Our thematic analysis shows that the current fee-for-service system, according to the respondents, does not provide the right incentives for the integration of stroke care. A number of solutions and different financing models were mentioned by the respondents, but there was no consensus amongst them. Several challenges, related to general factors (e.g. the foundations of the financing system and the issue of co-morbidity), or stroke-specific factors (e.g. diverse patient population and non-uniformity of patient pathways) were mentioned regarding the implementation of a more integrated financing mechanism for stroke care.
The finding that fee-for-service systems do not provide the right incentives for improving quality and efficiency in health care has been recognized by others as well [5, 15–17]. In our study, the respondents did not agree amongst each other about the best alternative for the fee-for-service system. Disagreement is also found in the literature. Evidence on pay-for-performance programs for instance seems to depend to a great extent on the specific circumstances and situations of the institutions involved [15, 18].
The respondents in our study came up with several suggestions for alternative financing mechanisms for stroke care (presented in Figure1). There have been financial experiments performed in different groups of chronic patients for several of the proposed payment systems. If we look at those experiments, can we predict the effect of using these systems instead of fee-for-service?
An experiment in Germany for instance, has provided positive evidence for implementation of cooperation fees . In this experiment, sickness funds received higher payments when they set up certified disease management programs and induce patients to enrol while providers were able to receive additional funding if they established integrated care projects. After four years of implementation, a study reported better medical outcomes and significant lower overall costs for diabetes patients enrolled in a disease management program compared to usual care .
In the Netherlands, the first experiments with bundled payments for diabetes care have yielded mixed results . Bundled payments in theory rewards providers who have lower costs while penalizing higher-cost providers. In the United States, bundled payments are promoted as the most promising opportunity to control health care spending while encouraging high quality [6, 8]. But in accordance with our findings, Davis  pointed out that the problem with bundled payments lies in assigning accountability for care across different settings and over time. Care patterns for stroke patients are highly dispersed, there is a lack of continuity in the physician-patient relationship and many different professionals are involved.
Experiments with population-based financing or global payments have been performed in Germany in the Gesundes Kinzigtal experiment . Here, a regional management network together with a physician network and two insurance companies have set up a population-based financing system that is combined with shared-savings. Results so far have shown a reduction in costs of the Kinzigtal region compared to other regions, but future studies have yet to conclude that the decreased costs are indeed due to population health gain. Population-based financing for stroke care in the Netherlands and elsewhere has to be tried out before introduced nationwide. Both results in literature and experiments show mixed effects on the end goal: improving quality of care and reducing costs per patient. The priority now is to experiment with several modes of integrated financing on a small scale, before large national changes will be made.
Despite the challenges and problems mentioned by the participants, there have been significant improvements in the organisation of stroke care in the Netherlands in the last ten years, without any financial innovations . These developments however have not occurred in every stroke service to the same extent, and also not as fast as is thought necessary to deal with the financial constraints and the rapidly aging population. Therefore, it is important now to move beyond care innovations and to look into the options for financial reforms.
Strengths and limitations
The strengths of our study include the purposive sampling strategy from different stakeholder groups; in this way we ensured that multiple perspectives were captured through in-depth interviews of highly knowledgeable informants from the five stakeholder groups. The sample included those directly involved in integrated stroke care, and those who were knowledgeable about, but not directly involved in the day-to-day work of stroke services. Anonymisation ensured that respondents felt free to share their own personal opinions. The semi-structured interview technique allowed issues to be explored in a flexible manner. Also, respondents were free to raise any issue that they felt were relevant to the topic under investigation. As a result, it is believed that the information gathered was reflective of genuine concerns and views. Since the respondents were not familiar with the interviewer, we believe that the potential influence of the researchers on data collection is kept minimal. All researchers were involved in analysis and interpretation of the data to ensure that the conclusions accurately reflect the collected opinions and views of the participants.
The main limitation is that we did not include caregivers in the group of patient respondents. Inclusion of this group could have provided additional information. Another limitation could be that the research is performed in the Dutch context, and therefore less applicable to other countries. However, we believe that the Dutch case could be interesting for others because the Dutch system is a fragmented fee-for-service system, which is found in many other countries. Most of the results of this research are specifically applicable to integrated stroke care. However, many of the issues raised regarding stroke care are also applicable for financial and organisational issues in other types of integrated care.