gms | German Medical Science

Background: Evidence-based guidelines potentially improve healthcare and outcomes of chronic conditions. However in a patient with multimorbidity, i.e. two or more chronic conditions [1], the application of multiple guidelines may have undesirable effects such as disease-drug (Di-D-I) and drug-drug interactions (DDI) [2]. In chronic heart failure (CHF), more than 90% of the patients have at least one additional condition [3], and a mean number of six prescribed medications [4]. Our pilot study aimed to identify interactions of diseases and drugs in patients with CHF and common comorbidities.

Materials and Methods: We conducted a text analysis of evidence-based guidelines addressing either CHF or one of its commonly co-occuring conditions (prevalent in at least five percent of the CHF patients (N=18) [3]). We searched for the guidelines at Guidelines International Network Library, and at the websites of established guideline developing organizations (NICE, SIGN, and the German AWMF). For each condition, we selected two to four guidelines which should be evidence-based, comprehensive, up-to-date, relevant for primary care, and developed with high methodological standards. We extracted data on interactions between CHF and comorbidities, and key recommendations on diagnostic and therapeutic management. From a subset of data, we derived 13 subcategories within disease-disease (Di-Di-I), disease-drug (Di-D-I), drug-drug interactions (DDI), synergistic treatments, and uncertain drugs, providing the framework of the interaction matrix. We classified the interactions reported in the guidelines and tested the inter-rater reliability for main and subcategories, refined the framework, and agreed upon interactions of the matrix.

Results: We selected 48/771 guidelines from U.K. (n=21), Germany (n=16), U.S. (n=8), and one each from Europe, Australia and New Zealand. Two thirds of them provided information about comorbidities at least in a single section; 2/3 CHF and 1/45 non-CHF guidelines addressed comorbidities in separate paragraphs. In total, we identified N=247 interactions (on average 14 per co-morbidity): Of them, 68 were Di-Di-I (e.g., worsening prognosis or function), 115 Di-D-I (e.g., matters of caution), 12 DDI, and 52 synergisms. The most common subcategory was a disease-drug contra-indication/caution, followed closely by synergisms. We identified more interactions from the CHF guidelines alone (83) than the non-CHF guidelines alone (63). All 18 co-morbidities contributed at least one interaction, with asthma, COPD and hypertension contributing the most and ocular disorders contributing the least.

Conclusion: Our pilot study provides a systematic approach to address multimorbidity in clinical practice guidelines. The classification of interactions in multimorbidity, i.e. within and among multiple treatments and conditions was feasible in CHF. The application of the classification results in an interaction matrix presenting omissions, matters of caution, safe and synergistic principles, as well as gray zones of clinical management. The matrix may support weighing up potential treatment benefits and risks and therefore, rational prescribing in patients with CHF and common comorbidities. Major limitations were the limited number of included guidelines, the unknown validity of interactions reported in guidelines [5], the restriction to two-way interactions between CHF and comorbidities (but not between comorbidities), and the unknown clinical relevance of the identified interactions. The applicability of the classification to other chronic conditions, as well as feasibility and efficacy of the matrix as a clinical decision support tool to improve healthcare guidance and outcomes should be evaluated in future studies.