The number of patients with multiple chronic diseases is growing world-wide [1]. Multimorbidity affects the quality of life, utilization of health care services and mortality [2]. In general practice, multimorbidity is defined as the co-occurrence of two or more chronic conditions [3]. Little is known about the optimal management of patients with multimorbidity and interactions between multiple drugs and diseases within the framework of chronic illness care.

The Chronic Care Model (CCM) is an evidence-based approach to improving care for chronically ill patients with multimorbidity [4]. It was developed in the United States during the 1990s [5]. The Assessment of Chronic Illness Care (ACIC) is internationally used to determine whether and to what extent an implemented system is optimizing the quality of care according to the CCM [6, 7].

The ACIC assesses all six elements of the CCM (organization of the healthcare delivery system, decision support, delivery system design, self-management support, clinical information system, and community linkages). To measure how care is provided in German practices, researchers have started to translate-and tried to culturally adapt the ACIC instrument. During this process, we determined that the ACIC is not fully applicable to the German health care system. These limitations are due to a multifaceted understanding of terms, different levels of knowledge of the CCM, and fundamental differences between health care systems [8]. The CCM has proven benefits for chronically ill patients [9], and its implementation in primary care seems to be promising [10]. Therefore, a new instrument--the "Questionnaire of Chronic Illness Care in Primary Care" (QCPC)--was created by German researchers to measure the current status of care according to the CCM and potentials for improvement and further development. The QCPC includes the theoretical background of the CCM. Unlike using the ACIC, instrument, a physician in charge of a practice should be able to complete the QCPC without any special training in CCM. Here, we report the results of the psychometric properties and test-retest reliability of the QCPC.

In a first step, we collected data on the CCM and factors influencing the care of patients in Germany with multiple chronic conditions based on a literature search and our own hypotheses. The ACIC instrument and questionnaires such as Patient Assessment of Chronic Illness Care (PACIC) [11] and the Commonwealth Fund survey [12] were assessed. Based on the theory behind these questionnaires, new items were developed according to the rule to design a questionnaire study in health care [13]. Additional topics were collected and items drafted in a qualitative process of brainstorming and consensus [8]. Most additions concerned the structure of practices as former research has shown that factors such as practice size rather than chronic care organization determine outcomes [14–16].

Nine general practitioners were asked to evaluate the questionnaire to determine whether each item was understandable, whether the content of an item was important for care of chronically ill patients, and whether any important aspects were missing. Details of this process are published elsewhere [8].

Because one item of the "community linkages" subscale ("satisfaction with the exchange of information with other GP practices") was not stable, it was deleted from the questionnaire.

This report describes the results of the psychometric properties and test-retest reliability of the QCPC, which contains elements of the CCM. The results indicate that the QCPC might be the first instrument giving psychometric properties insight into the question of whether and to what degree German practices have already implemented CCM elements into daily care of patients with one or more chronic conditions. The importance of practice structure has been demonstrated in previous research [16, 20]. Almost two-thirds of our participating practices belong to one cluster and shows that a small inter-practice variability regarding measure of quality of care exists. This is an important aspect for caring patients with chronic conditions at primary care [21]. However, more then one-third of the items on the QCPC relate to aspects of practice structure. The QCPC does not contain items relating to the CCM element "health system/organization of health care" because of its limited applicability in the German context. Knowledge transfer from one country to another has known limitations [22]. Therefore a country specific process has to be intended in any case of knowledge transfer. However, QCPC does contain the other five elements of the CCM (decision support, delivery system design, self-management support, clinical information systems and community linkages). All of these subscales demonstrated moderate internal consistency and moderate test-retest reliability over the three-week test-retest interval.

For "delivery system design" the internal consistency was low. One reason for this result might be that especially reasons for referral are very little reflected in daily practice. Stability, as assessed using the Wilcoxon matched paired test, was very good. Because one item of the "community linkages" subscale was not stable, it was deleted from the questionnaire. A low test-retest reliability was observed for the item "disease management programs or guidelines" (r = 0.48). The implementation of guidelines in German primary care practice is basic yet. Decision making is rather a matter of former individual hospital pathways but evidence [23]. Therefore treatment is due to individual behavior in each reason for encounter and might explain the low reliability score. We also found a low test-retest reliability of two items of the "clinical information systems" subscale; namely, "documentation and patient files" (r = 0.29) and "access to data from hospitals" (r = 0.41). Because it currently is not possible to access data directly from German hospitals, we decided to delete that item. Conversely, we decided that "documentation and patient files" was important for German physicians in terms of billing by the German Association of Statutory Health Insurance Physicians. Overall, the results of our study showed moderate to low values for test-retest reliability, which is related to the lower variance of some items [24]. However results need to be discussed in the light of the low response rate. This might be partly due to little interest in redesigning care and the length of the questionnaire.

The QCPC is a questionnaire for assessment of care according to the CCM in Germany which was tested of their psychometric properties. Unlike the ACIC instrument, the QCPC can be used in a high number of practices as the respondents need no prior training in CCM. The QCPC can gauge the current state of care as well as areas for improvement of care. The ESTHER-cohort was the first study performed in the federal state of Saarland, Germany, in which the QCPC was used to survey whether structured care according to the CCM leads to better care of multimorbid (and frail) patients. The ESTHER-cohort includes almost 700 GPs and about 10,000 patients [27, 28]. Outcome data from this study will make it possible to complete the next validation step.