The effects of the introduction of a chronic care model-based program on utilization of healthcare resources: the results of the Puglia care program

Aging of the population is continuously increasing the prevalence of chronic illnesses. Co-morbidity is associated with the worst health outcomes and increased healthcare expenditure [1, 2]. This is even more true in Healthcare Systems that show an acute and intramural oriented assistance [3]; in which fragmentation, poor coordination of care and lack of patient involvement results in inadequate responses to the needs of older people with chronic illnesses [4]. The inadequate management of patients with chronicity??chronic conditions is putting pressure on the sustainability of Healthcare Systems [5]. For this reason, the promotion of proactive, integrated, and personalized care is indispensable to optimize the use of healthcare resources [6, 7]. Chronic Care Models (CCM) have been used to address the needs of frail people in the continuum of care, testifying to an improvement in health outcomes and more efficient access to healthcare services for these patients [8]. The elements that characterize the CCM framework are: 1) an appropriately organized delivery system that contributes to producing interaction between patients and community healthcare resources; 2) supporting and improving of self-management [9]; 3) promoting interaction with patients, integrating guidelines with patient preferences [10]; 4) timely access to clinical data about patients through a computerized registry; 5) care coordination between healthcare teams [11]; 6) regular interaction between caregivers and patients, preferring a “continuous healing relationship” to face-to-face visits [12]. The deployment of a program of management of chronic patients at regional level requires the clinical risk stratification of the entire regional population and the identification of patients to be included in the CCM. The development of an infrastructure framework able to record patients’ clinical data in the continuum of care and the use of a widely applicable risk stratification tool are indispensable steps in this process [13].

The impact of CCM deployment on health care outcomes and the utilization of health care resources have already been trialled in a selected cohort of patients affected by specific chronic illnesses [14‐18]. The deployment of an integrated framework based on a CCM have also testified to the improved quality of care and health outcomes in more heterogeneous cohorts of patients with chronic conditions [19]. In fact, the effect of the application of CCM on utilization and clinical outcomes had already been trialled in clinical and/or socially frail people [20‐22], but a quantitative analysis on direct healthcare expenditure and the modality of recourse to the healthcare system had not previously been performed.

To evaluate the impact of the deployment of a CCM on direct healthcare expenditures and the modality of recourse to the healthcare system in a heterogeneous regional population.

We conducted a dynamic retrospective population-based cohort study, using a record linkage analysis of administrative databases (ADs) and electronic medical records (EMRs), including data on the adult population (aged 40 years or over) in the 6 local area health authorities of Puglia in Italy (approximately 2 million people).

Overall, 1112 non-oncological patients aged > = 40 years old were identified in Puglia Care from January 01, 2012 to December 31, 2013. All these subjects presented at least one follow-up control every 6 months. For each subject, EMR data was linked with those of regional administrative databases produced from January 1, 2011 to December 31, 2014, identifying 1074 subjects aged 40 years or above who were present both in EMRs and the regional civil registry with a minimum pre-inclusion and follow-up period of 12 months. In parallel, 2126 controls with the same clinic-demographical characteristics (age, sex, healthcare district, DDCI class, number of unplanned hospitalization in the pre-inclusion period) at inclusion data were identified. Pre-matching characteristics differed significantly between subjects included in Puglia Care and the population in usual care, but the matching gave a good balance for all the considered characteristics (Table 1). The two samples also presented a similar median pre-inclusion period (22.0 months in the Puglia Care group and 21.9 months in the control group) and a similar duration of follow-up (25.8 and 25.9 months, respectively).

Our study shows that the implementation of CCM for frail patients is associated with less recourse to and a significant reduction in costs due to unplanned hospitalizations. These findings suggest that integrated care may lead to a reduced utilization of the emergency system by patients with co-morbidity. The effectiveness of the CCM has been tested in a few trials, having mixed results regarding the improvement in process and outcomes of care [29‐33]. Moreover, the great majority of these studies have analyzed the impact of CCMs in selected cohorts of patients, investigating outcomes relevant to specific chronic diseases (monitoring of diabetes, cardio-vascular risks, blood pressure, BMI,..) [34, 35]. Few randomized controlled trials have enrolled a nonspecific chronic disease cohort with multi-morbidity, and none have investigated the impact of CCM on modality of recourse to the health care system [36‐38]. The results of retrospective cohort studies, as well as case studies and case series, offer a greater numerical sample and an analysis of impact on healthcare practice. The vast majority of these studies included diabetic patients, assessing the effectiveness of CCMs for improving health outcomes [39‐41]; however, knowledge on CCM cost-effectiveness in diabetes care is nascent [42]. The only data available have testified that CCM reduced lifetime risks of blindness, end-stage renal disease, and coronary artery disease, in diabetic patients with a resulting increase in benefits in a cost range from $33,386/Quality Adjusted Life Years [43] to $42,051/Quality Adjusted Life Years [44].

A cost analysis on patients with chronic conditions was experimented in a large population of Tuscany Region through a Before-After study. This study showed that the application of activated diagnostic and therapeutic pathways, involving patients affected by the four most common chronic conditions (Diabetes Mellitus, Heart Failure, Chronic Obstructive Pulmonary Disease and stroke) reduces the recourse to hospitalization and A&E access, despite an increase in outpatient specialist visits and expenditure on drugs [16]. This study did not include a control group; for this reason it was not possible to determine whether modifications in recourse to hospitalization and A&E departments were caused by the disease management programs or by other factors. On the contrary, in our study we have analyzed the effects of CCM both in a before-after and in a control-group analysis, also evaluating modifications in access and costs of healthcare delivery in the control group. Moreover, a key strength of this study is the utilization of data for all patients of a regional healthcare system and the utilization of a matched pair analysis using demographic and clinical variables to match patients included in Puglia Care and subjects in a control group. These elements have allowed us to preserve a recruitment bias and compare outcomes between the two groups in a homogeneous case-mix population.

In the pre-inclusion period, the only significant differences between cases and controls were direct healthcare expenditure. In particular, data showed higher costs related to hospitalizations and drug consumption in the case group. This finding might not be related to differences in the case-mix of the population between the two groups, but rather it can be explained by the inclusion of patients in the control group who were low-users of healthcare resources with respect to patients with similar clinical characteristics included in the case group. In fact, GPs were called to select cases - including in this way patients who had regular contact with the healthcare service - while the selection of controls was random, also including low-compliant patients.

Although a greater increase in drugs costs and outpatients specialist visits was recorded during follow-up in the Puglia Care group compared with the control group, total expenditure due to unplanned hospitalizations registered an important reduction in the Puglia Care group during follow-up. The decrease in unplanned hospital admissions is only a part of the cost saving due to the reduction in the recourse to the emergency healthcare system; the overall impact of CCM on direct or indirect healthcare expenditure needs further studies, including the analysis of additional data.

Aging and the increased prevalence of chronic conditions are putting pressure on the sustainability of healthcare services. Modifications are indispensable to tailor healthcare delivery to patients’ needs, reducing the inappropriate recourse to emergency healthcare services and the consequent increase in care costs. The evaluation of the cost-effectiveness impact of alternative healthcare intervention is mandatory and nowadays information about the impact of CCMs on the healthcare pathways of chronic patients is emerging. Our study provides evidence documenting a reduction in days and costs of unplanned admissions in the CCM group when compared with subjects with similar demographic and clinical characteristics but persisting in usual care. On the other hand, the introduction of Puglia Care has produced and increased cost of drugs and out-patient visits; these costs are probably related to the definition of multi-specialist pathways of care and more appropriate pharmacological therapies. In conclusion, the deployment of a CCM-oriented strategy in frail subjects has proved its efficacy in the reduction of recourse to the emergency healthcare system, allowing a more appropriate redistribution of economic resources.