Incidence and Healthcare Resource Utilization Among Patients with Hypertrophic Cardiomyopathy Hospitalized for Heart Failure in Japan

This retrospective observational study used data from the Medical Data Vision (MDV) database in Japan, from January 01, 2011, to December 31, 2023. This study investigated the incidence rate of HF events in patients with HCM and evaluated treatment patterns and outcomes during hospitalization in patients with HCM who were either hospitalized or non-hospitalized for HF.

The MDV database is based on health claims data and administrative data from Japanese hospitals participating in the diagnostic procedure combination (DPC)/per-diem payment system (PDPS), and the database includes data from over 38.7 million individuals across 461 DPC hospitals, covering approximately 26% of all DPC hospitals in Japan as of January 2022 [19]. In Japan, a DPC hospital is a major healthcare facility that primarily focuses on treating acute-phase diseases and operates under the DPC system. This system combines elements of both a fee-for-service model and a per diem payment structure, aiming to enhance efficiency and transparency in larger hospitals' management. The database includes both inpatient and outpatient data and reports information such as patient demographics (age, sex), diagnoses, prescriptions, medical procedures, tests performed, departments visited, and medical costs (calculated only based on the fee-for-service payment system). The data recorded in the MDV database include International Classification of Diseases (ICD)-10 codes for diagnoses, disease names coded using Japanese-specific disease codes, and procedures and prescriptions coded using Japanese-specific receipt codes. Procedures and prescriptions are recorded with exact dates, and whether the patient was inpatient or outpatient is noted. For diagnoses, only the month and year are provided, but variables indicating whether a diagnosis was confirmed and whether it was the reason behind a DPC hospitalization are included. In-hospital mortality data are available in the electronic medical records (EMR) file, which contains information on DPC hospitalizations. The MDV database is not a closed system; consequently, visits incurred outside of the MDV network are not captured in the EMR data and may not have been captured in the claims data.

The cohort entry date was defined as the date of the first HCM diagnosis using all available data during the study period. Patients were followed up for the outcomes of interest starting from the cohort entry date and censored at the following events (whichever occurred first): death of the patient, the last record in the MDV database, or the end of the study period. The date of first hospitalization due to HF after the cohort entry date was defined as the index date and the incidence was evaluated. Patients with an eligible HCM diagnosis were required to have a lookback period of > 12 months before the cohort entry date to identify the first HCM diagnosis and to exclude those with past hospitalization due to HF (Fig. 1). A nested case–control design was applied to compare treatment patterns and patient characteristics prior to the event in patients hospitalized due to HF and those without hospitalization due to HF.

Bild vergrößern

Patients diagnosed with HCM in Japan according to the ICD-10 diagnosis codes for HCM (ICD-10-CM: I42.1 and 142.2) were included in the study. The inclusion criteria were as follows: (1) patients diagnosed with HCM during the study period (between July 01, 2011, and December 31, 2023) and (2) patients with a lookback period of 12 months without an HCM diagnosis prior to the date of first HCM diagnosis (Supplementary Material Table S1). The exclusion criteria were as follows: (1) Hospitalization due to HF during the lookback period; (2) diagnosis of end-stage renal disease (ESRD) at any time during the lookback period (N18.5); (3) diagnosis of HCM for the first time in the database during the same month as hospitalization due to HF; and (4) diagnosis with a dilated phase of HCM (D-HCM) at the entry date or during the lookback period.

The primary objective was to calculate the event rate of hospitalization due to HF in patients with HCM during the follow-up period. Definition of hospitalization due to HF is shown in Supplementary Material Table S2. Secondary objectives were to identify patient characteristics and medication treatment patterns in patients with HCM hospitalized and non-hospitalized with HF in Japan before the index date. Patient characteristics included comorbidities such as diabetes mellitus, hypertension, dyslipidemia, AF, myocardial infarction (MI), valvular heart disease, peripheral artery disease (PAD), ischemic stroke, malignant tumor, chronic kidney disease, and D-HCM. Medication classes included beta-blockers, non-dihydropyridazine (DHP) calcium (Ca) channel blockers, DHP Ca channel blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), angiotensin receptor-neprilysin inhibitors (ARNIs), thiazides, vasopressin V2 receptor antagonist, mineralocorticoid receptor antagonists (MCRA), digitalis, and sodium channel blockers (SCBs). Secondary objectives also evaluated HCRU associated with hospitalization due to HF, route after discharge, and outcome at discharge (Supplementary Material Table S2).

This study used de-identified claims/EMR data and, therefore, did not require ethics committee review. In addition, this study did not require informed consent to be obtained from patients. This study was conducted in accordance with the International Society for Pharmacoepidemiology Guidelines for Good Pharmacoepidemiology Practices and applicable regulatory requirements [20]. This article is based on information from an existing database and does not contain any new studies with human participants or animals performed by any of the authors [19]. Since the database is available in the public domain (https://en.mdv.co.jp/ebm/about-mdv-database/mdv-database-overview/), no permission was obtained to access and use the data from the database utilized in this study.

Continuous variables were described as the mean with 95% confidence intervals (CIs), standard deviation (SD), median, first and third quartiles, minimum, and maximum. Categorical variables were described as the number and percentage of patients in each category with 95% CI. The number of patients with missing data for each variable was reported. Only available data were summarized; no imputation methods were used for missing data for any endpoints. Person-years were calculated from the cohort entry date to the first of the following: death of the patient, the last record in the MDV database, end of the study period, or the first HF hospitalization. The incidence rate was calculated by dividing the number of cases of HF hospitalization by the total person-years and presented per 100,000 person-years. Time to event of first hospitalization with HF was analyzed using the Kaplan–Meier method for the overall population, and median time along with interquartile range was reported. For secondary outcomes, a nested case–control design was applied. A risk set of possible controls for each case (defined as patients with a first hospitalization due to HF during the follow-up period) was established, with at least an equal duration of observation periods at the case index date and remaining free of hospitalization due to HF until the index date of the respective case. Control patients were matched on age, sex, and calendar year based on the cohort entry date. For each case, four control patients were identified, with replacement. Crude and adjusted conditional logistic regression analyses were performed to estimate odds ratios (ORs) and respective 95% CIs. Variables for adjustment are described in Supplementary Material Table S2. Patient characteristics, HCRU, and hospitalization outcomes were calculated and reported using mean, SD, and median and interquartile range for continuous variables and frequencies and proportions for categorical variables. Analyses were performed using R version 4.2 or later (R Foundation for Statistical Computing, Vienna, Austria).

Overall, 40,316 patients with HCM were identified in the MDV database between January 01, 2011, and December 31, 2023. Of these, 13,475 patients had a 12-month lookback period prior to the first HCM diagnosis, and 13,136 had no hospitalization due to HF during the 12-month lookback period. Of these patients, 12,145 (30.12%) were included in the present study (Fig. 2; Table 1). A total of 525 (4.32%) patients experienced first hospitalization due to HF during the observational period. The mean (SD) age of the overall study population at the cohort entry date was 71.4 years (± 14.0), and 45.8% (n = 5566) were female patients. Left ventricular ejection fraction at the time of admission due to HF is presented in Supplementary Material Table S3. HF with preserved ejection fraction (HFpEF) constitutes the largest proportion (59.4%) among patients with available data.

Bild vergrößern

Unspecified HCM was most commonly reported (n = 7822; 64.4%), followed by apical HCM (17.0%), obstructive HCM (14.3%), nonobstructive HCM (3.7%), and mid-ventricular obstructive HCM (0.4%) in patients diagnosed with HCM at the index date (Table 1). The most common cardiovascular comorbidity during the lookback period was HF (17.0%), followed by arrhythmia (8.0%), MI (6.0%), valvular heart disease (5.5%), AF (5.4%), ischemic stroke (4.5%), and PAD (3.2%; Table 1). The most common non-cardiovascular comorbidity was diabetes mellitus (19.2%), followed by malignant tumors (17.9%). Patients were diagnosed with HCM at cohort entry date; therefore, medications for HCM were rarely prescribed (Ca channel blockers [6.5%], beta-blockers [6.0%], ACE inhibitors/ARBs [6.0%], diuretics [2.6%], SCBs [0.5%], and sodium-glucose cotransporter-2 inhibitor [SGLT2i; 0.2%].

The strengths of the present study were, firstly, that the MDV database covers a broad range of patients and hospitals in Japan; secondly, the study population included a large sample of patients diagnosed with HCM who were hospitalized with HF and treated in a real-world clinical setting in Japan; and finally, the MDV database contains valuable information about patients during hospitalization for HF, including diagnoses, administered medications and prescriptions, medical procedures, and outcomes of the hospitalization.

There are also some limitations to this study. First, the MDV is not a closed system; thus, while information is expected to be complete for the HF hospitalization, services received outside the MDV hospitals have not been captured, such as some hospitalizations post-discharge and outpatient visits. Second, given that the 12-month lookback period was established before HCM diagnosis, patients had been visiting DPC hospitals for the treatment or observation of other diseases. Therefore, patients might have had more comorbidities than the general population. Additionally, because the majority of patients in this study had received a diagnosis of HCM of an unspecified subtype, we could not fully assess differences among HCM subtypes. Third, as mentioned earlier, NYHA classification, echocardiographic data, and laboratory testing data were not available; therefore, important clinical factors associated with cardiac function and HF events were not included in the analysis. Additionally, the absence of data on the number of patients on implantable cardioverter-defibrillator (ICD) therapy and the number of ICD shocks is a limitation of our study. Fourth, only in-hospital deaths are captured in the MDV; thus, post-discharge mortality could not be analyzed. Fifth, due to the limited availability of laboratory data in the database, including biomarkers for HF such as BNP/N-terminal pro B-type natriuretic peptide (NT-proBNP), we were unable to assess their potential as prognostic indicators or treatment targets, nor analyze their relationship with patient outcomes. Given this lack of laboratory data, it is difficult to evaluate the impact on the length of stay and the costs of treatments. Sixth, we acknowledge that understanding the interplay between HF phenotypes or severity and other comorbidities is critical for approaching the prognosis of patients with HCM more comprehensively. Lastly, the retrospective study design is a limitation that may affect the findings. Such analyses would require prospective studies designed to collect robust datasets capturing information on HF subtypes, severity, and relevant comorbid conditions. These factors should be examined in future investigations to enhance the understanding of prognostic determinants in patients with HCM suffering from HF. Results of the analyses using the data from HCM registries in Japan are eagerly anticipated in this regard [14, 37‐39].