Muñoz et al. BMC Family Practice (2020) 21:145 https://doi.org/10.1186/s12875-020-01213-x RESEARCH ARTICLE Open Access Relationship between the place of living and mortality in patients with advanced heart failure Miguel-Angel Muñoz1,2,3* , Raquel Garcia1,2,3, Elena Navas2, Julio Duran4, José-Luis Del Val-Garcia1,2 and José-Maria Verdú-Rotellar1,2,5 Abstract Background: Social and environmental factors in advanced heart failure (HF) patients may be crucial to cope with the end stages of the disease. This study analyzes health inequalities and mortality according to place of residence (rural vs urban) in HF patients at advanced stages of the disease. Methods: Population-based cohort study including 1148 adult patients with HF attended in 279 primary care centers. Patients were followed for at least 1 year after reaching New York Heart Association IV functional class, between 2010 and 2014. Data came from primary care electronic medical records. Cox regression models were applied to determine the hazard ratios (HR) of mortality. Results: Mean age was 81.6 (SD 8.9) years, and 62% were women. Patients in rural areas were older, particularly women aged > 74 years (p = 0.036), and presented lower comorbidity. Mortality percentages were 59 and 51% among rural and urban patients, respectively (p = 0.030). Urban patients living in the most socio-economically deprived neighborhoods presented the highest rate of health service utilization, particularly with primary care nurses (p-trend < 0.001). Multivariate analyses confirmed that men (HR 1.60, 95% confidence interval (CI) 1.34–1.90), older patients (HR 1.05, 95% CI 1.04–1.06), Charlson comorbidity index (HR 1.16, 95% CI 1.11–1.22), and residing in rural areas (HR 1.35, 95% CI 1.09 to 1.67) was associated with higher mortality risk. Conclusions: Living in rural areas determines an increased risk of mortality in patients at final stages of heart failure. Keywords: Advanced heart failure, Rural health, Community healthcare, Health inequalities Background in health and disease patterns between rural and urban According to data published by the World Bank, the per- areas have not been widely studied [2]. centage of rural populations fluctuates from 19 to 59% de- Evidence concerning inequalities in health and mor- pending on the degree of development of the countries in tality between rural and urban settings, has reported question [1]. In spite of these figures, however, differences conflicting results with respect to the source and characteristics of the population [3, 4]. With respect to cardiovascular diseases, coronary * Correspondence: mamunoz.bcn.ics@gencat.cat 1 heart disease mortality has been observed to be more Institut Català de la Salut, Barcelona, Spain 2 prevalent in rural areas [5], and similar results have Fundació Institut Universitari per a la recerca a l’Atenció Primària de Salut Jordi Gol i Gurina (IDIAPJGol), Carrer Sardenya 375, Entresol, 08025 Barcelona, been published for heart failure (HF) although only Spain regarding men [6]. Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 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 in a credit line to the data. Muñoz et al. BMC Family Practice (2020) 21:145 Page 2 of 10 Whilst the HF incidence rate has declined in recent The database incorporates both administrative and years, its prevalence has increased, suggesting that sur- clinical data which are encrypted to guarantee the confi- vival over time is longer, probably due to better care and dentiality and anonymization of the information gath- treatment [7]. Nevertheless, this improvement in survival ered for research purposes and provides data about is modest [8], and research in HF is considered one of diagnoses, clinical characteristics, comorbidity, labora- the most important priorities [9]. Patients classified as tory and diagnostic tests, social and demographic vari- New York Heart Association (NYHA) functional class ables, and performance in activities of daily living, tests III/IV present almost four-fold greater rates of mortality, to evaluate functional physical and mental status, drug and up to 58.3% of NYHA functional class IV dies after prescriptions, and primary health care service utilization. a five-year follow-up [10, 11]. However, due to its irregu- Information regarding the patients’ vital status is also in- lar evolution, HF is not usually considered a terminal cluded and comes from the Central Insurance Register disease, whilst at advanced stages it could be compared (RCA) (ie, if patient is death or alive). to malignant neoplasms [12]. Patients with HF were identified using the Inter- Although many predictors of mortality among HF pa- national Classification of Diseases, Tenth Revision (ICD tients have been well identified [13], social and environ- 10), claim code I50. mental determinants are not usually included in predictive The ICD 10 codes selected to register comorbidities models, particularly in patients at terminal stages of the were: diabetes (E10-E14), hypertension (I10-I15), coron- disease. Some studies have shown worse prognoses re- ary heart disease (I20-I25), stroke (I63-I65), atrial fibril- garding socio-economic position [14], social risk [15], lation (I48), chronic kidney disease (N18), chronic lower health literacy [16] and urban areas of residence [17]. respiratory diseases (J40-J44), and cancer (C00-C97). Even though gaps in the availability of general pallia- Clinical variables, laboratory analyses, and tests asses- tive care in rural areas have been reported [18], informa- sing functional were obtained from the patient’s consult- tion about the differences in the lifetime of HF patients ation closest to the inclusion date, and missing values or health services utilization depending on their resi- were imputed. dence and socio economic level is scarce. The Barthel index, which has proved helpful in asses- This study was aimed at analyzing health inequalities sing the functional status of a patient [21], and the and mortality according to place of residence (rural vs Charlson index, used to predict ten-year mortality and urban) in HF patients at advanced stages of the disease healthcare resource utilization in patients with a range (NYHA functional class IV) attended in the community of comorbid conditions [22] were regularly collected by analyzing real world data. during the nurses’ consultations. To determine socio-economic status among the urban Methods patients we employed the MEDEA Index which is an ag- We followed a population-based cohort of adult patients gregated socioeconomic deprivation model which clas- presenting the most advanced stages of heart failure sify the population living in small geographical areas, (NYHA functional class IV) between January 1st, 2010, according to the percentage of unemployment, manual and December 31st, 2014. and temporary workers, and individuals with insufficient The inclusion date was taken to be when patients were education (less than primary school). The lowest quintile first registered as having NYHA functional class IV in (Urban areas 1) represents individuals with the most fa- their primary healthcare electronic medical records dur- vorable socio-economic position, and the upper one the ing the study period. NYHA functional class IV was con- worst (Urban areas 1). The unit of aggregation was the sidered when patients diagnosed from HF were unable census tracts, which is the smallest territorial unit for to carry on any physical activity without symptoms of which population data are available in our country. This HF, or symptoms of HF at rest [19]. The whole cohort index has been proven valid for urban areas although in was followed-up for at least 1 year from the inclusion rural ones it does not discriminate accurately [23]. date or until a fatal occurrence took place during the Family networks, as well as living conditions, were study period. assessed through an interview by a social workers or Information was collected from the primary care elec- nurses using the Gijon social-family scale, which in- tronic medical records, through the Information System cludes questions about housing conditions, family and for the Development of Research in Primary Care (SIDI social relationships and income [24]. AP). This database contains data from 5.8 million indi- Family network was explored by the following items: viduals attended in 279 primary healthcare centers Living with relatives (physically independent), living with which attend 80% of the whole population of Catalonia relatives (physically dependent), living with a partner, (north-east Spain), and it has already been validated for living alone (offspring close to their home), and living use in cardiovascular research [20]. alone (offspring far from their home). We grouped the Muñoz et al. BMC Family Practice (2020) 21:145 Page 3 of 10 last two variables and classified family network as living bivariate analyses (sex, age, Charlson Index, place of resi- alone or not. dence, diabetes mellitus, and number of visits to primary Regarding housing conditions, we considered them in- care urgencies) were included in the final model. adequate when the house had structural barriers, humid- Statistical analysis was conducted using R Software for ity or incomplete facilities, no elevator or telephone or Windows version 3.6.1, Vienna, Austria. patients were living in a slum. Health service utilization was computed as the number Results of consultations made with the family physician, primary Descriptive findings care nurse, or specific primary care emergency services. We analyzed data from 1148 HF patients in NYHA Rural residence was defined when patients lived in an functional class IV. Mean age was 81.6 (SD 8.9), 95% of area with less than 10,000 inhabitants, or the density of patients were older than 64 years, and 61% were women. population was lower than 150 inhabit/km2, according to Most of the population lived in urban areas (N = 972) the Catalan Healthcare Administration classification [25]. and more than 69% were older than 74 years (N = 793). The prescription of angiotensin converting enzyme in- Among the total number of patients included, 68% pre- hibitors, angiotensin II receptor blockers, beta blockers, sented high comorbidity according to the Charlson mineral corticoid antagonists, and loop diuretics was index (score > =3), and daily living activities were moder- also collected. ately to severely affected (Barthel Index < 60) in 56%. The primary outcome was all-cause mortality occur- The most frequently associated comorbidities were ring during the follow-up. hypertension (79%), atrial fibrillation (47%), diabetes (42%), and coronary heart disease (34%). Cardiovascular Statistical analysis comorbidity (coronary heart disease, atrial fibrillation, Continuous variables were described by means, Standard and stroke) was present in 69% of the patients, and 22% deviation, and median, and percentages were used to de- presented two or more of these conditions. scribe categorical variables. Regarding the number of chronic conditions included Variables included in the analyses were: sociodemo- in the analyses, the median was 3 (percentile 25–75, 2– graphic, clinical and laboratory test data, comorbidity, 4), and 62% of HF patients had simultaneously three or consultations made to primary healthcare professionals, more comorbidities. activities of daily living, presence of family networks, Beta-blockers, angiotensin converting enzyme inhibi- housing conditions and medication related to HF. tors/angiotensin II receptor blockers, and mineral corti- Living in urban or rural residence was the main inde- coid antagonists were prescribed in 45, 66, and 26% of pendent variable and global mortality was the end point patients, respectively. A combination of the three medi- of the study. cations was present in 10%, whilst loop diuretics were The Anova test was performed to compare the num- prescribed in 88%. ber of visits according to the different levels of MEDEA During the period of the study, patients consulted deprivation index as a categorical variable. their family physician on average 22 occasions, 20 with Characteristics related to mortality and differences be- their primary care nurse, and on three occasions the pri- tween urban and rural patients were first compared mary healthcare emergency centers. using chi-square test or Student t test. Odds ratios were estimated by logistic regression models Differences between urban and rural patients to evaluate the relationship between urban and rural resi- Mean follow-up until the end of the study, or the occur- dence and to evaluate the association between these char- rence of a fatal event, was 16 months (SD 12.4) for urban acteristics, the place of residence and mortality. and 14.5 months (SD 10.9) for rural patients, respectively. Multiple imputation models were performed to man- Although rural patients were, on average, 2 years age the missing data by using the “Mice” and “VIM” R older, they presented lower comorbidity (Charlson packages. The “mice” function is based on the condi- index). In spite of not being statistically significant, the tional specification, where each incomplete variable is urban cohort tended to live more commonly alone. imputed with a separate model. A total of 50 imputa- No differences in cardiovascular comorbidity, clinical var- tions were generated and the final data were obtained iables, performance in activities of daily living, or HF medi- with the “complete” function that generates a complete cation were reported between rural and urban patients. data set that combines the observed and imputed values. Table 1 describes how in analyses stratified by gender Cox multivariate logistic regression was used to ana- there were no outstanding differences regarding place of lyse the characteristics related to mortality and covari- residence in any of the variables, with the exception of ates included in the model. Socio demographic variables age (older women in rural areas) and the higher percent- and those statistically significant (at level p < 0.01) in the age of women living alone in the cities. Muñoz et al. BMC Family Practice (2020) 21:145 Page 4 of 10 Table 1 Characteristics of patients with heart failure at New York Heart Association IV functional class, according to their place of residence Overall Women Men Urban Rural P value Urban Rural P value Urban Rural P value N = 972 N = 176 N = 598 N = 110 N = 374 N = 66 Age* 81.3 (8.8) 83.0 (8.6) 0.020 82.9 (8.5) 84.5 (7.6) 0.039 79.0 (8.95) 80.5 (9.5) 0.228 Mortality 488 (51) 104 (59) 0.037 277 (46) 60 (54) 0.120 211 (56) 44 (67) 0.137 Cardiovascular Risk Body Mass Indexa 30.4 (6.8) 29.4 (5.8) 0.112 31.6 (7.1) 30.5 (6.3) 0.186 28.7 (5.9) 27.6 (4.6) 0.154 Smoker 36 (4) 5 (3) 0.404 8 (1) 0 (0) 0.490 28 (7) 5 (8) 0.596 Comorbidity Diabetes 422 (43) 63 (36) 0.062 248 (42) 38 (35) 0.210 174 (47) 25 (38) 0.243 Coronary heart disease 326 (34) 49 (28) 0.163 162 (27) 22 (20) 0.150 164 (44) 27 (41) 0.757 Stroke 141 (15) 22 (13) 0.465 79 (13) 12 (11) 0.612 62 (17) 10 (15) 0.914 Chronic kidney disease 331 (34) 60 (34) 0.999 217 (36) 36 (33) 0.543 114 (31) 24 (36) 0.420 Chronic obstructive pulmonary disease 322 (33) 56 (32) 0.738 121 (20) 24 (22) 0.803 201 (54) 32 (49) 0.512 Atrial fibrillation 453 (47) 83 (48) 0.897 305 (51) 51 (46) 0.429 148 (40) 32 (49) 0.222 Cancer 183 (19) 33 (19) 0.989 96 (16) 18 (16) 1.000 87 (23) 15 (23) 1.000 Hypertension 774 (80) 135 (77) 0.384 500 (83) 94 (86) 0.732 274 (73) 41 (62) 0.089 Clinical Variables Heart rate (beats/minute) a 77.0 (14.4) 76.0 (15.8) 0.456 76.8 (14.6) 75.5 (16.4) 0.450 77.1 (14.1) 76.8 (15.0) 0.895 Systolic blood pressure (mmHg)a 126.0 (18.6) 124.0 (18.4) 0.208 127.0 (19.0) 125.0 (18.1) 0.222 123.0 (17.9) 122.0 (19.2) 0.598 Diastolic blood pressure (mmHg)a 68.2 (10.6) 67.1 (10.6) 0.192 68.9 (11.0) 67.8 (9.8) 0.297 67.3 (9.9) 66.1 (11.9) 0.437 Potassium mEq/L a 4.6 (0.6) 4.5 (0.5) 0.780 4.5 (0.5) 4.5 (0.5) 0.711 4.6 (0.6) 4.5 (0.5) 0.402 Sodium mEq/L a 141.0 (3.7) 140.0 (3.9) 0.485 141.0 (3.6) 141.0 (3.8) 0.713 140.0 (3.8) 140 (4.0) 0.534 Glomerular filtration 46.7 (14.6) 46.6 (13.8) 0.894 44.7 (13.8) 44.3 (13.5) 0.796 49.9 (15.3) 50.3 (13.5) 0.853 a Haemoglobin (g/dL) 12.1 (1.8) 12.1 (1.7) 0.993 11.9 (1.6) 12.1 (1.6) 0.268 12.5 (1.9) 12.3 (1.9) 0.287 Charlson Indexa 3.6 (1.8) 3.3 (1.5) 0.031 3.4 (1.7) 3.1 (1.4) 0.058 3.9 (1.9) 3.6 (1.6) 0.194 Barthel Indexa 57.3 (27.9) 55.9 (31.1) 0.535 52.5 (27.3) 51.4 (29.6) 0.707 64.9 (27.2) 63.3 (32.5) 0.701 Medication Beta-blockers 448 (46) 69 (39) 0.098 279 (47) 46 (42) 0.406 169 (45) 23 (35) 0.154 ACEi or ARB b 639 (66) 119 (68) 0.692 400 (67) 78 (71) 0.474 239 (64) 41 (62) 0.890 c MRA 263 (27) 41 (23) 0.343 143 (24) 19 (17) 0.161 120 (32) 22 (33) 0.954 Loop diuretics 859 (88) 156 (89) 0.929 522 (87) 99 (90) 0.524 337 (90) 57 (86) 0.485 Social Variables Living alone 301 (31) 42 (24) 0.071 195 (33) 26 (24) 0.079 106 (28) 16 (24) 0.591 Inadequate housing conditions 424 (44) 83 (47) 0.384 243 (41) 54 (49) 0.122 181 (48) 29 (44) 0.593 Mean (Standard deviation); b ACEi angiotensin converting enzyme inhibitors and ARB Angiotensin II receptor blockers),c MRA mineralocorticoid receptor antagonists a We compared primary healthcare service utilization by was statistically significant in the case of the later HF patients, according to place of residence and socio- (Fig. 1). economic deprivation it was observed that those in the most socio-economically deprived urban areas tended to Overall mortality use more frequently primary care emergency services Among the 592 patients (52%) who died during the period than the other urban individuals, and even three more of the study, 488 (82%) lived in urban and 104 (18%) in times compared to rural ones. Regarding consultations rural areas, respectively. Analyzing the number of fatal with the family physician and primary healthcare nurses, events during follow-up, the mean rate of survival from patients in the most deprived urban areas were the study inclusion was 10.6 (SD 9.9) and 9.8 (SD 9.8) months, greatest users of the former and this higher utilization for the urban and rural cohorts, respectively. Muñoz et al. BMC Family Practice (2020) 21:145 Page 5 of 10 Fig. 1 Primary healthcare services utilization by urban patients with heart failure at New York heart Association (NYHA) functional class IV according to the place of residence (U 1: urban less socioeconomically deprived area to U 5: urban most socioeconomically deprived area) Older age, Charlson comorbidity index, chronic kidney associated with higher risk of mortality (Fig. 2). The ex- disease, higher potassium levels, and cancer, were related cess of mortality risk for men was 60% whilst for pa- with a higher risk of dying. tients living in rural areas it was 35% (HR 1.35, 95% CI In contrast, greater body mass index, blood pressure, 1.09 to 1.67). glomerular filtration, Barthel index, and hemoglobin were protective with respect to mortality risk (Table 2). Discussion When analyzing sociodemographic variables related to In our study, based on a community cohort of HF pa- mortality we found that being men, older, living in rural tients at the most advanced stages of the disease, we ob- areas and having unfavorable housing conditions (struc- served that mortality was 35% higher in those residing in tural barriers and lack of facilities) were related to higher rural setting. Being men, aged, and presenting higher co- probability of dying, but no significant differences were morbidity were also found to be related to a greater risk found among urban patients regarding the area where of dying. We found an inverse gradient in primary they lived. healthcare resources utilization regarding socio- Table 3 shows the differences in mortality according to economic deprivation among urban patients, those resid- the fact of living in urban and rural areas. Older age, ing in the most deprived socio-economic areas had the chronic kidney disease, lower levels of blood pressure, highest health services utilization. lower levels of haemoglobine and lower body mass index Compared to the rest of Europe, the population dens- were more commonly related to mortality, regardless the ity in the study region is among one of the highest, the place of residence. We observed that in patients living in distribution is, however, irregular with more than 40% of rural areas the prevalence of stroke among those patients the inhabitants concentrated in the metropolitan area of who died was higher, whilst in urban patients mortality Barcelona. Although a number of small hospitals are dis- was higher in those with higher prevalence of cancer, tributed across the territory, most tertiary hospitals are lower Barthel Index and worse levels of creatinine. located in this large metropolitan area which can limit accessibility of rural patients to highly specialized care. Multivariate analyses With respect to place of residence, we did not observe Cox multivariate regression model, adjusted by the vari- differences in clinical characteristics or HF medication ables significantly associated both with the mortality and prescribed. Although rural women were slightly older, the place of living, confirmed that being men, older, hav- urban patients presented higher comorbidity and re- ing higher comorbidity, and living in rural areas were ported more social isolation. This last finding concurs Muñoz et al. BMC Family Practice (2020) 21:145 Page 6 of 10 Table 2 Characteristics related to the overall mortality of patients with heart failure at New York heart Association IV functional class, according to gender Overall Women Men Alive Dead P value Alive Dead P value Alive Dead P value N = 556 N = 592 N = 371 N = 337 N = 185 N = 255 Age 79.7 (9.3) 83.3 (8.0) < 0.001 81.4 (8.6) 85.0 (7.7) < 0.001 76.5 (9.9) 81.2 (7.7) < 0.001 Body Mass Index 31.5 (6.7) 28.9 (6.4) < 0.001 32.4 (6.8) 30.1 (7.0) < 0.001 30.0 (6.2) 27.5 (5.2) < 0.001 Smoker 22 (4) 19 (3) 0.698 7 (2) 1 (0) 0.112 15 (8) 18 (7) 0.009 Comorbidity Diabetes 234 (42) 251 (42) 0.971 154 (41) 132 (39) 0.577 80 (43) 119 (47) 0.538 Coronary heart disease 178 (32) 197 (33) 0.782 100 (27) 84 (25) 0.597 78 (42) 113 (44) 0.725 Stroke 78 (14) 85 (14) 0.854 50 (14) 41 (12) 0.683 28 (15) 44 (17) 0.644 Chronic kidney disease 162 (29) 229 (39) < 0.001 119 (32) 134 (40) 0.040 43 (23) 95 (37) 0.003 Chronic obstructive pulmonary disease 170 (31) 208 (35) 0.089 76 (21) 69 (21) 1.000 94 (51) 139 (55) 0.502 Atrial fibrillation 254 (46) 282 (48) 0.707 184 (50) 172 (51) 0.758 70 (38) 110 (43) 0.309 Cancer 81 (15) 135 (23) < 0.001 54 (15) 60 (18) 0.284 27 (15) 75 (29) < 0.001 Hypertension 449 (81) 460 (77) 0.218 313 (84) 281 (83) 0.800 136 (74) 179 (70) 0.513 Clinical Variables Heart Rate (beats/minute) a 76.6 (15.2) 76.8 (14.0) 0.735 76.7 (15.4) 76.5 (14.3) 0.798 76.4 (14.9) 77.4 (13.7) 0.487 Systolic blood pressure (mmHg)a 127.7 (18.4) 123.3 (18.5) < 0.001 129 (18.3) 125 (19.2) 0.007 126 (18.4) 121 (17.5) 0.009 a Diastolic blood pressure (mmHg) 69.5 (10.7) 66.7 (10.3) < 0.001 70.2 (11.1) 67.1 (10.3) < 0.001 68.2 (9.8) 66.3 (10.4) 0.051 Potassium mEq/L a 4.5 (0.5) 4.6 (0.6) 0.033 4.4 (0.5) 4.5 (0.5) 0.133 4.5 (0.6) 4.6 (0.6) 0.370 a Sodium mEq/L 140.8 (3.5) 140.5 (3.9) 0.060 141 (3.4) 141 (3.7) 0.351 141 (3.5) 140 (3.9) 0.309 Glomerular filtration 1.2 (0.6) 1.4 (0.8) < 0.001 1.1 (0.4) 1.3 (0.5) < 0.001 1.3 (0.8) 1.5 (1.0) 0.007 Hemoglobin (g/dL)a 12.3 (1.7) 11.9 (1.8) < 0.001 12.2 (1.5) 11.7 (1.7) < 0.001 12.8 (1.9) 12.3 (1.9) 0.009 Barthel Index a 59.8 (27.8) 54.4 (28.6) < 0.001 55.9 (27.2) 48.5 (27.6) < 0.001 67.8 (27.5) 62.4 (28.2) 0.043 Charlson Index a 3.2 (1.6) 3.8 (1.8) < 0.001 3.2 (1.5) 3.5 (1.6) 0.024 3.4 (1.6) 4.2 (1.9) < 0.001 Medication Beta-blockers 271 (49) 246 (42) 0.011 182 (49) 143 (42) 0.091 89 (48) 103 (40) 0.130 ACEi or ARB b 397 (71) 361 (61) < 0.001 262 (71) 216 (64) 0.077 135 (73) 145 (56) 0.001 c MRA 145 (26) 159 (27) 0.883 83 (22) 79 (23) 0.803 62 (33) 80 (31) 0.711 Loop diuretics 479 (86) 536 (91) 0.083 315 (85) 306 (91) 0.023 164 (89) 230 (90) 0.714 Social Variables Living alone 164 (29) 179 (30) 0.988 121 (33) 100 (30) 0.446 43 (23) 79 (31) 0.093 Inadequate housing conditions 220 (40) 287 (48) 0.002 131 (35) 166 (49) < 0.001 89 (48) 121 (47) 0.968 Mean (Standard deviation); b ACEi angiotensin converting enzyme inhibitors, ARB Angiotensin II receptor blockers);c MRA mineralocorticoid receptor antagonists a with a previous study reporting that in a rural environ- Mortality ment it is easier to rely on family members [26]. The higher mortality described for men, older HF pa- Although the bivariate analyses found statistical differ- tients, and in those with decreased body mass index, ences in the characteristics associated with mortality de- chronic kidney disease, and lower blood pressure levels, pending on rural or urban setting, these differences were concurs with previous studies, particularly in the ad- not clinically relevant. vanced stages of the disease [28–30]. Moreover, higher We found a considerable number of consultations, mortality in rural patients is in agreement with a previ- particularly with primary care nurses, among the urban ous study [6]. patients living in the most deprived areas. In this regard, It has been hypothesized that populations residing it has already been reported that low income and other in urban areas have better health due to easier acces- psycho-social disadvantages imply greater healthcare re- sibility to health services, in addition to better jobs source utilization [27]. and income [3]. Muñoz et al. BMC Family Practice (2020) 21:145 Page 7 of 10 Table 3 Characteristics related to the mortality of patients with heart failure at New York Heart Association IV functional class according to their place of residence Overall Urban Rural Alive Dead P value Alive Dead P value Alive Dead P value N = 556 N = 592 N = 484 N = 488 N = 72 N = 104 Age 79.7 (9.3) 83.3 (8.0) < 0.001 79.7 (9.5) 83.0 (7.9) < 0.001 79.9 (8.5) 85.1 (8.0) < 0.001 a Body Mass Index 31.5 (6.7) 28.9 (6.4) < 0.001 31.7 (6.8) 29.1 (6.5) < 0.001 31.0 (5.7) 28.0 (5.6) 0.004 Smoker 22 (4) 19 (3) 0.698 20 (4) 16 (3) 0.532 2 (3) 3 (3) 0.801 Comorbidity Diabetes 234 (42) 251 (42) 0.971 204 (42) 218 (45) 0.466 30 (42) 33 (32) 0.233 Coronary heart disease 178 (32) 197 (33) 0.782 160 (33) 166 (34) 0.804 18 (25) 31 (30) 0.597 Stroke 78 (14) 85 (14) 0.854 74 (15) 67 (14) 0.549 4 (5) 18 (17) 0.037 Chronic kidney disease 162 (29) 229 (39) < 0.001 146 (30) 185 (38) 0.013 16 (22) 44 (42) 0.009 Chronic Obstructive Pulmonary disease 170 (31) 208 (35) 0.089 150 (31) 172 (35) 0.180 20 (28) 36 (35) 0.428 Atrial fibrillation 254 (46) 282 (48) 0.707 224 (46) 229 (47) 0.891 30 (42) 53 (51) 0.289 Cancer 81 (15) 135 (23) < 0.001 70 (15) 113 (23) 0.001 11 (15) 22 (21) 0.432 Hypertension 449 (81) 460 (77) 0.218 388 (80) 386 (79) 0.739 61 (85) 74 (71) 0.056 Clinical Variables Heart ratea 76.6 (15.2) 76.8 (14.0) 0.735 76.5 (14.7) 77.3 (14.2) 0.413 77.5 (18.8) 75.0 (13.4) 0.329 Systolic blood pressure (mm Hg)a 127.7 (18.4) 123.3 (18.5) < 0.001 128 (18.3) 124 (18.9) 0.002 128 (19.5) 121 (17.1) 0.008 a Diastolic blood pressure (mm Hg) 69.5 (10.7) 66.7 (10.3) < 0.001 69.6 (10.7) 67.0 (10.3) < 0.001 69.2 (10.7) 65.7 (10.4) 0.031 Potassium (mEq/dL)a 4.5 (0.5) 4.6 (0.6) 0.033 4.5 (0.5) 4.5 (0.6) 0.082 4.5 (0.4) 4.6 (0.5) 0.246 Sodium (mEq/dL)a 140.8 (3.5) 140.5 (3.9) 0.060 141 (3.4) 141 (3.8) 0.180 141 (3.8) 140 (3.9) 0.455 Creatininea 1.2 (0.6) 1.4 (0.8) < 0.001 1.2 (0.6) 1.5 (0.8) < 0.001 1.2 (0.5) 1.3 (0.6) 0.171 Haemoglobin (gr/dL)a 12.3 (1.7) 11.9 (1.8) < 0.001 12.3 (1.8) 11.9 (1.8) < 0.001 12.5 (1.5) 11.9 (1.8) 0.013 Barthel Index imputeda 59.8 (27.8) 54.4 (28.6) < 0.001 59.7 (27.3) 54.9 (28.3) 0.007 60.8 (31.6) 52.5 (30.5) 0.083 Charlson Indexa 3.2 (1.6) 3.8 (1.8) < 0.001 3.3 (1.6) 3.8 (1.8) < 0.001 2.9 (1.2) 3.5 (1.6) 0.005 Medication Beta-blockers 271 (49) 246 (42) 0.011 235 (49) 213 (44) 0.142 36 (50) 33 (32) 0.022 ACEi or ARBb 397 (71) 361 (61) < 0.001 339 (70) 300 (62) 0.006 58 (81) 61 (59) 0.004 MRAc 145 (26) 159 (27) 0.883 130 (27) 133 (27) 0.947 15 (21) 26 (25) 0.644 Loop diuretics 479 (86) 536 (91) 0.083 416 (86) 443 (91) 0.025 63 (88) 93 (89) 0.878 Social Variables Living alone 164 (29) 179 (30) 0.988 146 (30) 155 (32) 0.639 18 (25) 24 (23) 0.909 Inadequate housing conditions 220 (40) 287 (48) 0.002 187 (39) 237 (49) 0.002 33 (46) 50 (48) 0.889 Mean (Standard deviation); b ACEi angiotensin converting enzyme inhibitors, ARB Angiotensin II a receptor blockers; c MRA mineralocorticoid receptor antagonists a Regarding cardiovascular diseases, inequalities in rural large cities, offering services such as radiology and la- patients have been reported with respect to access to boratory tests, which may improve the diagnosis and treatment, such as percutaneous coronary interventions, treatment of complications whilst in the rural areas which require a longer travelling distance [31]. more complex assistance needs to be provided in small Another possible factor contributing higher mortality hospitals. among rural HF patients in our sample could be the dif- Many rural patients lack daily access to their family ferences reported in the use of health services, although doctors which can lead to delay in the treatment of de- we cannot affirm that highest primary care services compensations. Urban primary healthcare centers are utilization by urban patients determines longer survival. opened between 8 am and 8 pm every day, whilst many There is a widespread network of primary healthcare rural villages do not have general practitioner accessible out-of-hours centers, nevertheless, its distribution and all days, and population must travel long distances to be characteristics vary. These centers are more complete in attended, especially in out of hours. Muñoz et al. BMC Family Practice (2020) 21:145 Page 8 of 10 Fig. 2 Hazard ratio and 95% confidence interval for variables predicting mortality adjusted by sex, age, Charlson comorbidity index, diabetes mellitus and place of living Other factors might be involved in explaining in these comorbidities or treatment, rural HF patients had the differences such as the decision to die at home allows pa- highest mortality rates. tients to maintain control over their lives. The approach to the end of life could be culturally different in a rural Strengths and limitations setting compared to an urban one. A recent systematic re- Although different approaches have been employed to de- view oriented to analyse the socio-economic factors deter- fine rurality, our definition concurs with others used in mining the place of death found that dying at home was similar articles. Nevertheless, due to data limitations, we associated with living in rural areas but, potential causes could not fully discriminate the analyses between patients remain unsolved [32]. Mean survival time among patients living in the most isolated areas from the other rural ones who died in our study was slightly higher among the to ascertain whether differences in accessibility related to urban ones. Nevertheless, considering the considerable mortality are proportional to distance and frequency of limitations such patients undergo in their daily activities, healthcare service provision. The deprivation index used it may be questioned whether this increased survival time to study urban socioeconomic differences assumes homo- is cost effective in terms of quality of life. geneity among the population living in the same geo- Evidence regarding socioeconomic deprivation among graphical area, but may imply an ecological fallacy because urban residents is controversial. Hawkins et al., employ- it is possible to find both poor and affluent individuals in ing a geographical composite deprivation index similar the same areas, sometimes divided by only one street. to ours, and analyzing data from 2000 to 2007, did not Moreover, we lack information in order to discriminate describe differences regarding outcomes in HF patients the presence of socio-economic differences within rural [33]. In a more recent article, Witte et al., using the populations and thus considered them homogeneous in same deprivation index as Hawkins, reported that socio- terms of social status. economic deprivation was linked to an increased risk of Since administrative databases are used for clinical pur- death in HF patients, but only as a consequence of non- poses can lead to missing data. In the case of HF some cardiovascular causes [34]. In addition, a study per- variables such as ejection fraction are not always available formed with the same population in our country showed to have proper diagnoses according to guidelines [35]. a protective, although not significant, effect regarding Nevertheless in our study this fact is not relevant because mortality in the most deprived urban patients [17]. all patients were at final stages of the disease. Regarding The National Health Service in our country provides uni- other possible missing values we performed multiple im- versal healthcare, which may reduce social inequalities in putation models to minimize such an effect. health by facilitating access to primary care, prescriptions It would be advantageous to possess data regarding and hospitals to populations lacking economic resources, quality of life in order to analyze differences among rural but the distribution of healthcare premises in rural and re- and urban patients. mote areas is conditioned by geographical limitations. The results of our study are potentially valid for areas Nevertheless, future research will be needed to explain of the rest of our country with similar characteristics, why, with no differences in either cardiovascular concentrating population in large cities. Nevertheless, in Muñoz et al. BMC Family Practice (2020) 21:145 Page 9 of 10 Spain and in Europe there are many areas where the Availability of data and materials rural depopulation is remarkable, and the accessibility to The datasets generated and/or analysed during the current study are not publicly available due to the restrictions by the data owner (IDIAP Research healthcare services could be different. Institute), but could be available from the corresponding author on Since our results are based on information coming from reasonable request. more than 80% of the Catalan population, it is expected to Ethics approval and consent to participate think that they could be extrapolated to the rest. The study protocol was approved by the Primary Healthcare University Prospective studies would be needed, not just to ob- Research Institute IDIAP- Jordi Gol ethics committee (reference number: P13/ serve the presence of differences depending on the place 052). Confidentiality of data was guaranteed in all the process of the study, and data available for research purpose was anonymous. Since the data of living but oriented towards specifically analyse their obtained underwent a double process of anonymization, no interventions causes, to be more precise in measuring the differences were made and the research team had not access to the patient in the accessibility, the characteristics of patients, explor- information, ethics committee considered not necessary the patient informed consent either verbal nor written. To obtain the access to the ing their perceptions and needs as well as their beliefs database an external scientific committee from IDIAP Jordi Gol assessed and and preferences regarding the treatment received at the approved the quality and appropriateness of the study protocol, before its end of life. submission to the ethics committee. We had permission from our Research Institute (IDIAP Jordi Gol) to accede to the database and publish this To our knowledge, evidence about differences between manuscript. rural and urban heart failure patients, particularly at ad- vances stages of the disease is scarce. This article specif- Consent for publication ically analyses evidence from electronic health records Not applicable. from elderly HF patients at advanced stages of the dis- Competing interests ease through a large database. The authors declare that they have no competing interests. Author details 1 Conclusions Institut Català de la Salut, Barcelona, Spain. 2Fundació Institut Universitari per a la recerca a l’Atenció Primària de Salut Jordi Gol i Gurina (IDIAPJGol), Mortality in elderly patients with heart failure at the Carrer Sardenya 375, Entresol, 08025 Barcelona, Spain. 3Department of final stage of HF is higher among those living in rural Pediatrics, Obstetrics and Gynecology and Preventive Medicine, Universitat setting. Accessibility and inequalities in the healthcare Autònoma de Barcelona, Bellaterra, Spain. 4Clínica Sant Antoni. Institut Mèdic i de Rehabilitació, Barcelona, Spain. 5Department of Medicine, Universitat provided with respect to the place of residence may con- Autònoma de Barcelona, Bellaterra, Spain. tribute to such differences. The increased healthcare ser- vices utilization by urban patients living in most Received: 30 January 2020 Accepted: 7 July 2020 socioeconomic deprived areas is not followed by a re- duction in mortality. References Health policies should face with social and geograph- 1. United Nations Population Division's World Urbanization Prospects: 2018 ical inequalities, to ensure that most part of population Revision. https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS. Accessed 28 Aug 2019. have similar access to healthcare provision, especially to 2. Moreno-Lostao A, Guerras JM, Lostao L, de la Fuente L, Martínez D, primary care services, which are essential in improving Rodríguez-Artalejo F, et al. Cardiovascular mortality and risk behaviours by health and reducing mortality. degree of urbanization before, during and after the economic crisis in Spain. BMC Public Health. 2019;19(1):1109. 3. House JS, Lepkowski JM, Williams DR, Mero RP, Lantz PM, Robert SA, et al. Abbreviations Excess mortality among urban residents: how much, for whom, and why? HF: Heart failure; NYHA: New York Heart Association; HR: Hazard ratios; Am J Public Health. 2000;90(12):1898–904. CI: Confidence interval; RCA: Central Insurance Register; SIDIAP: Information 4. Eberhardt MS, Pamuk ER. The importance of place of residence: examining System for the Development of Research in Primary Care. health in rural and non rural areas. Am J Public Health. 2004;94(10):1682–6. 5. Kulshreshtha A, Goyal A, Dabhadkar K, Veledar E, Vaccario V. Urban-rural differences in coronary heart disease mortality in United States: 1999-2009. Acknowledgements Public Health Rep. 2014;129:19–29. We thank to the Information System for the Development of Research in 6. Gamble JM, Eurich DT, Ezekowitz JA, Kaul P, Quan H, McAlister FA. Patterns Primary Care (SIDIAP) professionals their support with data collection. of care and outcomes differ for urban versus rural patients with newly diagnosed heart failure, even in a universal healthcare system. Circ Heart Authors’ contributions Fail. 2011;4(3):317–23. MAM, RG, JLV, EN, JD, and JMVR participated in the interpretation of results, 7. Heidenreich PA, Albert NM, Allen LA, Bluemke DA, Butler J, Fonarow GC, the elaboration and revision of the drafts of the manuscript and in the et al. Forecasting the impact of heart failure in the United States: a policy approval of the final version. EN participated in the data management, statement from the American Heart Association. Circ Heart Fail. 2013;6:606– analysis and elaboration of the last draft and the approval of the final 19. version. All the authors gave final approval of the version to be published. 8. Taylor CJ, Ordóñez-Mena JM, Roalfe AK, Lay-Flurrie S, Jones NR, Marshall T, et al. Trends in survival after a diagnosis of heart failure in the United Kingdom 2000-2017: population based cohort study. BMJ. 2019;364:l223. Funding 9. Drake C, Wald HL, Eber LB, Trojanowski JI, Nearing KA, Boxer RS. Research This study has been granted by the Primary Healthcare University Research priorities in post-acute and long-term care: results of a stakeholder needs Institute IDIAP-Jordi Gol. By this grant, the Institute assumed the costs origi- assessment. J Am Med Dir Assoc. 2019;20(7):911–5. nated to the SIDIAP database, related to the elaboration and data manage- 10. Muntwyler J, Abetel G, Gruner C, Follath F. One-year mortality among ment to develop this study. unselected outpatients with heart failure. Eur Heart J. 2002;23(23):1861–6. Muñoz et al. BMC Family Practice (2020) 21:145 Page 10 of 10 11. Ahmed A, Aronow WS, Fleg JL. Higher New York heart association classes illness? A systematic review and appraisal of methodological rigour. Palliat and increased mortality and hospitalization in patients with heart failure Med. 2019;33(8):900–25. and preserved left ventricular function. Am Heart J. 2006;151(2):444–50. 33. Hawkins NM, Scholes S, Bajekal M, Love H, O'Flaherty M, Raine R, et al. 12. Jaarsma T, Beattie JM, Ryder M, Rutten FH, McDonagh T, Mohacsi P, et al. Community care in England: reducing socioeconomic inequalities in heart Advanced Heart Failure Study Group of the HFA of the ESC. Palliative care failure. Circulation. 2012;126(9):1050–7. in heart failure: a position statement from the palliative care workshop of 34. Witte KK, Patel PA, Walker AMN, Schechter CB, Drozd M, Sengupta A, et al. the heart failure Association of the European Society of Cardiology. Eur J Socioeconomic deprivation and mode-specific outcomes in patients with Heart Fail. 2009;11(5):433–43. chronic heart failure. Heart. 2018;104(12):993–8. 13. Alba AC, Agoritsas T, Jankowski M, Courvoisier D, Walter SD, Guyatt GH, 35. Muñoz MA, Mundet-Tuduri X, Real J, et al. Heart failure labelled patients et al. Risk prediction models for mortality in ambulatory patients with heart with missing ejection fraction in primary care: prognosis and determinants. failure: a systematic review. Circ Heart Fail. 2013;6(5):881–9. BMC Fam Pract. 2017;18(1):38. 14. Schockmel M, Agrinier N, Jourdain P, Alla F, Eicher JC, Coulon P, Druelle J, Mulak G, Danchin N, Thilly N. Juillière Y; ODIN cohort participants. Socioeconomic factors and mortality in diastolic heart failure. Eur J Clin Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in Investig. 2014;44(4):372–83. published maps and institutional affiliations. 15. Checa C, Abellana R, Verdú-Rotellar JM, Berenguera A, Domingo M, Calero E, et al. Social risk and mortality: a cohort study in patients with advanced heart failure. J Cardiovasc Nurs. 2019;34(1):E8–E15. 16. Moser DK, Robinson S, Biddle MJ, Pelter MM, Nesbitt TS, Southard J, et al. Health literacy predicts morbidity and mortality in rural patients with heart failure. J Card Fail. 2015;21(8):612–8. 17. 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(9):845–51. 18. Lynch S. Hospice and palliative care access issues in rural areas. Am J Hosp Palliat Care. 2013;30(2):172–7. 19. McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Böhm M, Dickstein K. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: the task force for the diagnosis and treatment of acute and chronic heart failure 2012 of the European Society of Cardiology. Eur J Heart Fail. 2012;14(8):803–69. 20. Ramos R, Balló E, Marrugat J, Elosua R, Sala J, Grau M, et al. Validity for use in research on vascular diseases of the SIDIAP (information system for the development of research in primary care): the EMMA study. Rev Esp Cardiol (Engl Ed). 2012;65(1):29–37. 21. Formiga F, Chivite D, Casas S, Manito N, Pujol R. Functional assessment of elderly patients admitted for heart failure. Rev Esp Cardiol. 2006;59(7):740–2. 22. Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP. The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol. 2008;61(12): 1234–40. 23. Domínguez-Berjón MF, Borrell C, Cano-Serral G, Esnaola S, Nolasco A, Pasarín MI, et al. Constructing a deprivation index based on census data in large Spanish cities (the MEDEA project). Gac Sanit. 2008;22:179–87. 24. García González JV, Díaz Palacios E, Salamea García A, Cabrera González D, Menéndez Caicoya A, Fernández Sánchez A, et al. An evaluation of the feasibility and validity of a scale of social assessment of the elderly. Aten Primaria. 1999;23(7):434–40. 25. Domínguez Amorós M, Monllor Rico N, Simó Solsona M (2010). “Món rural i joves. Realitat juvenil i polítiques de joventut als municipis rurals de Catalunya”. Col·lecció Estudis, 31. Generalitat de Catalunya. 26. Henning-Smith C, Moscovice I, Kozhimannil K. Differences in social isolation and its relationship to health by rurality. J Rural Health. 2019;35(4):540–9. 27. Buja A, Toffanin R, Rigon S, Lion C, Sandonà P, Carraro D, et al. What determines frequent attendance at out-of-hours primary care services? Eur J Pub Health. 2015;25(4):563–8. 28. Takiguchi M, Yoshihisa A, Miura S, et al. Impact of body mass index on mortality in heart failure patients. Eur J Clin Investig. 2014;44(12):1197–205. https://doi.org/10.1111/eci.12354. 29. Okabe T, Yakushiji T, Kido T, Oyama Y, Igawa W, Ono M, et al. Relationship between worsening renal function and long-term cardiovascular mortality in heart failure patients. Int J Cardiol. 2017;230:47–52. 30. Vidán MT, Bueno H, Wang Y, Schreiner G, Ross JS, Chen J, et al. The relationship between systolic blood pressure on admission and mortality in older patients with heart failure. Eur J Heart Fail. 2010;12(2):148–55. 31. Timonin S, Kontsevaya A, McKee M, Leon DA. Reducing geographic inequalities in access times for acute treatment of myocardial infarction in a large country: the example of Russia. Int J Epidemiol. 2018;47(5):1594–602. 32. Neergaard MA, Brunoe AH, Skorstengaard MH, Nielsen MK. What socio- economic factors determine place of death for people with life-limiting