This study was performed to identify the differences in patient outcomes and their relationship with chronic care management of OAT. Although the Netherlands manage OAT with specialized clinics as is recommended by the ACCP guidelines and achieve a good quality of care according to the criteria of the Dutch National Network of ACs, remarkable differences exist. The percentage of patients in the correct INR ranges differed with more than 20%-point. Furthermore, differences existed in the application of chronic care management which was measured with elements based on the components of the chronic care model (i.e. the health care organization, self-management support, delivery system design, decision support and clinical information system). Two chronic care management elements, i.e. patient orientation and the ratio of specialized nurses versus doctors, were associated with some patient outcomes. Moreover, the overall use of chronic care management elements is positively associated with the patient outcomes.
To our knowledge, no other comparative study of the differences on the quality of OAT between regions was performed on a national level before. It was already shown in previous research that the quality of OAT provided by ACs is higher than OAT provided by routine medical care. For example, Ansell and colleagues showed that the percentage of patients within the correct INR ranges and time-in-range is higher in anticoagulant clinics in Spain and Italy than in routine medical care in France, U.S. and Canada [9
]. Our study showed that differences in quality of OAT exist between ACs which suggests that even the ACs that already provide a relatively high quality of care compared with routine medical care [9
], could improve their quality of care. Furthermore, our results showed differences in the percentage patients below the appropriate therapeutic range (range 1.8% to 13.4%) which should be further studied as the relative risk for adverse events was suggested to be frequently underestimated [16
The identified differences in chronic care management offered an opportunity to explore whether these could be associated with the quality of OAT. Only two chronic care management elements, i.e. patient orientation and the ratio of specialized nurses versus doctors, were significantly associated with patient outcomes while the other elements were not. It could be hypothesized that patient orientation and ratio of specialized nurses versus doctors results in more patient centered care since there is more time spent per patient and therefore more likely to deliver good care. However, the data of this study were not specific enough to test the hypothesis and should be further studied.
Furthermore, we found an association between the use of more CCM components and better patient outcomes. This is in line with results of earlier published meta-analyses of clinical trials [8
], that showed an association between the differences in patient outcomes and chronic care management. Moreover, our results seem to confirm that only when more components of the CCM are used, this will result in better care, while fragmentary use of the CCM is unlikely to improve care.
As a consequence, additional insight in the use and validation of a construct variable for the overall application of chronic care management can be useful since the effect is assumed to be achieved by the combination of initiatives instead of certain elements. Some instruments were developed to assess the overall extent of chronic care management, however, the validation of these instrument are limited and not performed for OAT [20
]. A validated instrument to measure the construct variable would be useful for the professionals to gather insight in the needs for quality improvement regarding chronic care management.
Meanwhile, the complexity of the chronic care management should be taken into account. Chronic care management is a social construct which effectiveness is influenced by more factors than the number of components such as the implementation and the integration of the chronic care model components [22
]. As chronic care management seemed to be associated with the patient outcomes for OAT, more insight in the working mechanism of chronic care management is needed for quality improvement. In particular, qualitative studies are required to explore the association between chronic care management and patient outcomes in more detail. First, as chronic care management is complex and underlying mechanisms are not fully understood, additional qualitative research should be performed to identify the true needs for quality improvement [23
]. Second, oral anticoagulants is not a static field, but is evolving over time. For instance, two recently developed ACs only include relatively healthy patients who are performing self-management. For comparability these ACs were not included in our analysis. Although these two ACs provide care to only a relatively small number of patients, this new organizations could inspire the traditional ACs.
Findings of this study must be interpreted in the light of several limitations. First, the questionnaire was sent out by the Health Care Inspectorate (IGZ) which could cause bias especially social desirability bias which might have resulted in an underestimation of the association between chronic care management and patient outcomes. However, the respondents were aware that this questionnaire aimed to explore the chronic care management of OAT on regional level instead of focusing on the ACs alone. Next, the highly developed documentation of the ACs on national level is in contrast with the scarcely developed documentation in the clinical setting. As a consequence, gaps exist in the follow-up of patients (e.g. INR values around hospitalization). These gaps could not be analysed and controlled for. Thus they may have affected the association between patient outcomes and elements of the chronic care model under study. Furthermore, chronic care management could not be measured by a validated instrument since these are not yet available for OAT. However, we selected and made the elements measurable based on the literature and the expert opinions of about twenty professionals working in the thrombosis field. Finally, the analyses of this study were limited to quality measures reported on national level. This implies that analyses were performed on the organization level without case mix corrections due to a lack of information about patient characteristics such as age and co-morbidity which are not systematically registered by the ACs. We only could correct for the type of reagent, other variations in data gathering for the reported outcome measures could not be eliminated. Yet, the used data-registration of the Dutch National Network of ACs on national level is unique worldwide and gives opportunities to gather more insight in the needs for quality improvement. Therefore, AC regions should be stimulated to gain more insight in the delivery of chronic care management and their influences on the patient outcomes.