2024 年 88 巻 5 号 p. 672-679
Background: This study determined the incidence of hospitalization-associated disability (HAD) and its characteristics in older patients with heart failure in Japan.
Methods and Results: Ninety-six institutions participated in this nationwide multicenter registry study (J-Proof HF). From December 2020 to March 2022, consecutive heart failure patients aged ≥65 years who were prescribed physical rehabilitation during hospitalization were enrolled. Of the 9,403 patients enrolled (median age 83.0 years, 50.9% male), 3,488 (37.1%) had HAD. Compared with the non-HAD group, the HAD group was older and had higher rates of hypertension, chronic kidney disease, and cerebrovascular disease comorbidity. The HAD group also had a significantly lower Barthel Index score and a significantly higher Kihon checklist score before admission. Of the 9,403 patients, 2,158 (23.0%) had a preadmission Barthel Index score of <85 points. Binomial logistic analysis revealed that age and preadmission Kihon checklist score were associated with HAD in patients with a preadmission Barthel Index score of ≥85, compared with New York Heart Association functional classification and preadmission cognitive decline in those with a Barthel Index score <85.
Conclusions: This nationwide registry survey found that 37.1% of older patients with HF had HAD and that these patients are indicated for convalescent rehabilitation. Further widespread implementation of rehabilitation for older patients with heart failure is expected in Japan.
The incidence of heart failure in Japan is increasing with aging of the population.1 Older patients with heart failure often have physical frailty in addition to various comorbidities.2 Therefore, older patients are more likely to have irreversible physiological changes or declines in physical function at discharge due to bed rest or inactivity during hospitalization.3,4 Such declines in physical function after hospitalization are referred to as “iatrogenic disability”,5 or hospitalization-associated disability (HAD),6 and have recently become a topic of concern. In fact, the incidence of HAD in the elderly aged >65 years who have received acute medical or surgical care is approximately 30% (95% confidence interval [CI] 24–36%),7 highlighting that hospitalization has a significant impact on functional independence in the elderly.
In recent years there have been several studies of HAD in Japanese patients with heart disease. Associations have been reported between HAD and mid-term all-cause mortality in older patients undergoing transcatheter aortic valve implantation,8 as well as between HAD and post-discharge outcomes in older patients undergoing cardiac surgery.9 It has also been reported that HAD is an independent risk factor for heart failure-related rehospitalization and death in older patients with heart failure.10,11 Ogawa et al investigated HAD in Japan using the Japanese Registry of All Cardiac and Vascular Disease Diagnosis Procedure Combination (JROAD-DPC) database and found that the prevalence of HAD was 7.43% and that the risk of HAD increased with lower body mass index.12 However, these reports are not consistent; for example, some are single-center reports, patients did not necessarily undergo rehabilitation during hospitalization, the number of cases included was small, and definitions of HAD differed (standard activities of daily living [ADL] scores were not used).
The Japanese Society of Cardiovascular Physical Therapy conducted a nationwide multicenter registry study, the Japanese PT Multi-center Registry of Older Frail Patients With Heart Failure (J-Proof HF), to investigate the characteristics and prognosis of older patients with heart failure who undergo physical rehabilitation during hospitalization after the onset of heart failure. Using this registry database, the present study aimed to determine the incidence of HAD and its characteristics in older patients with heart failure in Japan.
J-Proof HF is a prospective nationwide multicenter cohort registry of patients with heart failure in Japan. Ninety-six institutions across Japan employing members of the Japanese Society of Cardiovascular Physical Therapy participated in the registry. From December 2020 to March 2022, consecutive patients aged ≥65 years who were prescribed physical rehabilitation by their physician during hospitalization after the onset of heart failure were enrolled in the J-Proof HF Registry. Exclusion criteria were as follows: death during hospitalization, invasive treatment (transcatheter aortic valve implantation, MitraClip, surgery) during hospitalization, and acute coronary syndrome. Patients who were bedridden prior to admission were also excluded.
Basic information about the J-Proof HF Registry, the study description, and the primary and secondary endpoints can be found in the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (ID: UMIN000047893).
Patient data from each institution were stored on password-protected USBs and transported to and from the data center by special mail with verifiable tracking information. At the data center, omissions, input errors, and out-of-range values were checked by a person not involved in the study on a computer with restricted access that was not connected to the Internet, and each institution was asked to correct the data if necessary. The data were then re-entered and double-checked by 2 physical therapists for additional verification.
Research Assessment ItemsPatient demographic data, diagnoses, comorbidities, cardiac function based on echocardiography at admission, New York Heart Association (NYHA) classification at admission, laboratory data at admission (B-type natriuretic peptide [BNP], estimated glomerular filtration rate, albumin, hemoglobin concentration) were recorded. The prehospital Barthel Index and the Kihon checklist of the Ministry of Health, Labour and Welfare,13 were also recorded. Prescribed medications, the Short Physical Performance Battery (SPPB),14 grip strength, upper arm and lower leg circumference, and the Japanese Cardiovascular Health Study Index15 (to assess physical frailty) were assessed at hospital discharge. Cognitive function was assessed at baseline using either the Hasegawa Dementia Scale (HDS-R),16 Mini-Mental State Examination (MMSE),17 Mini-Cognitive Assessment Instrument (Mini-Cog; https://mini-cog.com/download-the-mini-cog-instrument/), or the Japanese version of the Montreal Cognitive Assessment (MoCA-J).18 In this study, cognitive decline was defined as <21 points on the HDS-R, <24 points on the MMSE, <3 points on the Mini-Cog, and <26 points on the MoCA-J.
We then examined rehabilitation content during hospitalization (duration of rehabilitation, number of rehabilitation units [20 min/unit], and physical therapy content) and the discharge destination. The Functional Independence Measure (FIM)19 was also measured at discharge whenever possible.
Inpatient Physical Rehabilitation ProgramAll participating hospitals provided standardized inpatient cardiac rehabilitation for patients with heart failure according to the guidelines of the Japanese Society of Cardiology and the Japanese Association for Cardiac Rehabilitation.20,21 Physical rehabilitation started with individualized low-intensity exercise training in bed after hemodynamic stability was confirmed. Then, patients undertook standing beside their bed, walking around their bed, and walking around the ward. In addition, patients walked or performed endurance training on a stationary bicycle or treadmill, and resistance training was performed on the upper and lower extremities using rubber bands, strength training machines, and body weight. Physical rehabilitation was performed according to the contraindications and precautions for exercise training in patients with heart failure.20
Definition of HADADL were assessed in this study using the Barthel Index,22 which consists of 10 items, including eating, transferring, and toileting, and is scored on a 4-point scale (15, 10, 5, and 0 points) according to ability. In general, a score on the Barthel Index of ≥85 out of a possible 100 points is considered independent, 60 points is considered partially independent, 40 points indicates major assistance is required, and 0 points indicates total assistance is required with ADL. In the present study, with reference to previous studies,23,24 HAD was defined as a decrease of at least 5 points on the Barthel Index at discharge compared with the Barthel Index in a stable condition before hospitalization.
Patient ConfidentialityThe research protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and the Japanese ethics guideline Ethical Guidelines for Life Science and Medical Research Involving Human Subjects (https://www.lifescience.mext.go.jp/bioethics/seimeikagaku_igaku.html [in Japanese]), and was approved by the Ethics Committee of the Juntendo University School of Health Sciences, Tokyo, Japan (19-005). The study was also approved by the ethics committees of all 96 participating institutions (see Supplementary File). Patient consent to participate in the study was obtained by either the opt-out or informed consent method, depending on the judgment of the ethics committees of the participating institutions.
Statistical AnalysisBaseline clinical characteristics, social background, and outcomes were compared between patients exhibiting HAD (a lower score for ADLs than before admission; HAD group) and patients without HAD (non-HAD group). Continuous variables are presented as the median and interquartile range (IQR); categorical variables are presented as numbers and percentages. Categorical variables were compared using χ2 tests when appropriate or Fisher’s exact test otherwise; continuous variables were compared using Student’s t-test or the Wilcoxon rank-sum test depending on distribution. To clarify the clinical characteristics of patients with HAD, patients were divided into 2 subgroups according to the presence or absence of disuse syndrome before admission, and the change in Barthel Index score during hospitalization was determined. Multivariate logistic regression analyses were performed to assess for factors associated with the occurrence of HAD in the group with a preadmission Barthel Index score of ≥85 and in the group with a preadmission Barthel Index score of <85.
From December 2020 to March 2022, there were 10,062 heart failure patients aged ≥65 years with a prescription for cardiac rehabilitation during hospitalization at 96 institutions across Japan. Of these patients, 341 who died in hospital were excluded. In addition, patients were excluded if they had a hospital stay of ≤3 days (n=42) and had missing Barthel Index data at admission and discharge (n=276); this left 9,403 patients for analysis in the present study (see Figure). The median patient age was 83.0 years (IQR 77–88 years), and 50.9% of patients were male. The overall median preadmission Barthel Index and Kihon checklist scores were 100.0 and 11.0 points, respectively, and 56.8% of patients did not have long-term care insurance. The median number of rehabilitation days during hospitalization was 7.0, the median number of total rehabilitation units was 15.0 (where 1 unit=20 min), the median length of hospital stay was 16.0 days, and 85.4% of patients were discharged home.
Subjects flow diagram.
Of the 9,403 patients included in the study, 3,488 (37.1%) had HAD. Table 1 presents patient characteristics in the HAD and non-HAD groups. Compared with the non-HAD group, the HAD group was older and had slightly higher rates of hypertension, chronic kidney disease, musculoskeletal disease, and cerebrovascular disease comorbidity. The HAD group also had a significantly lower Barthel Index score and a significantly higher Kihon checklist score before admission than the non-HAD group.
Characteristics of All Patients and Those With and Without Between HAD Separately
| All patients | Non-HAD | HAD | P value | |
|---|---|---|---|---|
| No. (%) patients | 9,403 | 5,914 (62.9) | 3,489 (37.1) | |
| Age (years) | 83.0 [77.0–88.0] | 81.0 [75.0–87.0] | 85.0 [80.0–90.0] | <0.01 |
| Male sex | 50.9 | 54.0 | 45.6 | <0.01 |
| BMI (kg/m2) | 22.5 [20.1–25.2] | 22.9 [20.4–25.4] | 22.1 [19.6–24.8] | <0.01 |
| Ischemic heart disease | 29.7 | 30.0 | 29.2 | 0.42 |
| HT | 69.1 | 68.2 | 70.4 | 0.03 |
| DM | 34.8 | 35.4 | 34 | 0.18 |
| CKD | 40.3 | 39.1 | 42.4 | <0.01 |
| MSD | 26.7 | 24.8 | 30.0 | <0.01 |
| CVA | 15.6 | 14.3 | 17.9 | <0.01 |
| On admission | ||||
| NYHA functional class | <0.01 | |||
| I | 3.7 | 4.1 | 3.0 | |
| II | 19.3 | 21.4 | 15.8 | |
| III | 40.8 | 42.7 | 37.7 | |
| IV | 36.2 | 31.8 | 43.5 | |
| LVEF (%) | 48.0 [34.0–60.2] | 47.0 [34.0–60.0] | 50.0 [35.0–61.0] | <0.01 |
| Hb (g/dL) | 11.4 [10.0–13.0] | 11.6 [10.1–13.2] | 11.1 [9.8–12.6] | <0.01 |
| eGFR (mL/min/1.73 m2) | 41.0 [28.0–55.8] | 42.3 [29.0–56.2] | 39.0 [26.3–54.0] | <0.01 |
| BNP (pg/mL) | 556.7 [308.6–991.9] | 527.4 [287.2–931.6] | 610.6 [343.6–1,108.0] | <0.01 |
| NT-proBNP (pg/mL) | 4,798.0 [2,236.0–10,311.0] | 4,257.0 [2,071.5–9,247.5] | 6,212.5 [2,677.4–12,734.0] | <0.01 |
| β-blockers | 66.1 | 68.2 | 62.6 | <0.01 |
| ACE inhibitor | 21.1 | 21.8 | 19.9 | 0.03 |
| ARB | 26.4 | 27.6 | 24.5 | <0.01 |
| MRA | 31.4 | 31.0 | 32.0 | 0.31 |
| SGLT2 inhibitor | 16.1 | 17.3 | 14.1 | <0.01 |
| Prehospital Barthel Index score | 88.3±19.4 | 89.0±20.1 | 87.0±18.2 | <0.01 |
| Prehospital Kihon checklist score | 11.00 [7.00–14.00] | 10.00 [6.00–14.00] | 13.00 [9.00–16.00] | <0.01 |
| Grip strength at start of rehabilitation (kg) | 17.8 [12.8–23.8] | 19.0 [14.0–25.3] | 15.2 [11.2–20.3] | <0.01 |
Unless indicated otherwise, data are given as percentages, median [interquartile range], or mean±SD. ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; BMI, body mass index; BNP, B-type natriuretic peptide; CKD, chronic kidney disease; CVA, cerebrovascular disease; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HAD, hospitalization-associated disability; Hb, hemoglobin; HT, hypertension; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonists; MSD, musculoskeletal disease; NT-proBNP, N-terminal pro B-type natriuretic peptide; NYHA, New York Heart Association; SGLT2, sodium-glucose cotransporter 2.
ADL scores and physical function at discharge are presented in Table 2. Compared with the non-HAD group, the HAD group had a significantly lower Barthel Index score and FIM at discharge, a higher percentage of patients with cognitive decline at the start of physical therapy (31.1% vs. 54.3%, respectively), and significantly lower SPPB scores at discharge. In the HAD group, the median length of stay was 18.0 days and the discharge to home rate was 70.2%. In the HAD group, physical therapy was provided for a median of 9.0 days during hospitalization, and the median number of total rehabilitation units was 17.0, with a mean of 1.9 rehabilitation units per day.
Functional Characteristics of All Patients and Those With and Without Between HAD Separately at Discharge
| All patients | Non-HAD | HAD | P value | |
|---|---|---|---|---|
| No. (%) patients | 9,403 | 5,914 (62.9) | 3,489 (37.1) | |
| Barthel Index | ||||
| At discharge | 81.8±24.2 | 90.5±18.7 | 67.1±25.2 | <0.01 |
| At discharge–on admission | −6.4±15.5 | 1.5±5.3 | −19.9±17.6 | <0.01 |
| FIM | 111.0 [90.0–123.0] | 120.0 [106.0–126.0] | 95.0 [73.0–110.0] | <0.01 |
| Cognitive decline (%) | 39.7 | 31.3 | 54.3 | <0.01 |
| Grip strength (kg) | 16.1 [10.3–22.8] | 18.0 [12.0–25.0] | 13.4 [8.0–19.0] | <0.01 |
| Comfortable walking speed (m/s) | 0.74 [0.55–0.94] | 0.82 [0.62–1.00] | 0.60 [0.45–0.77] | <0.01 |
| Maximum upper arm circumference (cm) | 23.0 [21.0–25.8] | 23.5 [21.5–26.0] | 22.5 [20.0–25.0] | <0.01 |
| Maximum right lower leg circumference (cm) | 30.0 [27.3–32.5] | 30.5 [28.0–33.0] | 28.7 [26.1–31.2] | <0.01 |
| SPPB | ||||
| Balance score | 3.00 [2.00–4.00] | 4.00 [2.00–4.00] | 2.00 [1.00–4.00] | <0.01 |
| Walking time (s) | 5.38 [4.22–7.34] | 4.84 [3.97–6.48] | 6.69 [5.19–8.97] | <0.01 |
| 5-times rise time (s) | 12.1 [9.7–16.1] | 11.4 [9.2–14.9] | 14.4 [11.2–19.4] | <0.01 |
| Total score | 7.0 [3.0–11.0] | 9.00 [5.00–11.00] | 5.00 [1.00–8.00] | <0.01 |
| LOS (days) | 16.0 [11.0–23.0] | 15.0 [11.0–21.0] | 18.0 [13.0–27.0] | <0.01 |
| Individual cardiac rehabilitation days | 7.0 [4.0–11.0] | 6.0 [4.0–10.0] | 9.0 [5.0–14.0] | <0.01 |
| Total rehabilitation unitsA | 15.0 [10.0–24.0] | 15.0 [9.0–23.0] | 17.0 [10.0–27.0] | <0.01 |
| Home discharge rate (%) | 85.4 | 95.4 | 70.2 | <0.01 |
Unless indicated otherwise, data are given as percentages, median [interquartile range], or mean±SD. AOne rehabilitation unit is defined as 20 min. FIM, Functional Independence Measure; HAD, hospitalization-associated disability; LOS, length of hospital stay; SPPB, Short Physical Performance Battery.
Of all the 9,403 patients, 2,158 (23.0%) had a preadmission Barthel Index score of <85 points (i.e., they had a preadmission disuse syndrome). In the HAD group, the median change in Barthel Index score during hospitalization in patients with a preadmission Barthel Index score of ≥85 and <85 points was 19.6 points (IQR 96.1–76.5 points) and 20.5 points (IQR 62.7–42.2 points), respectively.
Binomial logistic analysis was performed with the object variable being the occurrence of HAD in the groups with a preadmission Barthel Index score of ≥85 vs. <85 (Table 3). Binomial logistic analysis revealed that age and preadmission Kihon checklist score were associated with the occurrence of HAD in patients with a preadmission Barthel Index score of ≥85, whereas NYHA functional class and preadmission cognitive decline were associated with HAD in those with a preadmission Barthel Index score of <85 points (Table 3). Among HAD patients, the home discharge rate for those with a Barthel Index score of ≥85 and <85 was 75.4% and 57.3%, respectively.
Binomial Logistic Analysis With the Object Variable Being the Occurrence of HAD
| Preadmission Barthel Index score ≥85 | Preadmission Barthel Index score <85 | |||||
|---|---|---|---|---|---|---|
| log(OR) | 95% CI | P value | log(OR) | 95% CI | P value | |
| Age | 0.037 | 0.015, 0.059 | <0.01 | 0.011 | −0.007, 0.030 | 0.2 |
| Sex | −0.048 | −0.432, 0.333 | 0.8 | 0.065 | −0.242, 0.373 | 0.7 |
| BMI | 0.003 | −0.037, 0.043 | 0.9 | 0.000 | −0.031, 0.030 | >0.9 |
| DM | −0.034 | −0.344, 0.272 | 0.8 | 0.204 | −0.053, 0.462 | 0.12 |
| HT | −0.125 | −0.435, 0.191 | 0.4 | 0.097 | −0.161, 0.355 | 0.5 |
| CKD | 0.142 | −0.155, 0.436 | 0.3 | −0.045 | −0.292, 0.201 | 0.7 |
| MSD | −0.239 | −0.610, 0.113 | 0.2 | 0.095 | −0.170, 0.360 | 0.5 |
| CVD | 0.242 | −0.199, 0.654 | 0.3 | 0.156 | −0.164, 0.478 | 0.3 |
| NYHA functional class | ||||||
| 1 | – | – | – | – | ||
| 2 | −0.706 | −1.43, 0.084 | 0.07 | 0.300 | −0.434, 1.07 | 0.4 |
| 3 | −0.636 | −1.31, 0.114 | 0.08 | 0.432 | −0.273, 1.18 | 0.2 |
| 4 | −0.322 | −0.997, 0.430 | 0.4 | 0.841 | 0.137, 1.58 | 0.02 |
| LVEF | −0.004 | −0.013, 0.006 | 0.5 | −0.003 | −0.011, 0.005 | 0.5 |
| Hb | −0.034 | −0.104, 0.036 | 0.3 | 0.038 | −0.021, 0.096 | 0.2 |
| Prehospital Kihon checklist | 0.034 | 0.002, 0.066 | 0.03 | 0.015 | −0.012, 0.043 | 0.3 |
| Grip strength when rehabilitation starts | −0.024 | −0.052, 0.004 | 0.09 | −0.004 | −0.029, 0.021 | 0.7 |
| Cognitive function decline | −0.009 | −0.367, 0.334 | >0.9 | 0.417 | 0.178, 0.656 | <0.01 |
CI, confidence interval; CVD, cerebrovascular disease; OR, odds ratio. Other abbreviations as in Table 1.
Of the 9,403 patients in the study, 4,870 (51.8%) had a Barthel Index score of <85 at the time of discharge. FIM was also assessed at discharge in 5,802 patients; 55.5% of these patients had a discharge FIM score of <115.
The aim of this study was to determine the incidence and characteristics of HAD in older patients with heart failure in Japan. In summary, of the 9,403 patients included in the study, 3,488 (37.1%) had HAD. In addition, 51.8% patients had a Barthel Index score of <85 at the time of discharge. Compared with the non-HAD group, the HAD group was older and had more comorbidities. The HAD group also had significantly lower preadmission Barthel Index scores and significantly higher Kihon Checklist scores. The HAD group also had significantly lower discharge Barthel Index, FIM, and SPPB scores than the non-HAD group. Furthermore, a higher percentage of patients in the HAD group had cognitive decline at the start of rehabilitation. In the group with a preadmission Barthel Index score of ≥85, factors associated with HAD were older age and frailty; however, in the group with a preadmission Barthel Index score of <85, severity of heart disease and cognitive decline were associated with HAD (i.e., this group had disuse syndrome before admission).
Representative epidemiologic studies of heart failure in Japan, in chronological order of implementation, include the Heart Institute of Japan-Department of Cardiology Heart Failure (HIJC-HF) study,25 the Japanese Cardiac Registry of Heart Failure in Cardiology (JCARE-CARD) study,26 and the Acute Decompensated Heart Failure Syndromes (ATTEND) registry.27 These Japanese registries enrolled all patients hospitalized for heart failure. The J-Proof HF Registry is also a Japanese registry study, but only enrolled consecutive patients aged ≥65 years who were prescribed physical rehabilitation by their physician during hospitalization after the onset of heart failure; therefore, the patients included in the present study were older than those in previous Japanese registry studies. The mean (±SD) age of patients in REHAB-HF,28 a multicenter randomized controlled trial conducted in the US to evaluate a transitional, tailored, progressive rehabilitation intervention, was 72.7±8.1 years. The median age of patients in the J-Proof HF Registry is 83 years, 10 years older than the mean age of patients in the REHAB-HF registry, making the J-Proof HF Registry the oldest large cardiac rehabilitation registry in the world. The SURUGA-CARE study,29 a multicenter prospective cohort study conducted by Kato et al in collaboration with 5 hospitals and one university in Shizuoka Prefecture, Japan, had a median age of 82–84 years, which is comparable to the median age of patients in the J-Proof HF Registry. This may be because both studies included older patients hospitalized and rehabilitated for acute heart failure.
In addition, the aforementioned HIJC-HF, JCARE-CARD, and ATTEND registries have a female proportion of approximately 40%. Due to the older age of the cohort in the present study, the percentage of female patients is approximately 10% higher than in these other registries. Hypertension was the most common comorbidity in the J-Proof HF cohort (50–70%), similar to other Japanese registries.
In the present study, HAD was found in 37.1% of patients. In recent years, relevant studies of HAD in Japanese patients with heart disease have reported HAD incidence rates ranging from 7.4% to 25.0%.8–12 This difference is due primarily to differences in the regions and number of institutions covered, differences in whether rehabilitation is provided during hospitalization or not, and differences in the definition of HAD. Takabayashi et al reported that in their study of 923 patients with heart failure (mean age 75.7 years), 10.6% developed HAD.10 Takabayashi et al classified ADLs into 4 categories (independent outdoor walking; independent indoor walking; indoor walking with assistance; or abasia both before admission and at discharge) and defined HAD as the change in walking ability before and after hospitalization. No mention was made as to whether rehabilitation was performed during hospitalization.10 Saitoh et al reported that 25.0% of 1,941 heart failure patients (mean age 81 years) developed HAD,11 and Kato et al reported that 21.0% of 411 heart failure patients (mean age 82.7 years) developed HAD.29 Although these studies are similar to the J-Proof HF Registry in terms of the mean age of subjects, the incidence of HAD is slightly lower. This may be because the reports are from limited areas of Tokyo and Shizuoka. Ogawa et al investigated hospital-acquired disability using the Japanese Registry of All Cardiac and Vascular Disease Diagnosis Procedure Combination (JROAD-DPC) database and found that HAD occurred in 7.4% of 238,160 patients with heart failure (mean age 81.0 years).12 Although the study by Ogawa et al had the largest number of cases, it included a large number of people who did not receive rehabilitation, and ADLs were assessed after rather than before admission. This may explain why the incidence of HAD was 7.4%, which is significantly lower than that in the J-Proof HF Registry.
The target population of the J-Proof-HF Registry reflects the current status of heart failure rehabilitation in Japan, a superaging society, with a very large number of patients (37.1%) requiring ongoing cardiac rehabilitation.
Among patients with HAD, when grouped according to the presence or absence of disuse syndrome (Barthel Index score <85 points) prior to admission, the magnitude of the decline in the Barthel Index during hospitalization was similar, approximately 20 points. The Barthel Index is a 100-point scale, with scores of ≥85, ≥60, and <40 indicating independent ADL, partial independence with ADL, and total assistance required with ADLs, respectively. Patients who did not have a disuse syndrome prior to hospitalization, even if HAD occurred, had a mean score of 76.5 at discharge, a decline to the level of partial independence. Conversely, when patients who had a disuse syndrome prior to admission developed HAD, their Barthel Index score fell from a mean of 62.7 points prior to admission (indicating partial independence) to a mean of 42.2 points at discharge (indicating they required assistance with ADL). The meaning of HAD differed greatly depending on the presence or absence of a preadmission disuse syndrome.
In the multivariate analysis, factors associated with HAD differed according to whether the patient had disuse syndrome prior to hospitalization. The group that did not have disuse syndrome prior to hospitalization was more likely to have HAD because of older frailty (Kihon checklist score of ≥8), whereas the group that had disuse syndrome prior to hospitalization was more likely to have HAD because of higher disease severity and cognitive decline. That is, the factors that should be considered for HAD prevention differ depending on the presence or absence of disuse syndrome prior to hospitalization.
Among the patients included in this study, a high percentage (55.5%) had an FIM score of ≤115 at discharge, and 39.7% had cognitive decline. This means that the heart failure patients in the present study had ADL impairments similar to stroke patients and that prevention and improvement of functional decline through rehabilitation is essential. In an epidemiologic study of stroke, it was reported that rehabilitation started on average 3.6 days after hospitalization, with a mean of 2.7 rehabilitation units per day.30 However, in the heart failure patients who developed HAD in the present study, the number of rehabilitation days was approximately half the length of the hospital stay, with a mean of 1.9 rehabilitation units per day. It is possible that the amount of rehabilitation performed during the acute phase was inadequate, leading to HAD.
Kamiya et al reported that of 51,323 patients admitted for heart failure at 288 Japanese facilities, 60% did not receive inpatient or outpatient cardiac rehabilitation, 33% received inpatient cardiac rehabilitation only, and only 7% of all patients able to continue outpatient cardiac rehabilitation after inpatient cardiac rehabilitation.31 This shows the difficulty of continuing cardiac rehabilitation after discharge from hospital in Japan. In addition, patients may not be eligible for long-term care insurance and may not receive home rehabilitation after discharge from hospital. Transfer to a convalescent rehabilitation hospital is considered necessary, but current problems in the Japanese healthcare system make transfer to a convalescent rehabilitation hospital difficult.
Rehabilitation for heart failure was covered by public insurance in 2006. In 2022, “acute myocardial infarction, angina pectoris, and postoperative heart surgery” were added as conditions for patients to receive convalescent rehabilitation, and inpatient convalescent cardiac rehabilitation began in earnest in Japan. Conversely, as of 2023, heart failure will not be recognized as a patient condition requiring convalescent rehabilitation. The convalescent rehabilitation unit is an intensive rehabilitation unit whose main goal is to prevent patients from becoming bedridden and to return them to their homes by improving their ADL. Although ADL is an independent prognostic factor for older patients with heart failure32 and is considered an indication for hospitalization, the actual number and characteristics of patients who are rehabilitated in acute care hospitals and subsequently require further improvement in ADL have not been determined in Japan. In the present study, 3,488 (37.1%) of 9,403 patients were found to have HAD. In addition, the criteria for admission to a convalescent rehabilitation unit include patients who are eligible for rehabilitation for disuse syndrome (i.e., “a condition in which a certain degree of decline in basic functional ability, applied functional ability, speech and hearing, and [ADL] has occurred due to disuse syndrome caused by bed rest associated with acute illness, with or without treatment”; https://www.mhlw.go.jp/topics/2008/03/dl/tp0305-1d_0014.pdf [in Japanese]). The term “a condition in which a certain degree of decline in basic functional ability, applied functional ability, speech and hearing, and [ADL] has occurred” indicates conditions such as those with an FIM score of ≤115 and a Barthel Index score of <85 at the start of rehabilitation. Of the 9,403 patients in this study, 4,870 (51.8%) had a Barthel Index score of <85 at the time of discharge. FIM was also assessed at discharge in 5,802 patients, 55.5% of whom had a discharge FIM score of <115. That is, more than half the older patients with heart failure who were hospitalized for acute heart failure and underwent cardiac rehabilitation during their hospitalization met the criteria for admission to a convalescent rehabilitation unit (i.e., they were patients who should continue rehabilitation after discharge).
Japan is a super aging society, and the population of patients with heart failure is also aging. Seamless rehabilitation is desirable based on the recognition that older patients with heart failure are individuals with multiple diseases, overlapping disabilities, and ADL impairments.
Study LimitationsThis study has several limitations. First, although J-Proof HF was conducted with the cooperation of 96 institutions in 37 prefectures throughout Japan, it did not cover all regions of Japan or all hospitals offering cardiac rehabilitation.
Second, this study included only patients who were prescribed rehabilitation after hospitalization for acute heart failure. Because we did not investigate detailed treatment in the emergency department or coronary care unit, we could not determine the relationship between acute treatment and HAD.
Third, the content of cardiac rehabilitation includes basic activity exercises, walking training, strength training, aerobic exercise using a bicycle ergometer and treadmill, and resistance training using weight machines as physical therapy during hospitalization. Usually the content of these exercise changes during hospitalization, depending on the patient’s recovery status and medical condition. Therefore, it was difficult to analyze changes in the content of highly individualized exercises for each participating facility in the study. In the present study, only the number of individual cardiac rehabilitation days and total rehabilitation units (20 min/unit) are presented as the total amount of rehabilitation performed.
Finally, the data collection period for this study was during the COVID-19 pandemic. The relationship between the COVID-19 pandemic and HAD is unknown because there are no large registry data for the control, non-pandemic period. Therefore, we could not determine whether the COVID-19 pandemic affected HAD.
HAD occurred in 37.1% of patients, and 55.5% of patients had an FIM score <115 points at discharge. Factors associated with HAD varied according to preadmission ADL, and the magnitude of the decline also varied. Factors associated with admission-related decline included age and the Kihon checklist score for patients with a preadmission Barthel Index score ≥85 points, and heart failure severity and cognitive decline at the start of physical therapy for those with a preadmission Barthel Index score <85 points.
The authors are grateful to Kaori Taniguchi (Faculty of Health Science, Juntendo University) and Yuji Kanejima (Department of Rehabilitation, Kobe City Medical Center General Hospital) for their skillful technical assistance. During the preparation of this manuscript, the authors used DeepL to verify their written English.
This work was supported by research funding of Japanese Society of Cardiovascular Physical Therapy.
The authors declare that there are no conflicts of interest.
This study was approved by the Ethics Committee of the Juntendo University School of Health Sciences, Tokyo, Japan (Reference no. 19-005).
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-23-0722
