Developing the 2013 WHO consolidated antiretroviral guidelines

Abstract

Introduction The 2013 ‘Consolidated guidelines on the use of antiretroviral (ARV) drugs for treating and preventing HIV infection’ [1], released in July 2013, are the latest and most comprehensive of a series of important guidelines on antiretroviral therapy (ART) over the last decade from the World Health Organization (WHO). They were developed in response to important advances in the science and practice of HIV care since publication of the 2010 WHO guidance for adults and adolescents [2], pregnant women [3] and children [4]. This includes evolving evidence on the preventive and individual clinical benefits of earlier ART, innovations in service delivery such as the progressive decentralization of HIV testing and care, and the more widespread availability and affordability of once-daily fixed-dose combinations (FDCs) ART regimens [5]. In this special supplement of AIDS, we present a series of thirteen articles and five commentaries covering key aspects of the consolidated guidelines: the process, evidence base, recommendations and guidelines implementation. In this first article, we describe the WHO process and methodology of developing these guidelines. This is followed by seven selected systematic reviews [6–12] that provided the evidence base for specific recommendations. They are presented under the section heading of the relevant guidelines population or topic: Adults and adolescents (when to start ART) [6]; Pregnant women (safety of efavirenz in pregnant and breastfeeding women) [7]; Children (what ART regimen to use in children under 3 years) [8]; ART monitoring (how to monitor treatment response in adults and children) [9,10]; Service delivery (evaluation of effectiveness of service delivery innovations of decentralization and integration) [11]; and different strategies to improve treatment adherence [12]. The systematic reviews in each section are prefaced by commentaries written by the co-chairs and/or members of the Guideline Development Groups (GDGs) that highlight key recommendations and their rationale, and provide additional context for guidance [13–17]. The International HIV/AIDS Alliance and the Global Network of People Living with HIV (GNP+) report on the findings (and lessons learnt) from their consultation on community values and preferences that also informed many of the recommendations [18]. The three concluding articles all address different aspects of the critical phase of country-level adaptation and implementation of the guidelines. This includes what is known about the current status of national adoption of the recommendations in WHO ARV guidelines [19], projections of the global impact and cost of implementation of new recommendations [20] and country-level implications of implementing these guidelines for policy makers, such as diversification of service delivery models, generation and use of data, healthcare financing, human resource capacity and supply chains for drugs and diagnostics [21]. This supplement is not intended to be an exhaustive collation of all the evidence that informed the consolidated guidelines. Several of the commissioned systematic reviews as well as modelling studies have already been published in the peer reviewed literature or are in development [22–28], and a companion AIDS supplement published in January 2014 has already collated other modelling work that contributed to the guidelines process [29]. A comprehensive summary of all supporting evidence is provided as Web Annexes in the guidelines website (http://www.who.int/hiv/pub/guidelines/arv2013/annexes/en/index.html). Distinctive features of the 2013 WHO consolidated guidelines The 2013 consolidated guidelines were distinctive from previous WHO ART guidelines, or other international ART guidelines in several ways. Providing guidance across the entire continuum of HIV care More than fifty new recommendations are provided across the cascade of HIV care, from HIV testing and diagnosis, linkage to care, using ART for prevention, ART initiation, monitoring for treatment failure and ART toxicity, and retention in care. This comprehensive approach responds to the needs of programme managers who are responsible for delivery of care across all of these steps. Expanding guidance: clinical, operational and programmatic In addition to the usual clinical recommendations, there is operational guidance on how to improve delivery of HIV care (with recommendations on task shifting, decentralization, integration and adherence and improving retention in care). The guidelines also provide a framework and tools for programme managers to consider in prioritizing implementation of recommendations according to their national context, including HIV epidemiology, levels of ART uptake, health workforce capacity and available financial resources. This integrated approach better reflects the complex interplay between clinical recommendations and implementation at facility and programme level. Addressing all ages and populations Instead of separate ART guidelines for adults, pregnant women, adolescents and children, as in previous years, guidance is provided across all age groups and populations of adults, pregnant and breastfeeding women, adolescents, and children, as well as those coinfected with tuberculosis (TB), hepatitis B and/or hepatitis C. This enables a more harmonized approach to ART regimen choice, simplifying both the role of healthcare providers, and procurement and supply chain management. Target audience: programme managers in low-income and middle-income countries As for other WHO guidelines, the primary target users are country policy makers and programme managers responsible for national and regional policy and planning decisions on ART scale-up, and in settings with limited resources. Incorporating the key guiding principles of the public health approach and health equity in ART scale-up The 2013 guidelines, as with previous WHO ART guidance, are based on a public health approach to ART scale-up that promotes simplified and standardized approaches to treatment and monitoring that facilitates the widest possible access to high-quality care at the population level [30]. This in turn requires innovations in service delivery to maximize the efficiency of HIV programmes such as through integration of HIV care with other services (e.g. maternal and child health, TB and drug dependence), improved treatment adherence and retention in care, harmonized ART regimens and more affordable diagnostics. Another key guiding principle underpinning implementation of the guidelines is promotion of human rights and health equity in national HIV policies and programmes, so that expanded access is fair and equitable; priority for ART initiation is given to those most in need; and care is provided in a supportive and responsive environment, free of stigma and discrimination. Linking new recommendations with existing guidance New recommendations on the use of ART for treatment and prevention have been harmonized with relevant selected recommendations from existing WHO guidance on HIV testing, prevention and management of coinfections. WHO guidelines development process and the GRADE approach The revision process for the 2013 guidelines was initiated in early 2012, and conducted in accordance with procedures established by the WHO Guidelines Review Committee, to ensure that WHO guidelines are developed using a transparent, evidence-based, decision-making process [31]. Since 2008, WHO has used the internationally agreed standard of the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluation) to assess the quality of a body of evidence, formulate recommendations and rate their strength [32–38] (GRADE working group: http://www.gradeworkingroup.org). Quality of evidence and strength of recommendation GRADE classifies the quality of evidence into one of four levels: high, moderate, low and very low. The rating of quality of evidence from randomized controlled trials starts as high, but may be decreased because of risk of bias, inconsistency in results across studies, indirectness of evidence, imprecision and publication bias [34–36]. The rating of evidence based on observational studies starts as low, but may be increased if the magnitude of the treatment effect is very large, there is evidence of a dose–response relationship, or if residual biases would underestimate the effect size [37]. In addition to the quality of the evidence, other considerations in formulating recommendations and rating their strength include the overall balance of benefits and harms to the individual and at a population level, community values and preferences, resource use, cost-effectiveness, feasibility and constraints to implementation in multiple settings, equity and human rights implications [38] (Table 1).Table 1: Key domains considered in formulating recommendations and determining their strength (strong or conditional).GRADE also classifies strength of recommendations as either ‘strong’ or ‘conditional’ [38]. A strong recommendation is one for which the GDG was confident that the desirable effects of the recommendation outweigh the undesirable effects, while a conditional recommendation is used when it is concluded that the desirable effects probably outweigh the undesirable effects, but there is uncertainty about these trade-offs. The higher the quality of evidence, the more likely a strong recommendation can be made. A conditional recommendation is more likely when high-quality evidence is absent, the estimates of effect are imprecise, there is uncertainty or variability in how individuals value the outcomes, or the benefits are either small or not considered worth the costs. The implications of a conditional recommendation are that, although most people or settings would adopt the recommendation, some would do so only under certain conditions. The following sources of evidence and supporting material were used to inform the development of the new recommendations. Systematic reviews Systematic reviews were commissioned on forty-six topics across the continuum of HIV care, including nine on when to start ART; eleven on what ART to start; four on monitoring the response to ART; six on monitoring toxicity; and eleven on operational aspects of service delivery. The questions were framed using the Population, Intervention, Comparison and Outcome (PICO) format [33], and outsourced to seven different research teams and organizations through a process of competitive tendering. These groups then developed search protocols and conducted reviews of the available scientific evidence. A standardized GRADE evidence table was used to present quantitative summaries of the evidence and assessment of its quality for each PICO question by outcome [32]. The full list of review questions, search protocols, GRADE tables and evidence summaries for each topic are available at http://www.who.int/hiv/pub/guidelines/arv2013/annexes/en/index.html. Seven systematic reviews are included in this supplement [6–12], and others have been published elsewhere [22–28], or are in development. Consultations on community values and preferences An assessment of community values and preferences on key ARV guideline topics was coordinated by the International HIV/AIDS Alliance and the Global Network of People Living with HIV (GNP+) through both an online e-survey (n = 1088), and moderated e-forum discussions with civil society networks (n = 955) [18,39] in six languages (Arabic, Chinese, English, French, Russian and Spanish). Key topics included community preferences regarding possible recommendations (e.g. which ART regimens to use and when to initiate ART for adults, adolescents, pregnant and breastfeeding women, and children), as well as ART service delivery considerations. Four focus group discussions were also held in Uganda and Malawi on the experiences of pregnant women with lifelong ART (option B+). Finally, two e-surveys of health workers caring for HIV-infected adults (n = 98) and children (n = 342) were undertaken on similar topics covered in the community consultation through clinical networks of eight global implementing partner organizations. In addition to the report in this supplement [18], a full consultation document is available [39]. Mathematical modelling of impact and cost effectiveness We commissioned two key modelling projects on health impact (measured using disability-adjusted-life-years (DALYs)) and cost-effectiveness from the HIV Modelling Consortium (http://www.hivmodelling.org) to support the 2013 guidelines. The first examined various HIV testing strategies and criteria for ART initiation in different populations (adults, pregnant women and HIV serodiscordant couples) on the basis of data from countries representative of different HIV epidemic types (generalized, concentrated and mixed) and level of ART coverage [26]. A second project examined different strategies for monitoring treatment response (clinical, CD4+ T-cell count and viral load) and switching to second-line ART [27]. A key strength of these analyses was their use of multiple independently developed models to compare different scenarios, for which there were limited data in the literature. Additional modelling work commissioned included a causal modelling analysis of the impact of starting ART at different ages in children, based on data from the IeDEA South African collaboration [28]. A recent article has highlighted some of the challenges in using modelling data in guidelines development, including the lack of standardized criteria for rating the quality of modelling, and clarity on the positioning of modelling within the GRADE framework for decision-making [40]. It concludes with some key considerations to guide the future use of modelling in guidelines development. Feasibility surveys Reports were commissioned on country implementation experiences, including adoption of lifelong ART in pregnant and breastfeeding women (option B+) in Malawi; introducing tenofovir (TDF) in first-line ART regimens in Zambia; phasing out stavudine (d4T) in Zimbabwe; and scaling up viral load monitoring in Médecins Sans Frontières programmes in southern Africa. These are summarized in the consolidated guidelines web annexes (http://www.who.int/hiv/pub/guidelines/arv2013/annexes/en/index.html). Impact assessment of implementation of recommendations An impact assessment was undertaken using the AIDS Impact Model (AIM) and Goals model within the Spectrum modelling system [41] to estimate the number of adults and children newly eligible for ART, based on the new treatment recommendations [20]. It also examined the cost and impact that would result if ART coverage expanded to 80% of those eligible for ART. End-user survey to inform guidelines presentation and dissemination strategies An additional preparatory activity was the conduct of an internet-based survey of country-level end-users of recent WHO HIV-related guidelines. This was undertaken to understand better how WHO ART guidelines are used, and identify areas for improvement in format, presentation and dissemination of the new guidelines. The survey targeted WHO National Program Officers and Ministry of Health HIV focal persons, and was administered in English, French, Russian and Spanish between June and September 2012. Overall, there were 78 respondents from 44 countries across all regions (28% South East Asia and Western Pacific, 26% from Central and Eastern Europe, 10% Latin America and Caribbean, 28% Sub-Saharan Africa, 8% Middle East). All respondents had used at least one of twelve WHO HIV guidelines, and the majority (75%) had used them primarily in the development of national guidelines. Although the response rate was limited, and not fully representative of all end-users, there was a good geographic spread of respondents, and several consistent observations emerged. There was a broad agreement that the most critical guideline components were clearly stated recommendations with brief evidence summaries and a clear rationale supporting the recommendation (with inclusion of GRADE tables only as part of web annexes). The value of best practice examples from a wide range of different settings to support implementation was also highlighted. Specific suggestions to enhance readability included reduced length, larger font size, and greater use of colour, summary tables and algorithms. Accessibility and user engagement in dissemination of new WHO HIV guidelines were highlighted as critical factors influencing effective country-level adaptation and implementation. Specific activities to facilitate regional and country-level dissemination of the guidelines activities were in-country workshops and webinars; the availability of guidelines in all UN languages, particularly Arabic, Chinese and Russian; the continued need for printed in addition to electronic versions of the guidelines; and an improved notification system for new guidelines using e-mail together with conference and website announcements. Guideline development groups (GDGs) and process of formulating recommendations The development of recommendations was undertaken by four separate, external technical and and Service and but as a process to ensure an integrated guidelines document for adults, pregnant and breastfeeding women and There were more than GDG members across the four HIV country HIV programme guideline from or other development and of civil society and/or networks of people with HIV on the basis of four technical and regional previous with guidelines We a balance of by and All external members of the and external peer review group WHO of that included in and research support and financial There was also a at the GDG of by members within or clinical trials on either the of ART, or of specific ART Overall, the WHO and of each GDG were that there had been a of and that from the Four GDG were held in between and January The decision-making process and of recommendations first a critical review of the evidence based on systematic reviews of randomized clinical trials observational It also considered of the overall balance of benefits and harms to the individual and at a population level, community values and health preferences, resource use, cost-effectiveness, feasibility and constraints to implementation in multiple settings, and of equity and human The both the of the recommendations and the rating of its strength (strong or All decisions were by and on the recommendations, including their strength if the to be to the recommendations. were through e-mail and recommendations and of of the guidelines were to GDG and a full of the guidelines was to GDG members and peer for A group including of the four was held in 2013 to ensure and of recommendations across the guidelines. Key strength of recommendations and quality of evidence In July 2013, the guidelines were at the International AIDS conference held in and as a printed and electronic including a policy brief in seven languages (Arabic, Chinese, English, French, with an additional web that includes and all supporting and evidence as There were a of new recommendations in the 2013 WHO ‘Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection’ were recommendations on adults pregnant and on were on service delivery and four on HIV The most important new clinical recommendations were earlier ART initiation, starting ART in all adults with a CD4+ T-cell count of or prioritizing those with clinical or a CD4+ count than ART initiation of CD4+ count in pregnant and breastfeeding women, children under of HIV-infected in serodiscordant and those coinfected with or hepatitis B a first-line ART regimen of or efavirenz as a once-daily fixed-dose for adults, pregnant women and children 3 and and the use of viral load testing as the approach to monitoring ART response and treatment There were four recommendations on of testing, and also recommendations on improving the efficiency of HIV through ART delivery to primary healthcare and community ART services within child health and other to address in health and strategies to improve retention in care, and adherence to ART. the strength of recommendation and quality of evidence for all recommendations, and then according to population and of the recommendations were as and based on low or very low quality of evidence. The recommendations were of which were based on low and on very low quality evidence. were according to population and All of the service delivery and HIV testing recommendations were as with of the and of the recommendations. all strong service recommendations were based on low or very low quality evidence, this was for the twelve strong recommendations, and only of the strong recommendations in The evidence base in HIV care and for service delivery is well and these were as priority areas for operational and implementation research the guidelines More there are key challenges in consistent adherence to the GRADE guidance on rating of recommendations as strong than There is also a need to address a and some guideline group members that a conditional recommendation may not be and by strong recommendations are based on low quality evidence, it is critical that a clear rationale is A recent of recommendations from different WHO guidelines that had used the GRADE approach also that strong recommendations based on low or very low quality evidence were of strong so this is not specific to HIV care of strength of recommendations and rating of quality of evidence of recommendations in 2013 consolidated dissemination and the of the guidelines, WHO and regional have with national of health and in-country to support national and adaptation through a series of regional dissemination workshops South and held between July and The consolidated guidelines be reviewed and two as new evidence and practice in the use of ART. In there be through technical and programmatic guidance, with one in early 2014 on early diagnosis, scale-up of viral load monitoring and drug and in July 2014 on management of important including HIV-related and hepatitis use of linkage and retention in care, and community ART An science was held with key in and of key research in the 2013 consolidated guidelines and the research to inform development of future WHO also consolidated guidelines in 2014 in two other Key and which provide a of quality for HIV prevention and treatment The the of all members of the Guidelines Development and in the of the four and and for AIDS and of and Network for the of Children by AIDS, and Service for and and of South of and of Health and The WHO of the and and and We also and for with the conduct of the WHO guidelines of There are of