Each year, approximately 17 million HIV antibody tests are performed at private and public health clinics in the United States.(1) HIV voluntary counseling and testing (VCT) traditionally has comprised a large component of the Centers for Disease Control and Prevention (CDC) budget for HIV prevention services.(2) This investment in counseling and testing is premised on the notion that test clients receive personalized counseling to identify and reduce risk behavior. For those receiving positive results, HIV antibody testing can serve as a gateway to clinical care, support services, and counseling to reduce the chance of transmitting HIV to others. However, the role of VCT in changing the behavior of those receiving negative test results is less clear. Research on testing behavior suggests that many repeat testers do not reduce their risk behaviors.
This chapter will provide an overview of current models of VCT practice with links to guidelines, curricula, and research protocols that describe them. Each of these guidelines emphasizes a particular theoretical approach to behavior change that was developed and tested in a research context. This chapter also will examine some of the discontinuities between research-based approaches to counseling and what we know about real-world practice of VCT. This chapter will describe a new approach to evaluating and improving VCT practice using process analysis.
| Evolution of VCT Practice|
VCT is still a relatively new field of clinical practice and there is little evidence about which models of practice are most appropriate and effective for various populations.(3,4) Models of VCT practice have also evolved in response to changes in the volume and demographics of test seekers at publicly funded test sites. The earliest model of VCT was developed in San Francisco by the University of California San Francisco (UCSF) AIDS Health Project in early 1985.(5) At that time, nearly 20% of the test clients, mostly gay men, were testing HIV positive. As a result, VCT centered on crisis management and explaining to clients what little was known about the significance of the antibody test for disease prognosis. Following basketball star Earvin "Magic" Johnson's disclosure that he had tested HIV positive in November 1991, testing volume nearly doubled in the United States as predominantly low-risk, heterosexual clients flooded testing centers.(6) Demand for testing has remained relatively steady since that time, as "the test" has become incorporated into mainstream American culture as a routine part of dating rituals and clinical practice.(7,8) After the influx of low-risk testers began in 1992, the positivity rate at publicly funded test sites in the United States declined to <2%.(9) As research developed new treatments and a better understanding of the routes of HIV transmission, VCT in publicly funded tests sites evolved into a more didactic, health education format providing basic information on routes of transmission and advice on safer sex.(10) Although HIV testing has become a routine procedure in private practice, little is known about the quality or scope of the counseling, if any, that is provided with the test in such settings.(11)
| Client-Centered Counseling|
In response to mounting evidence of poor outcomes from VCT evaluation studies, the CDC revised VCT guidelines in 1993.(12) The new guidelines instructed counselors to employ a personalized, "client-centered" approach and to "negotiate a risk reduction plan" with each test client. Aside from these vague recommendations, the new client-centered guidelines offered few practical suggestions on how to engage clients in a more interactive discussion (see Table 1). Despite the CDC's promotion of client-centered guidelines over the last decade, the degree to which current practice reflects this model is unclear. This is because little research has examined what VCT interaction actually consists of in practice.
By definition, "counseling" assumes a helping relationship in which a client, having identified a problem or concern, seeks the help of a mental health professional.(13) VCT differs fundamentally from other counseling relationships in two respects.(4) First, VCT counselors in the United States are provided minimal formal training and generally are not mental health professionals. Given the clinical context in which VCT often occurs, test counselors are mostly nurses, medical assistants, and paraprofessionals such as outreach workers and volunteers with little more than a few days of formal training in HIV prevention counseling.(14) Second, the "counseling" relationship in VCT is not requested by the client but imposed unilaterally by state laws regulating the provision of HIV testing. Counseling is thus a condition for receiving the test. Because clients are primarily seeking the test result, not a counseling session, any discussion about risk with the counselor is experienced as an unpleasant but necessary requirement for getting the test. In this light, test counseling in a public clinic may be more accurately described as a service encounter in which the client provides personal information in exchange for a clinical screening test that is provided free of charge or for a nominal fee.
Although the term client-centered counseling has been used as a catch-all phrase in literature on VCT, the context in which the approach was developed is fundamentally different from the public health/clinical context of test counseling. Client-centered counseling was developed by Carl Rogers in the 1940s. Rogers' approach was a radical rejection of the dominant model of counseling practice of the time that was concerned largely with psychometric testing. In Rogers' new, humanist model of counseling, each individual is endowed with an organic capacity for growth and change.(15) The counselor's role, according to Rogers, is not to assess or evaluate the client, but to feed this natural capacity for change through empathy and unconditional acceptance. The primary technique of client-centered counseling is to actively listen and reflect the client's statements in a nondirective, nonjudgmental manner, thereby providing a safe environment for the client's self-exploration. Client-centered counseling hinges on the development of a counselor-client relationship based on unconditional regard, often over multiple hour-long sessions. This relationship enables the counselor to clarify the client's feelings without imposing external assessments or values.(16)
| Data Collection Forms|
Although lacking specifics when describing how to be client centered, the CDC guidelines were clear on what client-centered counseling should not sound like:
HIV counseling is not a lecture. An important aspect of HIV counseling is the counselor's ability to listen to the client in order to provide assistance and to determine specific prevention needs. Although HIV counseling should adhere to minimal standards in terms of providing basic information, it should not become so routine that it is inflexible or unresponsive to particular client needs. Counselors should avoid providing information that is irrelevant to their clients and should avoid structuring counseling sessions on the basis of a data-collection instrument or form.(12)
The last phrase refers to the tendency of counselors to structure sessions similar to an assessment of risk behaviors listed on a standardized form. These data-collection instruments can be simple forms consisting of a few risk categories that the counselor can fill in after the client is gone, or they can be quite elaborate, such as the one used in California, where the form has become so complex and lengthy that test counseling sessions have increasingly come to resemble research interviews.(17)
Counseling guidelines rarely mention the form, often instructing counselors only to use it sparingly in favor of a more client-centered approach. Yet the guidance that accompanies the most recent California risk assessment form asserts that completing the Client Information Form (CIF) is synonymous with client-centered counseling:
With the exception of a few administrative items, the content of this form is essential for adequate client-centered HIV counseling. HIV counseling cannot be client centered unless the counselor has a complete understanding of the client's risks and current issues. The CIF has been designed in intensive collaboration with the HIV counselor training curriculum development staff, HIV counselor trainers and senior counselors. The information is recorded to insure that it is obtained and available for reference during HIV risk assessment, disclosure and post disclosure HIV counseling sessions. It is the basis for service documentation and reimbursement. It also provides program planners with information about the HIV counseling process and our clients. This information is critical to the continuous improvement of primary HIV prevention in California....
...While strict reliance on the CIF results in poor counseling, many counselors glance at it occasionally for support, prompts and recording information at convenient points during the interview....
...Blanks represent incomplete risk assessments and can affect the level of payment for counseling and testing services and reflect the adequacy of the service provided.(18)
Rather than facilitate "client-centered" counseling, the form enforces a standard of performance based on thoroughness in interrogating the client's risk.
Sheon used conversation analysis, a method of process analysis further described below, to examine 40 VCT sessions recorded in 1996 and another 30 recorded in 2003.(5) These data revealed how counselors' use of the state-mandated risk assessment form to probe for past risks can result in a highly depersonalized, bureaucratic interaction that reduces the counseling relationship to an exchange of personal data for a "free" test.(5) Interviews with test clients about their testing experiences suggested that ritually recording one's transgressions on an official piece of paper can produce a juridical dynamic within the session. Counselors manage this dynamic by treating the data collection form as a routine and bureaucratic formality, thereby mitigating the moral upshot of the questions. This depersonalized approach can also be seen as a strategy for managing the repetitive and tedious nature of form-based counseling. As a result of these constraints, counselors employed a routine script with a fixed sequence of topics that varied little from client to client. Despite the emphasis on accurate data collection in the guidance provided for the form cited above, recordings of test sessions revealed that counselors combine, phrase, and omit questions on the form in creative ways, thus compromising the validity of the data collected. The point here is not to criticize test counselors but to suggest that the roles of client-centered counselor and data collector are incompatible.
By spending so much of the session interrogating clients about past risk behaviors, actions that clients are powerless to change, counselors may unwittingly create an awkward confessional dynamic in which clients feel compelled to emphasize their contrition for "sins" elicited and inscribed on the risk assessment form.(5) More importantly, counselors may miss crucial opportunities to discuss how the results of current and subsequent tests might impact future risk taking. Plans for change should be realistically assessed in terms of potential sources of support for change as well as obstacles to change and ways to overcome them. Despite the emphasis on negotiating risk reduction plans in CDC guidelines and local training curricula, few of the recorded VCT sessions include any discussion of risk reduction plans aside from future testing to account for the antibody window period.
Besides limiting the scope of counseling about future behavior, the form's focus on past risk behavior may unwittingly re-enforce risk behavior by highlighting the discontinuity of past risks divulged on the data collection form and a client's current negative test result. Clients receiving a negative result after engaging in high-risk behavior may conclude that they are somehow lucky, that they are immune to HIV, or that the risk associated with their behaviors has been overstated.(19) The latter conclusion is only strengthened by the often-conflicting messages clients receive about the relative risks of various practices, such as oral sex.(20) In this way, receiving several negative test results after divulging all their risks to the counselor may actually dilute clients' motivation to reduce HIV risk behavior.
| Research on the Effects of VCT|
Evidence for the effects of VCT on behavior change comes from 2 types of research studies.(13,21) Outcome studies typically use quantitative survey methods to measure self-reported behavior and sometimes biological markers such as sexually transmitted disease (STD) rates among large samples of test clients before and after VCT. Process studies focus on the actual testing experience of a small sample of test clients, usually by audio-recording the test counseling interaction or by interviewing clients about their motivations to seek testing. Outcome and process studies complement each other by posing the question of effectiveness in different ways. Outcome studies use controls to isolate the effects of the intervention, whereas process studies raise fundamental questions about the nature of the interaction between counselor and client. Outcome studies tell us whether or not an intervention succeeds, whereas process analysis tells us how an intervention succeeds or fails. While randomized control trials are based on the assumption that counselors have adhered to a "counseling" protocol, process analysis measures the counseling "dose" empirically.
To date, most research on test counseling has relied on outcome measures with very little attention to the counseling process.(22,23) Unlike placebo-controlled trials measuring the effect of a specific drug on a specific biological factor, research on the effects of counseling upon behavior is hampered by a lack of consensus on how to define "counseling" and how to measure change in self-reported risk behavior. Little research has examined different counseling styles and approaches, or the degree to which these approaches resemble the client-centered model outlined in CDC guidelines and trainings. This lack of process research has left policy makers, counselors, trainers, and test clinic administrators with little information about appropriateness and feasibility of research-tested approaches in real-world clinical and outreach settings.
| VCT Interventions for Heterosexuals|
The evidence for the effects of VCT on risk behavior is mixed. For example, Weinhardt et al conducted a meta-analysis of 27 outcome studies conducted between 1986-1996 and their analysis confirmed the conclusions of previous reviews of the literature.(23,24,25) Clients receiving a positive test result tended to reduce their risk behavior, but those receiving a negative test result did not. In fact, some clients receiving a negative test result increased their risk behavior. The authors point out that it is difficult to draw firm conclusions from this body of work because studies used different approaches to counseling as well as different criteria for measuring risk. Two multisite outcome trials published after the meta-analysis conducted by Weinhardt et al were successful in reducing risk behavior and STD rates among heterosexual couples in Kenya, Tanzania, and Trinidad, as well as heterosexual patients at STD clinics in 6 U.S. cities.(26,27) These interventions were based on the CDC's client-centered counseling guidelines and provided some of the only evidence for the effectiveness of that approach. Protocols of these interventions are available online.(28,29)
| Interventions for Men Who Have Sex With Men|
Although evidence of the effectiveness of the client-centered approach among heterosexual couples in Africa and heterosexual STD clients in the United States is strong, few randomized controlled trials of VCT interventions for men who have sex with men (MSM) have been conducted. Research on this population is badly needed as it comprises some of the heaviest users of VCT services and epidemiological trends show a steep rise in STD rates among some subgroups in this population, suggesting that, in time, HIV incidence may also rise.(30,31) An innovative approach to VCT was assessed in a study involving 120 high-risk, repeat-testing MSM in San Francisco.(32) Gold's distinction between "online" cognitive function in the moments before sex occurs and "offline" reflections about the incident served as the basis for a 40- to 60-minute counseling intervention with a mental health professional that occurred between the pre- and post-test sessions.(33) Clients were asked to recall a specific incident of unprotected sex, complete a self-administered questionnaire consisting of common self-justifications, and then narrate the story in detail to a counselor trained to probe for self-justifications that may have facilitated the risk behavior. This approach differs markedly from traditional VCT in that the client is asked to construct a detailed account about a single incident of unprotected sex and the self-justifications that supported the decision to have risky sex. The results of this trial suggested that the intervention was both effective in reducing risk and acceptable to clients and counselors.
A second trial, called RED2, is currently under way to determine whether a streamlined version of the self-justifications questionnaire and cognitive counseling intervention can be effective when incorporated into a pretest session with paraprofessional test counselors. A combination of process and outcome measures (described further in the following section on time charts) is being used to compare the intervention against the standard client-centered model and describe the most effective way for counselors to select and elicit a narrative account of a particular risk incident during the counseling session. Our process analysis suggests that focusing the session around a detailed narrative of a risk incident ensures that the intervention counselors listen attentively to the clients' own assessments of their behavior. In standard VCT sessions, counselors rely more on the risk assessment form and thereby elicit less-personalized accounts of the clients' risks. Should the trial show effectiveness, process measures will also provide tools to develop training and supervision guidelines so that the intervention can be successfully adapted in real-world VCT practice.
| Process Analysis of VCT|
Although it provides a measure of the effectiveness of an intervention, outcome research is of limited use for refining counseling practice because it tells us little about what specific techniques or client experiences produced the behavioral effects measured. For example, outcome research does not reveal how much of the effect is due to the counselor's particular style, the client's readiness to change, or a reaction to the test result itself. Accordingly, Beardsell and Coyle have called for process research on VCT in order to illuminate the black box of counseling and examine what it is and why it succeeds or fails to produce the desired outcomes.(22) Process research can help translate effective counseling approaches developed in research settings for everyday use at test sites. Specifically, we need to examine how counselors working in real-world clinics allocate time to counseling tasks in the face of pervasive workload issues such as time pressures, surveillance forms, and clients' resistance to counseling they did not expect or request. These issues are less likely to be factors in a research study wherein the counselor's task is made easier by having willing participants and a more clearly defined role thanks to more elaborate training and supervision.
| Conversation Analysis|
Process research examining recorded VCT sessions has identified a number of barriers to implementing the client-centered approach. One method of process research is conversation analysis (CA). CA is an approach to linguistic analysis that illuminates the sequential, turn-taking structure of talk to examine how the counselor and the participant interpret each other's utterances and display their interpretations in subsequent turns.(34,35) CA research has examined the narrative strategies participants and counselors employ to jointly construct the participants' accounts for their decision to engage in risky sex and seek testing, as described below in the section titled Clients' Motivations for Testing.(5,36)
Another major area of inquiry in CA studies of HIV counseling and testing has been counselors' strategies for delivering safer-sex advice to participants.(37,38) Silverman's landmark study of VCT interaction examined recordings of approximately 93 VCT sessions from the United States and the UK.(37) The study revealed a very common pattern across the sites. Counselors commonly packaged advice as information delivered in depersonalized monologues, often framed in terms of "We tell everyone that it's best to...." This strategy clearly is not "counseling" because clients remained largely passive and silent recipients of generic advice/information. However, such an approach permitted counselors to avoid the negative implications of personally directing advice to clients who had not requested any, while allowing the discussion to quickly cover a wide range of topics within a short period of time.
In another CA study of VCT sessions recorded in the United States, Kinnell and Maynard found that counselors' advice sequences frequently glossed over clients' actual behaviors and instead provided information that was clearly not relevant to them.(38) For example, counselors frequently provided didactic information concerning the risks of sharing of sex toys or injection drug equipment even when clients specifically stated that these were not relevant risk behaviors. Counselors justified this practice by viewing the clients as conduits for risk reduction information to the broader community.
CA has proven particularly useful as a method for analyzing the effectiveness of various counseling strategies by examining participants' responses during the sessions.(5,37) The participants' degree of engagement can be observed in their activity or passivity, for example, in the length of the participants' turns at speaking and the degree to which the participants direct the topic of discussion. Silence or minimal uptake by participants, such as the occasional "m hm" or "yeah" between counselors' advice sequences, displays resistance to advice deemed irrelevant by the participants. By contrast, participants responded to relevant and empathic advice sequences with overlapping speech that completed or elaborated upon counselors' utterances.(39,40) By examining how counselors and clients cope with the delicate, moral implications of talk about sexual risk and HIV testing, process research provides numerous insights into why actual practice falls short of the goals set forth in guidelines for client-centered counseling.
| Time Charts|
It is difficult to visualize patterns in audio-recorded data unless the information is indexed and tabulated in some manner. Sheon developed time charts as a way to display the chronological sequence and distribution of tasks performed during a VCT session. Time charts provide a measure of how time and effort are allocated in the session, and offer a valuable tool for training and supervision. Time charts provide an overview or map to the counseling session data, and help reveal outliers in the data set. Time charts are created by indexing digital recordings of VCT sessions using qualitative analysis software called Atlas.ti and then entering the time values of various tasks into an Excel spreadsheet. Patterns of interaction can be further examined using conversation analysis. By coding the tasks across the data set, we can easily compare different counselor strategies or participant responses, for example how sessions start and end, what types of referrals are offered, how the risk assessment form is administered, or how reasons for testing are elicited.
The time chart in Figure 1 depicts the sequence of tasks for 10 pretest sessions recorded by 2 counselors, C1 and C2. Each color represents a task, such as reason for testing, filling out paperwork, making referrals, and discussing oral sex risk. The tasks in blue and violet shades are "counseling" tasks, defined here as interaction where the client is talking about personal views, feelings, concerns, questions, reasons for testing, etc. The yellow, orange, and green tasks relate to risk assessment and health education tasks during which the client listens passively or provides only short responses to close-ended questions. The lengthy stretches of orange represent paperwork tasks such as the CIFs and STD test forms that contain data collected for county and state health departments for surveillance purposes.
| Figure 1. Time Chart Displaying the Sequence of Tasks for 10 Pretest Sessions by 2 Counselors|
(see larger image of Figure
Time charts allow us to tabulate and display the contents of VCT sessions in different ways. For example, if we reduce the myriad tasks displayed in Figure 1 into only 3 categories, we can show the distribution of time among clinical, counseling, and surveillance or data gathering tasks. This ratio provides a useful measure of the degree to which sessions are "client centered," or focused on the personal concerns, interests, and meanings that the client assigns to risk behavior. Figure 2 contrasts the same sessions displayed in Figure 1 in terms of the distribution of time for the 3 categories of tasks.
| Figure 2. Time Chart Displaying Percentage of Clinical, Counseling, and Surveillance Tasks|
(see larger image of Figure
The 2 counselors have similar levels of training and experience. Yet, as we see in Figure 1, counselor C1 employs a relatively fixed and shorter sequence of tasks that includes very little counseling (with the exception of client PACT 01, who was particularly distressed and talkative) when compared with counselor C2, who devotes a greater proportion of the sessions to counseling tasks. Figure 2 also highlights the enormous proportion of the sessions that counselors spend on clinical and data-gathering tasks. As the CA studies reviewed above suggest, not only do these tasks take up precious time, they also create a bureaucratic dynamic whereby clients may feel less able or willing to articulate their personal concerns.
Time charts provide a valuable training and supervision tool by showing--in a nonthreatening manner--how counselors allocate effort within sessions. This enables counselors to think critically about how they manage the session and transition from one communication format to another. Time charts provide an objective way to compare the approaches of different counselors or those of one counselor who has different types of clients. Counselors also can compare their approaches in light of client-centered guidelines that suggest they focus on developing custom-tailored plans for future risk reduction, while minimizing time spent providing clinical lectures and completing surveillance forms.
| Clients' Motivations for Testing|
Regular HIV testing, particularly for MSM, has been encouraged by media campaigns and by test counselors. A survey conducted in California in 2000 showed that more than two thirds of MSM presenting for testing had previously tested 3 or more times, with half of that group having previously tested 5 or more times.(41) A particularly puzzling finding from VCT outcome studies has been that clients receiving repeated negative results report increasing levels of risk behavior. Analysis of HIV testing records in San Francisco and London reveals that MSM testing 3 or more times reported significantly higher rates of HIV risk behavior than MSM who had tested 2 or fewer times.(42-44) A similar trend was noted among a 7-city sample of 3,430 young MSM in the United States.(45) Some researchers have suggested that there may be a relationship between the number of times men test and their propensity to engage in risky sex.(43) The relationship between increased risk behavior and rates of repeat testing is indeed troubling because it may suggest that VCT actually contributes to risk behavior among a population already at high risk of infection. In light of these findings, it is crucial to understand why high-risk clients test and how testing relates to subsequent risk behavior.
Research on motivations for behavior is difficult to conduct because there is no way to judge the accuracy of an account for a particular action, such as seeking an HIV test. Accounts are always mediated by the social context of the interaction.(34,36) For example, media campaigns suggest reasons for testing such as "knowledge is power," "respect yourself," and "how do you know what you know?" Questionnaires with close-ended questions, such as the risk assessment form, tend to suggest other motivations, often leading clients to choose reasons for testing suggested by the counselor. It is therefore difficult to extricate reasons for testing suggested by public health messages from clients' "own" motivations. One way to explore motivations for testing is to examine recordings of test sessions to hear how clients describe their reasons for testing to test counselors. This can be supplemented with in-depth qualitative research interviews to elicit accounts of past testing behavior and descriptions of how the decision to test related to clients' HIV prevention strategies, risk behaviors, relationships, or other factors. At best, we can describe some of the interactional and moral constraints on replies to the question, "What brought you in to get tested?"
Two CA studies that examined in detail how counselors elicited their clients' motivations for testing demonstrate how motivations are artifacts of VCT interaction, not pre-existing mental constructs retrieved from clients' minds.(5,36) The discussion of a client's reasons for testing at the start of the session has important implications for subsequent counseling interaction. The following is a typical excerpt from the first few moments of a test session recorded by Sheon. "CO" is the counselor and "CL" is the client.
1 CO: Okay (0.6) what brings you in to be tested for HIV?
2 CL: Uh:: (0.4) it's been a while since my last test (0.5)
and I just wanna be certain
3 CO: Okay. And uhm .hhhh any particular thing that you can
think of that might have
4 put you at risk for (0.4) exposure to HIV?
5 CL: No. But I've been with my partner for about four years
and it's been monogamous
6 but (0.6) you know (0.7) you never know
7 CO: Okay
8 CL: Y I (1.5) I understand it can lie dormant for quite a while
For a guide to transcription symbols commonly used in conversation analysis, see http://www-staff.lboro.ac.uk/~ssca1/notation.htm
Sensitive issues bearing on the client's character are often raised obliquely and then quickly dropped in favor of more neutral topics such as the demographic questions on the risk assessment form or factual information on how the test works. For example, when asked the ostensibly neutral question of "What brings you in?" in the context of a "free" testing service, the client is placed in a bind. He or she must present enough risk to warrant a test, yet simultaneously avoid the implication of irresponsible behavior. As we see from the client's first response (line 2), clients typically account for their tests as a part of a routine practice, not as a response to a particular risk incident. In this way, the client's minimal and somewhat vague reply invites the counselor to specify the preferred type of response through a follow-up question about specific risks. The client's initial "No" in line 5 is immediately followed by an account relating to his monogamous relationship with his partner. In addition to providing an account for testing, the client's statement serves to portray him as both faithful and ever-vigilant against HIV risk. This statement can be seen as an example of identity work, ie, self-characterizations that serve to bolster a person's moral status. However, the client's response "it's been monogamous" is a problematic response to the question about specific risk, and that is demonstrated by the counselor's continuer in line 7 and later in the session when the counselor continues to probe about the client's possible infidelity (see Table 2). The client produces another account for his testing in line 8 that demonstrates a possible misconception about the length of the antibody window period. That leads the counselor to digress into an explanation of the window period, thus temporarily shifting the format from a morally charged interrogation of fidelity to a more neutral and didactic format. Attention to the turn-by-turn construction of motivations at the start of the counseling session reveals the moral implications embedded in any account about a client's decision to test.
The session excerpt above reveals an assumption that has dominated much of the existing literature and practice of VCT: Clients seek testing in response to specific risk incidents. This assumption is most evident at the start of almost every VCT session when clients are asked to describe what specific risk incidents prompted them to seek testing. Initially, clients almost always describe the stimulus for testing not in terms of a specific risk incident but in terms of calendrical cycles ("it's that time," or "it's been about a year since my last test"). Routine testing is reinforced by the uncertainty created by the window period, and counselors often recommend regular testing every 3 to 6 months. The fact that clients generally do not see their decision to test as indicative of a problem requiring counseling presents a fundamental barrier to engaging clients in a counseling discussion. If there is no problem or concern about risk, there is no basis for a counseling relationship. As a result, counselors spend much of the sessions in search of problems or patterns of risk behavior that can serve as foundations for counseling relationships. Such a search often encounters passive or active resistance from clients who may become defensive and circumspect about their behavior. Counselors' probes to identify a specific incident that precipitated the decision to test can sometimes resemble police interrogations. In interviews about their recorded counseling sessions, counselors perceived clients' nonspecific motivations for testing--such as "it's been a while"--as evasive, obfuscating the "real" reasons for testing. In this context, the data collection form provides a useful warrant to query about a wide range of risks. As the sessions progress, clients' reasons for testing are reconstituted to reflect the risks disclosed for the form.
Process research on clients' motivations for testing suggests that routine testing may also serve as a way for clients to manage anxiety around sexual intimacy and relationships.(5,8,46) Clients often test at the start or end of a relationship. In the case of a new relationship, the act of testing offers a symbol of commitment. New couples may test specifically to justify having unprotected sex on the basis of HIV seroconcordance, a strategy frequently encouraged by counselors. When couples break up, testing serves as a purification ritual, a "clean slate," that helps clients manage the sense of anxiety over betrayal of negotiated safety agreements with primary partners. Clients may perceive routine testing as a preventive measure itself, much like routine screening tests for cancer, particularly among fatalistic clients who perceive little control over their risks. In either case, the motive to test is not necessarily linked to a particular concern about past risks or a desire to reduce unprotected sexual behavior. Prior research suggests that counselors tend to interpret clients' reasons for testing as responses to past risk behavior. This may obscure other, more salient, reasons for testing and thus limit the relevance of the session for the client.
| Rapid Testing|
The low rate of return for HIV test results has raised questions about the cost effectiveness of the current VCT model based on 2 visits. Rates of return as low as 50% in STD clinics suggest that the problem may lie not so much in client apathy about HIV status as in feelings among clients that they were pressured into testing or in feelings of dissatisfaction with the test counseling.(47,48) Very little research has examined why people fail to return for their results or avoid testing altogether.(49) The approval of OraQuick rapid HIV test kits by the U.S. Food and Drug Administration may revolutionize VCT practice in the United States. Long used in the developing world, rapid testing combines the standard pre- and post-test sessions, normally separated by a week or more, into a single visit. The same-day result eliminates the stressful wait and ensures that clients receive either a negative or a "preliminary" positive result. (A positive result from a rapid test must be confirmed, thus requiring a follow-up visit.) While it is too early to know whether counselors and clients will prefer single-session testing to the current 2-session format, the shift to the new testing technology provides an opportunity to reassess current counseling approaches and their relevance to reducing risk behavior.
Among the CDC's aims in promoting rapid testing are maximizing the number of people tested and providing more flexibility to health care workers in structuring the test counseling.(50) In a departure from previous guidelines, CDC interim guidelines on rapid testing downplay the need for individualized pretest counseling.(51) For example, the guidelines for rapid testing in nonclinical settings suggest pretest "group counseling" and "assembly line" rapid testing, with a single counselor providing post-test counseling while clients wait for test results. Prenatal HIV testing is another area that could benefit from greater flexibility in VCT procedures. For mothers who have not received prenatal care, rapid testing can be used during labor to determine whether the administration of nevirapine to prevent mother-to-child transmission is indicated. When attempting to provide the client with important information during labor, counseling must be streamlined to deal with highly complex issues of consent to test and treat both mother and child.(52)
How rapid testing will affect both testing outcomes and process remains to be studied. Preliminary results from the RESPECT-2 trial comparing STD incidence rates following rapid, single-session vs standard 2-session VCT among STD patients showed that rapid testing was slightly less effective than the traditional, 2-session model.(53) Anecdotal evidence from pilot programs in San Francisco and reports from the RESPECT-2 trial suggest that rapid testing VCT procedures are much more intense than traditional pretest counseling, perhaps because counselors accompany their respective clients through the entire test process and become more emotionally invested in the results.(54) The UCSF AIDS Health Project, which is contracted by the state of California to train hundreds of test counselors to conduct rapid testing, has developed rapid test guidelines based on the RESPECT-2 protocol. Unlike the scaled-back CDC guidelines, the UCSF guidelines re-emphasize client-centered counseling in response to trainers' observations of current counseling practice that, yet again, revealed both an over-reliance on the data collection form and a tendency to prematurely provide advice or referrals before adequately eliciting the clients' concerns.(55) The authors of the protocol see rapid testing as an opportunity for a more comprehensive exploration of risk and behavior change and have listed more than 30 counseling tasks (see Table 3). This comprehensive approach is viewed as necessary for enabling counselors to engage clients in a discussion for the full 20 to 30 minutes required for the test results to develop. Unfortunately, as with previous client-centered guidelines, there is no discussion of how to minimize the impact of the risk assessment form and other paperwork so that it does not jeopardize the rapport needed to establish a more comprehensive counseling relationship.
| VCT Is Not a Magic Bullet|
Evaluations of the CDC's VCT programs have consistently concluded that HIV testing was never intended to stand alone as a prevention intervention.(56) To be effective, testing must comprise one part of a larger portfolio of HIV prevention services. Just as counselors must link HIV-positive clients to clinical services providing "combination therapy," high-risk HIV-negative clients must be connected to "combination prevention" services such as ongoing individual or group counseling.(10) Unfortunately, many of these complimentary prevention services were never implemented because of funding restrictions imposed on explicit safer sex workshops targeting high-risk populations such as MSM and injection drug users. As a result, media campaigns promoting HIV prevention have tended toward euphemistic vagueness, emphasizing abstinence, monogamy, and routine HIV testing. To finance the new rapid testing initiative, the CDC recently redirected funds away from community-based outreach programs that were established to reach all high-risk minority populations, not just those willing to be tested. Current models of VCT are unlikely to have an impact on HIV prevention and actually may contribute to the trend toward complacency about HIV infection among populations that test routinely. New approaches to VCT clearly are needed because people will continue to seek testing. But these must be supplemented with other services such as individual and small group counseling, community organizing, and other interventions that have proved effective.(57)
| Future Directions for VCT Research|
More research on the relationship between repeat testing and risk behavior is needed so that VCT can be tailored to meet the needs of specific high-risk populations not benefiting from the current model. VCT research should combine outcome and process measures to identify specific counseling skills and approaches that are effective for specific high-risk populations that already seek testing. Given the current funding initiatives and large-scale efforts to introduce rapid testing within and outside clinical contexts, research examining different ways to structure rapid test counseling and compare its effectiveness with the traditional 2-session approach is badly needed. Research is also needed to examine the workload implications of single-session VCT for counselors and to determine how the quality of counselor-client interaction is affected by the waiting period for "rapid" results. Innovative approaches to counseling that have shown success in changing behavior among high-risk populations should also be tested against the client-centered approach in the context of single-session rapid testing. Finally, epidemiological research is best left to epidemiologists and trained survey interviewers. Test counselors should not be expected to collect surveillance data from their clients because data collection instruments produce a quality of interaction that is antithetical to truly client-centered counseling.
| Table 1. General Principles of Client-Centered Counseling|
Client-centered counseling: The clients we see will be best served by client-centered counseling, where the focus is on the client's concerns and interests. These techniques explore the personal meaning a client gives to the issues being discussed.
Context: The impact of counseling will be enhanced when counselors are able to explore and assess the physical and emotional circumstances under which clients' lives, and consequently their HIV risk behaviors, take place.
Individualized sessions: The impact of counseling will be enhanced when counselors tailor sessions based on the specific needs and unique situations of individual clients.
Information alone does not lead to behavior change: Behavior change is a complex process that requires interventions based on a client's personal circumstances. Providing information as the sole, or main, intervention generally is not sufficient to lead a person to change behaviors.
Neutral stance: It is appropriate for counselors to take a neutral stance when addressing ambiguous information with clients and to maintain a nonjudgmental manner when discussing sexual practices, substance use, and other personal behaviors.
Limited role: It is important for HIV counselors to recognize the limitations of the counselor role.
Source: Ed Wolf. UCSF Enhanced Counseling Skills Training: The Single Session Risk Assessment and Test Disclosure. September 2003.
| Table 2. Excerpt from an Extended Discussion of Reason for Testing|
CO: So yeah (.) I think you are very very very low (.)
uhm (.) in fact I'd almost wonder if your being in here for
this test (.) since you haven't had one in (.) what over
[four years why (.) why are you in here(.)
CL: [yeah three or four
CO: cause your risks are real real (.) your risks are almost
non-existent (1) can can you give me the information that
maybe this questionnaire hasn't (.) addressed that would
be more of a reason
CL: Uhm (.) I guess it would be mostly because I'm I've I don't
know if I would call it paranoid (.) but I'm overly (.) um
(.) overly cautious (.) and umm (.) as I mentioned earlier if
there's any chance that I might be with somebody I'd rather
be absolutely certain (.) I couldn't live with the knowing that
perhaps I'd infected somebody without my knowledge
Note that the client has just offered the reason for testing that had been suggested by the counselor reading from the choices listed on the CIF, "worried about infecting others." The client had previously rejected this reason since it implied he was actively seeking other partners. However, the counselor presses on in an attempt to extract a confession of truth about past risks.
CO: Okay (.) but if you haven't done any transmission
of bodily fluids
CL: M hm
CO: I wonder where that paranoia is coming from
CL: I don't know (.) Oh
CO: I mean have you engaged with any heavy petting with ah (.) I
mean I'm not going to go out and tell your partner (.) but
have you engaged in any heavy petting or or hanky panky with
any others since you've been up here?
CL: No but about oh I about two years ago (.) I guess what's been
bothering me is because about two years ago or about a year and
a half [ago actually um (.)
CO: [M hm
CL: I was involved in a fight
CO: [M hm
CL: [and ah we were both bleeding so
CO: And there was a mixture of blood (.) your blood and his
CL: [I don't know if there was any mixture of blood but there was
definitely blood on my shirt and on his shirt and I had some (.)
cuts and (1) the other person did also (.) and I don't know the
other person's history.
CO: Okay. The other person was another male?
CL: And so (.) I guess I just want to be absolutely sure
CO: Cause in as far as the sexual arena is concerned
CL: Oh yeah (.) I'm very careful
CL: I've always been very careful
CO: Okay (.) alright (.) that sounds like a good reason to get tested
For a guide to transcription symbols commonly used in conversation analysis, see http://www-staff.lboro.ac.uk/~ssca1/notation.htm
| Table 3. Checklist Version of Single-Session Risk Assessment and Test Disclosure--Negative Result|
Make introductions, conduct orientation to the session, and confirm informed consent.
Administer the test.
Address any client concerns about waiting for the result.
Enhance the client's perceptions of personal risk for HIV/STDs.
Listen for and identify behaviors that put the client at highest risk for HIV/STDs.
Assess the client's level of concern about having or acquiring HIV/STDs.
Discuss the client's history of HIV testing and behavioral changes in response to results.
Discuss the specifics of the most recent risk incident by exploring context--who, what, where, when, and how--of what happened.
Assess the level of risk that is acceptable to this client.
Assess communication of HIV/STDs concerns with partners.
Identify circumstances or situations that contributed to the risk incident.
Assess the client's patterns of risk behavior.
Identify risk reduction successes and obstacles.
Explore triggers and situations that increase the likelihood of higher-risk behaviors.
Reassess the client's level of acceptable risk.
Summarize the risk incident and risk patterns.
Explore behaviors that the client will feel most motivated to change or be most capable of changing and negotiate a risk-reduction step.
Problem-solve any obstacles in achieving the risk-reduction step.
Identify sources of support and provide referrals.
Assess client's readiness to receive test result.
Retrieve test result, complete appropriate paperwork, and strategize disclosing the result to the client.
Disclose negative test result. (a) Ask the client if he or she has any questions or concerns regarding the result. (b) Follow client's lead in determining when to disclose. (c) State result in a direct, neutral tone. (d) Wait for client's response before proceeding.
Assess client's emotional state and counseling needs.
Explore feelings. Assess and acknowledge client's feelings in response to negative result.
Explore information and thoughts. (a) Assess client's understanding of test result and clarify misconceptions. (b) Assess need for retesting based on window period and recent risks.
Explore behavior. Reassess client's risk-reduction step that was discussed earlier in the session.
Explore interpersonal issues. Assess client's support network and disclosure-related issues.
Evaluate clients who are at greater or lesser risk.
Identify sources of support for the risk-reduction step. Assess strength of social support for client's intended risk-reduction step.
Revisit referrals from risk assessment. Reassess client's risk-reduction step that was discussed earlier in the session.
Let the client know you have confidence in his or her ability to complete the step.
Close the session but not the door.
Source: Ed Wolf. UCSF Enhanced Counseling Skills Training: The Single Session Risk Assessment and Test Disclosure. September 2003.
|| ||Marcus A. FDA approves 20-minute HIV test: government hails device as big step in infection detection. Health On the Net Foundation Web site. November 7, 2002. Available at: http://www.hon.ch/News/HSN/510143.html. Accessed January 27, 2003.|
|| ||Institute of Medicine. No time to lose: getting more from HIV prevention. National Academy Press Web site. 2001. Available at: http://www.nap.edu/books/0309071372/html.|
|| ||Doll LS, Kennedy MB. HIV counseling and testing: what is it and how well does it work? In: Schochetman G, George JR (eds). AIDS Testing. New York, NY: Springer-Verlag; 1994:302-319.|
|| ||Sheon N. Sacraments of Surveillance: Ethnography of an HIV Test Clinic. [dissertation]. Berkeley, Ca: University of California; 1999. Parts available at: http://www.managingdesire.org/AboutConfessionandHIVTC.html.|
|| ||Roiphe K. Last Night in Paradise: Sex and Morals at the Century's End. Boston, Ma: Little, Brown; 1997.|
|| ||Silverman D. Strategies for evaluating HIV test counselling. In: Discourses of Counselling: HIV Counselling as Social Interaction. London: Sage; 1997. Edited version available at: http://www.managingdesire.org/SilvermanStrategies.html.|
|| ||California Department of Health Services. Evaluation of Testing and Counseling Program. Office of AIDS; 1996.|
|| ||Rogers C. Client-Centered Therapy: Its Current Practice, Implications, and Theory. Boston, Ma: Houghton Mifflin; 1951.|
|| ||Rogers C. Empathic: an unappreciated way of being. The Counseling Psychologist. 1975;5(2):2-10.|
|| ||California State Office of AIDS. HIV Counseling Information Form (HIV6). 2003. Available at: http://www.managingdesire.org/hivsurv/HIV6.pdf [PDF, 444K].|
|| ||California State Office of AIDS. Guidance for Completing the HIV Counseling Information Form (HIV6). 2003. Available at: http://www.managingdesire.org/hivsurv/HIV6guide.pdf [PDF, 65K].|
|| ||Polisher C. Is oral sex safe? 1993. Available at: http://www.qrd.org/qrd/aids/1993/oral.sex.and.hiv-preferred.stock-08.12.93|
|| || Kamb ML, Fishbein M, Douglas JM, Jr., Rhodes F, Rogers J, Bolan G, Zenilman J, Hoxworth T, Malotte CK, Iatesta M, Kent C, Lentz A, Graziano S, Byers RH, Peterman TA. Efficacy of risk-reduction counseling to prevent human immunodeficiency virus and sexually transmitted diseases: a randomized controlled trial. Project RESPECT Study Group. Jama 1998; 280:1161-7.|
|| ||Centers for Disease Control and Prevention. Project RESPECT Web page. Available at: http://www.cdc.gov/hiv/projects/respect/default.htm.|
|| ||Center for AIDS Prevention Studies. The Voluntary HIV Counseling and Testing Efficacy Study. Available at: http://caps.ucsf.edu/resources/project-websites/voluntary-counseling-and-testing-project/voluntary-hiv-counseling-and-testing-efficacy-study/|
|| ||McFarland W. Increases in Sexual Risk Behavior & STD's among MSM: Trends in the Era of HAART. May 13, 2003. San Francisco Department of Public Health. Available at: http://184.108.40.206/Powerpoints/NewHampshire6.pdf. Accessed August 12, 2003.|
|| ||Heritage J. Conversation analysis and institutional talk: analyzing data. In: Silverman D, ed. Qualitative Research: Theory, Methods, and Practice. London, England: Sage; 1997:161-182.|
|| ||Sacks H, Schegloff E, Jefferson G. A simplest systematics for the organization of turn-taking in conversation. Language. 1974;50(4):696-735.|
|| ||Kinnell A. So why are you here? Assessing risk in HIV prevention and test decision counselling. Sociology of Health and Illness. 2001;23(4):447-477.|
|| ||Silverman D. Discourses of Counselling: HIV Counselling as Social Interaction. London, England: Sage; 1997. Two chapters from this book are available at: http://www.managingdesire.org/ConferencePapersIndex.html.|
|| ||Kinnell AM, and Maynard DW. The delivery and receipt of safer sex advice in pretest counseling sessions for HIV and AIDS. Journal of Contemporary Ethnography 1996;24:405-437.|
|| ||Heritage J and Sefi S. Dilemmas of Advice: Aspects of the delivery and reception of advice in interactions between health visitors and first-time mothers. In: Drew P and Heritage J, eds. Talk at Work: Interaction in Institutional Settings. Cambridge, England: Cambridge University Press; 1992.|
|| ||Ferrara K. Therapeutic Ways with Words. Oxford, England: Oxford University Press; 1994.|
|| ||California State Office of AIDS. California HIV Testing and Counseling Annual Report, January--December 2000. California Department of Health Services; 2002. Available at: http://www.dhs.cahwnet.gov/aids/Reports/PDF/2000%20HIV%20C&T%20Annual%20Report.pdf.|
|| ||Dilley JW, Kellogg T, Rinaldi J. Risk behavior and HIV incidence among repeat testers at San Francisco's anonymous test sites, 1997-1999. Presented at: San Francisco HIV Prevention Summit; November 22, 2000; Half Moon Bay, California.|
|| || MacKellar DA, Valleroy LA, Secura GM, Bartholow BN, McFarland W, Shehan D, Ford W, LaLota M, Celentano DD, Koblin BA, Torian LV, Perdue TE, Janssen RS. Repeat HIV testing, risk behaviors, and HIV seroconversion among young men who have sex with men: a call to monitor and improve the practice of prevention. J Acquir Immune Defic Syndr 2002; 29:76-85.|
|| ||Beardsell S. Sexual Risk Taking and HIV Testing: A Qualitative Investigation. In: Aggleton P, Hart G, Davies P. Families and Communities Responding to AIDS. London, England: UCL Press; 1999.|
|| ||Flowers P, Church S. To test or not? HIV antibody testing amongst gay men. Culture, Health & Sexuality. 2002;4(1):43-65.|
|| ||Centers for Disease Control and Prevention. Rapid HIV Testing in Non-Clinical Settings. Advancing HIV Prevention: Interim Technical Guidance for Selected Interventions. August 1, 2003. Available at: http://www.cdc.gov/hiv/partners/Interim-Guidance.htm.|
|| ||Metcalf C, Cross, H, Dillon B, et al for the RESPECT-2 Study Group. Randomized controlled trial of HIV counseling with rapid and standard HIV tests (RESPECT-2) preliminary results. Presented at: 2002 National STD Prevention Conference; March 4-7, 2002; San Diego, California. Available at: http://www.cdc.gov/hiv/projects/respect-2/Bibliography.htm.|
|| ||Padilla SM, Dillon B, Iatesta M, et al and the RESPECT-2 Study Group. Counselor perceptions and impressions of counseling using a rapid HIV test: RESPECT-2. Presented at: 2001 National HIV Prevention Conference; August 12-15, 2001; Atlanta, Georgia. Available at: http://www.cdc.gov/hiv/projects/respect-2/Bibliography.htm.|
|| ||Wolf E. Personal communication at the UCSF Enhanced Counseling Skills Training: The Single Session Risk Assessment and Test Disclosure; San Francisco, California; September 2003.|
|| ||Centers for Disease Control and Prevention. External Review of CDC's HIV Prevention Strategies. CDC Advisory Committee on the Prevention of HIV Infection, Subcommittee on Promoting Knowledge of Serostatus; 1994.|
|| ||Centers for Disease Control and Prevention. Compendium of HIV Prevention Interventions with Evidence of Effectiveness. CDC HIV/AIDS Prevention Research Synthesis Project; August 2001. Available at: http://www.cdc.gov/hiv/resources/reports/hiv_compendium/.|