Disparities in Medication Therapy in Patients with Heart Failure across the State of Hawai‘i

Abstract

The purpose of this study is to evaluate if heart failure patients in Hawai‘i are receiving recommended standard therapy of a select beta-blocker in combination with an ACE inhibitor (ACEI) or angiotensin receptor blocker (ARB), and to determine if a gap in quality of care exists between the different regions within the state. A retrospective claims-based analysis of all adult patients (age > 18 years of age) with CHF who were enrolled in a large health plan in Hawai‘i was performed (n = 24,149). Data collected included the presence of pharmaceutical claims for ACEI, ARBs and select β-blockers, region of residence, gender, and age. Multivariable logistic regression was used to examine whether there were regional differences in Hawai‘i related to medication usage, after adjustment for age and gender. Results showed that only 28.4 % of patients were placed on the recommended therapy of an ACEI or ARB and a select β-blocker with significant differences being found between different regions. Further research is needed to better understand factors affecting regional differences in prescribing patterns.

Introduction

Despite recent advances in medical care, congestive heart failure (CHF) continues to be a major contributor to patient mortality and cost of health care with nearly $29 billion/year being spent in the United States on heart failure related hospitalizations.^1,2 Multiple studies have shown decreased morbidity and mortality when patients with impaired left ventricular ejection fraction (LVEF), commonly referred to as systolic heart failure, received a combination of a select β-adrenergic receptor blockers (β-blocker)with either an angiotensin converting enzyme inhibitor (ACEI) or an angiotensin receptor blocker (ARB).^3–7 Current guidelines from both the American College of Cardiology/American Heart Association (ACCF/AHA) and the Heart Failure Society of America (HFSA), recommend that patients with impaired LVEF be prescribed an ACEI or an ARB along with a β-blocker, specifically metoprolol succinate (long-acting).^8–9 In patients with isolated diastolic heart failure (preserved LVEF), which accounts for approximately 1/3 of patients diagnosed with heart failure it is also recommended that patients be placed on an ACEI or an ARB and a β-blocker, although the particular β-blocking agent is not specified.⁸ Both ACEI and ARBs exert their effects through modulation of the renin-angiotensin system that has shown to reduce cardiac remodeling.¹⁰ Long term therapy on ACEI and ARBs with β-blockers has also been shown to decrease symptoms, improve clinical status, and enhance the patient's overall feeling of well-being in addition to reducing the risk of hospitalization and/or death.

Several studies have demonstrated the impact of ethnic and socioeconomic factors on the prevalence and mortality associated with CHF.¹¹ An epidemiological study conducted by the Hawai'i State Department of Health found definite geographic and ethnic disparities in cardiovascular disease mortality in Hawai‘i, with Pacific Islanders having a greater overall risk of developing heart failure compared to other ethnic groups.^1,12 Data on native Hawaiians demonstrate a 68% higher incidence of heart disease compared to the national average.¹³ Reviews of epidemiological data have also demonstrated that demographic differences such as age, gender, ethnicity, and socioeconomic environment may influence primary medication adherence which may be a major contributing factor to the increased risk of mortality, hospitalizations, and increased health care costs observed in these patient groups. ^15–17

The goal of this study was to examine regional prescribing patterns for CHF patients in Hawai‘i. Findings from this study can lay the foundation for targeted interventions to improve the quality of care for patients with heart failure in Hawai‘i.

Methods

A retrospective analysis of administrative data from January 1 to December 31, 2010 for adult patients (age > 18 years of age) who were enrolled in a large health plan in Hawai‘i was performed and all patients with a medical diagnosis of CHF and enrollment with medical and drug coverage were included in our analysis (N = 24,149). CHF was identified by the health plan based on an ICD-9 code of 428. Pharmaceutical claims data for calendar year 2010 for ACEIs, ARBs, and select β-blockers (metoprolol succinate, bisoprolol, and carvedilol) were obtained from the health plan; the short acting form of metoprolol (metoprolol tartrate) has not shown decreased mortality and is therefore not recommended and was not included in our analysis.

Information on patient age, gender, and region of residence were obtained from the health plan and linked using a unique patient identifier to claims data. Patients were stratified based on billing zip codes into six geographic regions: Hawai‘i- East; Hawai‘i- West; Kaua‘i; Maui County (includes Lana‘i and Moloka‘i islands); O‘ahu- Honolulu Metropolitan Statistical Area (MSA), and O‘ahu-Other than Honolulu MSA. Patients with out of state billing zip codes (n = 621) were excluded. Within each demographic region, patients were evaluated based on the presence of pharmaceutical claim(s) for (1) an ACEI or an ARB (regardless of β-blocker fill), (2) one of the selected β-blockers (regardless of ACEI or ARB fill), (3) both a β-blocker and an ACEI or an ARB and (4) pharmaceutical claim(s) for neither ACEI, ARB, nor β-blockers.

Pearson's chi-squared tests were used to examine differences by region, while analysis of variance was used to examine regional differences in age. Multivariable logistic regression was used to examine whether there were regional differences related to medication usage, after adjustment for age and gender. Age was categorized as: (1) 18–44 years; (2) 45–65 years; (3) 65–84 years; (4) 85+ years. Odds ratios and their 95% CIs for the association of age, gender, and region with recommended medication use were calculated. This research study was deemed exempt by the University of Hawai‘i Committee on Human Subjects. All analyses were conducted using Stata 11.0 (StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP).

Results

Age and gender differed significantly by region (Table 1). Mean patient age ranged from a low of 68.7 (SD 14.0) years in Maui County compared to a high age of 75.9 (SD 13.8) years in O‘ahu-Honolulu MSA. Proportion of patients who were men was highest in Maui County at 62.7 % and lowest in Hawai‘i- East (50.5%).

Table 1.

Patient Age and Gender by Region of Hawai‘i

Region	Age [Years; Mean (SD)]	Female	Male
Overall (N = 24,138)	73.0 (14.3)	44.4%	55.6%
Hawai‘i- East (n = 3151)	75.2 (13.6)	49.5%	50.5%
Hawai‘i- West (n = 1581)	70.2 (13.2)	44.9%	55.1%
Kaua‘i (n = 1339)	72.8 (13.9)	39.5%	60.5%
Maui County (n = 2143)	68.7 (14.0)	37.3%	62.7%
O‘ahu- Honolulu (n = 7220)	75.9 (13.8)	45.2%	54.8%
O‘ahu- Other (n = 8704)	71.4 (14.7)	44.4%	55.6%
P-value	P < .001a	P < .001b

^aAnalysis of variance was used to examine differences in age across regions.

^bPearson's chi-squared test was used to examine differences in gender across regions.

Table 2 describes the unadjusted compliance rates with recommended medication therapy for patients with CHF, of a β-blocker and an ACEI or ARB. Overall results revealed that 28.2% of patients with heart failure were on the appropriate dual drug therapy of both a β-blocker and an ACEI or an ARB. Between the different Hawai‘i regions, rates ranged from a low of 21.0% in Hawai‘i-West to a high of 35.2% on Kaua‘i. Only 41.2% of patients were placed on one of the three recommended β-blockers. West Hawai‘i island had the lowest rate of prescription for the three select β-blockers, at 30.6%. Overall 11.0% of patients with CHF were not on either agent, with Kaua‘i having the lowest rate of 9.3%.

Table 2.

Use of ACEI^a, ARBs^b, and select β- Blockers by Region, Unadjusted (N = 24,138)

	Fully Treated as Recommended	Receiving a Recommended Medication		Not Treated as Recommended
Region	ACEI/ARB + select β- Blocker	ACEIa/ARBb	Select β- Blocker	Neither
Overall (N = 24,138)	6,993 (28.2%)	18,845 (76.1%)	10,202 (41.2%)	2,714 (11.0%)
Hawai‘i- East (n = 3151)	926(29.4%)	2,278 (72.3%)	1,427(45.3%)	372(11.8%)
Hawai‘i- West (n = 1581)	333(21%)	1,233 (78%)	484 (30.6%)	197(12.5%)
Kaua‘i (n = 1339)	472(35.2%)	1,076 (80.3%)	610 (45.6%)	125 (9.3%)
Maui (n = 2143)	619(28.9%)	1,614 (75.3%)	918 (42.8%)	230 (10.7%)
O‘ahu- Honolulu (n = 7220)	2,015(27.9%)	5,413 (75%)	3,038 (42%)	684(9.5%)
O‘ahu- Other than Honolulu (n = 8704)	2,500(28.7%)	6,805 (78.2%)	3,517 (36.3%)	882(10.1%)
P-valuec	P < .001	P < .001	P < .001	P = .01

^aACEI = Angiotensin Converting Enzyme Inhibitors;

^bARBs= Angiotensin Receptor Blocker;

^cBased on Pearson's chi-squared test.

In adjusted multivariable analyses, patients aged 18 to 44 with CHF were significantly less likely to be filling prescriptions for select β blockers, ACEI, or ARBs, either alone or in combination and more likely to have filled neither prescription, relative to patients aged 45 to 64 (Table 3). Patients over age 85 were also less likely to be on both medications and one of the two recommended medications, relative to patients aged 45 to 64. Prescription fill rates for all medications were similar for patients between ages of 65 and 84 to those aged 45 to 64 years. Women were slightly less likely to be taking select β-blockers than men but did not differ in terms of fill rates for the other medication groups (Table 3).

Table 3.

Adjusted Odds Ratio (OR) of Receiving Treatment as Recommended by Age, Gender, and Region (N = 24,138)^*

	Fully Treated as Recommended		Receiving a Recommended Medication				Not Treated as Recommended
	ACEI/ARB & β blockers		ACEI or ARB		β blockers		Neither
	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Age
45–65	1		1		1		1
18–44	0.73	[0.62, 0.87]	0.58	[0.50, 0.68]	0.82	[0.71, 0.95]	2.14	[1.76, 2.59]
65–84	0.97	[0.90, 1.03]	1.03	[0.96, 1.11]	0.96	[0.90, 1.02]	0.97	[0.88, 1.08]
85+	0.73	[0.67, 0.80]	0.77	[0.71, 0.84]	0.91	[0.84, 0.98]	1.12	[0.99, 1.26]
Sex
Male	1		1		1		1
Female	0.96	[0.91, 1.02]	1.01	[0.95,1.07]	0.94	[0.89, 0.99]	1.06	[0.97, 1.15]
Region
Honolulu	1		1		1		1
East Hawai‘i	1.07	[0.98, 1.17]	0.86	[0.79, 0.95]	1.14	[1.05, 1.24]	1.10	[0.97, 1.26]
West Hawai‘i	0.66	[0.58, 0.75]	1.13	[0.99, 1.29]	0.60	[0.53, 0.67]	1.18	[1.00, 1.40]
Kaua‘i	1.37	[1.21, 1.55]	1.33	[1.15, 1.54]	1.14	[1.01, 1.28]	0.86	[0.70, 1.04]
Maui county	0.99	[0.89, 1.10]	0.98	[0.87, 1.09]	1.01	[0.92, 1.12]	0.99	[0.85, 1.17]
O‘ahu-other than Honolulu	1.01	[0.95, 1.09]	1.18	[1.09, 1.27]	0.93	[0.87, 0.99]	0.92	[0.83, 1.01]

^*Findings significant at P < .05 are in bold.

There were also significant regional differences (Table 3). Compared to patients from O‘ahu-Honolulu MSA, those from Kauai‘i were significantly more likely to be prescribed one or both recommended medications. In contrast, patients from Hawai‘i-West were significantly less likely to be fully compliant with combination ACEI/ARB and β-blocker therapy. Patients from O‘ahu-other than Honolulu MSA were more likely to be on ACEI or ARBs (OR = 1.18 95% CI [1.09,1.27]) but less likely to be on select β-blockers (OR = 0.93 95% CI [0.87,0.99]) than patients living in O‘ahu-Honolulu MSA. Patients from Hawai‘i-West were less likely to be using select β-blockers (OR = 0.57 95% CI [0.50,0.66]), less likely to be using both ACEI or ARBs and select β-blockers (OR = 0.60 95% CI [0.53,0.67]), and more likely to be using neither(OR = 1.18 95% CI [1.00,1.40]). Medication use in Maui County did not differ from O‘ahu-Honolulu MSA.

Discussion

Our study investigated compliance with the nationally recommended guidelines for pharmacological management of patients with CHF. This analysis uncovered a rather low rate of compliance overall and demonstrated differences between the various regions of Hawai‘i. There are three identified limitations to this analysis with the first limitation being that only the use of select β-blockers (carvedilol, bisoprolol, and metoprolol succinate) was evaluated. The analysis was limited to these three β-blockers because these agents are specifically recommended for systolic heart failure due to their proven benefits in reducing morbidity and mortality.⁸ Our analysis found that the proportion of patients on β-blockers is significantly lower than those who are on either an ACEI or ARB, and the lack of pharmaceutical claims for the select β-blockers we queried for appears to be the limiting factor for CHF patients in receiving recommended dual therapy. Unlike systolic heart failure, current guidelines do not recommend a specific β-blocker for patients with isolated diastolic heart failure. By only investigating the use of the specific β-blocking agents recommended for systolic heart failure, our analysis may inappropriately label patients with isolated diastolic heart failure as “non-compliant.” Patients with isolated diastolic heart failure account for approximately one third of the heart failure population. Another possible explanation for the relatively low rates of compliance with appropriate β-blocker therapy is that there may be some confusion among prescribers in prescribing the appropriate formulation of metoprolol (short versus the long acting). The long acting formulation metoprolol succinate was included in our analysis as it has demonstrated decreased mortality in patients with CHF and is recommended by national guidelines,⁶ The short acting form (metoprolol tartrate) has not shown any decreased mortality and is therefore not preferred and was not included in our analysis. The initiation or dose titration of β-blocking agents may also worsen symptoms of CHF such as fatigue and shortness of breath. These adverse experiences may encourage noncompliance in both patients and prescribers.

The second limitation is that pharmaceutical claim data were only present when a patient filled a prescription. This limitation prevents any conclusions from being made regarding physician prescribing patterns, as we are unable to differentiate between prescriber and patient non-compliance. Also, claim data does not ensure medication adherence or even that the ACEI or ARB classes of medications were simultaneously taken with a β-blocker. This limitation introduces the potential for false- positives where patients may have filled claims for medications but were not adherent and/or did not take the medications concurrently. Moreover, patients may have had more than one form of drug coverage and filled prescriptions through a plan for which we do not have information. This would also cause an underestimation of compliance.

Factors such as lower socioeconomic status or ethnicity may influence patient's propensity to fill a prescription. Previous studies have demonstrated that lower socioeconomic status was associated with decreased primary medication compliance.¹⁸ The findings in our analysis however were not entirely consistent with previously published patterns. Hawai‘i-West demonstrated the lowest proportion of CHF patients who were on recommended therapy and the highest proportion of patients on neither an indicated β-blocker or an ACEI or ARB. However all four areas of Hawai‘i-West (South Kona, North Kona, South Kohala, and North Kohala) are below the state average in terms of percentage of residents living below the federal poverty level than other Hawai‘i regions. In contrast Hawai‘i-East demonstrated the second highest rates of patients receiving recommended medication therapy despite demographic data that indicates that the areas of Hilo, Puna, Ka‘u, and Hamakua are above the state average for percentage of residents living below the federal poverty level.¹⁴

Another limitation of this analysis was the inability to detect patient factors such as bradycardia, hypotension, renal instability, or allergies that would preclude the use of the recommended agents. Finally, our findings most likely overstate compliance with the recommended therapy as we require only one fill of a medication to be considered adherent.

Further investigation is needed to determine the key contributing factors that have led to overall low rates of compliance with the recommended medication therapy and the discrepancies between the different regions.

Conclusion

In the population of patients with CHF who were enrolled in a major health plan in Hawai‘i, there was a significant difference in compliance rates with ACCF/AHA recommended medication therapy within the state of Hawai‘i. Overall it was much more common for patients to be on an ACEI or an ARB compared to the recommended β-blockers. Hawai‘i-West demonstrated the lowest rates of compliance with dual therapy (ACEI/ARB + β-blocker), and had the lowest rates of pharmaceutical claims for the recommended β-blockers. Further investigation is needed to determine the causative factors.

Conflict of Interest

None of the authors identify a conflict of interest.