Developing WHO guidelines with pragmatic, structured, evidence-based processes: A case study.

Many guidelines, including those produced by the World Health Organisation (WHO), have failed to adhere to rigorous methodological standards. Operational examples of guideline development processes may provide important lessons learned to improve the rigour and quality of future guidelines. To this end, this paper describes the process of developing WHO guidelines on prevention and care interventions for adults and adolescents living with HIV. Using a pragmatic, structured, evidence-based approach, we created an organising committee, identified topics, conducted systematic reviews, identified experts and distributed evidence summaries. Subsequently, 55 global HIV experts drafted and anonymously submitted guideline statements at the beginning of a conference. During the conference, participants voted on statements using scales evaluating appropriateness of the statements, strength of recommendation and level of evidence. After review of voting results, open discussion, re-voting and refinement of statements, a draft version of the guidelines was completed. A post-conference writing team refined the guidelines based on pre-determined guideline writing principles and incorporated external comments into a final document. Successes and challenges of the guideline development process were identified and are used to highlight current issues and debates in developing guidelines with a focus on implications for future guideline development at WHO.

Trends and predictors of quality of life among HIV-infected adults taking highly active antiretroviral therapy in rural Uganda.

We examined trends and predictors of quality of life (QOL) over 12 months among a prospective cohort of 947 HIV-1-infected adults initiating highly active antiretroviral therapy (HAART) between May 2003 and May 2004 in rural Uganda. Participants provided clinical, demographic and psychosocial data at baseline and every three months thereafter. Outcome measures included physical and mental health summary scores based on the Medical Outcomes Study-HIV Health Survey (MOS-HIV). Generalised estimating equations were used to assess magnitude of change in summary scores and factors associated with QOL. Of 710 women and 237 men enrolled, the mean age was 38.7 years and mean baseline CD4 cell count was 124.1 cells/microL. At enrollment, physical and mental health summary scores were 39.2 and 40, respectively. By 12 months of HAART, scores increased by 11.2 points (p <0.001) and 7.4 points (p <0.001), respectively. For both scores, most gains were achieved by the third month of therapy. While several clinical, psychosocial and sociodemographic factors predicted QOL at HAART initiation, financial dependence on others was the only remaining predictor after controlling for time on HAART. Interventions to enhance the economic and employment opportunities of patients taking HAART in rural Africa may help maximise gains in QOL.

Living with discordance: knowledge, challenges, and prevention strategies of HIV-discordant couples in Uganda.

Prevalence of HIV-discordance among couples in sub-Saharan Africa is high. Negative partners are at high risk of HIV infection but few HIV/AIDS service providers have developed effective counseling messages for HIV-discordant couples. To identify clients' explanations for discordance, challenges, and prevention strategies, 24 in-depth interviews and 4 focus group discussions were conducted with 32 female and 35 male members of HIV-discordant couples who sought HIV voluntary counseling and testing (VCT) in Uganda. In addition, counselor explanations for discordance were gathered from 62 counselor trainers during 3 interactive workshops. Misconceptions about discordance were widespread among clients and counselors. Common explanations included: the concept of a hidden infection not detectable by HIV tests, belief in immunity, the thought that gentle sex protected HIV-negative partners, and belief in protection by God. Such explanations for discordance reinforced denial of HIV risk for the negative partner within discordant couples and potentially increased transmission risk. Couples identified negotiation of sexual relations as their most formidable challenge. Prevention strategies included condom use, abstinence and separation of beds, contractual agreements for outside sexual partners, and relationship cessation. Discordant couples represent a critical risk group and improved counseling protocols that clearly explain discordance, emphasize high risk of transmission, and support risk reduction are urgently needed.

Leaving sex work: barriers, facilitating factors and consequences for female sex workers in northern Thailand.

Factors facilitating or inhibiting women's ability to leave sex work are still poorly characterized, and little is known about women's lives after they leave the profession. This paper presents findings from a qualitative study about factors affecting women's ability to leave sex work and influencing their lives after leaving. We interviewed 42 current and former female sex workers (FSWs) drawn from a cohort study of 500 FSWs in northern Thailand. All but one of the participants had quit sex work at least once. The majority experienced one or more quit-re-entry-quit cycles. Women's ability and decisions to leave sex work were determined primarily by four factors: economic situation, relationship with a steady partner, attitudes towards sex work and HIV/AIDS experience. Economic concerns, ranging from survival needs to materialistic desires, had the strongest influence. Most women perceived their risk for HIV infection to be lower after leaving sex work, but three of the 17 HIV-infected women acquired infection after having left, presumably from their steady partners. Prevention efforts should guide women as they transition out of commercial sex work. Interventions aimed at assisting women wanting to leave sex work need to address the role of economic factors.

High prevalence and incidence of sexually transmitted diseases in urban adolescent females despite moderate risk behaviors.

To better understand the prevalence, incidence, and risk factors for sexually transmitted diseases (STDs) among female adolescents, a prospective 6-month cohort study was conducted at four teen clinics in a southeastern city. At enrollment, 260 (40%) of 650 sexually active females ages 14-19 years had an STD: chlamydia, 27%; herpes simplex virus type 2 (HSV-2), 14%; gonorrhea, 6%; trichomoniasis, 3%; and hepatitis B, 2%. At follow-up, 112 (23%) of 501 participants had an incident infection: chlamydia, 18%; HSV-2, 4%; gonorrhea, 4%; and trichomoniasis, 3%. At either enrollment or follow-up, 53% had >/=1 STD; of those with 1 lifetime partner, 30% had an STD. Having a new partner (odds ratio [OR], 2.2; 95% confidence interval [CI], 1. 1-4.2) or friends who sell cocaine (OR, 1.6; CI, 1.0-2.6) was independently associated with incident infection. STD incidence and prevalence were extremely high in this population, even in teenagers with only 1 lifetime partner. Individual risk behaviors appeared less important for STD risk than population factors.

HIV-1 seroprevalence among childbearing women in northern Thailand: monitoring a rapidly evolving epidemic.

OBJECTIVES: To describe trends in prevalence of HIV-1 infection among women giving birth at Chiang Rai Hospital (CRH) and to assess risk factors associated with HIV infection in this population. DESIGN: Analysis of hospital registry data for all deliveries at CRH from 1990 to mid-1997. METHODS: From 1990 to mid-1997, women giving birth at CRH were tested for HIV-1 infection using enzyme immunoassay (EIA); positive sera were confirmed using a different manufacturer's EIA. Demographic and clinical data were abstracted from delivery-ward log books. RESULTS: Data from 40723 deliveries indicated that overall HIV-1 seroprevalence increased sharply, from 1.3% in 1990 to a peak of 6.4% in 1994, and then declined to 4.6% in the first 6 months of 1997. Prevalence was highest, at 7.0%, among young (age < or = 24 years) primigravidas, compared with 2.4% among older (age > or = 25 years) multigravidas. When we controlled for age, prevalence declined 40% from 1994 to 1997 among young primigravidas (95% confidence interval for percentage reduction, 16-57). Amongst older multigravid women, prevalence was consistently lower but increased steadily from 2.7% in 1994 to 3.4% in 1997. CONCLUSIONS: A rapid rise in HIV prevalence in childbearing women was followed by a sharp decline among young primigravidas. In each year, the prevalence was highest among young primigravidas. They may be the best subgroup of pregnant women for monitoring HIV epidemic trends, but they also represent a challenging prevention priority that will require its own targeted interventions.