Background
Polypharmacy is a common problem among patients with acute decompensated heart failure (ADHF) who often have multiple comorbidities.
Objective
The aim of this study was to define the number of medications at hospital discharge and whether it is associated with clinical outcomes at 1 year.
Methods
We evaluated the number of medications in 2578 patients with ADHF who were ambulatory at hospital discharge in the Kyoto Congestive Heart Failure Registry and compared 1-year outcomes in 4 groups categorized by quartiles of the number of medications (quartile 1, ≤ 5; quartile 2, 6–8; quartile 3, 9–11; and quartile 4, ≥ 12).
Results
At hospital discharge, the median number of medications was 8 (interquartile range, 6–11) with 81.5% and 27.8% taking more than 5 and more than 10 medications, respectively. The cumulative 1-year incidence of a composite of death or rehospitalization (primary outcome measure) increased incrementally with an increasing number of medications (quartile 1, 30.8%; quartile 2, 31.6%; quartile 3, 39.7%; quartile 4, 50.3%; P < .0001). After adjusting for confounders, the excess risks of quartile 4 relative to those of quartile 1 remained significant (P = .01).
Conclusions
In the contemporary cohort of patients with ADHF in Japan, polypharmacy at hospital discharge was common, and excessive polypharmacy was associated with a higher risk of mortality and rehospitalizations within a 1-year period. Collaborative disease management programs that include a careful review of medication lists and an appropriate deprescribing protocol should be implemented for these patients.
KEY WORDS: acute decompensated heart failure, older, polypharmacy, prognosis
Polypharmacy is becoming more prevalent in the contemporary, guideline-directed clinical practice in patients with heart failure (HF).1,2 Multiple medications and eventual polypharmacy entail risks such as adverse drug reactions and a decline in medication adherence, especially in older patients.3,4 Many of the individual medications for HF have been demonstrated to positively influence patient outcomes when tested against a placebo in randomized controlled trials.5 However, multimorbid older patients, who are more likely to receive polypharmacy, were excluded from many evidence-generating clinical trials.
In a cross-sectional study of 1.4 million patients in primary care in Scotland, authors reported that HF patients with left ventricular systolic dysfunction had significantly greater comorbidity and polypharmacy.6 A post hoc analysis of ROCKET-AF study, in which 60% of patients had HF, revealed that 10 or more medications in patients with atrial fibrillation was associated with a higher risk of bleeding but not stroke.7 However, no previous large-scale study has reported the impact of polypharmacy on clinical outcomes in acute decompensated HF (ADHF) patients.
Thus, in this study, we aimed to evaluate medication use in real-world clinical practice and analyze the association of the number of medications at discharge with 1-year clinical outcomes among hospitalized patients with ADHF using a large Japanese observational registry.
Methods
Study Design
The Kyoto Congestive Heart Failure (KCHF) registry is a physician-initiated, prospective, observational, multicenter cohort study enrolling consecutive patients admitted to hospitals because of ADHF for the first time between October 2014 and March 2016 in 19 secondary and tertiary hospitals in Japan. Details on the study design and patient characteristics in the KCHF registry have been reported previously.8,9 Briefly, we enrolled all consecutive patients with ADHF as defined by the modified Framingham criteria and those who underwent HF-specific treatment involving intravenous drugs within 24 hours after admission in each participating center. Clinical follow-up information was collected in October 2017. The attending physicians or research assistants at each participating facility collected data on clinical events after the index hospitalization from hospital medical records or patients, relatives, or referring physicians. Patient records were anonymized before the analysis.
The investigation conforms with the principles outlined in the Declaration of Helsinki. The study was approved by the institutional review boards of Kyoto University Graduate School of Medicine and from all the hospital facilities and academic centers involved in this project (please find listed this information in Annex 1). This study was registered with University Hospital Medical Information Network (UMIN identifier: UMIN000015238). A waiver of written informed consent from each patient was granted by the institutional review boards of Kyoto University and each participating center, because the study met the conditions of the Japanese ethical guidelines for Medical and Health Research Involving Human Subjects.10 We disclosed this study's details to the public as an opt-out method, and this notice informed patients of their right to refuse enrollment.
Definitions and Outcome Measures
The number of oral medications at the time of discharge from the index hospitalization was assessed by drug class according to the Anatomical Therapeutic Chemical Classification System.11 Therefore, a prescription of 2 types of loop diuretics in 1 patient was counted as 1 medication. Combined products were measured by counting each single-ingredient product separately. We did not collect data on medications administered via injection; medications administered via eye drops, suppositories, ointments, and plasters; or medications that were not taken routinely. We did not collect data of drug schedule. Although we collected the dose of some cardiovascular drugs, we did not collect dose of drugs other than cardiovascular drugs; hence, doses and schedules were not included in the analysis. Detailed definitions of baseline clinical characteristics have been described previously.8,9
The primary outcome measure in this study was a composite of death from any cause or rehospitalization at 1 year after hospital discharge. The secondary outcome measures were all-cause death, cardiovascular death, any rehospitalization, and rehospitalization due to HF.
Study Patients
Among the 4056 enrolled patients in the KCHF registry, 3785 patients (93.3%) were discharged alive after hospitalization for ADHF. We excluded 196 patients for missing data on medications at hospital discharge and 38 patients without follow-up data (Figure 1). We further excluded 942 patients who had a walking disability (ie, wheelchair-bound or bedridden patients) at hospital discharge and 31 patients who had no data on the functional status at discharge because the presence of disability and impairment can reduce patient ability to adhere to recommendations and alter patient preference for treatment or study outcomes.12 Accordingly, this study's population consisted of 2578 patients with ADHF who were ambulatory at discharge.
FIGURE 1.
Study flowchart. ADHF, acute decompensated heart failure; KCHF, Kyoto Congestive Heart Failure.
Statistical Analysis
Among the 2578 study patients, 81.5% received a prescription of more than 5 medications and 27.8% received more than 10 medications. The median number of medications was 8 (interquartile range, 6–11; range, 0–24) (Figure 2). As most patients received 6 or more medications, that is, the most commonly cited definition for polypharmacy, we did not use this criteria. Instead, we categorized patients into 4 groups based on the quartiles of the number of medications at hospital discharge (quartile 1, ≤5; quartile 2, 6–8; quartile 3, 9–11; and quartile 4, ≥12). We compared the baseline characteristics and clinical outcomes of patients across the quartiles using the χ2 test for categorical variables and 1-way analysis of variance for continuous variables. To assess the trend across the quartiles, we used the Cochran-Armitage trend test for categorical variables and the Jonckheere-Terpstra test for continuous variables.
FIGURE 2.
Number of medications at discharge from the index hospitalization.
We regarded the date of hospital discharge as time 0 for the clinical follow-up. Cumulative incidences were estimated by the Kaplan-Meier method and compared using the log-rank test. We used a multivariable Cox proportional hazards model to estimate the risk of quartiles 2, 3, and 4 relative to that of quartile 1 (reference) for primary and secondary outcome measures. Results were expressed as hazard ratios (HRs) with the corresponding 95% confidence intervals (CIs). To adjust for potential confounders, we included both the quartiles based on the number of medications and the 20 clinically relevant risk-adjusting variables: being 80 years or older, women, body mass index (BMI) less than 22 kg/m2, acute coronary syndrome, nonacute coronary syndrome, atrial fibrillation or flutter, hypertension, diabetes mellitus, previous stroke, chronic lung disease, current smoking, living alone, systolic blood pressure less than 90 mm Hg at admission, heart rate less than 60/min at admission, estimated glomerular filtration rate less than 30 mL/min per 1.73 m2, anemia, serum albumin less than 3 g/dL, and HF with reduced ejection fraction, consistent with a previous study.8 Proportional hazard assumptions for quartiles were assessed on plots of log (time) versus log [−log (survival)] stratified by the risk variables and were verified to be acceptable. Missing values were handled without imputation and excluded from the fully adjusted model. As a sensitivity analysis, we also explored whether there is a stepwise increase in the risk for clinical outcome measures from quartile 1 to quartile 4 using continuous variables (0 for quartile 1, 1 for quartile 2, 2 for quartile 3, and 3 for quartile 4) in the multivariable model.
We also performed the subgroup analysis for the primary outcome measure stratified by age, sex, BMI, left ventricular ejection fraction (LVEF), and presence or absence of anemia. We evaluated the interaction between the subgroup factors and the effects of the quartiles of the number of medications on the primary outcome measure.
All statistical analyses were conducted by physicians (N.O. and T.K.) and a statistician (T.M.) using JMP 14.0 or SAS 9.4 (both SAS Institute Inc, Cary, North Carolina). Two-tailed P values less than .05 were considered significant.
Results
Patient Characteristics
Among the 2578 study patients, median age was 77 (interquartile range, 69–84; 18–103) years, and women accounted for 40%; 2004 patients (77.7%) received renin-angiotensin-aldosterone system (RAAS) inhibitors, and 1843 patients (71.5%) received β-blockers (Tables 1 and 2). Patients who received more medications had a greater BMI; more often had a history of HF admission, coronary artery disease, atrial fibrillation or flutter, diabetes mellitus, chronic kidney disease, anemia, and chronic lung disease; were more often unemployed; and more often used public assistance and long-term care insurance at hospital discharge than those who received fewer medications (Table 1). No significant difference was found in the LVEF, presence or absence of dementia, and living status according to the number of medications between these patients.
TABLE 1.
Patient Characteristics
Variables | No. Medications at Discharge | |||||
---|---|---|---|---|---|---|
All Patients (N = 2578) | Quartile 1 (≤5) (N = 478, 18.5%) |
Quartile 2 (6–8) (N = 859, 33.3%) |
Quartile 3 (9–11) (N = 736, 28.5%) |
Quartile 4 (≥12) (N = 505, 19.5%) |
P for Trend | |
Clinical characteristics | ||||||
Age, y | 77 (69–84) | 75 (65–83) | 78 (67–84) | 78 (71–84) | 77 (70–83) | .014 |
Age ≥ 80 ya | 1078 (41.8) | 163 (34.1) | 381 (44.4) | 327 (44.4) | 207 (41.0) | .06 |
Womena | 1026 (39.8) | 209 (43.7) | 334 (38.8) | 300 (40.8) | 183 (36.2) | .06 |
BMI | 23.3 ± 4.5 | 22.7 ± 4.2 | 23.3 ± 4.8 | 23.4 ± 4.4 | 23.8 ± 4.6 | <.0001 |
BMI < 22 (kg/m2)a | 1053 (42) | 213 (46.2) | 358 (43.0) | 304 (42.5) | 178 (35.7) | .002 |
Body weight | 58.6 ± 14.8 | 56.9 ± 14.3 | 59.1 ± 15.9 | 58.3 ± 13.8 | 60.0 ± 14.6 | .006 |
Origin | ||||||
Coronary artery disease | 850 (33.0) | 56 (11.7) | 240 (27.9) | 292 (39.7) | 262 (51.9) | <.0001 |
ACSa | 146 (5.7) | 18 (3.8) | 49 (5.7) | 53 (7.2) | 26 (5.2) | .2 |
Non-ACSa | 704 (27.3) | 38 (8.0) | 191 (22.2) | 239 (32.5) | 236 (46.7) | <.0001 |
Previous myocardial infarction | 581 (22.5) | 30 (6.3) | 144 (16.8) | 206 (28.0) | 201 (39.8) | <.0001 |
Hypertensive heart disease | 640 (24.8) | 145 (30.3) | 232 (27.0) | 169 (23.0) | 94 (18.6) | <.0001 |
Valvular heart disease | 456 (17.7) | 100 (20.9) | 162 (18.9) | 125 (17.0) | 69 (13.7) | .002 |
Cardiomyopathy | 433 (16.8) | 106 (22.2) | 166 (19.3) | 100 (13.6) | 61 (12.1) | <.0001 |
Medical history | ||||||
Previous heart failure hospitalizationa | 875 (33.9) | 88 (18.4) | 229 (26.7) | 291 (39.5) | 267 (52.9) | <.0001 |
Atrial fibrillation or fluttera | 1073 (41.6) | 154 (32.2) | 361 (42.0) | 336 (45.7) | 222 (44.0) | <.0001 |
Hypertensiona | 1852 (71.8) | 289 (60.5) | 617 (71.8) | 570 (77.5) | 376 (74.5) | <.0001 |
Diabetes mellitusa | 992 (38.5) | 78 (16.3) | 281 (32.7) | 342 (46.5) | 291 (57.6) | <.0001 |
Previous strokea | 335 (13.0) | 32 (6.7) | 103 (12.0) | 122 (16.6) | 78 (15.5) | <.0001 |
Malignancy | 359 (13.9) | 67 (14.0) | 103 (12.0) | 114 (15.5) | 75 (14.6) | .25 |
Chronic lung diseasea | 336 (13.0) | 52 (10.9) | 91 (10.6) | 97 (13.2) | 96 (19.0) | <.0001 |
Current smokinga | 388 (15.3) | 92 (19.6) | 132 (15.6) | 97 (13.4) | 67 (13.5) | .005 |
Social background | ||||||
Dementia | 236 (9.2) | 36 (7.5) | 79 (9.2) | 79 (10.7) | 42 (8.3) | .46 |
Living alonea | 577 (22.4) | 108 (22.6) | 191 (22.2) | 172 (23.4) | 106 (21.0) | .72 |
Unemployed | 2133 (82.7) | 370 (77.4) | 683 (79.5) | 629 (85.5) | 451 (89.3) | <.0001 |
Public assistance | 152 (5.9) | 20 (4.2) | 39 (4.5) | 47 (6.4) | 46 (9.1) | .0002 |
Use of long-term care insurance at discharge | 568 (22.0) | 67 (14.0) | 151 (17.6) | 206 (28.0) | 144 (28.5) | <.0001 |
Vital signs and symptoms at admission | ||||||
Systolic blood pressure <90 mm Hga | 66 (2.6) | 14 (2.9) | 22 (2.6) | 13 (1.8) | 17 (3.4) | .95 |
Systolic blood pressure (continuous) | 149.4 ± 35.5 | 153.0 ± 36.2 | 150.3 ± 36.3 | 149.2 ± 34.8 | 144.8 ± 34.0 | .0006 |
Heart rate < 60/mina | 159 (6.2) | 42 (8.9) | 45 (5.3) | 35 (4.8) | 37 (7.4) | .35 |
NYHA class III/VI | 2194 (85.4) | 391 (82.5) | 738 (86.1) | 634 (86.3) | 431 (85.7) | .20 |
Tests at admission | ||||||
eGFRa < 30 mL/min per 1.73 m2 | 597 (23.2) | 59 (12.4) | 175 (20.4) | 190 (25.8) | 173 (34.3) | <.0001 |
Anemiaa | 1588 (61.7) | 261 (55.0) | 473 (55.1) | 491 (66.7) | 363 (72.0) | — |
Serum albumin < 3 g/dLa | 256 (10.3) | 36 (7.7) | 95 (11.5) | 67 (9.3) | 58 (11.9) | |
LVEF, % | 45.5 ± 16.1 | 46.3 ± 16.6 | 45.2 ± 15.5 | 45.5 ± 16.0 | 45.3 ± 16.7 | |
Classification of HF according to LVEF | .0006 | |||||
HFrEF (LVEF, <40%)a | 1015 (39.5) | 177 (37.3) | 338 (39.4) | 299 (40.7) | 201 (40.0) | .35 |
HFmrEF (LVEF, 40%–49%) | 487 (19.0) | 83 (17.5) | 168 (19.6) | 137 (18.6) | 99 (19.7) | .202 |
HFpEF (LVEF, ≥50%) | 1068 (41.6) | 215 (45.3) | 351 (41.0) | 299 (40.7) | 203 (40.4) |
Continuous variables are presented as mean ± standard deviation or median with interquartile range. Categorical variables are presented as number (percentage).
Abbreviations: ACS, acute coronary syndrome; BMI, body mass index; eGFR, estimated glomerular filtration rate; HFmrEF, heart failure with midrange ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; HR, heart rate; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association.
aRisk-adjusting variables selected for the multivariable Cox proportional hazards models.
TABLE 2.
Medication at Hospital Discharge
Variables | No. Medications at Discharge | P for Trend | ||||
---|---|---|---|---|---|---|
All Patients (N = 2578) | Quartile 1 (≤5) (N = 478, 18.5%) |
Quartile 2 (6–8) (N = 859, 33.3%) |
Quartile 3 (9–11) (N = 736, 28.5%) |
Quartile 4 (≥12) (N = 505, 19.5%) |
||
Medications at discharge | ||||||
No. medications | 8 (6–11) | 4 (3–5) | 7 (6–8) | 10 (9–11) | 13 (12–15) | — |
RAAS inhibitors | 2004 (77.7) | 318 (66.5) | 669 (77.9) | 601 (81.7) | 416 (82.4) | <.0001 |
MRAs | 1217 (47.2) | 179 (37.5) | 421 (49.0) | 366 (49.7) | 251 (49.7) | .0004 |
ACEi or ARB | 1615 (62.7) | 250 (52.3) | 529 (61.6) | 492 (66.9) | 344 (68.1) | <.0001 |
ACEi | 687 (26.7) | 127 (26.6) | 238 (27.7) | 185 (25.1) | 137 (27.1) | .79 |
ARB | 942 (36.5) | 126 (26.4) | 294 (34.2) | 308 (41.9) | 214 (42.4) | <.0001 |
BB | 1843 (71.5) | 288 (60.3) | 607 (70.7) | 556 (75.5) | 392 (77.6) | <.0001 |
Diuretics | ||||||
Loop diureticsa | 2113 (82.0) | 334 (69.9) | 683 (79.5) | 643 (87.4) | 453 (89.7) | <.0001 |
Thyazide diuretics | 140 (5.4) | 12 (2.5) | 30 (3.5) | 47 (6.4) | 51 (10.1) | <.0001 |
Tolvaptan | 239 (9.3) | 10 (2.1) | 49 (5.7) | 75 (10.2) | 105 (20.8) | <.0001 |
Vasodilators | ||||||
CCBsa | 897 (34.8) | 116 (24.3) | 255 (29.7) | 304 (41.3) | 222 (44.0) | <.0001 |
Nitrates | 297 (11.5) | 9 (1.9) | 52 (6.1) | 111 (15.1) | 125 (24.8) | <.0001 |
Inotropic agents | ||||||
Digitalis | 151 (5.9) | 13 (2.7) | 35 (4.1) | 53 (7.2) | 50 (9.9) | <.0001 |
Pimopendane | 129 (5.0) | 7 (1.5) | 23 (2.7) | 40 (5.4) | 59 (11.7) | <.0001 |
Antidysrhythmics | ||||||
Amiodarone | 197 (7.6) | 12 (2.5) | 55 (6.4) | 62 (8.4) | 68 (13.5) | <.0001 |
Antidysrhythmics other than amiodarone | 150 (5.8) | 14 (2.9) | 38 (4.4) | 50 (6.8) | 48 (9.5) | <.0001 |
Antithrombotic agents | ||||||
Triple antithrombotic agents | 386 (15.0) | 9 (1.9) | 90 (10.5) | 139 (18.9) | 148 (29.3) | <.0001 |
Dual antithrombotic agents | 1013 (38.3) | 68 (14.2) | 290 (33.8) | 352 (47.8) | 303 (60.0) | <.0001 |
Aspirin | 1013 (39.3) | 68 (14.2) | 290 (33.8) | 352 (47.8) | 303 (60.0) | <.0001 |
Other antiplatelet agents | 646 (25.1) | 32 (6.7) | 178 (20.7) | 226 (30.7) | 210 (41.6) | <.0001 |
Warfarin | 678 (26.3) | 54 (11.3) | 204 (23.8) | 230 (31.3) | 190 (37.6) | <.0001 |
DOAC | 578 (22.4) | 114 (23.9) | 216 (25.2) | 158 (21.5) | 90 (17.8) | .005 |
Glucose-lowering drugs | ||||||
Insulin | 205 (8.0) | 9 (1.9) | 48 (5.6) | 71 (9.7) | 77 (15.3) | <.0001 |
Metformin | 76 (3.0) | 4 (0.8) | 15 (1.8) | 29 (3.9) | 28 (5.5) | <.0001 |
DPP4 inhibitors | 444 (17.2) | 15 (3.1) | 104 (12.1) | 167 (22.7) | 158 (31.3) | <.0001 |
SGLT2 inhibitors | 13 (0.5) | 1 (0.2) | 2 (0.2) | 6 (0.8) | 4 (0.8) | .07 |
NSAIDs | 52 (2.0) | 3 (0.6) | 12 (1.4) | 18 (2.5) | 19 (3.8) | .0001 |
Others | 3 (2–5) | 1 (0–1) | 3 (2–3) | 4 (3–5) | 7 (6–9) | <.0001 |
Categorical variables are presented as number (percentage). Dual and triple antithrombotic agents at hospital discharge were defined as taking 2 and 3 antithrombotic drugs (aspirin, other antiplatelet, warfarin, and direct oral anticoagulant), respectively.
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; BB, β-blockers; CCBs, calcium channel blockers; DOAC, direct oral anticoagulants; DPP4, dipeptidyl peptidase-4 inhibitors; MRAs, mineralocorticoid receptor antagonists; NSAIDs, nonsteroidal anti-inflammatory drugs; RAAS, renin-angiotensin-aldosterone system; SGLT2, sodium glucose cotransporter 2.
aDrugs counted as 1 medicine.
Clinical Outcomes
The median length of follow-up was 492 (interquartile range, 374–666) days. The cumulative 1-year composite incidence of death or rehospitalization increased incrementally with an increasing number of medications (quartile 1 [≤5], 30.8%; quartile 2 [6–8], 31.6%; quartile 3 [9–11], 39.7%; quartile 4 [≥12], 50.3%; log-rank P < .0001) (Figure 3A). After adjusting for confounders, the excess risk of quartile 4 relative to that of quartile 1 remained significant for the primary outcome measure (HR, 1.30; 95% CI, 1.04–1.61; P = .01) (Table 3). The cumulative 1-year incidence of all-cause death was significantly higher in quartile 3 and quartile 4 than in quartile 1 (Figure 3B). However, the excess adjusted risk of quartile 3 and quartile 4 relative to that of quartile 1 was no longer significant for all-cause death (Table 3). The cumulative 1-year incidences of any rehospitalization and HF rehospitalization were also significantly higher in quartile 3 and quartile 4 than in quartile 1 (Figure 3C and 3D). The excess adjusted risk of quartile 4 relative to that of quartile 1 remained significant for any rehospitalization, and the excess adjusted risk of quartile 3 and quartile 4 relative to that of quartile 1 remained significant for HF rehospitalization (Table 3). Although detailed information of the causes of readmissions and deaths was not available in this study, we found that cardiovascular death was accounted in 162 subjects (6.5%) so, approximately in two-thirds (61.3%) of this group, the mortality may be attributable to cardiac causes (Table 3).
FIGURE 3.
Kaplan-Meier curves for clinical outcomes according to the quartiles based on the number of medications at discharge from the index hospitalization. A, All-cause death or any rehospitalization. B, All-cause death. C, Any rehospitalization. D, Heart failure rehospitalization.
TABLE 3.
Postdischarge Clinical Outcomes by Number of Medications at Discharge
No. Patients With Event and Cumulative 1-y Incidence, % | Crude HR | 95% CI | P | Adjusted HR | 95% CI | P | |
---|---|---|---|---|---|---|---|
All-cause death or any rehospitalization | |||||||
Quartile 1 (≤5) | 146 (30.8) | Reference | Reference | ||||
Quartile 2 (6–8) | 268 (31.6) | 1.01 | 0.84–1.21 | .85 | 0.89 | 0.73–1.09 | .46 |
Quartile 3 (9–11) | 289 (39.7) | 1.36 | 1.13–1.62 | .0008 | 1.09 | 0.89–1.33 | .26 |
Quartile 4 (≥12) | 251 (50.3) | 1.81 | 1.50–2.18 | <.0001 | 1.30 | 1.04–1.61 | .01 |
All-cause death | |||||||
Quartile 1 (≤5) | 39 (8.3) | Reference | Reference | ||||
Quartile 2 (6–8) | 82 (9.7) | 1.22 | 0.89–1.68 | .21 | 1.10 | 0.78–1.56 | .55 |
Quartile 3 (9–11) | 82 (11.3) | 1.43 | 1.04–1.97 | .03 | 1.21 | 0.85–1.74 | .28 |
Quartile 4 (≥12) | 66 (13.3) | 1.72 | 1.23–2.40 | .001 | 1.30 | 0.88–1.92 | .18 |
Cardiovascular death | |||||||
Quartile 1 (≤5) | 22 (4.7) | Reference | Reference | ||||
Quartile 2 (6–8) | 53 (6.3) | 1.02 | 0.91–1.15 | .62 | 1.02 | 0.90–1.17 | .66 |
Quartile 3 (9–11) | 49 (6.9) | 1.06 | 0.97–1.20 | .29 | 1.06 | 0.93–1.22 | .21 |
Quartile 4 (≥12) | 38 (7.9) | 1.16 | 1.02–1.33 | .02 | 1.10 | 0.94–1.29 | .22 |
Any rehospitalization | |||||||
Quartile 1 (≤5) | 136 (28.9) | Reference | Reference | ||||
Quartile 2 (6–8) | 242 (29.0) | 0.98 | 0.82–1.20 | .98 | 0.87 | 0.71–1.08 | .23 |
Quartile 3 (9–11) | 264 (36.9) | 1.33 | 1.10–1.61 | .002 | 1.07 | 0.86–1.32 | .53 |
Quartile 4 (≥12) | 230 (46.9) | 1.81 | 1.49–2.20 | <.0001 | 1.29 | 1.02–1.63 | .03 |
Rehospitalization for HF | |||||||
Quartile 1 (≤5) | 70 (15.1) | Reference | Reference | ||||
Quartile 2 (6–8) | 142 (17.2) | 1.18 | 0.90–1.54 | .22 | 1.03 | 0.76–1.37 | .88 |
Quartile 3 (9–11) | 171 (24.2) | 1.79 | 1.38–2.31 | <.0001 | 1.37 | 1.03–1.84 | .03 |
Quartile 4 (≥12) | 151 (31.2) | 2.29 | 1.76–2.99 | <.0001 | 1.50 | 1.10–2.05 | .009 |
Abbreviations: CI, confidence interval; HF, heart failure; HR, hazard ratio.
In a sensitivity analysis, we observed a significant excess risk for the primary outcome measure per quartile of the number of medications (HR, 1.12; 95% CI, 1.05–1.20; P = .0008). The excess risk for HF rehospitalization per quartile of number of medications was also significant (HR, 1.18; 95% CI, 1.05–1.29; P = .0005).
Subgroup Analysis
Significant interactions were found between those subgroup factors such as age and anemia and the association of the number of medications on the primary outcome measure. In patients younger than 80 years, but not in those 80 years or older, there was significant excess adjusted risk of quartile 3 and quartile 4 relative to that of quartile 1 for the primary outcome measure. In patients without anemia, but not in those with anemia, there was significant excess adjusted risk of quartile 4 relative to that of quartile 1 for the primary outcome measure. No significant interactions were observed between factors such as sex, BMI, LVEF, and the effect of the number of medications on the primary outcome measure (Figure 4).
FIGURE 4.
Forrest plots for the subgroup analyses on the primary outcome measure (all-cause death or any rehospitalization) at 1 year after discharge from the index hospitalization. BMI, body mass index; CI, confidence interval; HR, hazard ratio; and LVEF, left ventricular ejection fraction.
Discussion
The main findings of this real-world study, evaluating polypharmacy and clinical outcomes in patients hospitalized for ADHF, were as follows. First, the median number of medications at hospital discharge was 8, and 81.5% of patients received prescriptions for more than 5 medications. Second, patients receiving more medications had more complex medical history and social background than those receiving fewer medications. Finally, in patients receiving 12 or more medications, the adjusted risk for death or any hospitalization during the first year after discharge was significantly higher than in patients receiving 5 or less medications. However, it might be noteworthy that the excess risk of a greater number of medications for the primary outcome measure was not significant in patients 80 years or older and patients with anemia. Unmeasured factors such as poor compliance may be accountable for this discrepancy. As discussed in the Limitations section, compliance was not evaluated in this study.
The burden of HF falls disproportionately on older people, who are often simultaneously afflicted with many comorbidities. In this study, the median age of study patients was 77 years, and conditions such as diabetes (38.5%), hypertension (71.8%), renal failure (23.2%), chronic lung disease (13.0%), previous stroke (13.0%), previous myocardial infarction (22.5%), and atrial fibrillation or flutter (41.6%) were prevalent. Thus, practitioners typically face the challenge of managing not a single but multiple conditions. Consequently, multiple medications and polypharmacy are almost inevitable for these patients. Notably, the median age of 77 years in the present registry was much higher than that reported in previous large registries on HF. With the aging population, this scenario will become more common. Underuse of cardioprotective drugs such as RAAS inhibitors and β-blockers may increase hospital admissions or death because of exaggeration of HF, whereas excessive polypharmacy (ie, 10 medications) has been reported to be strongly associated with inappropriate medication use and adverse drug events.12,13 Adverse drug events include an increased risk of nonadherence, drug-drug interactions, adverse drug reactions, and preventable medication-related hospital admissions or death. Hypotension due to antihypertensive drugs, hyperkalemia and renal failure due to RAAS inhibitors, hyponatremia and/or hypokalemia due to diuretics, hemorrhagic events due to antithrombotic drugs, QT prolongation and ventricular dysrhythmia due to various drugs, and hypoglycemia due to hypoglycemic drugs are frequently observed adverse drug events in patients with HF. Notably, the number of drugs not listed in Table 2 are remarkably high in the study patients. Thus, noncardiovascular drugs may also cause adverse drug events in these patients. Use of the FORTA (“Fit fOR The Aged”) list has been reported to be helpful for improving pharmacotherapy in the multimorbid, older patients. This approach may also have the potential to improve future clinical outcomes in patients with HF.14
In addition to the medical history, we were interested in describing the social background of patients with HF. The presence of dementia, living without family support, and a low-income status complicate the management of HF. In this study, there was a substantial number of patients with dementia (9.2%), patients living alone (22.4%), and unemployed patients (82.7%). It is difficult to directly compare our findings with those of previous studies because of differences in the definitions; however, dementia, social isolation, and unemployment are generally associated with poor medication adherence.13,15 Approximately one-fifth of patients used long-term care insurance at discharge, a unique Japanese healthcare system to support the daily living of older patients. We could not clarify the effect of long-term care insurance itself in this study; however, such a healthcare system is crucial for older patients with HF to support their lives and maintain medication adherence. It may also have the potential to improve future clinical outcomes in these patients.
Our study emphasizes the need for physicians to be careful and judicious when caring for patients with HF. However, little evidence is available to guide polypharmacy in patients with HF and multiple comorbidities. Previous studies have indicated that it could be beneficial to decrease the number of medications among multimorbid patients to reduce the risk of medication-related harm.16,17 A survey among multimorbid older adults in Denmark revealed that 41% of patients 65 years or older with 10 or more prescribed medications were interested in a consultation at an outpatient clinic specializing in polypharmacy.16 To reduce inappropriate medication use, it would be required to provide medication reviews and to prioritize those drugs for possible discontinuation that have the lowest benefit-harm ratio and lowest likelihood of adverse withdrawal reactions or disease rebound syndromes in each patient before hospital discharge.18 Development of learning healthcare systems, which include cardiovascular disease care innovations in informatics, patient-clinician partnerships, incentives, and development of a continuous learning culture, has been drawing attention to improve the quality and efficacy of medication in patients with cardiovascular diseases.19 One of the learning healthcare systems available at present is outpatient cardiac rehabilitation; however, a poor participation rate is reported, with overall participation rates less than 50% during recent decades in Japan despite international recommendations.20,21 Efforts to increase the participation rate for cardiac rehabilitation and the development of new learning healthcare systems are necessary. Whenever possible, patients with HF, particularly those with multiple competing comorbidities and polypharmacy, need to be enrolled in such programs.
Limitations
This study has several notable limitations. First, there was potential for residual confounding due to the observational study design. Despite extensive adjustments, residual confounding may have influenced the observed association. Second, we did not have data on patient adherence to medications in this study. Although we excluded patients with walking disability at discharge, unmeasurable reasons for reducing patient ability to adherence to recommendations may affect the number of medications. A deprescribing protocol was not proposed in any hospitals that participated during the study period; however, it can be assumed that some of the attending physicians did not prescribe a prophylactic drug for very older patients or high-risk patients for ADEs.22 Third, the data were limited to prescriptions at hospital discharge, and we did not have data on the doses of medications or the prescriptions after hospital discharge. However, patients with polypharmacy driven by chronic medical conditions do not often have dramatic changes in the doses or number of medications they are taking. Fourth, the data did not include over-the-counter medicines, and complementary and alternative medicines after hospital discharge. However, because the public health insurance system in Japan is adopted for all citizens, patients with HF rarely need over-the-counter or complementary and alternative medicines because they are expensive. Fifth, although we investigated the number of all medications, we did not collect data to differentiate each drug except those listed in Table 2. Authors of a recent study who analyzed medication data using 558 older patients with HF hospitalization from the REGARD study indicated that most medications prescribed were noncardiovascular medications, such as proton pump inhibitors and electrolyte supplements.23 Sixth, we investigated the number of medications administered orally, but some types of drugs administered parenterally are often available in the current clinical practice. Therefore, the actual status and adverse effects of polypharmacy in patients with ADHF may have been underestimated in this study. Finally, as already mentioned in the Clinical Outcomes section, detailed information was not available for the specific cause of readmissions or death. Moreover, increasing medication use has been associated with a higher risk of ADEs, especially in older patients.24,25 Future studies are needed to examine the association of polypharmacy with ADEs in patients with ADHF.
Conclusions
In the contemporary cohort of patients with ADHF, polypharmacy at hospital discharge was common, and excessive polypharmacy was associated with a higher risk of mortality and rehospitalizations 1 year after discharge. Cardiac rehabilitation and collaborative disease management programs that include the careful review of medication lists and an appropriate deprescribing protocol should be implemented for these patients.
What’s New and Important
Patients with ADHF are prescribed multiple medications at hospital discharge in the real-word clinical practice, and the number of medications is independently associated with a higher risk of mortality and rehospitalizations. The excess risks of multiple medications were independently observed when the number of medications become 12 or more.
Our study underscore the importance in patients with ADHF to implement a comprehensive management program that includes a careful review of medication lists at discharge and perform an appropriate deprescribing protocol in the follow-up.
We consider that cardiac rehabilitation programs, regrettably underused in Japan and in many other countries, may be ideally used as an important component of this postdischarge program.
Annex 1:
The study was approved by the institutional review boards of Kyoto University Graduate School of Medicine (approval no. E2311), Shiga General Hospital (approval no. 20141120-01), Tenri Hospital (approval no. 640), Kobe City Medical Center General Hospital (approval no. 14094), Hyogo Prefectural Amagasaki General Medical Center (approval no. Rinri 26-32), National Hospital Organization Kyoto Medical Center (approval no. 14-080), Mitsubishi Kyoto Hospital (approved December 10, 2014), Okamoto Memorial Hospital (approval no. 201503), Japanese Red Cross Otsu Hospital (approval no. 318), Hikone Municipal Hospital (approval no. 26-17), Japanese Red Cross Osaka Hospital (approval no. 392), Shimabara Hospital (approval no. E2311), Kishiwada City Hospital (approval no. 12), Kansai Electric Power Hospital (approval no. 26-59), Shizuoka General Hospital (approval no. Rin14-11-47), Kurashiki Central Hospital (approval no. 1719), Kokura Memorial Hospital (approval no. 14111202), Kitano Hospital (approval no. P14-11-012), and Japanese Red Cross Wakayama Medical Center (approval no. 328).
Acknowledgments
The authors thank the contributions of all Kyoto Congestive Heart Failure study investigators.
They also thank Tomás Romero, MD, Associate Clinical Professor, University of California, and Member, Institutional Review Board, Sharp Health Care, San Diego, California, for his advice in reviewing and editing the manuscript.
Footnotes
This study was supported by grant 18059186 from the Japanese Agency for Medical Research and Development (Drs K. Kuwahara, T. Kato, and N. Ozasa). The founder had no role in the study design, collection, analysis, or interpretation of the data; writing the manuscript; or the decision to submit the article for publication.
T.M. reports lecturer's fees from Bristol-Myers Squibb, Daiichi Sankyo, Japan Lifeline, Kowa, Kyocera, Novartis, and Toray; manuscript fees from Bristol-Myers Squibb and Kowa; and being part of the advisory board for Sanofi. K.K. received grants and personal fees from Astra Zeneca and Otsuka Pharmaceutical; personal fees from Ono Pharmaceutical Co, Ltd; personal fees from Daiichi-Sankyo Co, Ltd; grants and personal fees from Mitsubishi Tanabe Pharma Corporation; personal fees from Eli Lilly Japan K.K.; grants and personal fees from Boehringer Ingelheim; personal fees from Bayer Yakuhin, Novartis Pharma, and Pfizer Japan, Inc; grants from EP-CRSU Co, Ltd; grants from Takeda Pharmaceutical Co, Ltd; grants from Medtronic Japan Co, Ltd; grants from Fukuda Denshi; and grants from Taisho Pharmaceutical Co, Ltd.
N. Ozasa and T. Kato had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Contributor Information
Takao Kato, Email: tkato75@kuhp.kyoto-u.ac.jp.
Takeshi Morimoto, Email: morimoto@kuhp.kyoto-u.ac.jp.
Hidenori Yaku, Email: yakuh@kuhp.kyoto-u.ac.jp.
Erika Yamamoto, Email: erkymmt@kuhp.kyoto-u.ac.jp.
Yasutaka Inuzuka, Email: inu3@kuhp.kyoto-u.ac.jp.
Yodo Tamaki, Email: ytamaki@tenriyorozu.jp.
Takeshi Kitai, Email: t-kitai@kcho.jp.
Koichi Washida, Email: washida_k@kuhp.kyoto-u.ac.jp.
Yuta Seko, Email: yuta1217@kuhp.kyoto-u.ac.jp.
Yusuke Yoshikawa, Email: yoshikaw@kuhp.kyoto-u.ac.jp.
Yukihito Sato, Email: cardioys@kuhp.kyoto-u.ac.jp.
Hirohiko Motoki, Email: hmotoki@shinshu-u.ac.jp.
Koichiro Kuwahara, Email: kkuwah@shinshu-u.ac.jp.
Takeshi Kimura, Email: taketaka@kuhp.kyoto-u.ac.jp.
REFERENCES
- 1.Boyd CM, Darer J, Boult C, Fried LP, Boult L, Wu AW. Clinical practice guidelines and quality of care for older patients with multiple comorbid diseases: implications for pay for performance. JAMA. 2005;294:716–724. [DOI] [PubMed] [Google Scholar]
- 2.Leendertse AJ, Egberts AC, Stoker LJ, Van Den Bemt PM. Frequency of and risk factors for preventable medication-related hospital admissions in the Netherlands. Arch Intern Med. 2008;168:1890–1896. [DOI] [PubMed] [Google Scholar]
- 3.Gheorghiade M Shah AN Vaduganathan M, et al. Recognizing hospitalized heart failure as an entity and developing new therapies to improve outcomes. Academics', clinicians', industry's, regulators', and payers' perspectives. Heart Fail Clin. 2013;9:285–290. [DOI] [PubMed] [Google Scholar]
- 4.Ho PM, Bryson CL, Rumsfeld JS. Medication adherence: its importance in cardiovascular outcomes. Circulation. 2009;119:3028–3035. [DOI] [PubMed] [Google Scholar]
- 5.Yancy CW Jessup M, et al. Writing Committee Members . 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association task force on practice guidelines. Circulation. 2013;128:e240–e327. [DOI] [PubMed] [Google Scholar]
- 6.Baron-Franco B McLean G Mair F, et al. Comorbidity and polypharmacy in chronic heart failure: a large cross-sectional study in primary care. Br J Gen Pract. 2017;67(658):e314–e320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Piccini JP Hellkamp AS Washam JB, et al. Polypharmacy and the efficacy and safety of rivaroxaban versus warfarin in the prevention of stroke in patients with nonvalvular atrial fibrillation. Circulation. 2016;133(4):352–360. [DOI] [PubMed] [Google Scholar]
- 8.Yaku H Ozasa N Morimoto T, et al. Demographics, management, and in-hospital outcome of hospitalized acute heart failure syndrome patients in contemporary real clinical practice in Japan: observations from the prospective, multicenter Kyoto Congestive Heart Failure (KCHF) registry. Circ J. 2018;82:2811–2819. [DOI] [PubMed] [Google Scholar]
- 9.Yamamoto E Kato T Ozasa N, et al. Kyoto Congestive Heart Failure (KCHF) study: rationale and design. ESC Heart Fail. 2017;4:216–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ito-Ihara T. Outline of ethical guidelines for medical and health research involving human subjects. Jpn Pharmacol Ther. 2015;43:s11–s21. [Google Scholar]
- 11.World Health Organization Collaborating Centre for Drug Statistics Methodology (WHOCC) . 2021. Available at: https://www.whocc.no/atc_ddd_index/. [Google Scholar]
- 12.Wong CY, Chaudhry SI, Desai MM, Krumholz HM. Trends in comorbidity, disability, and polypharmacy in heart failure. Am J Med. 2011;124:136–143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lau DT, Kasper JD, Potter DE, Lyles A, Bennett RG. Hospitalization and death associated with potentially inappropriate medication prescriptions among elderly nursing home residents. Arch Intern Med. 2005;165:68–74. [DOI] [PubMed] [Google Scholar]
- 14.Wehling M. How to use the FORTA (“Fit fOR The Aged”) list to improve pharmacotherapy in the elderly. Drug Res (Stuttg). 2016;66:57–62. [DOI] [PubMed] [Google Scholar]
- 15.Friedmann E Thomas SA Liu F, et al. Relationship of depression, anxiety, and social isolation to chronic heart failure outpatient mortality. Am Heart J. 2006;152:940.e1–940.e948. [DOI] [PubMed] [Google Scholar]
- 16.Schiøtz ML Frølich A Jensen AK, et al. Polypharmacy and medication deprescribing: a survey among multimorbid older adults in Denmark. Pharmacol Res Perspect. 2018;6:e00431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Achterhof AB Rozsnyai Z Reeve E, et al. Potentially inappropriate medication and attitudes of older adults towards deprescribing. PLoS One. 2020;15:e0240463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Scott IA Hilmer SN Reeve E, et al. Reducing inappropriate polypharmacy: the process of deprescribing. JAMA Intern Med. 2015;175:827–834. [DOI] [PubMed] [Google Scholar]
- 19.Maddox TM Albert NM Borden WB, et al. The learning healthcare system and cardiovascular care: a scientific statement from the American Heart Association. Circulation. 2017;135:e826–e857. [DOI] [PubMed] [Google Scholar]
- 20.Dalal HM, Doherty P, Taylor RS. Cardiac rehabilitation. BMJ. 2015;351:h5000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kamiya K Yamamoto T Tsuchihashi-Makaya M, et al. Nationwide survey of multidisciplinary care and cardiac rehabilitation for patients with heart failure in Japan—an analysis of the AMED-CHF study. Circ J. 2019;83:1546–1552. [DOI] [PubMed] [Google Scholar]
- 22.Mahé I Sterpu R Bertoletti L, et al. Long-term anticoagulant therapy of patients with venous thromboembolism. What are the practices? PLoS One. 2015;10:e0128741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Unlu O Levitan EB Reshetnyak E, et al. Polypharmacy in older adults hospitalized for heart failure. Circ Heart Fail. 2020;13:e006977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kripalani S Roumie CL Dalal AK, et al. Effect of a pharmacist intervention on clinically important medication errors after hospital discharge: a randomized trial. Ann Intern Med. 2012;157:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pretorius RW, Gataric G, Swedlund SK, Miller JR. Reducing the risk of adverse drug events in older adults. Am Fam Physician. 2013;87:331–336. [PubMed] [Google Scholar]