Care for HIV and AIDS patients has already overwhelmed the health services in many East African countries. Access to antivirus and prophylactic drugs is limited due to cost, inadequate supply and the need for close medical supervision. Currently the most feasible category of care is the provision of counselling services, as it is not dependent on drugs or medicines and can be offered at a lower cost. The costs associated with counselling, however, depend upon the setting where the counselling takes place. A growing body of literature recognises counselling as an integral part of services for people infected by HIV/AIDS (PHIV) and also those affected by these conditions [1-3]. Although counselling is not intended to be a substitute for medical interventions, it can ameliorate emotional problems of PHIV, which if ignored, can lead to greater alienation, deterioration of family cohesiveness, broader social problems and exacerbated medical problems [4-6].
As the epidemic continues, the need for psychological care and support services for PHIV and their families will increase. Recent estimates conclude that the per capita burden of mental health needs for PHIV soon will be relatively greater in Africa than in the rest of the world combined . Given the shortage of trained professional counsellors in Africa, paraprofessional counselling (PPC) may be a feasible way to address the growing need for psychological assistance among PHIV. However, the limited amount of training that PPCs receive has been insufficient to prepare them for the range of problems that PHIV present in counselling [1,8,9]. PPCs are typically trained over a period of weeks compared to the years of training that professional counsellors receive. Further, the content of training for PPCs often is not based on the types of problems they encounter in their counselling work . Effective training for PPCs is needed so that these counsellors are equipped to respond to the needs of people seeking counselling for HIV/AIDS.
Psychosocial problems of PHIV are complex and may be complicated by the individual‚s culture, context of infection, and disease stage. Several investigators point out the importance of identifying culturally appropriate counselling strategies to help PHIV cope with their infection [4,5,7,11-14]. Whereas there is a demonstrated need for psychological interventions to enhance AIDS care in Africa [7,12,13,15-17], there is still an insufficient understanding of the appropriate content and delivery of psychological care. Developing counselling interventions for PHIV in Africa is hampered by a gap in understanding the psychological needs of PHIV that can be addressed through counselling. Although there is a growing literature on the psychological care for PHIV, the emphasis has been largely on the Western client. Few empirical studies have addressed the psychological needs of PHIV in Africa [12,18].
In particular, little is known about the effectiveness of specific forms of counselling in the African setting. For example, it is not clear whether individual, family, or group counselling is more effective and whether intensive counselling is worth the cost of repeated follow-up . Likewise, there is little information on whether nurse, midwife, or trained paraprofessional can adequately handle the mental health needs of PHIV or if professional psychologists should take full responsibility . Further, counselling interventions typically have been discussed in terms of their role for fostering behaviour change to aid HIV prevention [11,21,22]. Rarely are counselling interventions evaluated for their impact on improving coping skills, easing suffering or improving the quality of life of HIV/AIDS patients in Africa .
However, before these issues can be considered, the psychological needs of African PHIV need to be defined. The counselling field currently lacks information on the issues and problems that PHIV themselves identify as priorities. Such data would be useful when developing curricula for PPCs‚ training for improving how they address the needs of their clients.
One of the authors developed this study in response to an operational problem of the Kenya AIDS Society (KAS), a local non-Governmental Organization (NGO) in Kenya. Kenyans with HIV/AIDS set up KAS in the early 1990s. The group initially offered peer support to other PHIV in an STD clinic in Nairobi. As the group expanded, it began operating support services in five other Nairobi clinics that functioned as HIV screening and treatment centres. The KAS offers counselling to individuals, families and communities affected by HIV/AIDS. KAS counsellors are all PPCs trained by the Kenya Red Cross Society on a 2-3week training course. Whereas their initial training was intensive, the counsellors found they were still not adequately equipped to respond to the issues that arose during HIV counselling. They felt both overwhelmed by their clients‚ problems and under-prepared to address them. During 1991 KAS counsellors sought technical assistance from one of the authors to improve their capability to provide HIV counselling services.
This study was conducted to define the psychosocial needs of people seeking HIV counselling and to identify and prioritise their associated problems. This information would provide a basis for constructing a curriculum for training PPCs in HIV counselling.
Materials and methods
Design and setting
This study was a descriptive, cross-sectional analysis of problems that PHIV presented during KAS counselling sessions in Nairobi, Kenya, over a 12-month period (January-December 1992). Sixteen KAS counsellors were instructed to record in their case notes the issues reported by all clients who sought follow-up counselling.
Study participants included 307 clients tested for HIV at six Nairobi clinics specialising in STDs, tuberculosis and other infectious diseases. The study population was split approximately equally between each of these clinics. Clients were provided pre-test and post-test counselling by 16 PPCs and were offered follow-up counselling as well. Of the 307 clients initially enrolled in the study, 125 (20 HIV-negative, one HIV status unknown) returned for follow-up counselling from the KAS during the study period (return rate =41%). The remaining 182 were lost to follow-up. An additional 43 individuals (one HIV-negative) who already knew their HIV status, also sought follow-up counselling from KAS counsellors during the study period. For this group the PPCs did not carry out pre- or post-test counselling. The total psychological data comprises information from 168 cohort individuals. All data for this study were collected during the follow-up counselling sessions.
The study team organised an additional 2weeks of intensive and specialised training by a professional HIV/AIDS psychologist, for the 16 PPCs. The KAS counsellors included nine male and seven female PPCs; all were HIV-positive. The tribal breakdown was: 10 Kikuyu; two Camba; one Luo; one Luya; one Coastal tribe; and one Swahili. As Kikuyus represented the largest tribe in Kenya, it is not surprising that most PPCs are from this tribe. As there was only a small number of PPCs, further stratification of the PPCs may give unreliable results. Nevertheless, PPCs were stratified as either being active (carried out >10 counselling sessions) or less active (carried out <10 sessions). A χ2 test revealed that female PPCs were no more active at counselling than males; and Kikuyus were no more active than PPCs from any other tribe. Gender or tribal affiliation are not thought to be related to the psychological events clients presented.
One author instructed the PPCs to note key words and phrases that clients said during the counselling sessions, and to use two questionnaires for this purpose: a demographic form and a counselling form. The PPCs recorded information on the demographic form after the first visit of each client, regardless of his/her HIV status. Information included age, sex, education, HIV status, marital status, employment status, wealth, number and age of children, knowledge of own HIV status, knowledge of partner‚s HIV status, and partner‚s knowledge of the client‚s HIV status. The only demographic data recorded for the 43 clients who joined the study without previous (pre- or post-test) counselling, included gender and HIV status.
The counselling form was completed during follow-up counselling sessions. Each counsellor transcribed his/her own notes onto a counselling form either during the session or immediately after it. They were instructed not to see another client until they had completed the task. Data recorded from the counselling notes included a client‚s: current physical complaints, psychological and financial problems, attitudes toward the family, perception of the family‚s attitudes toward the client, perception of societal values toward HIV infection, attitudes toward treatment regimens, and problems of material support for him/herself and his/her family. The time between the first visit, when the client learned his/her HIV status, and the follow-up counselling session ranged between 2weeks and 4months. Twenty clients returned for more than one session of follow-up counselling. In total, the study recorded data from 190 counselling sessions. All information collected by the PPCs was reported voluntarily by the clients and transcribed as reported.
Coding and analysis
The study team collected data forms daily from the 16 PPCs. Data were coded using an ipsative method so that each new counselling issue received a unique code . Each subsequent presentation of the same topic received the same code. By using an ipsative approach the full range of issues presented by clients could be studied. However, this observation is not meant to imply that the clients presented all their psychological problems. The study records only issues they elected to discuss with counsellors. The total number of psychological problems may be larger than what was found in this study. Clients may have had other serious problems that they did not share with counsellors and so are not included in the study. This limitation is inherent to a study of psychological problems that are reported during counselling. Additional investigations were not carried out to identify whether clients had other unreported or latent psychological problems.
The study data include 109 psychological variables and 56 other variables with which clients described their own health problems. These data represent the full range of issues brought to KAS counselling sessions. A full discussion is presented elsewhere . As the physical data were not confirmed by clinical examination (due to resource constraints) they are not analysed in an in-depth manner.
Analyses of psychological variables included factor analysis with a Varimax rotation to place the 109 psychological variables into logical groupings. Hence, factor analysis identified the underlying categories of psychological problems in the dataset and the variables included in each category. Then, for each client in the study, a factor score was created for each factor by counting the number of variables reported in each counselling session. For example, if a factor category included 10 variables, a client‚s score could be as low as zero or as high as 10. These factor scores were used in additional analysis. A χ2 test was used to examine associations among the demographic variables. Multiple regression examined associations between each factor and clients‚ demographic characteristics.
The following results stratify variables provided by the 168 clients attending follow-up counselling. Where the sum of data is less than 168 this is due to missing data.
Description of KAS counselling clients
Among the 168 clients attending follow-up counselling 87% were HIV-positive and 13% HIV-negative (Table 1). Approximately even proportions of male and female clients entered the study (49% women, 51% men); however, more women than men sought follow-up counselling after receiving their HIV test result (56% women, 44% men). Seventy-one percent of clients seeking counselling were unemployed and 86% had fewer than five children.
Most of the client population had received minimal education (66% had received less than 8years of education) and clients came from rural areas since most had chambas (small land holdings in rural areas). A wealth indicator was developed based on the size of these chambas, and the number and type of animals and crops owned. The weight of animals and crops were converted into their relative cash value. Clients were categorised into seven strata according to their wealth measurement (range =0-60 points). Nearly 70% of the counselling clients were in the lowest two wealth strata. The average male wealth score (23.89±17.8) was nearly twice as large as the average female wealth score (12.21±14.5).
Thirty-four percent of clients reported having received HIV counselling in the past. Yet, 89% of clients did not know their HIV status prior to this study. Among HIV-positive clients, 87% were previously unaware of their HIV status, which indicates that most of the seropositive clients in the study represented newly detected cases of HIV infection. Likewise, 83% of clients were unaware of their partner‚s HIV status. This suggests that large proportions of clients either have been placed at risk of HIV infection by their partners or are placing their partners at risk.
Case load of clients in HIV counselling
Clients tended to raise many physical and psychological concerns in a single session. Clients frequently discussed their health problems (males =90.5%, females =90.3%). Male clients had up to seven physical complaints with 84% reporting as many as four problems. Female clients reported up to 10 physical problems, with 84% presenting as many as four complaints. Since all physical problems were self-reported without any clinical confirmation, no in-depth analysis of these data was carried out. Also, as there are no clinical confirmations, the study cannot determine whether clients‚ reports match their actual physical condition or assess their willingness to talk about their physical status.
In any one counselling session, clients presented varying numbers of psychological problems. Women presented up to 21 psychological problems per counselling session; 35% (n=33) presented one to seven, whereas 65% (n=60) had eight or more. By comparison, male clients presented up to 23 psychological problems in a single session; 51% (n=38) reported one to seven complaints, and 49% (n=36) had eight or more. Whereas the range of psychological problems was wider for men, women tended to present more of them in any one session. This between-gender difference in distributions was statistically significant (χ2=4.25, P=.039). When these problems were analysed whilst controlling for clients‚ wealth gender differences were not found. This finding suggests that the difference between genders in psychological problem presentation is based on differences in wealth.
The most frequently reported worry of both men and women was their economic situation. They mentioned that they could afford neither the cost of food nor medications. Women also frequently worried about how they would pay for rent and school fees. In addition to material concerns, men also tended to worry about keeping the infection a secret from family and others. They frequently mentioned concerns about being socially isolated and the difficulty of telling others about their infection.
Further analyses investigated the relationship of psychological problems and employment status. A χ2 test revealed a significant relationship between clients‚ employment status (employed versus unemployed), and the two frequency distribution groupings of psychological problems (i.e., one to seven problems and ≥eight) (χ2 =10.67, P=0.001). Sixty percent of employed clients reported seven or fewer problems whereas the majority (72%) of unemployed clients presented eight or more. This result demonstrates a probable relationship between poverty and the worries of PHIV.
The ipsative method of coding  psychological problems resulted in 109 binary variables (problem mentioned=1, problem not mentioned=0). Factor analysis was used to reduce these psychological variables to a more manageable number by structuring the data. A principal components factor analysis with a Varimax rotation (SAS 6.12) reduced the psychological items to 15 factors [25,28]. A variety of alternative factor solutions were tested but each resulted in having variables loading in several factors at a time.
The selected factor solution produced a set of 15 factors that explained 42.60% of the variance in the raw counselling data. Table 2 shows the variables that loaded in each factor and the variance resolved by the factors. A total of 73 variables were included in the factors; 36 were excluded as they did not load in any factor. All factor solutions were unique; no variables loaded in more than one factor. The 15 factors included between three and nine variables, each with loadings ranging between 0.42 and 0.79. The variables comprising each factor were examined to determine a common theme. The authors sub-divided factors according to their apparent underlying structure, as follows. Factor 1: Communication and counselling, Factor 2: Rejection, Factor 3: Secrecy regarding HIV status, Factor 4: Meeting family responsibilities, Factor 5: Economic concerns about basic needs, Factor 6: Denial, Factor 7: Blame spouse/partner for HIV infection, Factor 8: Concern for my children‚s future, Factor 9: I am burden to others, Factor 10: Responding to treatment, Factor 11: Protecting my children from HIV infection, Factor 12: Self-hate, Factor 13: Gaining access to treatment, Factor 14: ‚Was I bewitched?‚, Factor 15: Problems with spouse or partner.
Of the 15 factors, Factor 14: ‚Was I bewitched?‚, was the most difficult to interpret. It consisted of three variables (i.e., client is confused about having contracted the disease and feels he/she may be bewitched, client believes that he/she is bewitched, client‚s family is going to send client to village to await death). The label was derived from the concept of being bewitched appearing in two of the three variables. The third variable tended to accompany the bewitched variables and may signal the feeling of futility of families when dealing with the PHIV family member in an urban setting away from their traditional location.
All but one of the 15 factors only showed positive relationships among the variables. Factor 13 had both positive and negative loadings; family support variables had a negative relationship with treatment concern variables suggesting that clients with more family support tend to have fewer concerns about receiving treatment for physical problems.
In conclusion, results suggest that the seemly wide variety of psychological problems can be simplified, as revealed through factor analysis. Although some factors may not appear to be psychological in nature, the data nevertheless represent the range of stress-producing problems that clients brought to counselling.
Predicting clients‚ psychosocial needs for HIV counselling
Multiple regression analysis was used to assess the relationships between clients‚ demographic characteristics and the 15 psychological factors. The regressor variables in Table 3 were selected using stepwise regression methods. Then for each factor variable, a regression model was devised in which all the independent variables with significant effects were included in the model simultaneously. Significant results were produced from 11 of the 15 models (Table 3). However, only three models explained >11% of the variance (R2) and are worth considering: Communication and counselling (R2=19.85%), Economic concerns about basic needs (R2=31.39%), and Blame spouse/partner for HIV infection (R2=21.04%).
Independent variables that yielded significant results for the first model included Living with one‚s spouse or partner, Client‚s level of education, and Client‚s knowledge of his/her HIV status prior to this study. Results suggest that issues of communication and the perceived need for counselling tend to be associated with individuals who live with their spouse or partner, have little education, and recently learned of their HIV status. Significant variables in the second model include Employment status, Level of education, Wealth, and an interaction term, HIV status×Wealth. This result suggests that economic concerns tend to be counselling issues for clients who are unemployed, have little education, have little accumulated wealth, or are HIV-positive and also have some accumulated wealth. The third model showing considerable variance included the following independent variables: Number of children and HIV-positive status. This result suggests that clients who blame their partner for their HIV infection tend to be married and have children. There may be a tendency for the HIV-negative partner to blame the HIV-positive partner for having become infected.
Of the significant independent variables, only Employment status was a powerful independent predictor explaining 19% of the variance in Economic concerns about basic needs. Two other covariates were somewhat less important independent predictors, each for a different counselling need. Clients living with their partners accounted for 11% of the variance in the Communication and counselling factor. Client‚s number of children explained 12% of the variance in Blame spouse/partner for HIV infection. Of the remaining independent variables none was an important predictor of clients‚ counselling needs.
The models for eight other factor variables explained small amounts of variance (i.e., <11%, see Table 3): Rejection, Secrecy, ‚I am a burden to others‚, Responding to treatment, Protecting my children from HIV infection, Gaining access to treatment, ‚Was I bewitched?‚, and Problems with spouse/partner. Independent variables associated with each of these factor variables are presented in Table 3. They are not discussed because of the small amount of variance explained by each one. Despite their statistical significance, not enough variance is explained by each to have sufficient public health relevance. The remaining four factor variables had no statistically significant regression models: Meeting my family responsibilities, Denial, Concern for my children‚s future, and Self-hate.
The conclusion of this section is that whereas demographic variables were statistically significant in explaining psychological factors presented in counselling, their effects covered only a small amount of variance (between 3.3% and 31.4% of the variance). Therefore, counsellors may not rely exclusively on the results of this analysis for anticipating the psychological problems of clients. Additional studies are needed to investigate other variables that may cover additional unexplained variance.
The psychological issues, reported by a cohort of 168 Kenyan PHIVs seeking follow-up counselling from KAS PPCs, revealed 15 categories of problems. KAS clients systematically share this set of 15 factors in common. The factors do not represent all psychosocial problems of PHIV, but only those issues clients presented in counselling. However, these are the issues that HIV counsellors need to be able to handle. Although statistical analysis revealed these 15 factors, we have further simplified them, using normative principles, into seven broad problem areas: financial worries (Meeting my family responsibilities, Economic concerns about basic needs), children (Concern for children‚s future, Protecting my children from HIV infection), dilemma in personal relationships (Hate and blame partner, Problems with spouse/partner), relations with others (Communication and Counselling, Rejection, Secrecy, and ‚I am a burden to others‚), self image (Self-hate, ‚I am a burden to others‚), treatment (Responding to treatment concerns, Gaining access to treatment), and denial (Denial, ‚Was I bewitched?‚).
Clients‚ counselling problems primarily concern relations with family and friends. In the first four groupings (financial worries, children, dilemma in personal relationships, and relations with others), clients‚ problems are social, or focused on other people, rather than themselves. This focus on relations with others has been documented elsewhere  where the authors noted that HIV-positive and HIV-negative individuals placed an emphasis on relationships or social problems rather than on medically-oriented problems.
The remaining three groupings are directed inward (self-image, treatment and denial). Attitudes that are prevalent in these groupings include anger, guilt, confusion and concern for one‚s health; these are personal issues. However, they also have a social dimension. Hate and blame for one‚s self could be a product of negative family attitudes or social stigma. Denial of one‚s HIV status could result from the pressure of social stigma or from a misunderstanding of the disease and transmission mechanisms.
AIDS impacts a community tied together by social networks and a common lifestyle. It disconnects the HIV-positive person from family, friends and from normal social institutions (e.g., job, healthcare). Stigma and fear can reinforce the sense of isolation and guilt among PHIV , which is conveyed in several psychological factors: Rejection, Secrecy, ‚I am a burden to others‚, Denial, and Problems with spouse/partner. These issues signal a need to integrate counselling with community support systems to break down social stigma [29,30] and to help PHIV lead socially productive lives . Further, the financial concerns and treatment issues mentioned by clients suggest a need for community organized material help. Support groups could help restore a sense of connectedness and therefore improve coping responses of PHIV and their families [29,30].
A second observation is that families of PHIV are just as likely to be supportive as unsupportive and, therefore, cannot be relied upon to provide care for them. Clients reported positive support from family members as well as a lack of sympathy from their family. Families also had a tendency to blame the clients for their infection. In Africa, the family usually takes on the responsibility for providing support and care for sick family members. However, fear and discrimination associated with HIV/AIDS in the community discourages this help . Even trusted family members may choose not to provide care or support for fear of becoming infected . Nevertheless, other accounts show that AIDS patients are increasingly accepted by relatives and partners and are increasingly cared for at home [34-37]. A recent review of TASO‚s work in Uganda suggests that community-based counselling has helped families to cope with PHIV, with the majority of clients revealing their HIV-positive status .
The disparity in family support may explain why education and counselling (Factor: Communication and counselling) for the family was a priority need of clients in the study. Family counselling could support the restoration of the natural and traditional care-giving role of the family to reduce the client‚s sense of isolation and alienation. It could also help the client establish a setting in which to reveal his/her serostatus. Improving relations with the family could help alleviate the client‚s concerns about the care of their children after his/her death. Family counselling can extend support and care to members of the client‚s family who may be grieving the loss or anticipated loss of a loved one . Their negative attitudes toward the client may be due to their difficulty in coping with this.
A third observation is that counsellors may use, albeit with reservation, the demographic characteristics presented in Table 3 for anticipating any of the 15 categories of psychological problems. Multiple regression identified weak but statistically significant independent predictors for 11 of the 15 psychological factors. Not much is yet known about the antecedents to these problems and what are accurate predictors of their presentation in counselling sessions. Whereas the regression results could be used to initially orient a counsellor to a client, a counsellor should not expect a client to present any category of psychological factor because s/he has a particular demographic profile. However, three independent variables produced stronger regression coefficients: employment status, living with one‚s partner, and number of children. Each explained between 11%-19% of the variance in a single factor, namely, Economic concerns about basic needs, Communication and counselling, and Blaming spouse/partner for HIV infection, respectively. Counsellors should nevertheless review the statistically significant variables in Table 3 with a view to expanding the models to include additional variables that in future studies might explain more variance.
Caring for PHIV in Africa should be discussed in the context of limited medical facilities and personnel, incomplete and incorrect knowledge of HIV among some care providers, and the lack of social support systems [40,41]. Meeting the psychosocial needs of PHIV generally falls on insufficiently trained paraprofessional counsellors . As the number of HIV cases increases, demand for the services of trained PPCs will also increase . However, the lack of information on the psychosocial needs of African PHIV impedes the development of training systems for PPCs.
This study attempts to establish an empirical basis for designing a training system for PPCs in East Africa. In practical terms, each of the 15 factors presented in this study can be treated as a separate module in a curriculum to train PPCs to address the problems that PHIV bring to counselling. This study was initiated because KAS counsellors felt under-prepared to handle the problems that their clients presented. Previous research indicates that HIV counsellors in the UK complained that their training had not equipped them to address their clients‚ needs . Counsellors were uncomfortable in their work in part because they were not prepared to discuss certain topics (e.g., suicide, death, dying, and bereavement) and in part because they were uncertain how their clients would react in counselling.
In a study in South Africa, hospital counsellors reported that they felt unprepared to handle several problems they encountered during HIV counselling: client‚s denial of a positive HIV test result; anger at being infected; concerns about their jobs, children and fertility; traditional or superstitious beliefs about the disease; and the need to encourage HIV-positive clients to reveal their status to their sexual partners . The counsellors requested additional information on how to best deal with these problem areas. Each of these areas is similar to a problem reported by clients in the current study, suggesting that focused training for PPCs could accommodate the needs of clients and counsellors alike.
Counselling training for paraprofessionals varies widely in content. Topics range from basic facts about HIV antibody testing, HIV transmission and prevention, to interpersonal and facilitation skills, to decisions about notifying sexual partners, to coping strategies for death, dying and bereavement [1,3,16]. The issues are numerous and complex. A PPC would require extensive training and/or experience in order to feel comfortable in his/her role as an HIV counsellor. PPCs could be better prepared if they received focused training in areas that are relevant to the issues that tend to arise in counselling. The typology developed in this study reduces the myriad of possible topics to a manageable and teachable set that is relevant to the experiences of PHIV and counsellors.
Whereas other studies have documented the multiple psychological and physical problems associated with HIV infection [30,39], the present study shows how training of paraprofessionals can be based on the diverse problems that clients present in counselling. This could help to reduce the uncertainty of what counsellors can expect to encounter in counselling. This may produce counsellors who are confident in their work and capable of assisting PHIV to cope with their infection.
Several factors limit the generalization of these results. First, the study is based on an assessment of one cohort of clients; findings may not be representative of other PHIV or people affected by HIV. Second, the data are cross-sectional which precludes an analysis of the progression of client attitudes over time. Given the uncertainty surrounding the onset and outcome of HIV and the symptoms of AIDS, client distress levels and coping abilities may change over the course of the infection . The original intent of the study design was to track the psychological problems clients brought to counsellors over time, and to associate these with clinically diagnosed physical problems. However, only 20 of 168 clients returned to more than one follow-up counselling session; at most they attended three sessions. Third, the study analyses psychological data reported only in counselling sessions. Clients may have had additional latent, unreported psychological problems that they might have revealed over several sessions, or through psychological testing. Despite this limitation, this study does depict the problems that clients chose to discuss with PPCs. Therefore the results do reflect problems actually presented in counselling sessions and may be pertinent for training counsellors. The findings of this study offer an inventory of problems that clients themselves identified as priorities for follow-up counselling. This study is an initial exploration of the psychosocial needs reported by HIV clients and may be helpful to PPCs working to improve the psychosocial counselling available to PHIV and people affected by HIV. Additional research is needed to validate the 15 factors we studied in other settings, and to determine the long-term psychosocial needs of PHIV and how these relate to their health status and disease stage.
The authors gratefully acknowledge all of our colleagues at the Kenya AIDS Society for their participation in this study. In particular, we thank Joe Muriuki (KAS President) for his support. Similarly, we thank colleagues at African Medical and Research Foundation who supported our efforts at the time of this study, in particular, Dr Al Henn and the Research and Evaluation Unit. We also gratefully acknowledge Dr Mandy Rose‚s important suggestions concerning the impact that the 15 factors could have on counsellor training.