The process developed by the Guideline Review Committee (GRC) of WHO was strictly followed. A WHO Guideline Steering Group was formed (see Annex 1), which identified, together with the chair of the EG (see below), the areas requiring evidence synthesis.
3.1. Expert Group meeting
An EG meeting was convened by the WHO Stop TB Department from 29th to 30th January 2013 to assess all available data on bedaquiline, and with a view to developing interim policy recommendations on its use, as appropriate. The EG (Annex 2) comprised researchers, epidemiologists, end-users (clinicians and national TB programme officers), community representatives and evidence synthesis experts. The EG meeting followed a structured agenda (Annex 3) and was chaired by a clinical epidemiologist/methodologist with expertise and extensive experience in evidence synthesis and guideline development.
The overall objective of the EG meeting was to evaluate the added benefit of bedaquiline for the treatment of MDR-TB and, if appropriate, to provide recommendations to WHO for interim guidance to countries on its use in conjunction with other second-line drugs used in MDR-TB treatment.
The specific objectives were:
To evaluate the efficacy and safety of bedaquiline in addition to currently WHO recommended MDR-TB treatment regimens.
To evaluate the balance between harms and benefits of the drug, its potential cost-effectiveness, patient and provider preferences and concerns, and the feasibility of introducing the drug into MDR-TB programmes.
To provide, as appropriate, recommendations on the use of the drug as part of WHO-recommended MDR-TB treatment regimens, including attention to concerns/constraints relevant to the use of a new drug for which Phase III clinical trial data are not yet available.
3.2. Management of conflicts of interest
WHO policies on conflicts of interest were developed and applied in consultation with the WHO Legal Department. Every member of the EG was asked to complete the WHO Declaration of Interest (DOI) form before their invitation was confirmed and data shared with them under non-disclosure agreements. All completed forms were reviewed by the WHO Guideline Steering Group in conjunction with the WHO Legal Department prior to the EG meeting. Particular attention was given to potential conflicts of interest related to the appraisal of evidence, the formulation of recommendations and the external peer review process. Particular attention was also given to assessment of financial as well as intellectual interests. In addition, individuals who were involved in clinical trials conducted by the bedaquiline manufacturer, or in any entity or committee related to the conduct of any trial conducted by the company (e.g. trial steering committee, data monitoring committee, scientific advisory board), even if not remunerated, as well as individuals who had been involved in development and testing of the new drug or other, potentially competing, drugs were not considered for inclusion in the EG.
DOI statements were summarized by the WHO/Stop TB Department (STB) secretariat at the start of the meeting. A summary is attached in Annex 4.
Technical resource consultants participated in the meeting and provided specific information on selected technical issues but were not involved in the decision-making process, or in the preparation of the actual recommendations. Observers participated only at the request of the Chair and did not contribute to the preparation of the recommendations. All participants signed a confidentiality agreement and were reminded of the need for confidentiality until the full WHO process had been concluded.
3.3. Review of evidence
Publicly available data on the pre-clinical and clinical development of the drug were assembled and reviewed to assess efficacy, safety and tolerability of the drug,10 and complemented by modelling work to assess the cost-effectiveness of implementation of the drug in MDR-TB programmes. Issues to be addressed in future research were also discussed. In addition, data on final outcomes of the pivotal proof-of-efficacy Phase II trial (that were not available at the time of US-FDA review) were provided to WHO by the manufacturer.
An independent consultant was contracted to review and synthesize all available data into a comprehensive document that was made available to all members of the EG, and prepare the draft GRADE11 evidence tables that were reviewed by the EG.
To comply with current standards for evidence assessment in formulation of policy recommendations, the GRADE system, adopted by WHO for all policy and guidelines development,12 was used. The GRADE approach, assessing both the quality of evidence and strength of recommendations, aims to provide a comprehensive and transparent approach for developing policy guidance. The GRADE process assesses the impact of a particular intervention on patient-important outcomes and the generalizability of results to the target population, taking into consideration the comparator used and whether comparison was direct or indirect.
A PICO (Population, Intervention, Comparator, Outcome) question was pre-defined in consultation with the WHO EG: “In MDR-TB patients, does the addition of bedaquiline to a background regimen based on WHO-recommendations safely improve patient outcomes?”
The following outcomes were selected by the EG for evaluation:
Cure by end of study – 120 weeks.
Serious adverse events during investigational 24 weeks treatment phase.
Mortality.
Time to culture conversion over 24 weeks.
Culture conversion at 24 weeks.
Acquired resistance to second-line drugs (fluoroquinolones, amino-glycosides and capreomycin) at 72 weeks.
In a first stage, experts evaluated the quality of evidence for each of the above outcomes according to the following criteria:
Study design: randomized trial(s), or consecutive selection of patients (observational), or selection of patients according to given reference standard (case-control).
Risk of bias or limitations in study design and execution.
Inconsistency: unexplained inconsistency in study endpoints or estimates.
Indirectness: absence of direct evidence of impact on patient-important outcomes and generalizability.
Imprecision: wide confidence intervals for treatment outcome estimates.
Other considerations: possibility of publication bias, etc.
A glossary of the GRADE terms used can be found in Annex 5.
In the second stage, as called for by GRADE, and based on the PICO question, the EG developed a recommendation and considered the strength of the recommendation (strong or conditional), based on a balance of effects (benefits weighed against harms), patient values and preferences, resources and equity. The system used to establish the strength and ranking of the recommendations involved assessing each intervention on the basis of: (1) desirable and undesirable effects; (2) quality of available evidence; (3) values and preferences related to interventions in different settings; and (4) cost options for different epidemiological settings.
3.4. Decision-making during the Expert Group meeting
The EG meeting was chaired by a recognized methodologist/evidence synthesis expert. Decisions were based on consensus (preferred option). Only exceptionally, when a consensus could not be achieved among members, did the EG proceed to a vote (with simple majority rule) – this was resorted to in only one instance (see page 27). Concerns and opinions of EG members were noted and included in the final meeting report. The detailed meeting report was prepared by the WHO Secretariat Steering Group and was revised based on input and sign-off by all EG members.
3.5. External peer review
An External Review Panel (ERP) independently reviewed the draft interim guidance prepared by the WHO Guideline Steering Group on the basis of the recommendations by the EG. The ERP was composed of 10 reviewers external to the EG, including content experts, end-users from high TB and HIV burden countries, and representatives from the WHO Strategic and Technical Advisory Group for TB (STAG-TB). The list of members of the ERP can be found in Annex 6. Comments made by the members of the ERP were reflected in the final version of the guidance document.
3.6. Financial support
Financial support for the EG meeting and related analyses were provided under the USAID consolidated grant to the WHO Stop TB Department (project number: US 2012 0392). The US Centers for Disease Control (CDC) completed the evaluation of sputum culture conversion as a surrogate marker of MDR-TB treatment outcome (work carried out by Ekaterina Kurbatova and colleagues).
