Background: The chronic care model (CCM) is an established framework for the management of patients with chronic illness at the individual and population level. Its application has been previously shown to improve clinical outcome in several conditions, but the prognostic impact of CCM-based programs for the management of patients with chronic heart failure (HF) in primary care is still to be elucidated. Methods: We assessed the prognostic impact of a primary-care, CCM-based project applied in Tuscany, Italy, in 1761 patients with chronic HF enrolled in a retrospective matched cohort study. The project was based on predefined working teams including general practitioners and nurses, proactively scheduled regular follow-up visitations for each patient, counseling for therapy adherence and lifestyle modifications, appropriate diagnostic and therapeutic pathways according to international guidelines, and a key supporting role of the nurses, who were responsible for the practical coordination of the follow-up. A matched group of 3522 HF subjects assisted by general practitioners not involved in the project was considered as control group. The endpoints of this study were HF hospitalization and all-cause mortality. Results: Over a 4-year follow-up period, HF hospitalization rate was higher in the CCM group than the controls (12.1 vs 10.3 events/100 patient-years; incidence rate ratio 1.15[1.05-1.27], p = 0.0030). Mortality was lower in the CCM group than the controls (10.8 vs 12.6 events/100 patient-years; incidence rate ratio 0.82[0.75-0.91], p < 0.0001). In multivariable analysis, the CCM status was associated with a 34% higher risk of HF hospitalization and 18% lower risk of death (p < 0.0001 for both). The effect on HF hospitalization was mostly driven by a 50% higher rate of planned HF hospitalization. Conclusions: Implementation of a CCM-based program for the management of HF patients in primary care led to reduced mortality and increased HF hospitalization. These findings support the hypothesis that the beneficial effects of CCM on survival might be extended to patients with chronic HF followed in primary care, but also support the need for further strategies aimed at improving the management of these patients in terms of hospitalizations. Keywords: Heart failure, Health services, Mortality, Hospitalization, Chronic disease * Correspondence: email@example.com Cardiology Unit, S. Maria Annunziata Hospital, via dell’Antella 58, Florence, Italy Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ballo et al. BMC Health Services Research (2018) 18:388 Page 2 of 8 Background this setting could have lead a positive impact on clinical Despite the progressive reduction in cardiovascular outcome. The rationale of this hypothesis was the as- death rates over the last decades, chronic heart failure sumption that a proactive approach, aimed at optimizing (HF) remains a major and growing health system patient-related and particularly system-related factors, challenge worldwide, and a leading cause of mortality, would have favoured a better adherence to guideline- recurrent hospitalization, and disability [1, 2]. Hospital recommended treatments. By investigating the potential discharges related to HF have progressively increased role of CCM in a chronic disease with large prevalence over the last decades in both US and Europe, and now and high economic burden, we also expected this study to exceed 1 million per year in the US [3–5]. To date, the contribute to the health service research field. incidence of HF in adults is 5–10 per 1000 persons per Therefore, the objective of this study was to investigate year in developed Countries, resulting in an overall the clinical utility of a CCM-based healthcare project for prevalence of 2-3% [6, 7]. The impact of HF is even the management of patients with HF. more evident in the elderly, exceeding a 10% prevalence among persons ≥70 years of age . The prognosis of Methods patients with HF also remains poor, with approximately Setting and intervention 50% of patients expected to die within 5 years and with This study was designed to explore the prognostic im- no significant trends towards improvement over the last pact of a regional healthcare project applied since 2010 two decades [9–12]. Moreover, despite the progressive in Tuscany, Italy, aimed at optimizing the clinical man- advances in the pharmacological therapy of HF, gaps be- agement of patients with HF in primary care and based tween guidelines and clinical practice in HF patients are on the implementation of the CCM. The “Project for still evident . All these factors contribute to impose proactive health care implementation at community a huge and progressively growing economic burden on level” was launched in 2008 by the Tuscan Regional healthcare systems . Health Ministry as a major target of the 2008-2010 These considerations raise the question of whether the Health Planning, and was based on the implementation development and the implementation of specific man- of the CCM in several diseases, including HF. The pro- agement programs could be effective in improving the ject involved the whole population living in Tuscany, clinical outcome of patients with HF. The Chronic Care Italy, and was applied to patients with chronic HF since Model (CCM) is a well-known model aimed at trans- 2010. The present analysis is a retrospective matched co- forming the health care system from simply reactive – i. hort study on this population. The project was designed e., responding in case of sickness - to a proactive one, by taking into account the characteristics of the local thus focusing on the maintenance of patients’ health by healthcare system. Italy has a tax-based universal health planned regular interventions at the community, system organized on several levels. The national level organization, practice, and patient levels [15, 16]. provides funding and dictates the fundamental services Although this model has been widely applied worldwide that must be provided to every inhabitant. The regional for the management of patients with chronic diseases, level receives the national funding and organizes the few analyses investigated its effectiveness in improving health system through a network of local health author- outcomes in HF patients, with considerable differences ities. Every inhabitant is entitled to choose a GP, who in the effects on hospitalization and quality of life across has a gatekeeper function and may have in charge a the studies [17–20]. A recent metanalysis confirmed that maximum of 1500 adult subjects. GPs can work either the CCM approach can probably be clinically useful for as single physicians or functionally associated with other the management of HF patients, but with substantial colleagues. Copayment of some health services may be heterogeneity in effectiveness . Moreover, most of requested, according to national or regional regulations. these studies were carried out in the US or in northern Local health authorities are further subdivided in health European Countries, whose health care systems are dif- districts that are homogeneous with respect to several ferent from the Italian one [22, 23]. Lastly, the majority characteristics (e.g., rural vs urban vs mountain areas), of evidences were obtained in hospital settings, so that and primary care is organized at the health district level. the clinical utility of CCM-based programs for the man- For this project, GPs were organized in teams agement of HF patients in primary care is still to be elu- comprising 5 to 15 physicians and at least a nurse per cidated. In this view, we sought to explore the effect of 10,000 patients. The project was specifically designed to CCM on the outcome of HF patients within the Italian implement the main principles of the CCM for the man- system, which is based on healthcare services provided agement of chronic HF patients in primary care. Accord- by a public system administered on a regional basis and ingly, regular follow-up visits were proactively scheduled hinges on a central role of the general practitioner (GP). for each patient and recalls were set up for patients who We hypothesized that the application of a CCM project in were not showing up. Particular care was given to Ballo et al. BMC Health Services Research (2018) 18:388 Page 3 of 8 provide adequate and systematic counselling for therapy responsible for updating the chronic disease registry, adherence and lifestyle modifications – including regular contacting patients for routine services, scheduling spe- physical activity, weight loss when appropriate, smoke cialist visits, managing patient counselling, providing cessation, and adequate dietary intake – and to establish additional self-management support and health behav- an effective patient-provider relationship based on col- iour counselling, and systematically recording weight laborative care and self-management education [24, 25]. and blood pressure. The GPs adhered to the project on a voluntary basis. All In each local health authority that participated to the the GPs who adhered were members of some form of project, diagnostic and therapeutic pathways were devel- association and adherence was always a groups’ decision. oped according to international guidelines compatible A pay for performance scheme was set up, based on the with the local available resources. GPs adhering to the following indicators: percentage of patients who were project enrolled patients over a predefined six-month enrolled, who were treated with ACE inhibitors/Angio- period, from January to June 2010. tensin receptor blockers and beta-blockers, who had cre- atinine and electrolyte blood tests, who attended Study sample individual or group counselling, and who had their For the purpose of this study, the exposed cohort (CCM weight measured. The pay for performance scheme was group) was selected among all patients enrolled in the the same throughout the study period. The dedicated CCM project by their GP because of chronic HF (Fig. 1, nurse had a key role in the project, as she was group A). Among them, only patients classified as Fig. 1 Top panel: Kaplan-Meier curves showing cumulative survival probability in the chronic care model (CCM, blue curve) group and in the control group (red curve). Bottom panel: Kaplan-Meier curves showing cumulative event-free survival probability in the chronic care model (CCM) group and in the control group Ballo et al. BMC Health Services Research (2018) 18:388 Page 4 of 8 having a definite diagnosis of chronic HF by administra- all-cause mortality, only the first event was considered in tive data (Fig. 1, group B) were included in the CCM survival analyses. The age classes used in survival analyses group (Fig. 1, group C). The following administrative were identified using predefined cut-offs (< 75, 75 to 85, data were considered: one or more hospital discharges and > 85 years). A p < 0.05 was considered significant. All with primary ICD9 code indicating HF (428, 3981, analyses were performed using STATA, ver. 12 (STATA 40201, 40211, 40291, 40401, 40403, 40411, 40413, Corp., College Station, TX, USA). 40491, 40493), exemption to payment because of chronic HF . In this group, complete data were available for Results 1761 (94.3%) patients, which formed the final CCM Main characteristics study cohort. The unexposed cohort (control group) was Main characteristics of exposed and unexposed subjects selected among all patients with a diagnosis of HF by are shown in Table 1. As a result of the matching pro- the same administrative data and who were assisted by cedure, the two groups showed equal percent distribu- GPs not adhering to the CCM project (Fig. 1, group D). tions of age class, gender, Charlson index, use of main To minimize the risk of selection bias, exposed and un- cardiovascular pharmacological classes, local health exposed subjects were exactly matched with a 1:2 ratio Table 1 Main characteristics for age class, gender, Charlson comorbidity index (a vali- CCM group Control group Prevalence dated and widely used prognostic score related to the (n = 1761) (n = 3522) number and severity of comorbidities), geographic area Female gender 763 1526 43.3% of living (defined as the local health authority), treat- Age class ment with ACE-inhibitors and/or ARBs, beta-blockers, < 75 559 1118 31.7% and diuretics, and history of hospitalization for HF dur- ing the previous 5 years (between 2005 and 2009). We 75-85 766 1532 43.5% used a frequentist matching method, randomly selecting > 85 436 872 24.8% two unexposed subjects in each stratum given by the Charlson comorbidity index matching variable combinations. The final CCM and 0 592 1184 33.6% control group included 1761 and 3522 patients, respect- 1 366 732 20.8% ively. All subjects were followed from the baseline 2 803 1606 45.6% (discharge) until death, readmission or end of follow-up period (4 years), whichever came first. No ethical approval Treatment at enrolment was needed for this type of study. For patients in both ACE-inhibitors or ARBs 1526 3052 86.7% groups, all data were extracted from an administrative Beta-blockers 1036 2072 58.8% archive (Health Informative Database of Tuscany Region, Diuretics 1484 2968 84.3% Italy), using an anonymous code to link subjects between HF hospitalization between 2005 and 2009 different databases (hospitalizations, drugs, mortality). 0 181 362 10.3% Endpoints 1 322 644 18.3% Two different endpoints were considered in this study: ≥ 2 1258 2516 71.4% 1) hospitalization for HF; 2) all-cause mortality. The Local Health Unit follow-up encompassed a 4-year period, from January 1, 103 2 4 0.1% 2010 to December 31, 2013. Hospitalizations for HF 104 121 242 6.9% were identified by considering all discharges for the 105 1 2 0.1% Aggregate Clinical Code 108, which is obtained by grouping the following ICD-9 codes: 39891, 4280, 4281, 106 124 248 7.0% 42820-42823, 42830-42833, 42840-42843, 429. 107 179 358 10.2% 108 268 536 15.2% Statistical analysis 109 267 534 15.2% Data for categorical variables were shown as numbers 110 687 1374 39.0% (percentages). Incidence rates and incidence rate ratios 111 74 148 4.2% (IRR, exposed vs unexposed) were calculated for each clin- ical endpoint. Event-free survival analysis was performed 112 38 76 2.2% by multivariable Cox regression, to assess the impact of Main characteristics of the chronic care model group. The P value was 1 for all comparisons, as a result of the matching procedure. ARBs angiotensin receptor CCM project on the risk of HF hospitalization and mor- blockers, ACE angiotensin converting enzyme, HF heart failure tality. Analyses were carried out adjusting for death com- a Equal prevalences in both groups as a result of the exact matching peting risk. For the endpoints of HF hospitalization and The majority of patients were treated with multiple medications at enrolment Ballo et al. BMC Health Services Research (2018) 18:388 Page 5 of 8 authority of living, and number of hospitalizations for Table 3 Event-free survival analysis HF during the 2005-2009 period. The median follow-up Hazard ratio 95% CI P value time was 1148 [1124-1174] days in the CCM group and CCM status 1.35 1.19-1.52 < 0.0001 1045 [1024-1066] days in the controls. During the period a Age class of study, an increase in the proportions of subjects who 75-85 1.40 1.21-1.63 < 0.0001 were treated with beta-blockers and with ACE inhibi- > 85 1.59 1.33-1.89 < 0.0001 tors/angiotensin receptor blockers, and who had creatin- Female gender 0.96 0.85-1.09 0.51 ine and electrolyte blood tests checked was observed in both groups (Table 2). After adjusting for pre- Charlson index 1.16 1.07-1.26 < 0.0001 intervention ones, the final values were significantly Treatment at enrolment higher in the CCM group for beta-blocker therapy (IRR ACE-inhibitors or ARBs 1.09 0.90-1.31 0.38 1.07 [1.03-1.12], p < 0.0001) and for creatinine and elec- Beta-blockers 1.24 1.09-1.41 0.0010 trolyte blood tests (IRR 1.20 [1.16-1.25], p < 0.0001). Diuretics 2.38 1.87-3.03 < 0.0001 Geographic area 1.00 0.97-1.03 0.99 HF hospitalization During the follow-up, there were 713 hospitalizations for Previous HF hospitalization 1.11 0.99-1.25 0.069 HF in 432 patients within the. Predictors of hospitalization in the overall study population, as identified by multivariable Cox regression analysis. ACE angiotensin converting enzyme, ARBs CCM group (12.1 events per 100 patient-years) and 1135 angiotensin receptor blockers, CCM chronic care model, CI confidence interval hospitalizations in 657 patients within the control group a Hazard ratios calculated vs age < 75 years as a reference (10.3 events per 100 patient-years), indicating a higher incidenceintheCCM groupthaninthe controls (IRR 1.15 significantly higher rate of planned hospitalizations (HR 1. [1.05-1.27], p = 0.0030). Mean length of stay was 8.8 days in 50 [1.15-1.94], p < 0.0001). This effect was more evident the CCM group and 8.1 days in the controls, corresponding than that observed on the rate of urgent hospitalizations to 1.07 and 0.84 days per patient-year, respectively (IRR 1. (hospitalizations for HF: HR 1.29 [1.13-1.46], p < 0.0001). 25 [1.21-1.29], p < 0.0001). Multivariable analysis (Table 3) showed that CCM status was independently associated Mortality with 35% higher probability of HF hospitalization (HR 1. There were 632 deaths in the CCM group (10.8 events per 35, 95% CI 1.19-1.52, p < 0.0001). The curves showing 100 patient-years) and 1393 deaths in the control group theadjustedcumulativesurvivalprobability inthetwo (12.6 events per 100 patient-years; IRR 0.82 [0.75-0.91], p groups are shown in Fig. 1,top panel. < 0.0001). Univariable Cox regression in the overall popu- After a first hospitalization for HF, no difference was ob- lation showed that CCM status was associated with 15% served between the two groups in the risk of further HF lower risk of death (HR 0.85, 95% CI 0.78-0.94, p =0.001). hospitalizations (IRR 1.001 [0.89-1.13], p = 0.98). The rate Multivariable analysis confirmed that CCM status was in- of 30-day HF readmissions after a first HF hospitalization dependently associated with a 18% risk reduction in mor- also showed no significant difference between the CCM tality (Table 4). Adjusted cumulative survival probabilities group and the controls (4.9% vs 5.9%, p =0.14). in the two groups are shown in Fig. 1,bottom panel. Inter- When planned and urgent HF hospitalizations were estingly, even after a hospitalization for HF, patients in the considered separately, CCM status was associated with a CCM group still showed a 16% lower risk of death than the controls (HR 0.84 [0.71-0.99], p <0.05). Table 2 Process and therapeutic indicators Discussion CCM group (n = 1761) Control group (n = 3522) HF is the most common cause of hospitalization in Creatinine and electrolyte tests Western countries, particularly in patients over the age Rate 2005-2009 55.8% 52.4% of 65, and represents a major challenge to the health Rate 2010-2013 80.7% 65.3% care systems. In outpatients with chronic HF, a Beta-blocker therapy hospitalization is one of the strongest prognostic predic- Rate 2005-2009 45.0% 43.0% tors for increased mortality, and unplanned readmissions Rate 2010-2013 65.1% 59.5% arouse a high financial burden . An adequate know- ledge of the precipitants of rehospitalisation in these ACE/ARBs therapy patients is therefore of major importance . Besides Rate 2005-2009 78.2% 76.7% classical clinical factors such as myocardial ischemia, Rate 2010-2013 81.0% 80.2% atrial fibrillation, uncontrolled hypertension, and exacer- Rates of diagnostic and therapeutic indicators in the periods before and after bations of chronic obstructive pulmonary disease or chronic care model (CCM) project implementation among exposed and unexposed subjects. ARBs angiotensin receptor blockers, ACE angiotensin converting enzyme infections, non-clinical determinants of hospitalization Ballo et al. BMC Health Services Research (2018) 18:388 Page 6 of 8 Table 4 Survival analysis our CCM program involved primary care physicians - therefore being somewhat different from disease-centric Hazard ratio 95% CI P value systems of care/interventions as we know them from the CCM status 0.82 0.75-0.91 < 0.0001 a hospital/specialist point of view - we can hypothesize Age class that they improved their awareness of HF patients and 75-85 1.89 1.67-2.15 < 0.0001 tended to assess their clinical status following clinical > 85 3.59 3.14-4.09 < 0.0001 pathways and using more facilities, including Female gender 0.87 0.79-0.95 0.0020 hospitalization. This hypothesis is supported by the evi- Charlson index 1.39 1.31-1.48 < 0.0001 dence that the CCM status was associated with a 50% increase in the rate of planned HF hospitalizations, Treatment at enrolment whereas the effect on the rate of urgent hospitalizations ACE-inhibitors or ARBs 0.77 0.68-0.86 < 0.0001 was considerably smaller. Also, while it cannot be Beta-blockers 0.81 0.74-0.89 < 0.0001 excluded that the adjustment for death competing risk in Diuretics 1.98 1.67-2.34 < 0.0001 our analysis was not able to completely remove the in- Geographic area 1.01 0.98-1.03 0.62 creased probability of hospitalization resulting from the Previous HF hospitalization 1.12 1.02-1.22 0.015 increased number of survivors, the possibility that not every hospitalization must be considered as a poor out- Predictors of death in the overall study population, as identified by multivariable Cox regression analysis. ACE angiotensin converting enzyme, ARBs angiotensin come - especially from the patient’s point of view - should receptor blockers, CCM chronic care model, CI confidence interval a be considered. In a study of intensive primary care follow- Hazard ratios calculated vs age < 75 as a reference up following a discharge for chronic diseases including (e.g., inadequate access to follow-up care or medications HF, admissions actually increased though patients rated and poor transitions of care) are progressively growing their health better . With this in mind, it is also inter- in importance . In this regard, the implementation of esting to observe that no differences in the risk of further strategies aimed at improving the quality of health care HF hospitalizations or the rate of 30-day HF readmissions delivery for patients with chronic HF may be of clinical were found between groups in our study. interest. Although the true prognostic impact of the Regardless of the factors underlying this divergent CCM still lacks consistent evidence of benefit across all trend, the finding of an increase in hospitalization rates medical conditions [15, 30], a potential positive effect on among patients enrolled in CCM project carried out in a clinical outcome has been reported in various chronic primary care setting – where the access to specialist diseases [31–35]. This study explored the effect of a consultations is not so direct as in hospitals or other fa- CCM-based, regional program for patients with chronic cilities – might support the potential clinical utility of HF applied in primary care over a 4-year follow-up. Our multidisciplinary approaches to be integrated within pri- findings show that patients enrolled in the program mary care CCM models. The role of multimorbidity in showed a lower risk of death but a higher risk of affecting the risk of hospitalization in patients with hospitalization for HF than a matched control population. chronic diseases is established, particularly in the elderly Previous studies, mostly performed in hospital set- [38, 39]. Though our CCM project was not designed to tings, reported that the application of the CCM for the directly involve specialists in patients’ management, it management of patients with chronic HF could lead to could be hypothesized that such a multidisciplinary potential beneficial effects on outcome, although with strategy, by means of providing a larger number of diag- some heterogeneity in effectiveness [17–21]. The im- nostic and therapeutic pathways to the referring general proved survival observed in this study adds to these evi- physician, might potentially reduce the need for dences by suggesting that these potential benefits might hospitalization. These concepts could also be applied be extended to a chronic HF population followed in pri- not only to the doctors involved in the management of mary care. This finding could suggest a higher quality of HF patients, but also to the nurses, who represent a key care for the patients in the CCM group, and potentially component in any CCM-based health care system . better cooperation between cardiologists at the hospital Our results might also have implications for the use of and the GP. However, the finding of an opposite trend both hospitalization and rehospitalization as measures of for HF hospitalization and mortality may be somewhat quality of care, suggesting that they should be used with surprising. Interestingly, a similar discrepancy was also caution. The limitations of these outcome measures are previously reported in the Veterans Affairs Health Care established, including the need of a time window that System, where mortality and HF hospitalization rates must be appropriate for the type of disease, the effect of showed a definite trend in opposite directions . case-mix factors, the competing risk of mortality, and Several explanations could be proposed for the lack of the fact that admissions must be avoidable and un- decline in HF hospitalization in our population. Since planned . Also, the possibility of a residual Ballo et al. BMC Health Services Research (2018) 18:388 Page 7 of 8 confounding role of measured and unmeasured individual covariates in multivariable analyses because we were in- factors that affect the likelihood of hospitalization – not terested in exploring the effect of each single variable. necessarily related to the quality of clinical care, such as The generalisability of the findings should also be con- social support, geographic location, and socioeconomics – sidered with caution. Lastly, our CCM program hinged should be taken into account [42, 43]. Accordingly, while on general physicians and dedicated nurses. As stated most efforts at discharge commonly focus on managing above, a multidisciplinary approach involving other congestion and close hemodynamic monitoring to reduce professional figures would be largely preferable for a early readmissions, broader strategies to treat HF-related complex disease such as chronic HF, particularly consid- comorbidities and patient-centered management may ering the clinical relevance of comorbidities in the probably be useful in the perspective of hospitalizations practical management of these subjects. Such a multidis- over a longer time [44, 45]. ciplinary strategy could also be useful to provide a larger Lastly, the complexity of the mechanisms underlying a number of diagnostic and therapeutic pathways to the hospitalization event should be taken into account. referring general physician, with a potential beneficial Interestingly, it has been previously written that “Pa- impact on the need of hospitalization. tients are readmitted, and not diagnoses” . Not only biological factors such as inadequate treatment, comor- Conclusions bidities, and progression of disease can deeply affect the In conclusion, the implementation of a regional health rate of hospitalization, but also a number of actors –hos- care program for patients with chronic HF, based on a pital and primary care physicians, outpatient caregivers, proactive CCM strategy carried out in a primary care the patients themselves, and specific interventions/ setting, finally yielded a higher risk of HF hospitalization organizational characteristics – can exert a strict impact and an improved survival. These findings might have im- on hospitalization rates. For example, in a large study plications for the potential improvement of similar about HF hospitalizations in the United States, where CCM-based programs, highlight the importance of a the total number of HF-related hospitalizations signifi- critical assessment of hospitalization as a measure of cantly increased from 2001 to 2009, primary HF hospi- outcome, and might support the need for multidisciplin- talizations steadily decreased, whereas the total number ary strategies aimed at optimizing the management of of secondary HF hospitalizations increased by nearly CCM patients in terms of hospitalizations. 400,000 . The authors argued about the hypothesis Abbreviations that these findings might be related to shifting in coding CCM: Chronic care model; GP: General practitioner; HF: Heart failure; practices, also because of incentives, as it happened for HR: Hazard ratio; IRR: Incidence rate ratio the downcoding of pneumonia hospitalizations . Funding Besides opportunistic behaviours by hospital coders, it This study received no specific grant from any funding agency in the public, must be recognized that “hospitalizations for HF does commercial or not-for-profit sectors. not equate to hospitalizations because of HF” . Availability of data and materials Some limitations should be highlighted in this study. The data that support the findings of this study are available from the authors Because the enrolment in the CCM program required upon reasonable request, subject to permission of the Regional Health Agency of Tuscany. that the patient was able to adequately follow the clinical visitations scheduled in the program, the possibility of a Authors’ contributions selection bias - related to the potential enrolment of pa- All the authors conceived of the study and established the methodology. FP and LP obtained and linked the data used in this study, and conducted tients with lower severity of HF - should be considered. the statistical analyses with inputs from PB, AZ, and PF. PB wrote the first This potential selection bias could have favoured a draft. FP, LP, LR, PF and AZ reviewed this for important intellectual content. reduction in mortality in the CCM group. Due to the ad- All the authors read and approved the final version. ministrative nature of the data, we were not able to con- Ethics approval and consent to participate sider some potentially relevant clinical variables in the Not needed for this type of study. analysis. Then, we cannot exclude that the observed differences in mortality and hospitalization between the Competing interests The authors declare that they have no competing interests. CCM group and the controls were influenced by other covariates. In this regard, information on HF degree Publisher’sNote measures, co-medications, other disease severity mea- Springer Nature remains neutral with regard to jurisdictional claims in published sures, adherence to therapy and discharge plans was not maps and institutional affiliations. available. It should also be highlighted that, in this study, Author details we did not adjust for the matched pair design, since we Cardiology Unit, S. Maria Annunziata Hospital, via dell’Antella 58, Florence, only would balance the two groups. Moreover, despite 2 3 Italy. Regional Health Agency of Tuscany, Florence, Italy. Tuscany Region, the matching procedure, we decided to use the Florence, Italy. Ballo et al. BMC Health Services Research (2018) 18:388 Page 8 of 8 Received: 21 January 2018 Accepted: 30 April 2018 26. International classification of diseases, Ninth revision, Clinical modification (ICD-9-CM). Available at: https://www.cdc.gov/nchs/icd/icd9cm.htm 27. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360:1418–28. References 28. Fonarow GC, Abraham WT, Albert NM, et al. Factors identified as 1. Rahimi K, Duncan M, Pitcher A, et al. Mortality from heart failure, acute precipitating hospital admissions for heart failure and clinical outcomes: myocardial infarction and other ischaemic heart disease in England and findings from OPTIMIZE-HF. Arch Intern Med. 2008;168:847–54. Oxford: a trend study of multiple-cause-coded death certification. 29. Gheorghiade M, Vaduganathan M, Fonarow GC, Bonow RO. J Epidemiol Community Health. 2015;69:1000–5. Rehospitalization for heart failure: problems and perspectives. J Am Coll 2. Nabel EG, Braunwald E. A tale of coronary artery disease and myocardial Cardiol. 2013;61:391–403. infarction. N Engl J Med. 2012;366:54–63. 30. Homer CJ, Forbes P, Horvitz L, Peterson LE, Wypij D, Heinrich P. Impact of a 3. Roger VL, Go AS, Lloyd-Jones DM. Heart disease and stroke statistics—2012 quality improvement program on care and outcomes for children with update: a report from the American Heart Association. Circulation. 2012;125: asthma. Arch Pediatr Adolesc Med. 2005;159:464–9. e2–e220. 31. Stellefson M, Dipnarine K, Stopka C. The chronic care model and diabetes 4. National Heart, Lung and Blood Institute. Morbidity and mortality: 2012 management in US primary care settings: a systematic review. Prev Chronic chart book on cardiovascular, lung, and blood diseases. 2012. Dis. 2013;10:E26. 5. Ambrosy AP, Fonarow GC, Butler J, et al. The Global Health and economic 32. Mipatrini D, Sinopoli A, Sestili C, Di Marcoberardino M, Giuliani P, Grasso G, burden of hospitalizations for heart failure: lessons Learned from HHF et al. Protocol for the evaluation of a chronic care model experience in registries. J Am Coll Cardiol. 2014;63:1123–33. Rome. Clin Ter. 2017;168:e317–9. 6. Roger VL, Go AS, Lloyd-Jones DM, et al. Executive summary: heart disease 33. Mangione-Smith R, Schonlau M, Chan KS, Keesey J, Rosen M, Louis TA, et al. and stroke statistics—2012 update: a report from the American Heart Measuring the effectiveness of a collaborative for quality improvement in Association. Circulation. 2012;125:188–97. pediatric asthma care: does implementing the chronic care model improve 7. Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93: processes and outcomes of care? Ambul Pediatr. 2005;5:75–82. 1137–46. 34. Vargas RB, Mangione CM, Asch S, Keesey J, Rosen M, Schonlau M, Keeler EB. 8. Curtis LH, Whellan DJ, Hammill BG, et al. Incidence and prevalence of heart Can a chronic care model collaborative reduce heart disease risk in patients failure in elderly persons, 1994–2003. Arch Intern Med. 2008;168:418–24. with diabetes? J Gen Intern Med. 2007;22:215–22. 9. Levy D, Kenchaiah S, Larson MG, et al. Long-term trends in the incidence of 35. Jeemon P, Harikrishnan S, Sanjay G, Sivasubramonian S, Lekha TR, and survival with heart failure. N Engl J Med. 2002;347:1397–402. Padmanabhan S, Tandon N, Prabhakaran D. A PROgramme of lifestyle 10. Zarrinkoub R, Wettermark B, Wändell P, et al. The epidemiology of heart intervention in families for cardiovascular risk reduction (PROLIFIC study): failure, based on data for 2.1 million inhabitants in Sweden. Eur J Heart Fail. design and rationale of a family based randomized controlled trial in 2013;15:995–1002. individuals with family history of premature coronary heart disease. BMC 11. Ross JS, Chen J, Lin Z. Recent national trends in readmission rates after Public Health. 2017;17:10. heart failure hospitalizations. Circ Heart Fail. 2010;3:97–103. 36. Heidenreich PA, Sahay A, Kapoor JR, et al. Divergent trends in survival and 12. Ohlsson A, Lindahl B, Hanning M, et al. Inequity of access to ACE inhibitors readmission following a hospitalization for heart failure in the veterans in Swedish heart failure patients: a register-based study. J Epidemiol affairs health care system 2002 to 2006. J Am Coll Cardiol. 2010;56:362–8. Community Health. 2016;70:97–103. 37. Weinberger M, Oddone EZ, Henderson WG. Does increased access to 13. Cook C, Cole G, Asaria P, Jabbour R, Francis DP. The annual global primary care reduce hospital readmissions? Veterans affairs cooperative economic burden of heart failure. Int J Cardiol. 2014;171:368–76. study group on primary care and hospital readmission. N Engl J Med. 1996; 14. Brouwer ED, Watkins D, Olson Z, Goett J, Nugent R, Levin C. Provider costs 334:1441–7. for prevention and treatment of cardiovascular and related conditions in 38. Leijten FR, de Wind A, van den Heuvel SG, et al. The influence of chronic low- and middle-income countries: a systematic review. BMC Public Health. health problems and work-related factors on loss of paid employment 2015;15:1183. among older workers. J Epidemiol Community Health. 2015;69:1058–65. 15. Coleman K, Austin BT, Brach C, et al. Evidence on the chronic care model in 39. Marshall A, Nazroo J, Tampubolon G, et al. Cohort differences in the levels the new millennium. Health Aff. 2009;28:75–85. and trajectories of frailty among older people in England. J Epidemiol 16. Zhang Y, Tang W, Zhang Y, Liu L, Zhang L. Effects of integrated chronic Community Health. 2015;69:316–21. care models on hypertension outcomes and spending: a multi-town 40. Thompson DR. Cardiovascular nursing: from Florence to Melbourne. Heart clustered randomized trial in China. BMC Public Health. 2017;17:244. Lung Circ. 2016;25:881–4. 17. Kastner M, Lillie E, Ashoor H, et al. Quality improvement strategies to 41. Fischer C, Anema HA, Klazinga NS. The validity of indicators for assessing optimise transition of patients with heart failure to independent living: quality of care: a review of the European literature on hospital readmission protocol for a scoping review. BMJ Open. 2014;4:e005711. rate. Eur J Pub Health. 2012;22:484–91. 18. Bosch M, van der Weijden T, et al. Structured chronic primary care and 42. Desai AS, Stevenson LW. Rehospitalization for heart failure: predict or health-related quality of life in chronic heart failure. BMC Health Serv Res. prevent? Circulation. 2012;126:501–6. 2009;9:104. 43. Joynt KE, Orav EJ, Jha AK. Thirty-day readmission rates for Medicare 19. van Lieshout J, Steenkamer B, Knippenberg M, Wensing M. Improvement of beneficiaries by race and site of care. JAMA. 2011;305:675–81. primary care for patients with chronic heart failure: a study protocol for a 44. Gorodeski EZ, Starling RC, Blackstone EH. Are all readmissions bad cluster randomised trial comparing two strategies. Implement Sci. 2011;6:28. readmissions? N Engl J Med. 2010;363:297–8. 20. Asch SM, Baker DW, Keesey JW, Broder M, Schonlau M, Rosen M, et al. Does 45. Dronavalli M, Thompson SC. A systematic review of measurement tools of the collaborative model improve care for chronic heart failure? Med Care. health and well-being for evaluating community-based interventions. 2005;43:667–75. J Epidemiol Community Health. 2015;69:805–15. 21. Drewes HW, Steuten LM, Lemmens LC, et al. The effectiveness of chronic 46. Drozda JP Jr. Readmission rates. Edging slowly towards a deeper care management for heart failure: meta-regression analyses to explain the understanding and ultimately better care for patients. BMJ. 2013;347:f7478. heterogeneity in outcomes. Health Serv Res. 2012;47:1926–59. 47. Blecker S, Paul M, Taksler G, et al. Heart failure–associated hospitalizations in 22. Murray CJL. The state of US health, 1990–2010: burden of diseases, injuries, the United States. J Am Coll Cardiol. 2013;61:1259–67. and risk factors. JAMA. 2013;310:591–608. 48. Lindenauer PK, Lagu T, Shieh MS, et al. Association of diagnostic coding 23. Schäfer W, Kroneman M, Boerma W, et al. The Netherlands: health system with trends in hospitalizations and mortality of patients with pneumonia, review. Health Syst Transit. 2010;12:1–228. 2003–2009. JAMA. 2012;307:1405–13. 24. Heredia NI, Lee M, Reininger BM. Exposure to a community-wide campaign 49. Roger VL. The changing landscape of heart failure hospitalizations. J Am is associated with physical activity and sedentary behavior among Hispanic Coll Cardiol. 2013;61:1268–70. adults on the Texas-Mexico border. BMC Public Health. 2017;17:883. 25. Bauman AE, Grunseit AC, Rangul V, Heitmann BL. Physical activity, obesity and mortality: does pattern of physical activity have stronger epidemiological associations? BMC Public Health. 2017;17:788.
BMC Health Services Research – Springer Journals
Published: May 30, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera