinic location	Brooklyn, NY	1.81* (1.13, 2.90)	1.01 (0.62,1.64)	0.68 (0.41, 1.12)
	Denver, Colorado	1.58* (1.01,.091)	1.57 (0.97, 2.53)	1.46 (0.92, 2.33)
	Kansas City, Missouri	2.97** (1.66, 5.31)	3.51** (1.88, 6.55)	3.41** (1.85, 6.29)
	Chapel Hill, North Carolina	1.79* (1.13, 2.82)	1.75* (1.06, 2.89)	1.37 (0.81, 2.30)
	Atlanta, Georgia (1)	1.23 (0.74, 2.05)	1.03 (0.61, 1.72)	0.99 (0.56, 1.74)
	Atlanta, Georgia (2)	9.49** (4.31, 20.87)	3.46** (1.91, 6.265)	5.07** (2.46, 10.45)
	Nashville, Tennessee	ref	ref	ref
Sexual orientation^{b, c}	Heterosexual women	0.95 (0.63,1.45)	0.88 (0.52, 1.35)	0.76 (0.51, 1.13)
	MSM	0.62* (0.41, 0.94)	0.55** (0.35, 0.85)	0.58* (0.39, 0.89)
	Heterosexual men	ref	ref	ref
Number of sexual partners in past 3 months^c	> 1 sexual partner1 sexual partner	0.95 (0.74, 1.22)ref	0.82 (0.64,1.05)ref	—ref
Time since HIV diagnosis	>8 years3 to 7 years	0.54** (0.35, 0.83)0.79 (0.49, 1.24)	0.58* (0.39, 0.88)0.77 (0.50,1.17)	0.62* (0.39, 0.96)0.76 (0.47, 1.22)
	< 2 years	ref	ref	ref
Assessment period	b (SE)	0.21 ( 0.06)**	0.20 ( 0.06)**	0.01 ( 0.05)

Notes: All of the variables listed in Table 1 were screened for inclusion in these multivariate models. Variables that did not meet a screening level of p < 0.20 in bivariate logistic regression were not included. The last entry in this table represents the adjusted trend in counselling across the three assessment periods; Dashes indicate that the variable did not qualify for inclusion in the model for that particular prevention topic; ^a Routine is defined as having received counselling at more than half of clinic visits in the past six months; ^b This variable was defined behaviourally by the gender of sex partners in the past 3 months; ^cTime-dependent variable. b = unstandardized regression coefficient; SE = standard error.

p < 0.05; ** p < 0.01.

Discussion

A four-hour group training session coupled with a short booster session increased HIV providers’ delivery of prevention counselling to their patients. Among the HIV patients in our cohort with complete longitudinal data, a larger percentage reported that they routinely received counselling on safer sex and disclosing their HIV infection to sexual partners after the providers received training in delivering prevention counselling compared to before they received that training. Previously we reported that HIV patients who participated in the ‘Positive Steps’ intervention showed a decline in the prevalence of sexual transmission risk behaviour at six and 12 months after the onset of the intervention.⁹ The findings reported in this paper strongly suggest that the increases in the frequency with which providers delivered prevention counselling, especially to higher-risk patients, was a factor underlying the sexual behaviour changes among the patients.

The upward trend in safer-sex counselling was especially strong among patients who were MSM, patients who reported on the ACASI that they had more than one sex partner in the past three months, and patients who had CD4 cell counts > 500cells/mm³. These three groups had a lower percentage (47–54%) of patients at baseline who received routine safer-sex counselling compared with their counterparts (60–66%). Although these three groups had more room for improvement, there was still opportunity for improvement in the subgroups. The medical providers may have targeted their counselling to patients whom they assumed were sexually active or had evidence of sexual activity from the screening tool or from conversations with patients. However, behaviours can change over time; thus it is important for medical providers to continue to keep the lines of communication open with their patients to screen for risk behaviours on an ongoing basis and discuss related challenges as they arise. In addition, we did not find a main-effect association between patients’ CD4 cell counts and receipt of prevention counselling, but we did find that, after providers received training, safer-sex counselling significantly increased among patients with CD4 counts greater than 500 cells/mm³, but not among patients with lower CD4 counts. Patients with higher CD4 counts are likely to be feeling well and, consequently, they may be sexually active. Further, providers may have more time to spend discussing prevention topics with these patients compared with patients with challenging medical issues.^12–15

We identified several correlates of receipt of prevention counselling apart from the time trends discussed earlier. Younger patients and African American patients were generally more likely than older patients and white patients, respectively, to report that their provider discussed safer sex with them. Also, those who were diagnosed eight or more years prior to enrollment and those who were MSM were generally less likely to report that their providers discussed each of the three prevention topics with them compared to their counterparts. We were surprised to find that MSM were generally less likely than heterosexual men to have received prevention counselling. This may be due to provider discomfort in discussing sexual issues with men who are known or presumed to be MSM,¹⁰ assumptions that HIV-positive MSM already know about the importance of safer sex, or other factors. It was encouraging, however, that after providers received training, a larger percentage of MSM received safer-sex counselling compared to baseline. Such counselling may reduce sexual transmission of HIV among MSM,⁹ who comprised 61% of new HIV infections in the US in 2009.¹⁶

This study has several limitations. The study was conducted without a concurrent control group because it was considered unethical not to deliver prevention counselling to patients at each participating clinic. We found significant pre–post changes in receipt of prevention counselling within clinics, but we were unable to compare our findings with clinics that did not implement the intervention. Although the analytic longitudinal sample did not differ at baseline from participants excluded from the analysis, a relatively large number of cohort participants responded ‘not applicable’ to the questions about whether their providers talked with them about alcohol/drug use and disclosure of HIV infection to sex partners. However, if these ‘not applicable’ responses truly reflected that the topic was not applicable to them and thus the provider did not broach the counselling issue, it is appropriate to omit these participants from the analysis. If the questionnaire forced them to respond ‘no’ or ‘yes’ about whether their provider delivered counselling on a specific topic, this could have resulted in a biased outcome (e.g. overestimating the percentage of patients who did not receive counselling on drugs/alcohol when such counselling was not needed). Another issue pertains to the providers’ use of the behavioural screening tool. The purpose of the behavioural screening tool was to quickly assess patient risks and use responses to guide delivery of prevention counselling. But we are unable to determine just how often the providers used the screening tool. Integrating the paper-based screening tool into the clinics’ electronic medical records (if available) would help prompt its use and provide a means of monitoring the type of prevention counselling delivered on an ongoing basis.

Conclusion

Brief training increased the frequency with which HIV medical providers delivered prevention counselling to their patients on a routine basis. The increases in receipt of safer-sex counselling were most pronounced among MSM, those with multiple sex partners, and those with CD4 cell counts greater than 500. The increases were modest, but even modest increases in prevention counselling targeted to higher-risk patients may prevent ongoing sexual transmission. Plus, the cost of training for medical providers is minimal and can be obtained through regional AIDS Education and Training Centres. HIV medical facilities should strive to integrate and sustain the delivery of prevention counselling as standard of care for HIV patients.

Footnotes

Acknowledgements

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the US Centres for Disease Control and Prevention.

References

Campsmith

Rhodes

Hall

Green

. Undiagnosed HIV prevalence among adults and adolescents in the United States at the end of 2006. JAIDS Journal of Acquired Immune Deficiency Syndromes, 2010: 53: 619–24.

Althoff

KRP

Gill

Horberg

Eron

Napravnik

Anastos

Bosch

Moore

Gange

. Trends in HIV Viral Load among Adults in Clinical Care in the US and Canada, 1997–2007 (#548). Presented at the 18th Conference on Retroviruses and Opportunistic Infections, Boston, 27 February—2 March 2011.

Attia

Egger

Müller

Zwahlen

Low

. Sexual transmission of HIV according to viral load and antiretroviral therapy: Systematic review and meta-analysis. AIDS, 2009: 23: 1397–404.

Crepaz

Marks

Liau

. Prevalence of unprotected anal intercourse among HIV-diagnosed MSM in the United States: A meta-analysis. AIDS, 2009: 23: 1617–29.

Cohen

Chen

McCauley

. Prevention of HIV-1 infection with early antiretroviral therapy. New England Journal of Medicine, 2011: 365: 493–505.

Golin

Marks

Wright

. Psychosocial characteristics and sexual behaviours of people in care for HIV infection: An examination of men who have sex with men, heterosexual men and women. AIDS and Behaviour, 2009: 13: 1129–42.

CDC. Incorporating HIV prevention into the medical care of persons living with HIV. Recommendations of CDC, the Health Resources and Services Administration, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR, 2003: 52: 1–24.

Ruiz

Gable

. No Time to Lose: Getting more from HIV Prevention. Atlanta, GA: Natl Academy Pr, 2001.

Gardner

Marks

O’Daniels

. Implementation and evaluation of a clinic-based behavioural intervention: Positive steps for patients with HIV. AIDS Patient Care and STDs, 2008: 22(8): 627–35.

10.

Thrun

Cook

Bradley-Springer

. Improved prevention counselling by HIV care providers in a multisite, clinic-based intervention: Positive steps. AIDS Education and Prevention, 2009: 21: 55–66.

11.

Richardson

Milam

McCutchan

. Effect of brief safer-sex counselling by medical providers to HIV-1 seropositive patients: A multi-clinic assessment. AIDS, 2004: 18(8): 1179–86.

12.

Grodensky

Golin

Boland

Patel

Quinlivan

Price

. Translating concern into action: HIV care providers’ views on counselling patients about HIV prevention in the clinical setting. AIDS and Behaviour, 2008: 12: 404–11.

13.

Koester

Maiorana

Vernon

Myers

Rose

Morin

. Implementation of HIV prevention interventions with people living with HIV/AIDS in clinical settings: Challenges and lessons learned. AIDS and Behaviour. 2007: 11: 17–29.

14.

Myers

Rose

Shade

. Sex, risk and responsibility: Provider attitudes and beliefs predict HIV transmission risk prevention counselling in clinical care settings. AIDS and Behaviour, 2007: 11: 30–8.

15.

Patel

Golin

Marks

. Delivery of an HIV prevention counselling programme in an infectious diseases clinic: Implementation process and lessons learned. AIDS patient care and STDs, 2009: 23: 433–41.

16.

Prejean

Song

Hernandez

. Estimated HIV incidence in the United States, 2006–2009. PloS one, 2011: 6: 1–13.