Skip to main content

Table 2 Key Topics/Questions to Inform Workshop

From: Bringing two worlds closer together: a critical analysis of an integrated approach to guideline development and quality assurance schemes

Key topic/question (topics were content domains for exploration and questions specific answerable questions)
1 What is the starting point for QI development? E.g., are the recommendations leading, or does one decide beforehand for which care aspects one wants to develop QI (see draft frameworks sections 1 and 2 in the GIN-McMaster checklist – draft framework)? What if there is only low certainty evidence for these aspects? (key topic)
2 We will need a GRADEpro QI development module: what would be the design? A module that runs parallel to the recommendation development, or integrated with the Evidence to Decision part? Or in addition to EtD? This is a fundamental question! (key question and topic)
3 How to move from recommendation to QI and from QI to performance measure? See Kahn 2014 and Shekelle 2013. When is a recommendation suitable for a QI, with attributes of good quality indicators in mind, and ‘translatability’, can the quality of evidence and EtD factors be of any help here? Which are applicable to QI development? (key topic)
4 How can GRADE help to make the process of QI development more transparent and explicit? (key question)
5 Based on what we learned and read, is there anything that should be added to the GIN reporting checklist for quality indicators and performance measures (Nothacker et al. – workbook documents). (key question)
6 Colorectal cancer screening can be done with many different approaches and sequence of tests and treatments. They should be evaluated in multiple intervention comparisons (of multiple tests and treatments). What is the impact of that on quality indicators (e.g. do we need quality indicators that can be applied to all test-treatment strategies)? (key question)
7 Does conflict of interest management differ in QA from developing the recommendations? (key question and key topic)
8 QI can assess structures, processes and outcomes of health care. Good outcome measures do not necessarily stand for good quality of care, while the application of process and structural indicators in daily practice can lead to better outcome measures. Impactful process and structure outcomes lead to better outcomes, but this should be underpinned by evidence. How can we include that in the framework, or is that somehow covered by certainty in the evidence (e.g. indirectness)? (key topic)
9 The involvement of professional societies, accreditation and certification bodies seems indispensable for implementation of guidelines and quality assurance. Are technical observers and consultants (non-voting on recommendations) appropriate to make informed decisions and here context? How and when should they be involved? (key question)
10 QI may relate to a) patient important outcomes addressed in recommendations; b) process issues (e.g. number of lost specimens in the laboratory pathway); c) diagnosis (tests correctly read by pathologists); d) informed choice (adequate use of processes of decision-making) that consider patient values and preferences. Which are the other domains that a joint guideline/QA needs to consider? Explain your considerations please. (key question and key topic)
11 We propose one working group instead of GDG and QASDG: what are the member criteria (expertise, which disciplines)? What are the training needs? (key question and key topic)
12 Regarding outcomes in quality assurance and quality indicators: do we need or want to define the clinical question/PICO/outcomes with QI development in mind? E.g. surrogate outcomes are often more feasible to measure than direct patient relevant outcomes, especially those that take longer time to occur. However, surrogate outcomes often lower the certainty in the evidence. How to rate the importance of outcomes – should quality indicators be determining the importance of outcomes (do outcomes become people important if they can be measured as performance indicators?)? (key question and key topic)
13 How would you expand the monitoring and evaluation section of the GRADE evidence to decision frameworks based on the lessons learned and the draft framework (click here for a description of the EtD templates)? (key question and key topic)
14 How should we integrate all QA aspects: quality indicators, performance indicators, requirements, certification, a quality assurance scheme? Quality indicators and performance measures determine the requirements for certification. Which aspects are logically related to development of recommendations and which aspects need a different route? (key topic)
15 How to assess certainty of the evidence of QI? Do we need such an assessment? E.g. by use of instruments for risk of bias appraisal of QI (QUALIFY or AIRE) extended to a full assessment? (key question)
16 How can we assess and express the potential benefit of proposed QI/performance measure (being one of the criteria for a good QI)? (key question and key topic)
17 Involvement of patients in QI development: how and in which stages? Results of the study by Schleedoorn 2016 et al. support the fact that there is poor correlation between patients’ and professionals’ perceptions regarding quality of care. Their set of key recommendations would have been different if only medical professionals were involved in the selection procedure. (key question)
18 For guideline and quality assurance development processes, expertise in many areas is required. How can groups function efficiently, how can subgroups of a panel interact with the main panel? E.g., should groups work separately to develop recommendations and quality indicators and ask a core panel to agree or vote on them? How can experts be involved on an ad hoc basis without being a full member of a panel? (key topic)
19 Should we provide guidance on how many QI to develop? Studies show an average number of 20–50 quality indicators per condition, but ideally the development of quality indicators results in a compact set of ≤5 indicators per clinical area (key question)
20 Reflecting on the table below, should quality indicators or performance indicators be based only on strong recommendations (see implications of strong and conditional recommendations)? How can we relate certainty of the evidence and strength of recommendations to QI development? What is the rationale (see Kahn, 2014)? How to best implement conditional recommendations as PI? (key topic)
21 When should modelling be integrated in guideline development? (Key question)
22 Should there be criteria for “retirement” of a guideline-based quality indicator?