Are degree of urbanisation and travel times to healthcare services associated with the processes of care and outcomes of heart failure? A retrospective cohort study based on administrative data
Jacopo Lenzi aff001; Vera Maria Avaldi aff001; Dario Molinazzi aff003; Carlo Descovich aff002; Stefano Urbinati aff004; Veronica Cappelli aff005; Maria Pia Fantini aff001
Authors place of work:
Department of Biomedical and Neuromotor Sciences, Alma Mater Studiorum – University of Bologna, Bologna, Italy
aff001; Department of Clinical Governance and Quality, Bologna Local Healthcare Authority, Bologna, Italy
aff002; Department of Management Control and Administrative Data, Bologna Local Healthcare Authority, Bologna, Italy
aff003; Department of Cardiology, Bellaria Hospital, Bologna, Italy
aff004; Directorate of Assistance, Technology and Rehabilitation, Bologna Local Healthcare Authority, Bologna, Italy
Published in the journal:
PLoS ONE 14(10)
A few studies have found that patients with heart failure (HF) living in less densely populated areas have reduced use of services and poorer outcomes. However, there is a lack of evidence regarding transport accessibility measured as the actual distance between the patient’s home and the healthcare facility. The aim of this study was to investigate if different urbanisation levels and travel times to healthcare services are associated with the processes of care and the outcomes of HF. This retrospective cohort study included patients residing in the Local Healthcare Authority of Bologna (2915 square kilometres) who were discharged from hospital with a diagnosis of HF between 1 January and 31 December 2017. Six-month study outcomes included both process (cardiology follow-up visits) and outcome measures (all-cause readmissions, emergency room visits, all-cause mortality). Of the 2022 study patients, 963 (47.6%) lived in urban areas, 639 (31.6%) in intermediate density areas, and 420 (20.8%) in rural communities. Most patients lived ≤30 minutes away from the nearest healthcare facility, either inpatient or outpatient. After controlling for a number of individual factors, no significant association between travel times and outcomes was present. However, rural patients as opposed to urban patients were more likely to see a cardiologist during follow-up (OR 1.42, 99% CI 1.03–1.96). These follow-up visits were associated with reduced mortality within 6 months of discharge (OR 0.53, 99% CI 0.32–0.87). We also found that multidisciplinary interventions for HF were more common in rural than in urban settings (18.8% vs. 4.0%). In conclusion, travel times had no impact on the quality of care for patients with HF. Differences between urban and rural patients were possibly mediated by more proximal factors, some of which are potential targets for intervention such as the availability and utilisation of follow-up cardiology services and multidisciplinary models of care.
Allied health care professionals – Cardiology – Critical care and emergency medicine – Health care facilities – Heart failure – Hospitals – Inpatients – Outpatients
Heart failure (HF) is a complex clinical syndrome with a prevalence of 1 to 2% in the adult population of Western countries, a value that exceeds 10% in individuals over 70 years of age [1,2]. The prognosis is poor, with an in-hospital mortality of 10% and a one-year mortality after discharge of 20 to 40% [3–8]. HF is one of the main reasons for hospital admission in Europe and the US , especially for patients over 65 years , and the readmission rate exceeds 20% and 50% within a month and a year of hospital discharge, respectively [11,12]. Patients require a care approach that takes into account all their complex care needs , and indeed it has been shown that easy access to care is a significant determinant of their outcomes .
A recent systematic review regarding accessibility in terms of distance and travel time from the patient’s home, showed that individuals living near healthcare facilities have better health outcomes or a higher rate of access to services than those living further away . However, this review included mainly cancer research studies relying on different data sources and variables.
With respect to HF, a few studies found that living in less densely populated areas was associated with reduced use of services and poorer outcomes [16–18]. These findings should be read keeping in mind that geographic position might have a close correlation with a large number of variables, such as socioeconomic status (SES), availability of appropriate services, and even peculiar clinical conditions. A potential limitation of these studies is the lack of information regarding transport accessibility measured as the actual distance between the patient’s home and the healthcare facility.
The Local Healthcare Authority (LHA) of Bologna, located in the Emilia-Romagna region of Italy, has various hospitals and primary care services in its territory. Over the past few years, an increasing number of healthcare homes and ambulatory care nursing practices have been established, where patients are medically examined and receive outpatient nursing interventions for lifestyle change or for the control and management of their own health problems. Many of the interventions included in the care pathways (CPs), such as the heart failure CP (HF-CP), are also provided within these care settings. This multidisciplinary and standardised CP promotes the integration of different services and professionals, and aims to improve the health status of the patients with HF through a direct and easy access to care.
Despite the large number of healthcare facilities and the efforts to improve patients’ access to services, the catchment area of Bologna has a relatively diverse geography and uneven population distribution. Therefore, the objective of this study was to investigate whether different travel times to healthcare services in the LHA of Bologna are associated with the processes of care and the outcomes of patients with HF. In keeping with existing literature, we also evaluated the impact of urbanisation levels on the study processes and outcomes.
Materials and methods
Setting and study population
This retrospective observational study included all hospital discharges in the LHA of Bologna with a primary diagnosis of HF (ICD-9-CM codes: 398.91, 402.x1, 404.x1, 404.x3, 428.xx) between 1 January and 31 December 2017. The LHA of Bologna is located in the northeast of Italy, has a population of about 876,000 and covers an area of 2915 km2, with a territory that is 29% mountains (the Apennines), 32% hills, and 39% plains (the Po Valley) (Figure A in S1 Fig).
Data were retrieved from the Hospital Discharge Records (HDRs) Database (this and the other data sources used in this study are described in S1 Table) . For patients with multiple eligible hospital admissions over the one-year study period, we considered the first one as the index admission. Repeated admissions within one day of discharge were regarded as one single episode of care, and the beginning of the follow-up was set at the discharge date of the episode of care. All patients were followed up to 6 months.
Patients were excluded if any of the following criteria were met (S2 Fig):
Permanent and/or current address outside the LHA catchment area
Registered at a general practitioner (GP) who practiced outside the catchment area
Age >100 years, because very old patients may have distinctive clinical features at diagnosis and survival
Planned hospital admission, to focus analyses on acute/urgent episodes of care
Transfer from another facility, to focus analyses on incident cases of HF
Daytime hospital care, i.e., one-day admissions to the hospital without overnight stay to perform diagnostic procedures and/or surgical, therapeutic or rehabilitative care
A secondary diagnosis of non-cardiogenic acute pulmonary oedema (ICD-9-CM 518.4), i.e., patients with symptoms probably related to causes other than HF
A secondary diagnosis of acute kidney failure (ICD-9-CM 584.x), i.e., patients whose reason for hospitalisation is likely not to be HF
Pregnancy, childbirth or puerperium (Major Diagnostic Category 14)
A major procedure on the cardiovascular system (ICD-9-CM 00.5x, 00.66, 35.xx, 36.xx, 37.31–37.66, 37.70–37.89, 37.94–37.98), i.e., patients with severe cardiac impairment as the main reason for hospitalization
Death during the index episode of care
Discharge against medical advice
Length of stay >90 days, i.e., very complex or unstable cases
Access to residential care facility for the elderly before index hospitalisation or during follow-up, i.e., patients being given end-of-life care.
The 6-month outcomes evaluated in this study included both process and outcome measures:
Cardiology follow-up visits provided in either outpatient or inpatient cardiology services, except for those booked prior the index admission or performed during hospital stays (source: Outpatient Care Database [OCD])
All-cause unplanned readmissions occurred at any hospital within 2 to 180 days of discharge, and lasting >1 day (source: HDRs)
Emergency room (ER) visits not related to injuries and not resulting in inpatient admission (source: ER database)
Using the Eurostat’s Degree of Urbanisation (DEGURBA) classification system (revised definition, 2014), the 45 municipalities (comuni) where the patients lived were subdivided into rural areas (alternative name: sparsely populated areas), towns or suburbs (intermediate density areas), and cities (densely populated areas). As illustrated in Figure B in S1 Fig, the city of Bologna was classified as urban (388,000 pop., 44%), the nearby comuni and other areas in the Po Valley were classified as towns or suburbs (304,000 pop., 35%), while the remaining comuni—both flat and mountainous—were classified as rural (184,000 pop., 21%).
Travel times to healthcare services
Because of the catchment area’s diverse geography, we calculated the travel times between the patients’ home addresses (source: civil registry) and a series of healthcare facilities that provide care for patients with HF. These included:
Emergency rooms (N = 12)
Cardiology wards (N = 5)
Outpatient cardiology services (N = 27)
Ambulatory care nursing practices (N = 34), 15 of which are located in the healthcare homes
GP practices (N = 739), run by a total of 525 GPs.
Due to proximity to the border, three hospitals providing emergency and/or cardiology care outside the LHA catchment area were included in the study.
After geocoding all locations in the WGS84 spatial reference by means of the Stata opencagegeo package, the travel times between geographic coordinates were computed using the georoute package [20,21]. georoute calculates how long it takes to drive the distance between two points under average traffic conditions .
To account for potential nonlinear relationships with the outcomes, travel times were split into five categories: ≤5 min (very short), >5–10 min (short), >10–20 min (medium), >20–30 min (long), and >30 min (very long).
In addition to the degree of urbanisation, different sets of routing distances were considered as the potential predictors of each study outcome. When we analysed cardiology follow-up visits, the travel time of interest was to the nearest cardiology service, either inpatient or outpatient. In all the other analyses, we considered three distinct travel times: to the nearest ER, the nearest patient’s GP practice, and the nearest outpatient service, either cardiologist or non-cardiologist. Moreover, follow-up cardiology visits were treated as potential predictors of hospital readmissions, ER visits and mortality [22–24]. An overview of these sets of variables is provided in S2 Table.
We collected some patient baseline characteristics to reduce the potential source of confounding. These included:
Length of stay
Provision of intensive care during hospital stay
Discharge from a cardiology ward
Thirty-one Elixhauser conditions identified in the index episode of care and in all hospital admissions occurring two-years prior to the index hospitalisation , plus four additional conditions not included in the Elixhauser’s list (myocardial infarction [ICD-9-CM 410.x, 412], cerebrovascular diseases [362.34, 430.x–438.x], dementia [290.x, 294.1, 331.2], leukaemia [204.x-208.x])
Use of 10 drug therapies one-year prior to the index admission (≥1 filled prescription) (source: Outpatient Pharmaceutical Database).
See Tables A and B in S3 Table for the detailed list of drug therapies and Elixhauser comorbidities. We also took into account the following information in the analyses [19,26–29]:
Use of first-line medications during follow-up, that is, angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs) and β-blockers (only for hospital readmission, ER visit and mortality analyses)
Registration at occasional or full-time general home-care services as a proxy of social and medical complexity
Registration at the HF-CP, a structured multidisciplinary care plan that promotes integration between primary and secondary care, and details essential steps in the care of patients. The GP remains the gatekeeper for the patients and coordinates with cardiologists and nurses for an easier access to consultation and counselling to improve lifestyle and optimise medication adherence. The HF-CP can involve either outpatient clinic-based interventions (clinic-based HF-CP) or home visits (home-based HF-CP), depending on the patient’s clinical condition and ability to move. There are no facilities specifically dedicated to the HF-CP: registered patients can access GP practices, cardiology services or ambulatory care nursing practices in case they need medical care or counselling.
The patient could already be registered at the beginning of the follow-up, or access the outpatient services following discharge. Information on outpatient care was collected from regional and LHA administrative databases. All the data sources used in this study are de-identified and linkable using the unique patient identifier.
As previously mentioned, cardiology visits during follow-up were included in regression analyses as potential predictors of hospital readmission, ER visits and mortality [19,22–24].
Continuous variables were summarised as mean ± standard deviation; discrete and categorical variables were summarised as frequencies and percentages. Comparisons across urbanisation levels were performed using one-way analysis of variance, Kruskal-Wallis test or chi-squared test, when appropriate. The spatial distribution of homes and healthcare facilities was graphically displayed with the aid of dot maps.
To ensure an equal time window for detecting and measuring time-varying covariates, such as outpatient care and medication use (see section above), the impact of urbanisation level and travel times on the study outcomes was assessed using a time-matched nested case-control design. Patients who experienced the study outcome were defined as cases, and 9 controls were randomly selected and matched to each case by gender, age group (defined using a decile split) and follow-up duration. This technique is called “incidence density sampling”. Odds ratios (ORs) were estimated by conditional logistic regression models to account for the matching of cases and controls .
All regression models included the potential confounders described earlier. However, to avoid overfitting and misclassification, not all comorbidities and previous drug therapies were included in the models. A subset of all candidate variables was preliminary chosen for inclusion using an automated selection method which is described in detail elsewhere [19,30]. In brief, a bootstrap procedure was adopted to determine which comorbidities were significantly associated with the outcomes. Using this approach, a backward elimination of potential confounders was applied in each replicated sample with a significance level or removal equal to 0.05, and only risk factors selected in at least 50% of the replicates were included as confounders in the final multivariable regression models. The confounders included in the final models are reported in table footnotes. The variance inflation factor, a measure of correlation among predictor variables (multicollinearity), was <3 for all of the predictors included in the models.
To control for type I error related to multiple testing, the significance level was set at 0.01. All analyses were carried out using Stata software, version 15 (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC).
Ethical approval to undertake this research was granted from the Comitato Etico di Area Vasta Emilia Centro (Submission Number 254/2019/OSS/AUSLBO).
This retrospective study was carried out in conformity with the regulations on data management with the Italian law on privacy (Legislation Decree 196/2003 amended by Legislation Decree 101/2018). Data were pseudonymised prior to the analysis at the regional statistical office, and each patient was assigned a unique identifier that eliminates the ability to trace the patient’s identity or other sensitive data. Pseudonymised administrative data can be used without a specific written informed consent when patient information is collected for healthcare management and healthcare quality evaluation and improvement (according to art. 110 on medical and biomedical and epidemiological research, Legislation Decree 101/2018).
Patients and the public were not involved in the design or planning of the study. All procedures performed in this study were in accordance with the 1964 Helsinki Declaration and its later amendments.
Of the 3138 patients discharged after HF, 2022 (64.4%) met the inclusion criteria (S2 Fig). Mean age was 82.2 ± 9.5 years and 1089 (53.9%) were females. A total of 963 (47.6%) patients lived in densely populated areas, 639 (31.6%) lived in intermediate density areas and 420 (20.8%) lived in rural areas. As shown in Table 1, patients living in rural areas were on average younger, more often registered at the HF-CP and more often discharged from internal medicine services. Specific comorbidities and previous drug therapies are summarised in Tables A and B in S4 Table.
Travel times are summarised in Table 1, while the spatial distribution of HF patients and healthcare services in the LHA catchment area is illustrated in Fig 1. Although outpatient healthcare services were relatively scattered as compared to inpatient services, most patients lived ≤30 minutes away from the nearest facility. Still, we found that travel times were significantly dependent on urbanisation, as patients living in rural communities were more distant from all healthcare services than those living in more densely populated areas.
The outcomes rates over the 6-month observation period, overall and by degree of urbanisation, are reported in Table 2. There was a lower rate of cardiology visits among patients living in Bologna. No other crude associations between urbanisation level and outcomes were found.
The impact of urbanisation levels and travel times resulting from multivariable regression analysis is presented in Tables 3 and 4. After adjusting for travel times and other patient characteristics (see table footnotes), we found that rural patients were more likely to see a cardiologist (OR 1.42, 99% CI 1.03–1.96), compared with urban patients. No other significant association between predictors and outcomes was present.
Other outcome predictors
The full multivariable regression models including travel times and processes of care, both inpatient and outpatient, are presented in Tables A and B in S5 Table. The main results can be summarised as follows:
Patients registered at the clinic-based HF-CP were more likely to be seen by a cardiologist during follow-up (OR 1.69, 99% CI 1.02–2.80)
As compared with patients with no cardiology follow-up visits, patients seen by a cardiologist during follow-up were less likely to die within 6 months of discharge (OR 0.53, 99% CI 0.32–0.87).
The main result of this retrospective cohort study was that urbanisation and travel times to healthcare services had no impact on the processes of care and the outcomes of patients with HF in the 6 months following hospital discharge. The only exception was that cardiology follow-up visits were more frequent among rural than among urban patients. We also found that patients registered at the clinic-based HF-CP were more likely to see a cardiologist during follow-up, and that outpatient cardiology care was associated with improved outcomes.
These results were obtained after controlling for several potential individual confounders. However, our administrative databases do not include very relevant clinical information, such as left ventricular ejection fraction and classification of disease severity, which can affect the patient outcomes.
To the best of our knowledge, this is the first study to evaluate whether the care of patients with HF is associated with transport accessibility to healthcare, measured as the travel time to a number of different services and facilities. A strength of our method is that the patient’s home address was the starting point to measure the travel times, while a potential limitation is that these calculations were made assuming that all patients would attend the nearest facility. Another strength is that our estimates of driving times were not only a function of driving distances [15,32], but also depended on average traffic volumes, road conditions and elevation. Nevertheless, travel time is only a component of the travel burden for these patients, who are elderly and can face a relevant combination of barriers due to disabilities, clinical conditions and comorbidities, and need for multiple medical assessments . Also, often elderly patients must be accompanied by their caregivers, whose availability depends, among other things, on the opportunity to take some time off work or to delegate the care of children. This important aspect would deserve further investigation, but administrative databases do not provide any information on informal caregivers.
In addition to travel times, we analysed the degree of urbanisation of the community where the patient lived, a measure adopted in many other studies [15,16,18,34]. As expected, we found that travel times were longer for patients living in less densely populated areas. However, contrary to other studies that found poorer outcomes among rural residents [15,16,18,35,36], we found that rural patients had more follow-up cardiology visits. The better access to follow-up care by rural patients might be explained by a seamless organisation of care that is easier to implement in non-urban communities than in metropolitan settings, where the higher number of providers and facilities can make paradoxically more complex to refer patients to the same professionals. To support this, we found that rural patients were more commonly registered at the HF-CP, whose aim is to standardise care and to define the reference professional for the patient.
Still, no other significant associations between degree of urbanisation and outcomes were present in this study. Although urbanisation level is commonly seen as a proxy for SES and other SES-related characteristics , one possible explanation is that such inequalities are not unevenly distributed across the LHA of Bologna. Another possible explanation is that the strong universal health coverage of Italy might play a role in reducing inequalities in access to healthcare, as suggested by a study evaluating the relationships between SES and HF outcomes in a universal national health service . However, we lack updated information on census track-level SES in our catchment area, and the available data are not equipped to disentangle the association between geography and SES among patients with HF.
Because geographic location is also a proxy for the availability of appropriate healthcare facilities, some studies hypothesised that the worst outcomes of rural patients are due to lack of appropriate healthcare and to limited access to specialist services [16,17,35,38]. However, a strength of our study is that we adjusted all analyses for provision of HF-CP and inpatient/outpatient cardiologist care, which are known (and we found) to correlate with better outcomes [14,19,24,29,38,39]. Involvement of cardiologists in the management of patients with HF might reduce mortality and readmissions and enhance adherence to guideline treatments by endorsing or refining GP recommendations. Moreover, the collaboration with GPs and outpatient nurses can provide additional monitoring of the patients’ concerns and adherence as planned in the CP [19,29,38]. Still, a shortcoming of our analysis is that the number of GP visits for each patient, either registered or not registered at the HF-CP, is not available. In Italy, most GPs use dedicated computer programmes to manage rosters, appointments and clinical data, but this information cannot be accessed by the local healthcare authorities.
Although rural patients had more follow-up cardiology visits (X → M) and these visits were associated with lower mortality (M → Y), we did not find evidence of reduced mortality among rural patients (X → Y). This should come as no surprise because, as a causal process becomes more distal, the size of the effect typically gets smaller and ultimately fails to achieve statistical significance . A possible interpretation is that the relationship between geographic location and mortality acts via a number of many other intermediary links, competing risks and random factors.
As mentioned above, there are a number of limitations to our study. First, we lack some relevant information, including clinical features, disease severity, SES, informal caregivers, and GP visits. Second, we assumed that all patients would attend the nearest facility. Third, travel times do not fully depict the travel burden of patients with HF, who are elderly and can be hindered by a relevant combination of disabilities and clinical conditions.
Our findings show that the travel times to healthcare services have no impact on the quality of care for patients with HF. Possible reasons for this result include the large number of healthcare services in the LHA of Bologna and the relatively short driving distances to the nearest facility (mostly <30 minutes), which can harm the generalisability of this study to rural areas with remote and isolated communities. We also found that the impact of the degree of urbanisation was possibly mediated by more proximal factors, some of which are potential targets for intervention such as the availability and utilisation of different types of care settings. More specifically, we found that cardiology services and multidisciplinary models of care had an impact on the quality of care for patients with HF.
When geographic accessibility is generally good, healthcare delivery and patient outcomes can be optimised by prioritising high-quality models of care, not only the quantity of available services being provided. This is particularly relevant considering the ageing of populations, the increase in disabilities and chronic diseases, and the downward trend of available healthcare resources for patients with complex needs.
Because the effectiveness of such composite programmes is context-specific, further research is needed to guide, tailor and improve healthcare settings and interventions to manage the clinical complexity and frailty of patients with HF.
S2 Table [pdf]
Six-month study outcomes, degree of urbanisation and travel times to healthcare.
S3 Table [pdf]
Comorbidities and drug therapies considered for inclusion in multivariable regression models.
S4 Table [pdf]
Distribution of comorbidities and previous medication use in the study population.
S5 Table [pdf]
Association of processes of care, urbanisation levels and travel times with the study outcomes of patients with heart failure.
1. Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137–46. doi: 10.1136/hrt.2003.025270 17699180
2. Dunlay SM, Roger VL. Understanding the epidemic of heart failure: past, present, and future. Curr Heart Fail Rep. 2014;11:404–15. doi: 10.1007/s11897-014-0220-x 25182014
3. Nieminen MS, Brutsaert D, Dickstein K, Drexler H, Follath F, Harjola VP, et al. EuroHeart Failure Survey II (EHEART FAILURES II): a survey on hospitalized acute heart failure patients: description of population. Eur Heart J. 2006;27:2725–36. doi: 10.1093/eurheartj/ehl193 17000631
4. Bueno H, Ross JS, Wang Y, Chen J, Vidán MT, Normand SL, et al. Trends in length of stay and short-term outcomes among Medicare patients hospitalized for Heart failure, 1993–2006. JAMA. 2010;303:2141–7. doi: 10.1001/jama.2010.748 20516414
5. Kociol RD, Hammill BG, Fonarow GC, Klaskala W, Mills RM, Hernandez AF, et al. Generalizability and longitudinal outcomes of a national heart failure clinical registry: Comparison of Acute Decompensated Heart Failure National Registry (ADHERE) and non-ADHERE Medicare beneficiaries. Am Heart J. 2010;160:885–92. doi: 10.1016/j.ahj.2010.07.020 21095276
6. Maggioni AP, Dahlstrom U, Filippatos G, Chioncel O, Leiro MC, Drozdz J, et al. EURObservational Research Programme: the Heart Failure Pilot Survey (ESC-HEART FAILURE Pilot). Eur J Heart Fail. 2010;12:1076–84. doi: 10.1093/eurjhf/hfq154 20805094
7. Maggioni AP, Dahlstrom U, Filippatos G, Chioncel O, Crespo Leiro M, Drozdz J, et al. EURObservational Research Programme: regional differences and 1-year follow-up results of the Heart Failure Pilot Survey (ESC-HEART FAILURE Pilot). Eur J Heart Fail. 2013;15:808–17. doi: 10.1093/eurjhf/hft050 23537547
8. Solomon SD, Dobson J, Pocock S, Skali H, McMurray JJ, Granger CB, et al. Influence of nonfatal hospitalization for heart failure on subsequent mortality in patients with chronic heart failure. Circulation. 2007;116:1482–7. doi: 10.1161/CIRCULATIONAHA.107.696906 17724259
9. Ambrosy AP, Fonarow GC, Butler J, Chioncel O, Greene SJ, Vaduganathan M, et al. The global health and economic burden of hospitalizations for heart failure: lessons learned from hospitalized heart failure registries. J Am Coll Cardiol. 2014;63:1123–33. doi: 10.1016/j.jacc.2013.11.053 24491689
10. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. Heart disease and stroke statistics—2016 update: a report from the American Heart Association. Circulation. 2016;133:e38–360. doi: 10.1161/CIR.0000000000000350 26673558
11. Ziaeian B, Fonarow GC. The Prevention of Hospital Readmissions in Heart Failure. Prog Cardiovasc Dis. 2016;58:379–85. doi: 10.1016/j.pcad.2015.09.004 26432556
12. Avaldi VM, Lenzi J, Castaldini I, Urbinati S, Di Pasquale G, Morini M, et al. Hospital readmissions of patients with heart failure: the impact of hospital and primary care organizational factors in Northern Italy. PLoS One. 2015;10:e0127796. doi: 10.1371/journal.pone.0127796 26010223
13. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JG, Coats AJ, et al. 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail. 2016;18:891–975. doi: 10.1002/ejhf.592 27207191
14. McAlister FA, Stewart S, Ferrua S, McMurray JJ. Multidisciplinary strategies for the management of heart failure patients at high risk for admission: a systematic review of randomized trials. J Am Coll Cardiol. 2004;44:810–9. doi: 10.1016/j.jacc.2004.05.055 15312864
15. Kelly C, Hulme C, Farragher T, Clarke G. Are differences in travel time or distance to healthcare for adults in global north countries associated with an impact on health outcomes? A systematic review. BMJ Open. 2016;6:e013059. doi: 10.1136/bmjopen-2016-013059 27884848
16. Gamble JM, Eurich DT, Ezekowitz JA, Kaul P, Quan H, McAlister FA. Patterns of care and outcomes differ for urban versus rural patients with newly diagnosed heart failure, even in a universal healthcare system. Circ Heart Fail. 2011;4:317–23. doi: 10.1161/CIRCHEARTFAILURE.110.959262 21430285
17. Cook NL, Lauer MS. The socio-geography of heart failure: why it matters. Circ Heart Fail. 2011;4:244–5. doi: 10.1161/CIRCHEARTFAILURE.111.962191 21586735
19. Avaldi VM, Lenzi J, Urbinati S, Molinazzi D, Descovich C, Campagna A, et al. Effect of cardiologist care on 6-month outcomes in patients discharged with heart failure: results from an observational study based on administrative data. BMJ Open. 2017;7:e018243. doi: 10.1136/bmjopen-2017-018243 29101146
20. Zeigermann L. OPENCAGEGEO: Stata module for forward and reverse geocoding using the OpenCage Geocoder API [software]. 2016 Mar 19 [cited 2019 Jul 15]. https://EconPapers.repec.org/RePEc:boc:bocode:s458155
21. Weber S, Péclat M. A simple command to calculate travel distance and travel time. Stata Journal. 2017;17:962–71.
22. Selim AM, Mazurek JA, Iqbal M, Wang D, Negassa A, Zolty R. Mortality and readmission rates in patients hospitalized for acute decompensated heart failure: a comparison between cardiology and general-medicine service outcomes in an underserved population. Clin Cardiol. 2015;38:131–8. doi: 10.1002/clc.22372 25694226
23. Uthamalingam S, Kandala J, Selvaraj V, Martin W, Daley M, Patvardhan E, et al. Outcomes of patients with acute decompensated heart failure managed by cardiologists versus noncardiologists. Am J Cardiol. 2015;115:466–71. doi: 10.1016/j.amjcard.2014.11.034 25637324
24. Avaldi VM, Lenzi J. The Role of Cardiologists in the Management of Patients with Heart Failure. Adv Exp Med Biol. 2018;1067:133–44. doi: 10.1007/5584_2017_120 29188455
25. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130–9. doi: 10.1097/01.mlr.0000182534.19832.83 16224307
26. Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr, Colvin MM, et al. 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America. Circulation. 2017;136(6):e137–e161. doi: 10.1161/CIR.0000000000000509 28455343
27. Vaduganathan M, Fonarow GC, Gheorghiade M. Drug therapy to reduce early readmission risk in heart failure: ready for prime time? JACC Heart Fail. 2013;1:361–4. doi: 10.1016/j.jchf.2013.04.010 24621940
28. Corrao G, Ghirardi A, Ibrahim B, Merlino L, Maggioni AP. Short- and long-term mortality and hospital readmissions among patients with new hospitalization for heart failure: A population-based investigation from Italy. Int J Cardiol. 2015;181:81–7. doi: 10.1016/j.ijcard.2014.12.004 25497525
29. Takeda A, Martin N, Taylor RS, Taylor SJ. Disease management interventions for heart failure. Cochrane Database Syst Rev. 2019;1:CD002752. doi: 10.1002/14651858.CD002752.pub4 30620776
30. Austin PC, Tu JV. Bootstrap Methods for Developing Predictive Models. The Am Stat. 2004;58:131–7.
31. Italian National Institute of Statistics [Internet]. Confini delle unità amministrative a fini statistici al 1 gennaio 2019; 2019 [cited 2019 May 8]. https://www.istat.it/it/archivio/222527
32. Konerding U, Bowen T, Elkhuizen SG, Faubel R, Forte P, Karampli E, et al. The impact of travel distance, travel time and waiting time on health-related quality of life of diabetes patients: An investigation in six European countries. Diabetes Res Clin Pract. 2017;126:16–24. doi: 10.1016/j.diabres.2017.01.014 28189950
33. Syed ST, Gerber BS, Sharp LK. Traveling towards disease: transportation barriers to health care access. J Community Health. 2013;38:976–93. doi: 10.1007/s10900-013-9681-1 23543372
35. Foraker RE, Rose KM, Suchindran CM, Chang PP, McNeill AM, Rosamond WD. Socioeconomic status, Medicaid coverage, clinical comorbidity, and rehospitalization or death after an incident heart failure hospitalization: Atherosclerosis Risk in Communities cohort (1987 to 2004). Circ Heart Fail. 2011;4:308–16 doi: 10.1161/CIRCHEARTFAILURE.110.959031 21430286
36. Teng TH, Katzenellenbogen JM, Hung J, Knuiman M, Sanfilippo FM, Geelhoed E, et al. Rural-urban differentials in 30-day and 1-year mortality following first-ever heart failure hospitalisation in Western Australia: a population-based study using data linkage. BMJ Open. 2014;4:e004724. doi: 10.1136/bmjopen-2013-004724 24793254
37. Garcia R, Abellana R, Real J, Del Val JL, Verdú-Rotellar JM, Muñoz MA. Health inequalities in hospitalisation and mortality in patients diagnosed with heart failure in a universal healthcare coverage system. J Epidemiol Community Health. 2018;72:845–51. doi: 10.1136/jech-2017-210146 29899056
38. Cook NL, Ayanian JZ, Orav EJ, Hicks LS. Differences in specialist consultations for cardiovascular disease by race, ethnicity, gender, insurance status, and site of primary care. Circulation. 2009;119:2463–70. doi: 10.1161/CIRCULATIONAHA.108.825133 19398667
39. Komajda M, Cowie MR, Tavazzi L, Ponikowski P, Anker SD, Filippatos GS, et al. Physicians’ guideline adherence is associated with better prognosis in outpatients with heart failure with reduced ejection fraction: the QUALIFY international registry. Eur J Heart Fail. 2017;19:1414–23. doi: 10.1002/ejhf.887 28463464
40. Shrout PE, Bolger N. Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychol Methods. 2002;7:422–45. 12530702