Neue Versorgungsstrukturen erproben: Ein ambulanter, interdisziplinärer Behandlungspfad für Patienten mit koronarer Herzkrankheit - Entwicklung und Evaluation im Mixed-Methods-Design

i. Background: Guidelines aim to improve the care of patients. However, they are yet insufficiently implemented in everyday practice in Germany. Another challenge in the German health care system is the limited gate-keeping role of General Practitioners (GPs). Using the example of coronary heart disease as common and relevant disease, we piloted the devel-opment of a treatment pathway as an implementation tool for guidelines. Treatment pathways standardize the care of a target disease and define tasks as well as interprofessional responsibilities. Such “complex interventions” should be examined in feasibility-studies. Our research questions were: Is it feasible to jointly develop a treatment pathway? Is the treatment pathway implemented in daily practice and why (not)? ii. Methods: GPs and cardiologists from the Marburg region jointly developed a treatment pathway in moderated group discussions covering medical treatment, documentation and appointments. In the subsequent pilot study, we differentiated two intervention groups (pathway-developers who participated in the development, and pathway-users who received a training only) and one control group with “usual care”. Per study arm, six GP-practices should recruit 150 patients. We obtained data by a patient-survey at study begin and after six (T1) and twelve (T2; intervention group) or nine months (T1; control group; assumption: no change), respectively. Additionally, we assessed clinical data and quality indicators by case-report forms. We investigated the feasibility of a randomized-controlled trial by sample size calculation, analysis of instruments and exploratory group comparisons (MANOVA) of the patient-relevant outcomes quality of life (EQ-5D index, VAS) and satisfaction (modified PACIC). In addition, we assessed quality indicators of pathway-implementation using case report forms and patient surveys. According to the “theory of planned behaviour” (TpB), we assumed that pathway-developers would hold a positive attitude, subjective norm and perceived behavioural control enhanced by a feeling of “ownership” due to pathway development or regional factors. In semi-structured interviews with all physicians of the intervention group, we explored the implementation process and triangulated qual-ity indicators and GPs’ self-assessment. Using content analysis and cross-case comparisons, we iden-tified implementation factors and ordered them into a model using an ecosystemic approach. iii. Results: Nine GPs and four cardiologists participated in the pathway development. One GP dropped out be-fore the pilot study. We conducted interviews with all remaining developers and cardiologists, as well as four of the six pathway-users. In the pilot study, 17 practices recruited 334 patients. At T1, 290 data sets could be analysed. The treatment pathway included a clear distribution of tasks between the physician groups and cov-ered the topics of medication, examinations and documentation. A patient-passport was developed for documentation and exchange of information. Both, physicians and patients, were satisfied with the pathway. However, about 25% of patients were less well or poorly informed about the pathway. Physicians criticized the effort required for development, but appreciated the constructive exchange. In the pilot study, the patients of the developer group had the best quality of life at all assessment points. The sample size calculation based on the EQ-5D-Index and resulted in 268 patients per study arm for an RCT and an inflation factor of four for a Cluster-RCT. Regarding patient satisfaction, we observed ceiling effects and some items proved to be problematic. Quality indicators of pathway implementation showed an increase in primary care appointments and a decrease in cardiological appointments in the intervention groups (consistent with the path-way). Contrary, fewer than 20% of patients received the recommended patient-passport. As cardiol-ogists used this passport to identify study-patients, this represented a key barrier regarding the study. According to the theory of planned behaviour, a positive attitude was associated with a good self-perception of implementation and a negative attitude was associated with a poor self-perception of implementation. Deviations could be explained by perceived behavioural control or barriers. Contrary to our theoretical assumption, “ownership” would favour pathway-implementation, devel-oper reported “not having to change anything” as their behaviour was “automatically” reflected in the pathway and therefore consistent with it (“ownership-trap”). We mapped further barriers and facilitating factors in relation to proximity to the individual and the intervention (ecosystemically), and added a time-level of behaviour change to an implementation model. iv. Discussion: The development of an ambulatory treatment pathway by GPs and cardiologists is feasible and the result was accepted by peers. However, the joint development did not automatically result in better implementation of the pathway or guideline recommendations. Several implementation factors influence implementation. Treatment pathways, however, can only influence a regional and interactional level. Our multilevel model can help to systematically analyse modifiable factors and thus support researcher in retrospectively explaining implementation as well as prospectively planning implementation. This pilot study has concrete implications for following studies. Such effectiveness studies are desirable in that the value of treatment pathways in the grow-ing research field of guideline implementation-tools remains unclear.