Dear Editor,
The paper by Vedel et al. [1] recently published in IJIC summarizes the evolution of integration in the province of Quebec. As the director of one of the leading teams involved in integrated care in Quebec, I wish to comment on, correct and add to some of the issues raised by the authors.
In their review of the different models of integrated care tried in Quebec over the last decade, the authors mention the PRISMA experiment and report that “the results on impact on functional decline were inconclusive” [2, p. 5]. This is not true since the main outcome of this four-year quasi-experimental study involving more than 1500 subjects was a statistically significant 6.2% reduction in the prevalence of functional decline (p<0.05) and a 13.7% decrease in the incidence of functional decline (p<0.01) in the experimental groups as compared to the comparison groups [3]. Functional decline in this study was defined as a five-point or more decrease on the disability scale, institutionalization or death. This indicator has been used previously in other studies [4, 5]. Moreover, participants with unmet needs, i.e. presenting disabilities not addressed by adequate resources or services, were 31.4% lower in the experimental groups in the fourth year of the study (p<0.001). These results are very conclusive since it was a population-based study designed to measure effectiveness not efficacy. The subjects were randomly selected from people 75 years and older at risk of functional decline, who did not necessarily benefit from the tested integrated care model during the study period. In fact, only 18% of the subjects were assessed and cared for by a case manager in the experimental groups during the study period.
The PRISMA experiment was the first attempt to operationalize and test a coordination-type model of integrated care [6]. The other types of integrated care described by Leutz (liaison and full integration) [7] have been tested in many countries to date [8]. The PRISMA model includes six components: 1) coordination between decision-makers and managers at the local level, 2) single entry point, 3) case management, 4) individualized service plan, 5) single assessment instrument coupled with a case-mix management system, and 6) computerized clinical chart. A coordination-type model like PRISMA is probably more adapted to publicly funded health systems since it does not duplicate (as full integration does) the current health care system but is embedded within it, putting all the existing providers into a coordinated network under an umbrella organization [9].
During the study, the Minister of Health was convinced of the adequacy of the model (even before the results were out) and decided to undertake a major health care reform that merged the different public organizations involved in caring for older people: hospitals, nursing homes and CLSCs (local community service centres) responsible for homecare. Rehabilitation centres were not included in this merger, contrary to what was stated in the Vedel et al. paper. This structural integration was seen by the Minister as providing strong support for improving the coordination of services. However, as demonstrated in other contexts, structural integration does not necessarily foster functional integration [10]. The reverse was actually observed in Quebec over the first four years of the reform. According to the Quebec Ministry of Health, the implementation rate of the PRISMA model, based on the indicators developed by our team [11] was only 38% on average in 2008 despite the fact that the generalization of the PRISMA model was included in the 2005–2010 action plan of the Ministry [12]. The newly created CSSSs (health and social service centres) were totally overwhelmed by the strategic planning process and the reorganization of services. The generalization of the PRISMA model was slowed down considerably and even stopped in many regions because, first, the CSSS’ different programs were in fact still working in silos and, second, this new big player in the system was no longer sitting at coordination tables and ignored the volunteer agencies, social economy enterprises and private providers also involved in providing services for frail older people. This natural experiment shows that the structural integration of different providers into a common organization is not necessary to implement a functional integration model like PRISMA. Nevertheless, after six years the implementation of the PRISMA model throughout the province is now back on track. The implementation of the computerized clinical chart, the sixth element of the PRISMA model, was delayed because the Ministry wanted to develop new, more powerful web-based software. After six years and an investment of many millions of dollars, the new software including the single assessment tool (the SMAF) is now being implemented across the province. This allows for the utilization of the management tool (Iso-SMAF Profiles) and completes the implementation of the fifth element of the PRISMA model. Now the only limiting factor for completing the implementation of the PRISMA model is the recruitment of case managers.
In PRISMA, a seventh component was not included in the model; financing is usually included as one of the components of integrated models [13]. This was not possible since the Quebec health care system is a universal, publicly funded, Beveridge-type system. Long-term care is included in the global funding of health and social services. Obviously this arrangement cannot prioritize long-term care and home care, especially during a period of budget restrictions. In the new CSSSs, most of the funding is directed toward the hospitals and nursing homes, which leaves the home care programs with insufficient funding to really make a difference in the way care is provided to frail older people suffering from chronic conditions and disabilities. To improve the efficacy of the PRISMA model and the case managers’ actions would require a specific funding scheme for long-term care modeled on the long-time care insurance programs in many European countries [14]. Japan has also implemented a long-term care insurance scheme associated with an integration of services and case management [15]. Following the needs assessment by the case manager, an allocation corresponding to the disability level of the frail older person could then be managed in order to contract out the appropriate services to the client. Such a financial incentive gives the case manager real power to obtain the necessary services from providers. Quebec and Canada will have to move toward this type of funding scheme, coupled with the integration of services, to adequately cope with the rapid aging of the population [16].
PRISMA is a good example of how a research innovation can actually be implemented by policymakers to improve health services. Right from the beginning, not only policymakers but also managers and clinicians were involved with the researchers in the project team. This was probably a critical factor in the relatively fast knowledge transfer to health policy.