Table 2

Practical example on cervical screening in Canada15

PhaseModuleUse of the ADAPTE process for the adaptation of a guideline on cervical cancer screening
I. Set-upPreparation
  • After confirming the need for a guideline on cervical cancer screening, a multidisciplinary guideline panel was assembled. Guideline adaptation was chosen over de novo development, as the organising committee was already aware of a number of credible cervical cancer screening guidelines produced by recognized guideline developers and currently in use by practitioners.

  • The organising committee prepared a short protocol outlining the process to be followed, which included an introduction and rationale for adaptation, the topic area, panel membership, the consensus process, the adaptation modules, and the funding source.

II. AdaptationScope and PurposeThe panel defined the health questions using PIPOH (Population, Intervention, Professional, Outcome, Healthcare setting). The organising committee decided on the following clinical question: What is appropriate cervical cancer screening (I) for average risk women (Pop) seen in primary care (Prof, H)? The optimal screening interval should result in improved survival at an affordable cost (O).
Search and Screen
  • The cervix cancer panel decided to retrieve as many guidelines as possible as opposed to selecting a single guideline to be adapted. The resource team searched MEDLINE and EMBASE from 1992 to 2004, combining the terms: practice guidelines, standards, consensus, cervix neoplasms and mass screening. They searched for guidelines on the internet (using Google) and known guideline developers' websites. From the 599 documents found, the chair and the methodologist selected 18 cervical cancer screening guidelines (9 from internet). Upon preliminary evaluation with the rigor dimension of the AGREE instrument (http://www.agreetrust.org), the chair and methodologist retained the seven guidelines with an average rigor score >40% for appraisal. Three poorly scoring guidelines were also retained because they were created for the panel's healthcare context and were all well known to panel members. The guidelines retained are represented by the dark bars on the graph.

  • Embedded Image

Assessment
  • Four of the seven cervical cancer screening guidelines had been published in 2003. One was published in 1993, one in 1995, and one in 1998. Only the source guideline published in 1993 was under consideration for updating, but at the time of the guideline adaptation, the developer had not yet begun the update. The adaptation panel eliminated the 1993 guideline from further consideration because it had not been updated for 10 years.

  • The panel created a recommendation matrix, also including the AGREE scores on Rigour of Development and overall assessment. After assessing the guidelines' consistency the group reclassified the levels of evidence for each recommendation, using their own system as the source guidelines used different classifications.

  • The ‘applicability’ and ‘acceptability’ of the recommendations to the local context were discussed in detail in light of potential local constraints, organisational barriers, legislation, policies, and/or resources in the healthcare setting.

Decision and SelectionThe panel summarised the information from the above assessment for each of the selected guidelines using a comparative table to facilitate the selection of relevant source guidelines and identification of recommendations that could be either adopted or modified. Review of the overall assessment scores of AGREE led to elimination of 2 guidelines. Information from the currency assessment led to elimination of 2 outdated guidelines. Based on AGREE scores, the panel then focused on the top 3 choices and their individual recommendations. After discussion, they accepted all five recommendations of Guideline 1. The panel did not agree with the annual screening schedule of Guideline 2 and did not support a rationale for 70 years as a stopping age for screening. The panel put aside Guideline 6 because of insufficient information about the developer and potential for conflict of interest. It did not add anything original to the other guidelines. The panel then focused on Guideline 1 to consider its adoption in its entirety. But the panel concluded that this guideline covered only the average-risk population. However, because the panel felt they needed more information and/or a comprehensive list of relevant guidelines for these special population groups, the panel decided to postpone the development of such recommendations. Eventually, the panel reached consensus to accept Guideline 1 in its current form.
CustomisationAfter completing their assessments, the panel decided to endorse one guideline without modification.
III. FinalisationExternal Review and AcknowledgementThe draft cervical cancer screening guideline was sent to family physicians/general practitioners for external review from across the county and working in both urban and rural practices. Practitioners were sent the draft guideline along with a short survey of questions about, for example, the practitioner's confidence in the process, the applicability of the guideline to the practitioner's patients and practice context, and whether the practitioner would use the guideline in practice. The organising committee sent a copy of the adapted guideline to the source developer for feedback. Once the adapted guideline was finalised, it was submitted to the national college of family physicians for review and official approval. The college of family physicians then posted the guideline on their Web site, and profiled the guideline at their annual conference.
Aftercare planning
  • The chair of the organising committee offered to take overall responsibility for deciding when a review and update of the adapted guideline might be necessary. He asked that those members of the panel with the relevant expertise assist with the actual work of updating and review when the time comes.

  • As the adapted guideline is only based on one guideline, the panel decided that the chair should keep in touch with the source guideline developers and monitor when they propose to review the evidence behind the source guideline and/or make substantive changes. The chair asked the resource team to monitor publication of new systematic reviews or health technology assessment reports, particularly those related to changes in technology. A plan for review was written up and put into the final adapted guideline.

Final ProductionThe final version of the adapted guideline was formatted to take into account the preferences of family physicians. A one-page summary of the recommendations prefaced the main document. As it has been shown that reminders targeted at both the practitioner and the patient improve screening rates, the panel produced a patient brochure that echoed the recommendations of the adapted guideline. The patient brochure was translated into the languages of those populations in that locale that are traditionally under-screened, for example, immigrant populations.