Our search did not yield any handbooks of which we were not already aware. Six handbooks for developing CPGs were included. The six handbooks were produced by the Council of Europe, the National Health and Medical Research Council of Australia (NHMRC), the National Institute for Health and Clinical Excellence in the UK (NICE), the New Zealand Guidelines Group (NZGG), the Scottish Intercollegiate Guideline Network (SIGN), and the WHO. Table 2 details the extent to which each handbook addressed the key elements. No additional key elements in the CPG development process were identified during data extraction.
Table 2. Comparison of guidance provided by CPG development handbooks
Five of the six handbooks provide clear guidance on the way in which CPG topics are or should be selected. For example the SIGN handbooks states: 'Guideline topics... are chosen on the basis of the burden of disease, the existence of variation in practice, and the potential to improve outcome.' The NICE handbook is the only handbook which does not address this aspect, the reason being that for NICE 'Guideline topics are selected by the Department of Health'.
Two of the handbooks (Council of Europe, WHO) make no mention of defining the scope of the CPG in terms of patient population, clinical condition, context or target audience. SIGN provides some mention and NICE, NZGG, and the NHMRC have very clear guidance and specific processes with the NHMRC stating: 'Before proceeding, the panel should clarify the purpose of and target audience for the guidelines. This will involve a careful specification of the following: the conditions and clinical problems that are at issue; the type of care providers for whom the guidelines are intended; the type of consumers for whom the guidelines are intended; a description of consumers not covered by the guidelines (where it might otherwise be assumed that the guidelines are relevant to these consumers); the types of settings in which the guidelines will be employed; and the interventions to be evaluated.'
Adaptation of existing CPGs is addressed in all except the WHO handbook. Varying levels of support for and details of the process are provided. The NZGG handbook states: 'If there is already a guideline in existence that addresses the same issue or problem, another guideline may not be needed. An evaluation and/or adaptation of that guideline may be more appropriate than developing a new guideline.' The NICE handbook differentiates between adaptation of NICE CPGs and adaptation of CPGs developed by other organisations. For NICE CPGs, the handbook suggests that the evidence summaries should be used in the form that they are found, and the wording of recommendations adapted as necessary. The handbook suggests that existing non-NICE CPGs can be used as evidence, but new evidence summaries should be created and new recommendations written.
The requirement for a multidisciplinary guideline development group is mentioned by all six handbooks, with all providing clear guidance, though the detail of the suggested makeup of the group varies, as does the emphasis on specific aspects of this process such as identification and management of conflicts of interest.
Consumer involvement is mentioned by all six handbooks. The degree to which consumer involvement is required, and the nature and level of detail in the advice on how this involvement should be facilitated vary widely. NHMRC, NICE, NZGG, and SIGN provide clear guidance, and WHO and The Council of Europe provide less detailed advice. The SIGN handbooks states: 'The involvement of patients, carers or their representatives in guideline development is therefore important to help ensure that guidelines reflect their needs and concerns.'
Formulation of clinical questions is only addressed in detail in the NICE, NZGG, and SIGN handbooks, though the other handbooks all mention the importance of defining the clinical problems to be addressed and provide limited guidance on how this should be done. While the NICE and NZGG handbooks provide useful guidance, for example, on how to format questions within a PICO (patient, intervention, comparison, outcome) framework, none of the handbooks address how to ascertain the clinical questions to be addressed by the CPG.
All six handbooks mention that systematic searches should be undertaken, and the NICE and NZGG handbooks provide clear guidance on how to do this, with practical details about which information sources to search and when to use filters or limits. Some handbooks seem to suggest that one systematic search may be enough for an entire CPG, others are clear that a systematic search should be undertaken for each clinical question addressed by the CPG. For example, the Council of Europe handbook states: 'At present, the preferred guideline development method is to search explicitly and systematically for pertinent evidence to answer each central question being addressed in the guideline.'
The NICE and SIGN handbooks provide clear guidance on processes for inclusion or exclusion of potentially relevant research evidence. The other handbooks provide more limited information.
All six handbooks address appraisal of evidence in some way, with NICE, NZGG and SIGN providing clear processes and templates.
Methods for drafting recommendations are at least mentioned by all six handbooks, with the Council of Europe providing more limited advice than the NHMRC, NICE, NZGG, WHO, and SIGN handbooks. NICE in particular provides detailed advice describing how to word recommendations including instructions that they should be 'standalone' and 'action-oriented'.
Similarly, all six handbooks mention the importance of facilitating consultation on the draft CPGs, for example, the WHO handbook states: 'Ensure peer review – Circulate widely to experts, professional organizations, regional offices, countries'. NICE, NHMRC, and SIGN have clear, well-defined consultation processes.
Writing of summaries is mentioned in all but the Council of Europe and WHO handbooks. The NZGG asserts that 'The majority of readers and users of CPGs will not have the time to read the whole document and it is essential that a summary of the guideline with recommendations is produced. Algorithms are also useful tools that can improve guideline utilisation.'
A process for developing an implementation strategy for inclusion in the CPG is addressed in all but the SIGN and NICE handbooks. Implementation is beyond the scope of the SIGN mandate, and the NICE handbook refers CPG developers to the NICE website for implementation materials. The Council of Europe, NHMRC, and NZGG handbooks provide particularly detailed practical guidance in this area.
Evaluation, review, and updating of CPGs is addressed by all of the handbooks, with the NHMRC pointedly stating: 'Evaluation of CPGs is essential. How will we know if guidelines are worth developing, disseminating and implementing if we do not know whether they make a difference to clinical practice and health outcomes?' The amount of practical detail available to enable guideline developers to carry out these processes varies between handbooks, and often discussion is quite limited.