Heart failure accounts for a great deal of the morbidity and mortality in the population, and the estimated 5-year age-adjusted mortality rate after a diagnosis of heart failure is 59% for men and 45% for women [2].

Evidence-based treatments that have been shown to decrease mortality rates [3,4] include angiotensin-2 receptor blockers (ARBs), angiotensin-converting enzyme inhibitors (ACE-I), beta-blockers, and aldosterone antagonists. However, several studies have shown that these treatments are underutilized and are often prescribed at lower dose levels [5-10]. Factors related to the underutilization of evidence-based treatments include a lack of initiation or early discontinuation. The causes of this lack of initiation include the contraindication of drug use, lack of knowledge, lack of expertise for using these drugs, lack of time, and economic restraints [9]. However, in many cases, the reasons for underutilizing evidence-based therapy are unclear. Some evidence has been collected from several randomized clinical trials (RCTs), but most RCTs have limited generalizability for treatment options because these trials typically have strict inclusion and exclusion criteria. Euro Heart surveys found that among 10,701 heart failure patients, only 1,346 were eligible to participate in the majority of RCTs [11]. However, these findings do not explain the underutilization of evidence-based treatment in eligible individuals, and elderly CHF patients could significantly benefit from evidence-based treatment [12,13].

The causes of treatment underutilization may include the fear of polypharmacy, inaccurate perceptions concerning adverse effects, and contraindications. However, the mortality rate of patients with renal insufficiency was lower for those who received ACE-I, beta-blockers, statins, and aspirin [14]. Beta-blockers can be prescribed safely to patients with diabetes [15-17] and chronic obstructive pulmonary disease [18-21], and this treatment significantly decreases the morbidity and mortality caused by heart failure. Also, the utilization of evidence-based treatment for patients with heart failure can significantly reduce heart failure-related hospital admission and mortality [12,22,23].

The evidence-based treatment group was more likely to be admitted to tertiary hospitals, and their healthcare providers were more specialized than those of patients in the non-use group. Although we did not find the subspecialty of internal medicine in our data set, we hypothesized that most internal medicine care providers would be cardiologists. The evidence-based treatment group was more likely to have coexisting cardiovascular diseases, such as angina, myocardial infarction, atrial fibrillation, and valvular heart disease, and comorbidities, such as diabetes and hyperlipidemia.

The non-use group was more likely to have been admitted to a primary or secondary hospital and treated by less specialized healthcare providers and was also more likely to have dementia and living in rural area. And rate of prescriptions from outpatient setting in non use group was 54%, which was lower than that of evidence based treatment group, where rate was 76%. Therefore, it could be inferred that the non-use group was more likely to be institutionalized. Shibata et al [10]., studied the utilization pattern of evidence-based treatment for heart failure in institutionalized elderly patients and found prescription rates to be low; the frequency of ACEi and ARB treatment was 51%, and the frequency of beta-blocker treatment was 16%.

When considering the associated factors for evidence-based treatment in elderly heart failure patients, the specialty of the treating healthcare providers and the type of hospital were important factors. This may be related to medication costs and reimbursements. If a patient is institutionalized due to dementia, the percentage of the fee that the national healthcare insurance system covers is limited, and the hospital should thus attempt to reduce medication and examination costs. Another factor related to the utilization of evidence-based treatment may be the healthcare providers. Beta-blockers have been traditionally misunderstood as aggravating chronic obstructive pulmonary disease or poor glycemic control in diabetes. However, several studies [26,27] have suggested that patients with chronic obstructive pulmonary disease tolerate selective beta-blockers well, although these medications should be administered at the lowest possible dose and require close monitoring. Carvedilol, the most frequently used beta-blocker for patients with heart failure, can also be safely prescribed to diabetic patients [28] because it has neutral effects on blood glucose levels. Many studies have shown that evidence-based beta-blockers, such as carvedilol or nebivolol [29], should be prescribed for diabetic heart failure patients because these treatments have significant health benefits for all causes of mortality. Our data suggest that many physicians appear reluctant to prescribe beta-blockers for patients with chronic lung disease or diabetes.

Many researchers from other countries have studied the utilization patterns of evidence-based treatment in heart failure. Research using the Medicare database [12] for elderly populations with heart failure found that the prescription rate of those prescribed only ACEi or ARB treatment was 27.9%. The rate of beta-blocker prescription was 15.7%, and patients were prescribed both treatments at a rate of 28.4%. Data from the Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF) [30] indicated that the prescription rate for ACEi or ARBs was 79.3%, and that for beta-blockers was 85.8%. This suggested that the utilization patterns of evidence-based treatment are dependent on the database that is used for the analysis. For comparability with our data, we searched articles using claims databases.

Gislason et al [31]., studied the persistent use of evidence-based pharmacotherapy in heart failure using the Danish National Patient Registry claims database. Prescription was defined as treatment initiation after discharge from a hospital admission due to heart failure, which was similar to our study design. The prescription rate for ACEi or ARB treatment was 43%. Beta-blockers were prescribed at a rate of 27%, while spironolactone was prescribed at a rate of 19%.

Another study was published using a population-based cohort (1999–2001) of 9,942 patients with heart failure who had been hospitalized in Ontario, Canada [32]. These researchers investigated the prescription rates of evidence-based treatment for patients admitted to the hospital due to heart failure after their discharge. The prescription rate of ACEi or ARB treatment was approximately 77%, while that of beta-blockers was 33%; however, the authors stratified prescription rates with respect to the validated heart failure risk scores. One study [9] used the Euro Heart Survey on Heart Failure, which included 46,788 heart failure patients from 115 hospitals in 24 ESC member countries. These data showed that the prescription rate for ACEi or ARB treatment was 62%, and the beta-blocker prescription rate was 37%.

The prescription rate of ACEi or ARB treatment was 54.7% for our total study population and 67.3% when the population using diuretics and digoxin was analyzed, as shown in Figure​Figure2.2. Similarly, the prescription rate of beta-blockers was 31.5% for the total study population and 35.4% in the population using diuretics and digoxin. Some of the underutilization in beta-blockers could be due to recent hospitalization of heart failure patients, because stabilization of patient’s status should be prior to initiating beta blockers.

Our data suggested that the non-use group may consist of institutionalized patients with dementia and that healthcare system factors, such as medication cost and reimbursement, healthcare provider factors, such as specialty or knowledge, and patient factors, such as comorbidities, could contribute to the underutilization of evidence-based pharmacologic treatment in heart failure.

We used the KHIRA claims database to ensure high generalizability and eliminate recall and selection bias. Our study results reflected the actual heart failure population, and because we used the National Health Insurance System, we were able to obtain detailed information concerning the medications that were used to treat diseases coded by the International Classification Codes. Thus, we were able to estimate the non-use group.

However, our study had several limitations. First, the diagnosis of heart failure in the KHIRA database was not validated, although a validity study that compared diagnoses by ICD-10 codes to clinical information obtained using the medical records showed a positive predictive value of approximately 70% [33].

Second, our study was limited because data concerning the left ventricular ejection fraction were not available. Therefore, our population had potentially considerably more diverse clinical characteristics, although we attempted to adjust for this limitation by including many comorbidities and medications. We also analyzed subgroups according to digoxin and diuretic use, which represented symptom-relieving heart failure treatment, and the results did not differ.