Objective:

To develop a clinical practice guideline on multimorbidity based on two methodological approaches: 1) systematic review of the literature clinically relevant to patients with multimorbidity; and, 2) development of an overarching algorithm (meta-algorithm) for primary care encounters of patients with multimorbidity.

Methods:

1) Clinical recommendations on multimorbidity were obtained from evidence reports and formal consensus.
2) Development of Meta-algorithm consisting of the following steps:
- Design of a generic meta-algorithm to structure and prioritise GP encounters of patients with multimorbidity based on 10 heterogeneous case vignettes. For each vignette specific management recommendations (N-of-one-guidelines) were developed, based on systematic guideline synopses and comments by practicing GPs. Using pattern recognition common decision nodes and care elements were identified from all N-of-one-guideline and combined into an overarching and generic meta-algorithm.
- Eliciting healthcare preferences of patients with multimorbidity from systematic literature analyses, complemented by qualitative interviews.
- Formal consensus of the meta-algorithm by the multidisciplinary guideline panel.

Results:

1) Key recommendations focus on the need of sufficient time for intensive doctor-patient-relationship and shared decision making, the encouragement of patients to present their personal preferences, the comparison of priorities of patients and doctors, medication review and consulting other professional groups. 2I) Our meta-algorithm for very complex patients illustrates a superordinate process which permanently keeps in view the entire patient and reflects the logic of a GP encounter with explicit decision situations, communication needs and priorities.

Conclusion:

Due to the complexity and heterogeneity of health problems in patients with multimorbidity guideline recommendations are best presented on a high abstraction level. Still, practitioners need concrete advice for action. We think the DEGAM guideline with the overarching meta-algorithm fulfills the two tasks. The guideline is currently being piloted in a German GP setting.