Treatment of Hypertension and Dyslipidemia or Their Combination Among US Managed-Care Patients

Authors


Michael P. Dutro, PharmD, BCPS, 5812 Valerian Place NE, Albuquerque, NM 87111
E-mail: michael.dutro@pfizer.com

Abstract

The authors examined treatment rates in managed-care patients with hypertension (HTN) only or dyslipidemia (DYS) only compared with patients who had both (HTN+DYS). A retrospective, cross-sectional claims analysis was performed in a 2002 US national managed-care database of 1.23 million continuously eligible members aged 18 years or older. Median age was 44.0 years, 8.8% were aged 65 years or older, and 53.2% were women. Study criteria identified 354,324 patients, 32.9% with HTN only, 34.7% with DYS only, and 32.4% with HTN+DYS. Overall, 49.7% of HTN patients had DYS and 48.3% of DYS patients had HTN. Patients with HTN+DYS were significantly older, more likely to have cardiovascular comorbidities, and more likely to use medications and hospital facilities than were patients with HTN only or DYS only (P<.01). About two-thirds of patients with HTN only received 1 or more prescription for an antihypertensive medication, compared with three-quarters of those with HTN+DYS. Fewer than half of patients with DYS only and approximately two-thirds with HTN+DYS received a cholesterol-lowering agent.

Cardiovascular disease (CVD) is the leading cause of death in the United States, affects >30% of the adult population, and had an estimated economic burden of $432 billion in direct and indirect costs in 2007.1 Hypertension (HTN) and dyslipidemia (DYS) are among the most important primary risk factors for coronary artery disease (CAD) and stroke.2 A majority of HTN patients have additional cardiovascular risk factors; DYS is one of the most co-prevalent risk factors.3,4 The risks associated with concomitant HTN or DYS are generally greater than the sum of the CVD risks from HTN and DYS alone.5,6 A recent study of HTN and DYS in US veterans found that the prevalence of CAD was more than doubled among patients with concomitant HTN and DYS compared with patients who had either condition alone.7

Clinical trials and meta-analyses have consistently demonstrated that pharmacologic treatment that lowers blood pressure (BP) and cholesterol levels reduces the risk of the clinical sequelae of CVD, such as myocardial infarction and stroke.8–11

Based on the results of such studies, clinical guidelines for the management of HTN and DYS have been developed that recommend therapeutic targets for BP and cholesterol.12,13 Despite these recommendations, <10% of patients with concomitant HTN and DYS from the National Health and Nutrition Examination Survey (NHANES) 2001–2002 are at goals currently established for both conditions.4 Similarly, a study conducted in the Veterans Affairs (VA) health care system demonstrated that among patients with concomitant HTN and DYS without evidence of CVD, 24% of nondiabetics and 15% of patients with concomitant diabetes are at both goals, and these proportions decreased to 14% and 13%, respectively, among patients with symptomatic CVD.14

Although recent guidelines have also emphasized the need for the concurrent management of multiple cardiovascular risk factors,13,15 few studies have assessed the management and quality of care of patients with concomitant HTN and DYS, especially within a managed-care population.

The objective of this study was to compare patients in a US managed-care population who have HTN only and DYS only with patients who have both conditions, with respect to key comorbidities, medication treatment rates, and hospitalizations.

METHODS

Data Source

Administrative claims data from 30 health plans were provided by Constella Health Strategies (now part of Ingenix, Inc). The data included eligibility information, facility claims, professional service claims, and outpatient medication claims. The regional distribution of members was as follows: Middle Atlantic, 30%; New England, 24%; South Atlantic, 12%; West North Central, 10%; Mountain, 8%; and other regions, 14%. Median age was 44.0 years, 8.8% of members were aged 65 years or older, and 53.2% were women. Plan designs represented in the data included preferred provider organizations (46.5%), health management organizations (36.6%), indemnity/point-of-service plans (13.7%), and other plan designs (3.1%). Approximately 2% of the membership was enrolled in Medicare risk plans. The database was compiled in accordance with all aspects of the Health Information Portability and Accountability Act of 1996 (HIPA).

Patient Selection and Classification

A total of 1.23 million members aged 18 years or older with continuous medical and pharmacy benefits in calendar year 2002 were extracted from the database. From this population, all patients who met the claims-based criteria for HTN or DYS were included in the study (Table I). Patients were then classified into 3 mutually exclusive analysis groups: patients with DYS but not HTN (DYS only), patients with HTN but not DYS (HTN only), and patients with both HTN and DYS (HTN+DYS). Table II outlines the inclusion criteria and assignment to analysis groups. All patients were identified using DTEC (Pfizer Inc, New York, NY) claims analysis software. DTEC is a proprietary software that integrates and analyzes managed-care organizations' enrollment data, professional medical claims, facility claims, and pharmacy claims to identify patients with respiratory or cardiovascular diseases, specifically asthma, chronic obstructive pulmonary disease, diabetes, DYS, heart failure, HTN, and ischemic heart disease, using predefined criteria.16

Table I. Inclusion Criteria: Definitions for Hypertension and Dyslipidemia
ConditionPatient Was Classified as Having Either Hypertension or Dyslipidemia If Any of the Following Conditions Were True During 2002
HypertensionThe patient filed an inpatient facility claim with a discharge diagnosis code for hypertension in any position or a professional service claim with a diagnosis code for hypertension in any position (ICD-9-CM codes 362.11; 401–401.9; 402–402.91; 403–403.91; 404–404.93; 405–405.99; 437.2; 642.0–642.04; 642.1–642.14; 642.2–642.24; 796.2)
 The patient filed a claim for one of the followings drugs used primarily for hypertension: rauwolfia, methyldopa, clonidine, or arterial/venous vasodilators
 The patient filed claims for drugs from ≥2 of the following 4 drug classes within 25 days of each other: diuretics, β-blockers, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, calcium channel blockers, or α receptor blockers
DyslipidemiaThe patient filed an inpatient facility claim with a discharge diagnosis code for dyslipidemia in any position or a professional service claim with a diagnosis code for dyslipidemia in any position (ICD-9-CM codes 272.0; 272.1; 272.2; 272.3; 272.4; 272.5; 272.7; 272.8; 272.9; 374.51)
 The patient filed a claim for one of the following cholesterol- or lipid-lowering drugs: statins, cholesterol absorption inhibitors, fibric acid derivatives, resins, or nicotinic acid/niacin
Table II. Inclusion Criteriaa and Analysis Groups
HTN≥1 inpatient or outpatient medical claims with a diagnosis claim for HTN, or
 ≥1 prescriptions for a drug used almost exclusively for HTN, or
 ≥2 prescriptions for a drug used commonly for HTN
DYS≥1 inpatient or outpatient medical claims with a diagnosis claim for DYS, or
 ≥1 prescriptions for a lipid-lowering drug
HTN onlyMet the criteria for HTN, but not DYS
DYS onlyMet the criteria for DYS, but not HTN
HTN+DYSMet the criteria for both HTN and DYS
aSee Table I for details. Abbreviations: Dys , dyslipidemia; HTN, hypertension.

Comorbidities, Medication Use, and Hospitalizations

Patients were evaluated for comorbidities, medication use, and hospitalizations based on their claims history during 2002. They were identified as having a selected comorbid condition based on the presence of at least one facility or professional claim with a specified diagnosis code for any of the following: diabetes, coronary artery disease, heart failure, heart rhythm/conduction disturbance, cerebrovascular disease, peripheral artery disease, therapeutic coronary artery procedure, acute myocardial infarction, angina pectoris, asthma, chronic obstructive lung disease, depression, kidney disease/abnormal kidney function, or osteoarthritis and/or rheumatoid arthritis. Medication use was identified based on the presence of at least one outpatient pharmacy claim for a drug in a specified drug class. Hospital admissions were identified based on the presence of at least one facility claim for an inpatient hospital stay. Although a hospitalization is seldom coded for HTN and rarely for DYS, disease specific hospitalizations were also evaluated. A hospitalization was classified as disease-specific if it included a diagnosis code for HTN or DYS in the first or second diagnosis positions.

It is important to point out, however, that the methods used for identifying patients with HTN and/or DYS are unavoidably imprecise because of the limited information contained in only 1 year of administrative claims (eg, claims data do not contain BP readings or provide numeric results of lipid testing). Although a very broad range of diagnosis codes and drugs were used to identify patients for the study, some patients who had HTN and/or DYS were likely not identified or, on the other hand, may have been misdiagnosed and therefore wrongly categorized.

Statistical Analyses

Univariate statistical tests were conducted for differences between the DYS only group and the HTN+DYS group and between the HTN only group and the HTN+DYS group. Chi-square tests were used to evaluate differences in proportions, and unequal variance t tests to evaluate differences in means. All statistical analyses were performed using Microsoft Access and Excel software (Microsoft Corporation, Redmond, WA).

RESULTS

Population and Demographics

A total of 354,324 patients who met the criteria for HTN or DYS in 2002 were included in the study. Of these, 116,502 (32.9%) were in the HTN only group, 122,925 (34.7%) were in the DYS only group, and 114,897 (32.4%) were in the HTN+DYS group. It should be noted that assignment to the groups in this study relied on diagnoses assigned or drugs prescribed by health care providers rather than specific clinical measures; therefore, patients may or may not have met guideline-defined criteria based on specific, defined BP or cholesterol levels. Mean ages for the HTN only and DYS only groups were 55.6 and 50.5 years, respectively, which were significantly lower than the mean of 58.7 years in the HTN+DYS group (P<.01). Members of the HTN+DYS group were more likely to be male compared with members of the HTN only group (54.7% vs 44.2%, respectively; P<.01). The proportion of men in the HTN+DYS group (54.7%) was only slightly higher than that in the DYS only group (54.1%; P<.01).

The prevalence of HTN in all dyslipidemic patients and DYS in all hypertensive patients was determined using the population sizes of the HTN only, DYS only, and HTN+DYS analysis groups. When examining the entire population of patients with HTN (n=231,399), the percentage who also had DYS was 49.7%. Similarly, when examining the entire population of patients with DYS (n=237,822), the percentage who had HTN was 48.3%.

Comorbidities

The percentages of patients in each of the 3 analysis groups who had a selected comorbid condition are reported in Table III. Patients with HTN+DYS had a significantly higher average number of comorbidities (3.5) than those in the HTN only or the DYS only groups (2.1 and 2.1, respectively; P<.01).

Table III. Comorbidities in Patients With HTN Only, DYS Only, or With Both HTN and DYS
ComorbidityHTN Only (n=116,502)aDYS Only (n=122,925)aHTN+DYS (n=114,897)
Diabetes, %12.89.624.4
Coronary artery disease, %9.88.225.3
Heart failure, %5.51.07.3
Heart rhythm/conduction disturbance, %13.48.916.0
Cerebrovascular disease, %4.62.27.2
Peripheral vascular disease, %4.01.96.1
Therapeutic coronary artery procedure, %0.91.15.7
Acute myocardial infarction, %0.70.63.4
Angina pectoris, %2.41.86.6
Asthma, %6.35.35.9
Chronic obstructive lung disease, %5.83.46.7
Depression, %7.27.66.4
Kidney disease/abnormal kidney function, %6.94.28.9
Osteoarthritis and/or rheumatoid arthritis, %11.97.812.6
Average No. of comorbidities per patient2.12.13.5
Abbreviations: DYS, dyslipidemia; HTN, hypertension. aAll differences between the HTN only and HTN+DYS groups and between the DYS only and HTN+DYS groups were statistically significant, at ≥1% significance level.

Except for asthma and depression, patients in the HTN+DYS group were more likely to have each of the selected comorbid conditions, including cardiovascular comorbidities, when compared with patients in the HTN only group or the DYS only group (P<.01). Further, these differences were greatest for diabetes, CAD (including therapeutic coronary artery procedure), and heart rhythm/conduction disturbance. The rate of depression among patients in the HTN+DYS was slightly lower than the rate among patients who independently had either condition, and these differences were statistically significant (P<.01).

Medication Use and Treatment Rates

The Figure shows the percentage of patients in each analysis group who were treated for their identified conditions: HTN, DYS or, in the case of the HTN+DYS group, either or both of these. Patients in the HTN+DYS group were more likely to be treated for HTN than those in the HTN only group (76.9% vs 66.4%; P<.01) and more likely to be treated for DYS than those in the DYS only group (64.2% vs 47.5%; P<.01). Treatment of both diseases occurred in only 56.9% of patients in the HTN+DYS group, however. Details on the percentages of patients who used selected medication classes by analysis group are included in Table IV. Patients in the HTN+DYS group were more likely than patients in the HTN only group to receive medications from each of the classes of antihypertensives evaluated (P<.01). Except for resins, patients in the HTN+DYS group were more likely than patients in the DYS only group to receive medications from each of the classes of lipid-lowering agents evaluated (P<.01). The rate of resin use among patients in the DYS only and the HTN+DYS group was the same.

Figure Figure.

Percentage of patients treated (ie, filled ≥1 prescriptions during 2002) in each analysis group. DYS indicates dyslipidemia; HTN, hypertension.

Table IV. Medication Use in Patients With HTN Only, DYS Only, or HTN and DYS
Medication, %HTN only (n=116,502)aDYS only (n=122,925)aHTN+DYS (n=114,897)
All antihypertensive agents (excluding diuretics and aldosterone blockers)66.4NC76.9
β-Blocker27.3NC37.8
ACE/angiotensin II inhibitor40.3NC52.8
CCB21.8NC25.4
α-Receptor blocker3.3NC4.1
Any diuretic32.73.234.2
Any dyslipidemia agent0.047.564.2
Statin0.043.560.1
Fibric acid derivative0.03.95.8
Resin0.01.71.7
Nicotinic acid/niacin0.01.21.8
Aldosterone blocker1.30.31.4
Potassium supplement5.20.75.8
Any antihypertensive and any dyslipidemia agentNC11.4b56.9
Any antihypertensive agent and statinNC10.6b53.3
ACE/angiotensin II inhibitor and statinNC3.9b36.5
β-Blocker and statinNC4.2b27.3
CCB and statinNC1.3b17.5
Abbreviations: ACE, angiotensin-converting enzyme; CCB, calcium channel blocker; DYS, dyslipidemia; HTN, hypertension; NC, not calculated. aAll differences between the HTN only and HTN+DYS groups and between the DYS only and HTN+DYS groups were statistically significant at the 1% significance level or better, with the exception of resins, where there was no difference in the percentage of drug use between the 2 groups. bThese values are not equal to zero because some DYS only patients were using an antihypertensive medication used to treat conditions other than HTN. Examples include ACE inhibitors for heart failure; β-blockers for angina pectoris, cardiac arrhythmia, and tremor; CCBs for angina pectoris; and α-receptor blockers for hypertrophy of the prostate gland.

Hospitalizations

The percentage of patients who had an any cause hospital admission was lower in patients in the HTN only group and in the DYS only group compared with patients in the HTN+DYS group (12.7%, 5.1%, and 13.7%, respectively; P<.01 for both comparisons). Patients in the HTN only group and the DYS only group also had lower disease-specific hospitalization rates compared with patients in the HTN+DYS group (2.5%, 0.4%, and 3.0%, respectively; P<.01). The percentage of patients who were hospitalized for stroke was lower in the HTN only group (0.3%) than in the HTN+DYS group (0.4%; P<.01); data were not available for the DYS only group. Patients in the DYS only group were less likely to have been hospitalized for a myocardial infarction than patients in the HTN+DYS group (0.1% vs 1.2%, respectively; P<.01). This figure was not available for the HTN only group, as the DTEC software does not provide this information.

DISCUSION

Few data exist comparing the prevalence and management of HTN only and DYS only with HTN+DYS in community practice. In this study of a large geographically diverse database of managed-care enrollees, we observed that HTN and DYS commonly occurred in the same patient, a finding that has been reported by others.4,7 More important, we found that effective therapies for treating these conditions appeared to be underutilized. Using a conservative metric for intervention, we observed that between 25% and 50% of patients with HTN or DYS or both failed to fill even a single prescription for therapy for these conditions. These data, therefore, highlight the current unmet needs in the management of these risk factors, at least in 2002. Newer clinical trial data published since 2002, which provide further evidence of the benefits of antihypertensive or lipid lowering therapy, may have resulted in a greater number of patients being treated more effectively.

This study specifically compares adult patients who had HTN alone or DYS alone with patients who had both conditions, in the context of care provided by US managed-care organizations in 2002. Our results showed that patients with HTN and/or DYS were distributed almost evenly among our 3 groups. Community-based prevalence estimates from 2 New England communities in the 1980s reported 23% of patients in an HTN only group, 46% in a DYS only group, and 30% in a HTN+DYS group.17 The differences between these findings and those of the present study are likely due to differences in estimating prevalences based on surveying the population of a community compared with claims generated by utilizers of health care in a national insured population. Because there are few comparative data in the managed-care database, it is uncertain whether these data are likely to be representative of what might be found in managed-care plans across the country. It is clear, however, that these risk factors for CVD are prevalent in the United States. A recent analysis of NHANES (2001–2002) demonstrated that the overall prevalence of HTN, DYS, and HTN+DYS was 30%, 47%, and 18%, respectively.4

Patients who had both conditions were on average older than patients who had either condition alone; this is consistent with existing literature4,7,18 and expected since increasing age raises the likelihood of having HTN or DYS.19,20 This study showed that HTN and DYS frequently occur together, suggesting that when one risk factor is diagnosed, the other should be checked for as well. Our data showed that 49.7% of all HTN patients had DYS and 48.3% of all DYS patients had HTN. A higher percentage of men were found to have both conditions. In contrast to this, Wong and colleagues4 observed that 16% of men and 20% of women in their NHANES sample had both conditions. This difference may be related to differences in the demographics of the 2 study populations; in the NHANES sample analyzed, ≈30% of patients were older than 60 years (59% female),4 whereas 8.8% of patients were aged 65 years or older (53.2% female) in the present analysis cohort. Also, the manner in which we identified HTN or DYS differed from previous studies, and this is likely to contribute to the difference found. We identified patients with HTN or DYS based on the presence of diagnosis codes in facility or professional claims or evidence of use of medications likely to be used for these conditions in pharmacy claims (Table I), whereas other studies used patient self-reported diagnoses or self-reported treatment with antihypertensive or lipid-lowering agents,4 together with BP measurements or lipid levels above guideline-defined cutoffs.4,7,18 It has previously been noted that lower rates of DYS in clinical vs survey data are related to the lower rates of testing in the clinic.18

There was a substantially higher prevalence of diabetes, CAD (including therapeutic coronary artery procedure), and heart rhythm/conduction disturbance in the HTN+DYS group compared with the HTN only or DYS only groups. This has previously been reported.7,18 The need for providers to treat each of these comorbid conditions to avoid adverse outcomes is particularly important.

The examination of medication use patterns suggests an underutilization of medications to control HTN or DYS in light of recent studies that confirm the reduction in CAD events associated with control of BP and DYS.9–11,21–23 Patients reported here were analyzed before some of these recent publications. Only two-thirds of patients in our study who had HTN only were receiving an antihypertensive medication. A recent national survey (Harris Poll) suggests that a much higher percentage of patients with HTN are receiving antihypertensive medication.24 Fewer than half (47.5%) who had DYS only received a medication for DYS, and only 56.9% of patients who had HTN and DYS received treatment for both of these conditions. These treatment rates are similar to those reported from 6 VA medical centers between October 1998 and September 200114 but less than reported recently.24 In this VA analysis, treatment rates were analyzed in patients with and without diabetes and symptomatic CVD. As expected, a higher proportion of patients with diabetes and symptomatic CVD received antihypertensive and lipid-lowering therapies than those without these conditions.14 The 2001–2002 NHAN ES analysis reported that only 28.5% of patients with HTN and DYS were treated for both conditions.4 This low treatment rate, compared with our data and the VA data, may reflect the difference between studies of medical records/claims data and the findings of surveys. In studies using medical records/claims data, patients are only included if they meet predefined criteria, such as a recorded diagnosis of HTN, DYS, or both conditions. In surveys, BP and lipid levels (as well as self-reported use of antihypertensive or lipid-lowering medications) are available for the whole sample. Surveys, therefore, may include patients with HTN, DYS, or both conditions who are not aware that they have these cardiovascular risk factors, since they have not been diagnosed or treated.

Nevertheless, our study defined treatment liberally, by the filling of ≥1 prescriptions for an HTN medication or a DYS medication during a 1-year period. If treatment had been evaluated based on reasonable persistency and adherence measures, treatment rates would have been much lower. Indeed, adherence to antihypertensive and lipid-lowering medication regimens has been shown to fall dramatically in the first year after initiation of therapy.25

The low utilization of appropriate medication treatment for each group is disturbing in light of the Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol In Adults (NCEP ATP II)13 and the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7)12 guidelines, which discuss the importance of treating patients for these risk factors as well as findings in landmark trials. The shortfalls today, however, may not be as striking as they were in 2002.

Of further interest is that the rank order for the various antihypertensive agents used was the same for HTN only and HTN+DYS patients, although slightly more HTN+DYS patients used each of the reported agents. Similarly, the rank order for the various DYS control agents used was the same in DYS only and HTN+DYS patients. Thus, the choice of medications for treating HTN and DYS was apparently independent of whether the patient had both conditions.

Limitations

The methods used to identify patients with HTN and/or DYS are unavoidably imprecise because of the limited information contained in only 1 year of administrative claims (eg, claims data do not contain BP readings or provide numeric results of lipid testing). Although a very broad range of diagnosis codes and drugs to identify patients for the study were used, some patients who had HTN and/or DYS were likely not identified or, on the other hand, may have been misdiagnosed. Information detailing the provider of the medications is not included in these analyses. Providers can have a significant effect on whether patients fill a written prescription. Furthermore, ethnicity was not accounted for in the present study, and it has been shown that African Americans are less likely than Caucasians to attain BP goals, despite receiving more antihypertensive medication prescriptions.26 Thus, a portion of the gap between diagnoses and filled prescriptions may be accounted for by ethnic differences. Moreover, it should also be noted that some patients with HTN attain their BP goal without using antihypertensive medications, potentially through lifestyle changes. For example, Riehle and associates26 noted that a substantial proportion of hypertensive patients achieved the BP goal without prescriptions for antihypertensive medications. The same phenomena may also occur in some patients with DYS.

In addition, because of the liberal inclusion criteria, it is also possible that a number of false-positives were identified. Because the size of each evaluation group was so large (each group consisted of >100,000 patients), even minor differences might prove to be significant.

Associations between treatment and hospitalizations cannot be fully established because of the cross-sectional nature of the design. Although prevalence of hospitalizations among the studied patients was observed, patient longitudinal data analysis would be needed to provide further insights into the effect of disease and treatment on hospitalizations.

A suggestion for future research would be to examine a population over a longer period to determine the effects of undertreatment in these at-risk populations. Further studies are also required to identify exactly where and when this breakdown in the management system is occurring. The results of this study and this additional work could be used by managed-care organizations to design and implement programs for addressing the gap in medication filling.

CONCLUSIONS

This study demonstrates that the percentage of managed-care patients in whom HTN and DYS are diagnosed is high. These patients are older and are hospitalized more often than their HTN only and DYS only counterparts. In addition, a low percentage of people diagnosed in 2002 were being medically treated for both diseases. This study suggests that managed-care and other health care organizations should continue to focus on these high-risk patients and develop additional programs that will identify and then effectively treat individuals who have both HTN and DYS.

Acknowledgements and disclosure:

Editorial support, which was funded by Pfizer, Inc, was provided by Jon Edwards and Karen Burrows of Envision Pharma as well as William Broderick of i3 Innovus. This research was funded and sponsored by Pfizer, Inc. Authors Dutro, Gerthoffer, and Tang are employees of Pfizer, Inc. Authors Peterson and Goldberg's contributions to this manuscript were underwritten by Pfizer, Inc.

Ancillary