Use of the ADAPTE Method to Develop a Guideline for the Improvement of Depression Care in Primary Care

We implemented two external panels whose members were selected on the basis of their experience in collaborative mental health and primary care approaches, one comprising 66 PCPs, charged with drafting an initial list of 40 clinical scenarios, and another comprising 12 members joined by seven PCPs, three psychiatrists, and two psychologists, who reviewed this draft and selected the 37 most relevant clinical scenarios. [A table providing a profile of the 66-member panel is available in an online supplement to this column.]

The clinical scenarios covered areas related to the definition and epidemiological aspects of depression, assessment, and detection and to general and specific treatment decisions [see the online supplement for a list of the clinical scenarios]. There was a high degree of consensus among PCPs on the selection of clinical scenarios, regardless of the PCPs’ depression expertise. On the suggestion of the GDG, the panel included two additional scenarios: depression during pregnancy and among older people.

We searched for depression clinical guidelines published in English or Spanish between 2004 and 2011. Sources included National Guidelines Clearinghouse, Canadian Medical Association InfoBASE, National Institute for Health and Care Excellence (NICE), Scottish Intercollegiate Guidelines Network (SIGN), GIN, New Zealand Guidelines Group, Institute for Clinical Improvement, National Health Service National Library of Guidelines, Turning Research Into Practice Database, and GuíaSalud, the National Catalog of Guidelines of the Spanish Health Care System.

We assessed the initial pool of clinical guidelines with AGREE, a 23-item appraisal instrument organized into six quality domains of established validity and reliability (8).

We considered eight clinical guidelines but selected five (9–13) after excluding three because of poor methodology or outdatedness. In accordance with AGREE criteria, we selected the NICE guideline as our main source for recommendations (11) but also used the SIGN guideline (14). [Scoring for the clinical guidelines is available in the online supplement.]

We elicited the opinion of 12 external experts—nine PCPs, two psychiatrists, and one psychologist—through a modified Delphi technique (15). Experts first individually rated the wording and feasibility of the recommendations, taking into account the Spanish clinical context on a 9-point Likert scale. Next, experts met at two face-to-face meetings, where recommendations with low consensus were discussed and voted on.

Recommendations made by the selected clinical guidelines were adopted without any modification if they were fully applicable to the Spanish PHC practice context; were adapted to better fit the Spanish context or resolve contradictions among guidelines [see the online supplement for examples of adapted recommendations]; or were rejected if unsuitable to the Spanish context. If no recommendation fully addressed any of the clinical scenarios, we conducted additional searches to identify recommendations, but none were adapted or developed on the basis of these new searches.

The resulting version included 121 recommendations, 116 of which were adopted without modifications from NICE and 17 of which included two statements. According to the GRADE classification system (16), 73 recommendations had strong evidence, and 30 had weak evidence.

None of the guidelines provided guidance on several areas of uncertainty (such as frequency and modalities of presentation of depression in PHC, screening in PHC, grief management, role of self-help groups and new technologies, management of treatment resistance, or intolerance).

We implemented a panel of 16 individuals from Spain, the United Kingdom, and the United States who belonged to a national research network on mental health, had international experience in depression research, or both. They included PCPs, psychiatrists, and a pharmacist. Their mission was to evaluate the relevance, methodological soundness, and clarity of the recommendations and to identify barriers and facilitators for implementation of the clinical guideline.

The reviewers assessed the clinical guideline as a high-quality guideline and suggested minimal changes to improve clarity. As such, the final version was included in GUIASALUD, a national guidelines catalog with stringent quality criteria (17).

The implementation process was monitored through five performance targets designed by the GDG to evaluate changes in detection, diagnosis, treatment, follow-up, and referral processes. The clinical guideline was implemented through a multicomponent program that included recruitment of PCPs with a special interest in mental health, responsibility for continuous medical education in their PHC centers, or both, who acted as internal facilitators; a training program that involved discussion of active cases; audit and feedback; dissemination of the guideline to all PHC centers that included the distribution of the clinical guideline (full, summary, and pocket versions) and support materials, such as a self-help guide (18); establishment of guideline-related performance targets for the health centers as well as an incentive program for PHC centers; enablement of access to mental health consultants; and identification of center-specific barriers to be addressed by guideline implementation “leaders.” Broader dissemination efforts included the endorsement of institutional and scientific societies and publicizing the clinical guideline though various media outlets.

Barriers and facilitators to adoption were identified through qualitative interviews with nine PCPs, five psychiatrists, and a pharmacist. Main barriers pertained to competence in depression care, continuity of care, and time constraints. Main facilitators pertained to quality of the clinical guideline and institutional commitment [see the online supplement for details].