Topic identification, selection and prioritisation for health technology assessment (HTA) - A report to support capacity building for HTA in low- and middle-income countries.
MetadataVis full innførsel
Executive summary: Aim: The aim of this project was to provide insight into the range of options for topic identification, selection and prioritisation (TISP) for health technology assessment (HTA) in low- and middleincome countries (LMICs). Background: HTA is an internationally accepted multidisciplinary approach to analysing and assessing evidence to inform health policy. HTA is recommended by the World Health Organization (WHO) to support universal health coverage (UHC). According to WHO, HTA should ideally be implemented in a formalised decision-support process or system. Deciding what topic(s) to assess represents the first step of this process. In the TISP approach, Topic Identification describes the step where topics potentially suitable for HTA are identified. Selection describes the step where details on the topics are collected and the identified topics are checked for conformity with the aims of the HTA process. Prioritisation describes the step where a decision is made to either initiate, reject or postpone the commencement of an assessment, taking into account the best use of limited resources and context-dependent values. This first step in the HTA process is very important as it has implications for subsequent steps. If TISP fails to work efficiently, this may jeopardise the value of the entire HTA process. In engaging in capacity building for strengthening HTA in LMICs, we (i.e. the Norwegian Institute of Public Health) emphasize approaches to TISP based on experiences from our own country. However, the concept of HTA is relatively new to most LMICs and, with limited institutional mechanisms, adapting complex TISP approaches may be difficult to operationalise. By describing a range of options to choose from, and by providing examples of TISP approaches adopted by countries with a formalised HTA process in Africa, Asia, Latin America and Eastern Europe, our aim in this report is to provide grounds for informed decisions on how best to proceed when planning to implement TISP. Methods: This report uses a systematic scoping review, a country survey and a stakeholder webinar as its methods. The systematic scoping review was guided by established scoping review methodology. A protocol of this review is published on our website. Articles from 2015 onwards were included, and the predefined elements analysed were: the TISP processes, criteria, methods or tools, collaborative networks or initiatives, governance and evaluation. In addition, we conducted a country survey to identify details of TISP approaches in selected African, Asian, Latin-American and Eastern European countries. We anticipated that survey results would supplement (or confirm) findings from the literature. Survey results would also be valuable, considering the paucity of research on LMICs in this field. Finally, we hosted a webinar to present preliminary findings of the scoping search and survey and to gather stakeholder feedback. Take-home messages from the webinar were noted. This report has been subject to internal and external peer review. Results: In presenting the results, we do not structure the narrative in accordance with our methodological chronology (i.e., scoping review, survey, and webinar feedback) but rather use the information from these sources to discuss i) the existing recommendations for TISP implementation guidance, ii) Topic Identification, Selection and Prioritisation, and lastly iii) selected country examples and other aspects of TISP. i)We identified and included five recent (2015 and later) guides on HTA implementation that included information on TISP. Findings suggest that TISP practices could not be translated to a set of common recommendations. In low resourced settings, the TISP process may be pragmatic at the beginning, as suggested by the International Decision Support Initiative, but it should be transparent and explicit. When adopting or adapting approaches and topics from other countries, it is important to take account of experiences from third countries. ii) Our research identified a gradient of TISP options from simpler to more complex processes. In a formalized HTA system, TISP includes a mix of reactive and proactive Topic Identification methods, whereas the more complex approaches (i.e., horizon scanning systems, and disinvestment strategies) are only used in a few HTA systems. Typically, HTA process stakeholders, including policy makers, clinical experts, health care workers, industry, donors and patient/users, are involved in the identification step to propose or nominate topics. Different TISP approaches may be used for different technologies. While experienced European HTA systems rely exclusively on industry submissions or industry solicitations to identify topics, countries new to HTA often rely on commissions from the Ministry of Health, and proposals or nominations made by stakeholders. For LMICs, it has been proposed that it would be beneficial to start with relatively simple, proactive approaches such as stakeholder involvement, adoption of topics from other HTA systems and identifying topics from essential technology lists, before moving to more complex approaches. Selection (also called filtration) is not clearly distinguished from prioritisation. During selection, details regarding the topic may need to be collected, and in some cases explicit yes/no selection criteria can be established. Selection is commonly informed by stakeholders, including clinical experts, industry, and the public or patients/users. Outputs of identification and selection such as lists of topics, short written vignettes or briefs, or even pragmatic assessments, are commonly used to inform HTA Prioritisation. The depth of information will depend on who is involved in prioritisation and how transparent the process is. Information included in the output can be categorised as: technology related; patient and setting related; policy related; evidence related; impact predictions; information on knowledge gaps. If all eligible topics are identified and all selected topics can be assessed (e.g., all new childhood vaccines, or all new medicines to be reimbursed), prioritisation at the level of the HTA process is not needed. In such cases, outputs of identification and selection can directly inform initiation of HTA. However, in most cases, prioritisation is needed, at least to ensure timeliness of the most important topics. Explicit criteria for prioritisation typically reflect: (unmet) needs, potential impact (on patient health, public health, costs, health service, and/or society), and alignment with national priorities. Scoring or ranking to prioritise identified topics is most commonly done by clinical experts, but may also involve other stakeholders, including policy makers, end users of the HTA and patient/user representatives. Ranking may be implicit or explicit. Tools to assist in prioritisation include the use of a Delphi panel, multi criteria decision analysis (MCDA), and on-line ranking tools. Involvement of industry stakeholders and donors in prioritisation is often avoided. Final decisions on prioritisation are commonly performed at governmental level or by a relevant government authority. iii) Our research identified three other aspects related to the TISP process: 1) governance and coordination, 2) evaluation and development, and 3) initiatives and networks for TISP. What is commonly understood as governance for TISP are the aims of the HTA process and funds for its conduct. Governance and coordination of TISP are typically defined politically by the HTA system owner (e.g. regional health authorities) or government-appointed institutions. Recommended evaluation methods include external and internal audits, surveys, interviews and focus groups. Survey respondents reported steps taken to improve the TISP process including: revising criteria and/or weighting, publication of tasks assigned by government authorities on websites, meetings with stakeholders and international partners, and support from external partners through training and capacity building. Collaboration and participation in scientific networks and bilateral capacity-building projects are good for the development of practices in HTA and TISP. We identified one scientific network (International Health TechScan (IHTS)) and one global initiative (The International Horizon Scanning Initiative (IHSI)) that engage specifically in TISP. Influential factors for the choice of TISP approach are contextual and similar to factors influencing other aspects of HTA. These include political support, the aims of the HTA process, experiences with TISP and HTA, national priorities, legislation, human resources and economic resource availability and values. Partnerships for capacity building and scientific networks are valuable, and also influential in choosing TISP approaches. As with the HTA process, the TISP approach can be evaluated and systematically improved to become more efficient and transparent. However, we found no evidence of comparative TISP evaluations. Discussion: This report is based on a systematic scoping review of TISP approaches, a country survey on TISP used in selected African, Asian, Latin America and Eastern European countries with a formalised HTA system, and feedback gathered during a webinar. This report aims to point to the range of TISP options, present examples and look critically at evidence, but makes no claims to be exhaustive. Rather, the results represent our understanding of how different approaches towards prioritising topics for HTA can be categorised by applying TISP. The results are intended to provide additional facts about TISP to supplement existing guidance on HTA implementation. Further work may include more detailed analysis of context-specific needs, comparisons of different approaches and structural limitations. Conclusions: Our findings suggest that: • As with the HTA process, it is important to ensure that TISP is transparent with regard to criteria, procedures and involvement of stakeholders. • The TISP approach should be carefully selected to acknowledge the relationship with the health system context (i.e. politics, needs, resources and values) to which it is applied. • For resourced limited settings, a simple TISP approach may be a starting point, but partnerships with more experienced countries, scientific networks and initiatives should be explored for solidification and comprehensiveness.