REPORTS Hypertension care

REPORTS

Improving hypertension care in a large

group-model MCO
PAUL GODLEY, ANH NGUYEN, KRISTA YOKOYAMA, JAMES ROHACK,
BILLY WOODWARD, AND TINA CHIANG

H
ypertension is a significant risk
Abstract: The effectiveness of a quality pertensive medications per day. Of the pa-
factor for cardiovascular mor-
improvement program for hypertension tients on monotherapy, 93% received an
bidity and mortality and may management practices and patient health angiotensin-converting-enzyme inhibitor
lead to stroke, myocardial infarction outcomes in a group-model managed care (27.3%), diuretic (26.6%), β-blocker (23.4%),
(MI), or congestive heart failure organization was evaluated. or calcium channel blocker (15.4%). The
(CHF).1 Elevated blood pressure (BP) Health-system pharmacists analyzed overall level of BP control significantly im-
contributed to over 200,000 deaths in medical and pharmacy claims data to iden- proved from 37.2% at baseline to 49.2% at
the United States in 1998. Direct costs tify hypertensive patients. Chart review was follow-up (p = 0.0007). BP control in the fol-
conducted on a random sample of these low-up evaluation was 22.2% in diabetic
(i.e., physician visits, hospital and
patients to validate a hypertension diagno- patients. For treatment of patients with co-
nursing home services, medications, sis and to obtain blood pressure (BP) con- morbid disease states, provider practice
and home health and other medical trol rates and prevalence of cardiovascular patterns were evaluated at baseline and
durables) and indirect costs (i.e., lost risk factors and comorbid conditions. The follow-up.
productivity resulting from morbidi- interventions consisted of educating health Improving the quality of hypertension
ty and mortality) of hypertension to- care providers and recommending appro- management increased BP control from
tal over $40 billion a year.2 priate pharmacotherapy for compelling 37.2% to 49.2%. Continued efforts to im-
An estimated 50 million Ameri- indications. Patient outcomes were com- prove hypertension management, particu-
pared with baseline hypertension data. larly in patients with concomitant diabetes
cans have hypertension, but nearly one After interventions were implemented, and in elderly patients with isolated systolic
third of patients are unaware of it.2-4 medical and pharmacy claims identified hypertension, are needed.
Approximately half of all hyperten- 30,721 hypertensive patients and chart re-
sive patients remain untreated (in- views were performed on a random sample Index terms: Angiotensin-converting-
cluding 20% of patients with known of 417 patients. Pharmacy claims revealed a enzyme inhibitors; Calcium antagonists; Di-
hypertension), and only one quarter total of 193,311 antihypertensive prescrip- abetes mellitus; Diuretics; Geriatrics; Hyper-
of hypertensive patients whose hy- tions. Approximately 47% of all hyperten- tension; Hypotensive agents; Interventions;
sive patients were managed with mono- Managed care systems; Pharmacists; Quali-
pertension is controlled with antihy- therapy, while 24% received dual therapy, ty assurance; Sympatholytic agents
pertensive pharmacotherapy.3,4 and 11% were taking three or more antihy- Am J Health-Syst Pharm. 2003; 60:554-64
The sixth report of the Joint Na-

PAUL GODLEY, PHARM.D., is Director of Clinical Pharmacy Services, study, she was Health Outcomes Research Fellow, Applied Health
Scott & White Hospital (SWH), Temple, TX. ANH NGUYEN, Outcomes, Tampa, FL.
PHARM.D., is Health Outcomes Research Fellow, Novartis Pharma- Address correspondence to Dr. Godley, Clinical Pharmacy Servic-
ceuticals Corporation, East Hanover, NJ; at the time of this study, she es, Department of Pharmacy, Scott & White Hospital, 2401 South
was Health Outcomes Research Fellow, SWH. KRISTA YOKOYAMA, 31st Street, Temple, TX 76508 ([email protected]).
PHARM.D., is Health Policy and Behavioral Research Manager, Well- Supported by a grant from Novartis Pharmaceuticals Corporation.
Point Pharmacy Management, West Hills, CA; at the time of this Presented at the ASHP Midyear Clinical Meeting, New Orleans,
study, she was Health Outcomes Research Fellow, SWH. JAMES LA, December 2001.
ROHACK, M.D., is Medical Director at Scott & White Health Plan,
Temple, TX. BILLY WOODWARD, B.S.PHARM., is Director of Pharma- Copyright © 2003, American Society of Health-System Pharma-
cy, SWH. TINA CHIANG, PHARM.D., is Director of Consulting Servic- cists, Inc. All rights reserved. 1079-2082/03/0302-0554$06.00.
es, Aequitas Consulting Group, San Diego, CA; at the time of this

554 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003

 REPORTS Hypertension care

tional Committee on Prevention, De- tions and evaluating the appropriate- Chart review was performed by an
tection, Evaluation, and Treatment of ness of pharmacotherapy for com- outside outcomes research organiza-
High Blood Pressure (JNC-VI) and pelling indications. tion to validate the hypertension diag-
the 1999 World Health Organization– nosis and to obtain follow-up mea-
International Society of Hyperten- Methods sures of the patients’ clinical status.
sion (WHO–ISH) Guidelines for the Study phases. The hypertension Data collection. Variables collect-
Management of Hypertension pro- management study was composed of ed during the chart review included
vide current guidelines on the man- three phases: phase 1 (September 1, BP, hypertension therapy, documen-
agement of hypertension.4,5 While 1998–August 31, 1999, in which tation of adverse effects, documenta-
JNC-VI includes recommendations baseline data were collected), phase 2 tion of self-care recommendations,
for specific patient populations, (April 1, 2000–September 30, 2000, the cardiovascular risk factors, comor-
guidelines for diabetic patients with intervention period), and phase 3 (Janu- bidities, and target organ damage. In
hypertension have also been provid- ary 1, 2000–December 31, 2000, in accordance with Health Plan Em-
ed by the American Diabetes Associ- which follow-up data were collected). ployer Data and Information Set
ation (ADA) and National Kidney Baseline assessment identified the (HEDIS) 2000 guidelines, the BP
Foundation (NKF).6,7 For patients hypertensive population in the health measurement used for analysis was
with concomitant high BP and CHF, plan, current treatment patterns, and the most recent measurement after
the American College of Cardiology opportunities for improving hyper- the diagnosis of hypertension during
(ACC) and American Heart Associa- tension care practices. This informa- the year of interest.10 BP must have
tion (AHA) provide the most current tion was compared with the outcomes been obtained during a physician of-
guidelines.8 data after the quality improvement fice visit or other nonemergency out-
In an effort to improve BP control interventions were implemented. patient facility visit. BP readings
in the health plan’s population, Scott Patient selection. Hypertensive from outpatient visits for surgical
& White Health Plan (SWHP), a patients were identified through procedures or diagnostic tests and
staff-model managed care organiza- medical and pharmacy claims proc- those self-reported were not consid-
tion covering approximately 180,000 essed during the respective phases of ered for analysis.
members in central Texas and oper- the study. All patients age 18 years Interventions. This hypertension
ating eight pharmacies, implemented and over with either a diagnosis of quality improvement program was
a hypertension management pro- hypertension or a pharmacy claim developed to enhance hypertension
gram. Scott & White Clinic is a mul- for an antihypertensive medication care practices in SWHP. The inter-
tispecialty practice, which employs were included. Patients with a diag- ventions were implemented over six
over 500 physicians, including 158 nosis of hypertension were defined as months (April–September 2000). The
family practice and internal medicine those having at least one code from interventions associated with this
physicians at 19 regional clinics. The the International Classification of Dis- project were primarily made by
methodology of the program and eases, 9th Revision, Clinical Modifica- pharmacy department personnel,
baseline patient characteristics were tion (ICD-9-CM) between 401.0 and but critical support was given by
previously described.9 401.9. Antihypertensive drugs were SWHP’s administrative and quality
The primary goal of the program identified according to Generic improvement (QI) staff members
was to improve BP control in this Product Identifier, Universal Sys- and the information systems (IS)
managed care population by (1) in- tems of Classification, or American staff. Components of the interven-
creasing awareness and understanding Hospital Formulary Service classifi- tion efforts included educating
of hypertension and its complications, cations. Patients with end-stage renal health care providers, identifying hy-
(2) identifying specific opportunities disease (ICD-9-CM code 585.x) or pertensive patients, and evaluating
for service improvement through de- patients receiving any antihyperten- the appropriateness of the pharma-
tecting discrepancies between JNC-VI sive medication for medical condi- cologic regimens.
guidelines and current care, and (3) tions other than hypertension were Educational efforts were aimed at
increasing appropriate antihyperten- excluded from the study. Table 1 lists increasing physicians’ awareness of
sive medication prescribing through the medical conditions and accom- hypertension, improving quality of
the application of JNC-VI guidelines panying pharmaceutical agents used care, and encouraging physician
for patients with compelling indica- as exclusion criteria. adherence to JNC-VI guidelines. Ed-
tions. This study measured the effec- From the identified hypertensive ucational tools included live interac-
tiveness of the hypertension man- population, a subsample of 500 pa- tive teleconferences, academic detail-
agement program by examining BP tients was randomly selected for medi- ing using physician profiling reports,
control rates in various subpopula- cal chart review in phases 1 and 3. direct mailings of physician profiling

Am J Health-Syst Pharm—Vol 60 Mar 15, 2003 555

 REPORTS Hypertension care

Table 1. ure 1), which included key guidelines

Exclusion Criteria for Eligible Patients by Pharmacy Claims from JNC-VI and the preferred SWHP
and Accompanying Medical Diagnoses formulary agents, was developed by the
pharmacy department. Total pharmacy
Drugsa Medical Diagnosis (ICD-9-CM Code)
department time involved in this project
β-blockers, CCBs, nitrates Ischemic heart disease (411.x, 413.x, (teleconferences, academic detailing,
β-blockers, ACE inhibitors, CCBs, nitrates 414.0x)
Diuretics, ACE inhibitors, ARBs, hydralazine Myocardial infarction (410.x, 412.x) and publications) exceeded 200 hours.
with nitrates, spironolactone, carvedilol Data analysis. Analyses were per-
(but not other β-blockers) formed to determine BP control, prev-
β-blockers, diltiazem, verapamil, clonidine Congestive heart failure (428.x, 398.91)
Diuretics Arrhythmia (427.0x, 427.3x, 427.6x) alence of risk factors and comorbidi-
α1-adrenergic-receptor blockers Lower-extremity edema (782.3x) ties, and drug treatment patterns. The
β-blockers, clonidine Benign prostatic hypertrophy (600.x) data were further analyzed by patient
Anxiety or panic disorder (300.0x)
a
demographics (age and sex). BP con-
ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification, CCBs = calcium-
channel blockers, ACE = angiotensin-converting enzyme, ARBs = angiotensin-receptor blockers. trol was analyzed according to JNC-VI
and HEDIS 2000 and 2001 guidelines.
According to JNC-VI and HEDIS
2000, adequate BP control is defined as
reports, spotlight cards summarizing and involved approximately 116 a systolic blood pressure (SBP) of <140
antihypertensive therapies, and news- hours in total. These meetings were mm Hg and a diastolic blood pressure
letters containing SWHP’s hyperten- conducted at all 14 clinics on two (DBP) of <90 mm Hg. JNC-VI defines
sion treatment algorithm (Figure 1). different occasions (each clinic was hypertension control for diabetic pa-
Physicians from the various clin- visited twice) and were held at the tients as <130/85 mm Hg, and HEDIS
ics participated in interactive presen- noon hour. The pharmacist pro- 2000 did not differentiate BP control
tations through teleconferencing. To vided a general presentation and for diabetic patients. The HEDIS 2001
enhance physician attendance, the then reviewed the individual profil- guidelines were more lenient, requir-
conferences were conducted during ing reports with the medical staff. ing BP control of ≤140/90 mm Hg in
the noon hour and lunch was pro- The general schedule and format of patients with or without diabetes.
vided using SWHP funds. These tele- these academic detailing sessions Statistical analysis. The sample size
conferences were moderated by a are detailed in Table 2. was based on the following statistical
clinical pharmacist from the SWHP Provider profiling reports en- assumptions: two-tailed test of signifi-
pharmacy department and featured a abled physicians to compare their cance between two proportions with α
review and update of JNC-VI guide- prescribing habits with national = 0.05 and a power of 80%. Descrip-
lines by a nationally-known physi- guidelines to treat hypertension. tive statistics were calculated for all
cian and hypertension specialist. The Data retrieval and analysis and in- variables measured. Phase 1 versus
teleconferences were immediately dividual physician reports were phase 3 outcomes comparisons were
followed by a 45-minute presenta- prepared by SWHP’s IS depart- made using a Student’s t test for con-
tion of phase 1 results conducted by a ment; provider profiling reports tinuous variables. The chi-square test
pharmacist. Typically, each telecon- and newsletters were mailed by was used to assess changes in the pro-
ference involved the participation of SWHP’s QI staff; approximately 40 portions of subjects of all variables an-
three pharmacists; it is estimated that and 10 hours were required for alyzed from baseline to follow-up.
the pharmacy department’s involve- each task, respectively. Similar ef-
ment in these teleconferences totaled forts had been successful at Results
60 hours. Teleconferences usually SWHP.11 Newsletters, distributed Demographics. Data from Septem-
linked two clinics per session, but in to health care providers, were used ber 1998 through December 2000 were
some cases three clinics were linked in to discuss the hypertension im- retrieved from SWHP medical and
a single conference, and a pharmacy provement program. The pharma- pharmacy databases. At baseline,
department representative was present cy and therapeutics (P&T) newslet- 23,285 medical and pharmacy claims
at each clinic to help moderate the dis- ter presented the methodology and and 374 medical charts were reviewed.
cussion. Teleconferences were held on results of phase 1, provided treat- During the follow-up period, 30,721
six different occasions and involved a ment guidelines, and encouraged medical and pharmacy claims and 417
total of 14 different clinics. the use of fixed-dose combination medical charts were reviewed. Patient
Academic detailing sessions using therapy. In addition, an SWHP demographics between baseline and
physician profiling reports were con- P&T committee-approved hyper- follow-up did not significantly differ,
ducted by three clinical pharmacists tension treatment algorithm (Fig- except for the age of patients randomly

556 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003

 REPORTS Hypertension care

Figure 1. Scott & White Health Plan treatment algorithm for hypertension. First-line thera- tients had both pharmacy and medi-
pies are in bold type. Adapted from reference 4. CCB = calcium-channel blocker, ARB = cal claims; and 5,641 patients had
angiotensin-receptor blocker, ACE = angiotensin-converting enzyme, MI = myocardial inf-
arction, BP = blood pressure, ISH = isolated systolic hypertension.
only a hypertension-related medical
claim.
Drug therapy. The pattern of drug
Hypertension therapy in phase 1 was previously re-
ported.9 After six months of interven-
No Initial drug tions, nearly one half of all hyperten-
coexisting therapy choices
diseases (if no contraindications exist) sive patients were on monotherapy
(46.8%), 24.2% were on dual therapy,
and 10.7% were taking three or more
b-blocker Coexisting diseases antihypertensive medications per day
Diuretic (Table 4). Of those patients receiving a
Angina
single antihypertensive agent, the four
most commonly used drug classes
b-blocker were angiotensin-converting-enzyme
CCB (concurrent use of diuretic strongly
recommended) (ACE) inhibitors (27.3%), diuretics
(26.6%), β-blockers (23.4%), and
Heart Failure calcium-channel blockers (CCBs)
(15.4%) (Table 5). Ninety-three per-
ACE inhibitor cent of the patients on monotherapy
Diuretic
ARB received an agent from one of these
b-blocker four drug classes. The most com-
monly used dual therapy regimen in-
Post-MI cluded a diuretic with a β-blocker,
ACE inhibitor, or CCB (Table 6). In
b-blocker (non-ISH) the dual therapy group, more than
ACE inhibitor (with systolic dysfunction)
57% of patients received a diuretic.
Diabetes (Goal BP <130/80 mm Hg)
Patients on multiple therapies
(three or more agents) totaled 3,287
ACE inhibitors (10.7%) of the 30,721 patients. The
ARB most common triple therapy combi-
nation was a β-blocker, an ACE in-
Isolated Systolic Hypertension
hibitor, and a diuretic (20.4%). Oth-
Diuretic er combinations included an ACE
Low-dose b-blocker (concurrent use of diuretic inhibitor, a CCB, and a diuretic
strongly recommended)
Long-acting CCB (e.g., verapamil) (15.9%) or a β-blocker, a CCB, and a
(concurrent use of diuretic strongly recommended diuretic (9.8%). In phase 1, 26.5% of
all patients in the multidrug group
Hyperlipidemia
had adequate BP control. After in-
a-receptor tervention, the multidrug group
ACE inhibitors showed the most improvement, with
CCB (concurrent use of diuretic strongly
recommended) 57.7% of patients achieving BP con-
b-blocker (with prior history of angina or MI) trol (Figure 2).
Low-dose diuretic
JNC-VI guidelines encourage the
use of ACE inhibitors and β-blockers
in combination with diuretics. Phase
selected for chart review (Table 3). total of 193,311 claims for antihy- 3 follow-up evaluation showed a
The mean age was older in phase 1 pertensive medications were filed at 12.5% relative decrease in CCB use,
chart review patients compared to SWHP. From pharmacy and medical while β-blocker use increased by 6.5%
those in phase 3 (66.2 and 63.1 years, claims, 30,721 hypertensive patients (Table 7). ACE inhibitor and diuretic
respectively) (p < 0.0001). were identified. There were 10,377 use remained the same. There was a
Patient identification. During patients with only a hypertension- relative increase of 21–100% in the use
the 12-month follow-up period, a related pharmacy claim; 14,703 pa- of fixed-dose combination agents

Am J Health-Syst Pharm—Vol 60 Mar 15, 2003 557

 REPORTS Hypertension care

Table 2.
Teleconferences and Academic Detailing Sessions To Improve Hypertension Care
Session Dates Description Duration (min)
Teleconferences (6 Apr–May 2000 Hypertension expert presentation 45
teleconferences for all 14 Pharmacy department presentation of 45
clinics) baseline results and discussion of results
1st academic detailing session Jun–Jul 2000 Pharmacy department presentation of 30
(1 session at each of 14 clinics) baseline results, review of hypertension
treatment algorithm, and introduction of
prescriber profiling reports
Individual review of physician profiling 60
reports
2nd academic detailing session Aug–Sep 2000 Pharmacy department presentation of 30
(1 session at each of 14 clinics) summary physician profiling reports
Individual review of physician profiling 60
reports with specific focus on JNC-VI
compelling indicationsa
a
JNC-VI = sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.

Table 3.
Patient Demographics (ACE inhibitor, angiotensin-receptor
% Claims/% CRa blocker [ARB], or β-blocker with di-
Phase 1 Phase 3 uretic or ACE inhibitor–CCB combi-
n = 23,285 (Claims) n = 30,721 (Claims) nation). However, since fixed-dose
Variable n = 374 (CR) n = 417 (CR) p combinations represented only 5.6%
Female 58.8/59.1 58.2/58.0 NS/NS of the total number of antihyperten-
Male 41.2/40.9 41.8/42.0 NS/NS sive prescriptions, these changes rep-
Mean age, yr 63.5/66.2 62.4/63.1 NS/<0.0001 resent overall small increases in the
a
CR = chart review, NS = not significant. use of combination products.
For treatment of hypertensive
patients with comorbid disease
Table 4.
states, provider practice patterns
Distribution of Antihypertensive Pharmacotherapy in Phase 3
were compared with practice stan-
(n = 30,721)
dards (JNC-VI guidelines for com-
pelling indications) for the patients
Therapy No. (%) Patients randomly selected for chart review.
a
None 5,641 (18.4) After the interventions, the follow-
Single therapyb 14,372 (46.8)
Dual therapy 7,432 (24.2) ing changes were observed (Table
More than three agents 3,276 (10.7) 8): an 8% increase in β-blocker use
a
Includes patients without Scott & White Health Plan pharmacy benefit. It cannot be determined if these in hypertensive post-MI patients
patients received antihypertensive medications. and a 9% increase in both ACE in-
b
Includes fixed-dose combination products (2,211 prescriptions [5.6%] of 37,839 pharmacy claims).
hibitor and ARB use in hyperten-
sive diabetic patients. A decline of
1.4% was found in ACE inhibitor
or ARB use in CHF patients with
Table 5. hypertension. A 2.3% decrease in
Frequency of Drugs Used in Monotherapy in Phase 3 (n = 14,372) ACE inhibitor, ARB, and CCB use
was also observed in hypertensive
Therapya No. (%) Patients
patients with nephropathy.
ACE inhibitors 3929 (27.3) Hypertension control. On the ba-
Diuretics 3819 (26.6)
β-blockers 3359 (23.4) sis of HEDIS 2000 guidelines, the
CCBs 2220 (15.4) overall level of BP control improved
α1-adrenergic-receptor blockers 569 (4.0) significantly from 37.2% to 49.2%
ARBs 291 (2.0)
Other hypertensive drugs 185 (1.3) (p = 0.0007) (Table 9) as judged from
a
ACE = angiotensin-converting enzyme, CCBs = calcium-channel blockers, ARBs = angiotensin-receptor
chart reviews. BP control increased
blockers. from phase 1 to phase 3, regardless of

558 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003

 REPORTS Hypertension care

Table 6. standards rather than those of

Frequency of Drugs Used in Dual Therapy in Phase 3 (n = 7432) HEDIS 2000 were used in the analy-
Therapya No. (%) Patients sis, the number of patients whose BP
was adequately controlled increased
Diuretic + β-blocker 1720 (23.1) from 48.4% to 60.4% (p = 0.0007).
Diuretic + ACE inhibitor 1682 (22.6)
Diuretic + CCB 880 (11.8) When patients were categorized into
β-blocker + ACE inhibitor 775 (10.4) different age groups and analyzed us-
CCB + ACE inhibitor 732 (9.8) ing HEDIS 2000 criteria, the 66–85
β-blocker + CCB 443 (6.0)
Other dual therapy 1200 (16.1) age group had a significant improve-
a
ACE = angiotensin-converting enzyme, CCB = calcium-channel blocker. ment in BP control from 30.9% to
50.3% (p < 0.0001) (Table 10).
The majority of patients not
Figure 2. Rate of blood presssure control according to Health Plan Employer Data and achieving BP control had an SBP of
Information (HEDIS) 2000 criteria by treatment type. Gray bars indicate baseline data, black >140 mm Hg. Since the frequency of
bars indicate follow-up data.
isolated systolic hypertension (ISH)
60.0 57.7
increases with age and more than
half of SWHP’s hypertensive popula-
tion were over age 60, the data on
50.0 46.4 47.3
45.8 ISH were also analyzed using the Sys-
43.3 tolic Hypertension in the Elderly
40.0 39.0 38.3 Program (SHEP) definition (≥160/
Patients (%)

<90 mm Hg) and the JNC-VI defini-

30.0 tion (>140/<90 mm Hg).12 Of the
26.5
417 patients selected for chart review
in phase 3, 37 patients (8.9%) had
20.0
ISH according to SHEP criteria ver-
sus 152 patients (36.5%) when using
10.0 JNC-VI criteria (Table 11). As ex-
pected, ISH prevalence in SWHP’s
0.0 population was highest in the 66–85
Single Dual Multiple Overall
(p = 0.20) (p = 0.77) (p = 0.02) (p = 0.05)
age group for both phases 1 and 3
(43.9% and 59.5%, respectively). In
Therapy Group
the 46–65 age group, phase 3 data
showed an increase in ISH from
phase 1 data. However, when using
Table 7. both SHEP and JNC-VI criteria, an
Frequency of Drug Classes Used to Manage Hypertension overall decrease in the number of pa-
No. (%) Patients tients with ISH occurred between
Phase 1 Phase 3 Relative phase 1 and phase 3.
Therapya (n = 27,571) (n = 37,839) Change (%) Approximately 20% of hyperten-
Diuretics 7,444 (27.0) 10,179 (26.9) –0.4 sive patients also had diabetes, placing
ACE inhibitors 6,010 (21.8) 8,287 (21.9) 0.5 them at greater risk for cardiovascular
β-blockers 5,514 (20.0) 8,060 (21.3) 6.5 disease and other complications.
CCBs 5,064 (18.4) 6,092 (16.1) –12.5
ACE inhibitor and JNC-VI guidelines recommend a
diuretic 523 (1.9) 908 (2.4) 26.3 pressure of <130/85 mm Hg for pa-
β-blocker and diuretic 386 (1.4) 643 (1.7) 21.4 tients with diabetes. From the phase
ARB and diuretic 110 (0.4) 227 (0.6) 50
ACE inhibitor and CCB 110 (0.4) 303 (0.8) 100 3 chart review data, 99 hypertensive
Other therapy 2,410 (8.7) 3,140 (8.3) –4.6 patients had diabetes, and only 22
a
ACE = angiotensin-converting enzyme, CCBs = calcium-channel blockers, ARB = angiotensin-receptor patients (22.2%) met the target BP, a
blocker.
slight decrease from the 25.8% who
had achieved BP control in phase 1
the number of pharmacotherapy sive agents had a significant improve- (Table 12). The proportion of un-
agents used (Figure 2). However, only ment (26.5% in phase 1 to 57.7% in controlled diabetic patients with BP
patients with multiple antihyperten- phase 3) (p = 0.02). If HEDIS 2001 130–139/85–99 mm Hg (borderline

Am J Health-Syst Pharm—Vol 60 Mar 15, 2003 559

 REPORTS Hypertension care

Table 8.
Conformance with JNC-VI Best Practice Standardsa
No. (%) Patients
Appropriate Relative
JNC-VI Compelling Indication Therapy Phase 1 Phase 3 Change (%)
History of myocardial infarction β-blocker 248 (57.4) 310 (62.1) 8.2
Diabetes ACE inhibitor or ARB 2186 (55.1) 3332 (60.1) 9.1
Congestive heart failure ACE inhibitor or ARB 650 (57.1) 880 (56.3) –1.4
Nephropathy ACE inhibitor, ARB, or CCB 191 (79.3) 320 (77.5) –2.3
a
JNC-VI = sixth report of the Joint National Committee on Prevention, Detection, Evaluation, Treatment of High Blood Pressure, ACE = angiotensin-converting enzyme,
CCB = calcium-channel blocker, ARB = angiotensin-receptor blocker.

Table 9. were 37.2% at baseline and 49.2% at

Rate of BP Control According to JNC-VI and HEDIS 2000 and 2001 follow-up (p = 0.0023) (Table 10).
Guidelinesa Compared with the Third Nation-
No. (%) Patients al Health and Nutrition Examination
Phase 1 Phase 3 Survey (NHANES III), which found
(n = 374 [CR]) (n = 417 [CR]) the national BP control rate to be
Guideline (n = 66 Diabetic Pts.) (n = 99 Diabetic Pts.) p 27.4%, SWHP had better results in
JNC-VI both phases of this study. However,
All pts. 127 (34.0) 175 (42.0) 0.0206 approximately 50% of the hyperten-
Diabetic pts. 17 (25.8) 22 (22.2) 0.6005
HEDIS 2000 139 (37.2) 205 (49.2) 0.0007 sive patients still have uncontrolled
HEDIS 2001 181 (48.4) 252 (60.4) 0.0007 BP. Aggressive therapy is needed to
a
BP = blood pressure, JNC-VI = sixth report of the Joint National Committee on Prevention, Detection, control BP, as hypertension is a mul-
Evaluation, Treatment of High Blood Pressure, HEDIS = Health Plan Employer Data and Information Set, CR =
chart review. tifactorial disease involving many or-
gan systems. For patients whose BP is
uncontrolled with a single antihyper-
high) was larger in phase 3 (18.2% BP control data were analyzed us- tensive agent, JNC-VI and WHO–ISH
and 30.3%, respectively). ing JNC-VI and HEDIS 2000 and 2001 guidelines recommend the appropri-
guidelines. HEDIS 2001 guidelines, ate use of multiple drug therapies.4,5
Discussion more lenient than those of 2000, take The use of two agents from different
The results from this quality into consideration that health care classes can improve the control of
improvement program for hyper- professionals may round up BP read- hypertension while minimizing
tension illustrated how education- ings (e.g., a SBP of 139 mm Hg would dose-dependent adverse effects.5,13,14
al efforts can significantly affect be rounded to 140 mm Hg). A signifi- A previous study found monothera-
the clinical outcomes of a disease cant improvement in BP from base- py to be effective in only 40–50% of
state. By implementing certain dis- line to follow-up was found using the hypertensive population.15
ease management techniques, the each of these criteria (Table 9). For In our study, BP control for pa-
quality of hypertension management simplification, we reported BP con- tients on multiple drug therapy in-
can be improved. Interventions such trol by pharmacotherapy (Figure 2) creased from 26.5% at baseline to
as encouraging physicians to use and by age group (Table 10) using 57.7% at follow-up (p < 0.02) (Fig-
fixed-dose combination products the stricter HEDIS 2000 guidelines. ure 2). In addition to improved effi-
for patients taking two separate The mean age of the baseline and cacy, a combination of products
products, increasing awareness of follow-up chart-review groups dif- from different classes allows for low-
the importance of BP control and its fered (66.2 and 63.1 years, respec- er doses of each drug, which may
impact on cardiovascular outcomes, tively). The 66–85 age group had the help minimize metabolic and other
advocating strict adherence to na- greatest improvement in BP control, adverse effects common to antihy-
tional guidelines, and providing with 30.9% controlled at baseline pertensive therapies.16 When appro-
physician prescribing profiles are ef- and 50.3% controlled at follow-up priate, fixed-dose combination anti-
fective. Through this quality im- (p < 0.0001) (Table 10); this repre- hypertensive products may enhance
provement initiative, SWHP in- sented the largest percentage of pa- compliance to therapy in those
creased BP control, according to tients of the four age categories eval- patients who require multiple drug
HEDIS 2000 guidelines, from 37.2% uated. Age did not influence overall regimens for BP control.17 Phase 3
to 49.2% (p = 0.0007) during the BP control rates. When the data were results found that fixed-dose combi-
follow-up period. controlled for age, BP control rates nation products accounted for only

560 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003

 REPORTS Hypertension care

Table 10.
Rate of BP Control by Age on the Basis of HEDIS 2000 Criteriaa
Phase 1 Phase 3
No. (%) Pts. with No. (%) Pts. with
Age Group (yr) All Pts. Controlled BP All Pts. Controlled BP p
18–45 38 20 (52.6) 53 30 (56.6) 0.7088
46–65 115 51 (44.3) 169 80 (47.3) 0.6204
66–85 194 60 (30.9) 181 91 (50.3) <0.0001
≥86 27 8 (29.6) 14 4 (28.6) 0.9444
Total 374 139 (37.2) 417 205 (49.2) 0.0023b
a
By chart review. BP = blood pressure, HEDIS = Health Plan Employer Data and Information Set.
b
Controlled for age groups.

Table 11. ticular interest in our study: diabetics

Prevalence of Isolated Systolic Hypertension (ISH) by Age Groupa and those patients with ISH.
Diabetes places patients at high risk
No. (%) Patients
for cardiovascular disease and mi-
ISH According to SHEPb ISH According to JNC-VIc crovascular and macrovascular com-
Age Group Phase Phase Phase Phase plications. JNC-VI guidelines for BP
(yr) 1 3 1 3
control in patients with concurrent di-
18–45 1 (2.4) 1 (2.7) 8 (4.6) 12 (7.9) abetes and hypertension recommend a
46–65 7 (17.1) 11 (29.7) 31 (17.8) 57 (37.5) SBP of <130 mm Hg and a DBP of <85
66–85 18 (43.9) 22 (59.5) 91 (52.3) 74 (48.7)
≥86 15 (36.6) 3 (8.1) 44 (25.3) 9 (5.9) mm Hg. The United Kingdom Pro-
a
By chart review. SHEP = Systolic Hypertension in the Elderly Program, JNC-VI = sixth report of the Joint spective Diabetes Study (UKPDS)
National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. found that strict control of BP (144/82
b
>160/<90 mm Hg (SHEP guidelines).
c
>140/<90 mm Hg (JNC VI guidelines). mm Hg) compared with less rigid con-
trol (154/87 mm Hg) resulted in a 44%
risk reduction in stroke, a 37% risk
5.6% of all antihypertensive agents and peripheral vascular disease. A re- reduction in microvascular endpoints,
dispensed, compared with 5.0% in duction of 10 mm Hg in DBP can and a 32% risk reduction in diabetes-
phase 1. There are many opportuni- greatly reduce the mortality risk asso- related deaths.20
ties for improving the use of fixed- ciated with cardiovascular disease.18 Patients with diabetes and hyper-
dose combinations; however, one of SHEP study results found that a reduc- tension are four times more likely to
the disadvantages of these agents is tion in SBP of 11–14 mm Hg or a have a cardiovascular event than pa-
difficulty in titrating doses. With reduction in DBP by as little as 3–4 tients without diabetes.21 In diabetic
more combinations of different dos- mm Hg could produce a relative risk hypertensive patients with nephrop-
es becoming available, the conver- reduction of a major CHD event by athy, the risk for a cardiovascular
sion from multiple drug therapies to 27%.12 Data from the Multiple Risk event increases 10-fold.22 In addition
fixed-dose combination products Factor Intervention Trial showed to the cardiovascular risk, other risks
may be easier to manage. In any that a reduction in SBP makes a exist for these patients. UKPDS found
health care decision, cost for the pa- greater difference in CHD-related that strict BP control is associated with
tient and third-party payer is always mortality than a reduction in DBP.19 a decrease in both microvascular and
a consideration. While fixed-dose Clinical cardiovascular disease and macrovascular complications and may
combination products may increase end-organ damage, such as retinopa- be more important than rigid glycemic
compliance and reduce payment thy, nephropathy, and peripheral arte- control in reducing clinical compli-
from two copayments to one for the rial disease, are also of major concern cations from diabetes. 20 JNC-VI
patient, the cost of fixed-dose brand- for hypertensive patients. Since most guidelines recommend a target BP
name products may be higher to hypertensive patients have multiple for this patient population of <130/
third-party payers when generics of risk factors (>80% of SWHP’s hyper- 85 mm Hg. New recommendations
the combined drugs are available. tensive patients have two or more from ADA and NKF also support the
It is well known that hypertension risk factors), aggressive control of target BP of <130/80 mm Hg.6,7 The
is a major risk factor of many mac- BP, as recommended by JNC-VI Hypertension Optimal Treatment
rovascular and microvascular compli- guidelines, is needed to prevent fur- trial found a 30% decrease in the
cations such as stroke, coronary heart ther deleterious effects. Two hyper- number of strokes for patients with a
disease (CHD), CHF, kidney failure, tensive subpopulations were of par- DBP of 80 versus 90 mm Hg.18 Ongo-

Am J Health-Syst Pharm—Vol 60 Mar 15, 2003 561

 REPORTS Hypertension care

Table 12.
Rate of BP Control According to JNC-VIa
No. (%) Patients
Phase 1 Phase 3
All Pts. Diabetic Pts. All Pts. Diabetic Pts.
Classification (mm Hg) (n = 374) (n = 66) (n = 417) (n = 99)
Controlled (<140/90) or diabetic (<130/85)b 127 (34.0) 17 (25.8) 175 (42.0) 22 (22.2)
Optimal (<120/80) 32 (8.6) 9 (13.6) 45 (10.8) 10 (10.1)
Other (120–129/80–85) 50 (13.4) 8 (12.1) 63 (15.1) 12 (12.1)
Borderline (130–139/85–89) 45 (12.0) 0 (0) 67 (16.1) 0 (0)
Not controlled (>140/90) or diabetic (>130/85) 247 (66.0) 49 (74.2) 242 (58.0) 77 (77.8)
Borderline (130–139/85–89) 12 (3.2) 12 (18.2) 30 (7.2) 30 (30.3)
Stage 1 (140–159/90–99) 153 (40.9) 24 (36.4) 144 (34.5) 32 (32.3)
Stage 2 (160–179/100–109) 55 (14.7) 8 (12.1) 51 (12.2) 13 (13.1)
Stage 3 (>180/110) 22 (5.9) 5 (7.6) 17 (4.1) 2 (2.0)
Undocumentedc 5 (1.3) 0 (0) 0 (0) 0 (0)
a
By chart review. JNC-VI = sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.
b
Includes patients in optimal group and patients with borderline high hypertension.
c
HEDIS guidelines specify classification of “not controlled” for undocumented blood pressure.

ing efforts at SWHP focus on con- gory (130–139/85–89 mm Hg) in such as asthma. In patients with CHF,
trolling BP and encouraging appro- phase 3 (30.3% versus 18.2% in ACE inhibitors alone or in conjunc-
priate pharmacotherapy in diabetic phase 1) may indicate possible im- tion with digoxin or diuretics are rec-
hypertensive patients. provement. Similarly, the percentage ommended. ARBs should be used in
JNC-VI guidelines recommend of patients with higher BP showed a patients who cannot tolerate ACE in-
ACE inhibitors, α 1-adrenergic- corresponding decrease (from 36.4% hibitors. ACC–AHA 2001 guidelines
receptor blockers, CCBs, or diuretics to 32.3% for stage 1 hypertension recommend the use of β-blockers in
in diabetic hypertensive patients. and from 7.6% to 2% for stage 3 hy- combination with ACE inhibitors,
ADA guidelines recommend ACE in- pertension) (Table 12). diuretics, and digoxin in CHF pa-
hibitors or ARBs as first-line therapy; The other population targeted by tients. In our study, a slight decrease
β-blockers or diuretics are also ac- JNC-VI was patients with ISH. From (1.4%) in the percentage of CHF pa-
ceptable therapies. Nondihydropyri- NHANES III, the BP of the majority tients receiving an ACE inhibitor or
dine CCBs may be used in patients of patients with ISH was not ade- ARB was observed. In hypertensive
who cannot tolerate ACE inhibitors quately controlled with current patients with nephropathy, for which
or ARBs. Practice patterns at SWHP pharmacotherapy.3,4 For older pa- ADA and NKF guidelines recom-
indicated an ongoing need for fur- tients, ISH poses a greater risk for a mend the use of ACE inhibitors or
ther emphasis on more aggressive BP cardiovascular event than does ele- ARBs as first-line therapy, there was
control and on the use of ACE inhib- vated DBP.23 SHEP revealed a 13% a 2.3% decrease in patients using an
itors or ARBs in diabetic hyperten- reduction in the risk of mortality and ACE inhibitor or an ARB. However,
sive patients. Despite a 9% increase a 33% reduction in the risk of stroke, overall use of ACE inhibitors or
in appropriate ACE inhibitor or ARB MI, or death caused by CHD in eld- ARBs in the nephropathy subpopu-
use in diabetic hypertensive patients, erly patients receiving antihyperten- lation was already at 79% during the
BP control actually decreased from sive therapies.12 While the prevalence baseline phase. Continued interven-
25.8% in phase 1 to 22.2% in phase 3 of ISH in the SWHP population de- tions are underway to further im-
(p = 0.6005) (Table 9). In phase 1, creased between phase 1 and phase 3, prove BP control in these patient
4337 patients were identified as hav- there remains room for improvement. populations with the appropriate use
ing both hypertension and diabetes, In hypertensive patients with other of antihypertensive therapy. JNC-VI
compared with 5542 patients in comorbid disease states, adherence to guidelines for compelling indications
phase 3. The increase of 1205 diabet- JNC-VI guidelines was mixed. In pa- continue to be emphasized in region-
ic hypertensive patients (28%) may tients with a history of MI, a β-blocker al clinic meetings and health care
reflect newly diagnosed diabetics. should be prescribed. Results from this provider newsletters.
These patients may not have benefit- study found an 8.2% increase in the This hypertension quality im-
ed from the study interventions or appropriate use of β-blockers. Some provement project was the largest
did not have enough time to control patients with a prior MI were not ide- disease management initiative in
their BP. The higher percentage of al candidates for β-blocker therapy which the department of pharmacy
patients in the borderline-high cate- because of coexisting conditions, has been involved. The high frequen-

562 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003

 REPORTS Hypertension care

cy of hypertension complicates the staff members, (3) enhancing the However, because these products ac-
logistics of data retrieval, report prepa- perception of the scope of pharmacy counted for such a small percentage
ration, and information management. practice by collaborations with other of the overall prescription volume
Relative to diabetes and CHF, the departments and medical staff, (4) (5.6% of 37,839 claims), this method
number of hypertensive patients at increasing the potential for pharma- of therapy classification did not sig-
SWHP is 6-fold and 12-fold greater, cy department involvement in simi- nificantly affect the overall results.
respectively. In addition, SWHP lar programs, and (5) identifying re-
physicians tended to underestimate lated areas for future projects (e.g., Conclusion
the importance of treating hyperten- diabetic hypertensive patients, pa- Through a hypertension quality
sion to the target goal. In this regard, tients with ISH). improvement initiative, BP control
there was some difficulty in encour- increased from a baseline of 37.2% to
aging physicians to consider modify- Limitations 49.2% at follow-up. Continued ef-
ing their clinical practice patterns in There were several limitations in forts to improve hypertension man-
hypertension management. Of the this study. It was assumed that the agement, particularly in patients
interventions conducted, the most information obtained from medical with concomitant diabetes and in
effective were the individual physi- and pharmacy claims databases was elderly patients with ISH, are needed.
cian profiling reports. Many medical correct and valid. While precautions
staff members took this information were taken in identifying hyperten- References
and reviewed their patients’ medica- sive patients on the basis of medical 1. Levy D, Larson MG, Vasan RS et al. The
progression from hypertension to con-
tion therapy relative to JNC-VI and pharmacy claims, some patients gestive heart failure. JAMA. 1996;
guidelines. The repeated educational taking antihypertensive medications 275:1557-62.
efforts at regional clinic meetings for conditions other than hyperten- 2. 2001 Heart and stroke statistical update.
Dallas, TX: American Heart Association;
also seemed to have a positive influ- sion may have been included in the 2000.
ence on increasing awareness and study population. In contrast, hyper- 3. Burt VL, Whelton P, Roccella EJ et al.
attentiveness to the medical staff’s tensive patients may have been ex- Prevalence of hypertension in the US
adult population. Results from the Third
management of their hypertensive pa- cluded if they received an antihyper- National Health and Nutrition Examina-
tients. Aggressive BP treatment goals tensive drug from a source that was tion Survey, 1988–1991. Hypertension.
were emphasized throughout the ed- not captured in pharmacy claims 1995; 25:305-13.
4. The sixth report of the Joint National
ucational interventions, but the most data (e.g., drug samples or self-paid Committee on Prevention, Detection,
effective element of this quality im- prescription medications). Evaluation, and Treatment of High Blood
provement program was the increased Patients also may have been ex- Pressure. Arch Intern Med. 1997; 157:
2413-46.
attention to and measurement of cluded if hypertension was not prop- 5. 1999 World Health Organization–
providers’ management of hyperten- erly documented or not appropriate- International Society of Hypertension
sive patients in an effort to adhere to ly treated, as no ICD-9-CM code guidelines for the management of hyper-
tension. J Hypertens. 1999; 17:151-83.
HEDIS and National Committee for would appear in the medical claims 6. American Diabetes Association. Treat-
Quality Assurance measures. data and no prescription claims ment of hypertension in adults with dia-
The size of this large-scale disease would be found. A small number of betes. Diabetes Care. 2002; 25(suppl 1):
S71-3.
management initiative made it diffi- patients may have received medical 7. Bakris GL, Williams M, Dworkin L et al.
cult to implement patient-based in- care from non-SWHP facilities where Preserving renal function in adults with
terventions. A patient-focused medi- out-of-area claims were not processed. hypertension and diabetes: a consensus
approach. Am J Kidney Dis. 2000; 36:646-
cation compliance intervention There is also the possibility that pa- 61.
could have greatly enhanced this tients controlling their BP through 8. Hunt SA, Baker DW, Chin MH et al.
project, but our program was time- diet and lifestyle modifications alone ACC/AHA guidelines for the evaluation
and management of chronic heart failure
consuming and labor-intensive and were excluded. These patients could in the adult: executive summary. A report
could not support a patient-level ini- have been identified only if there was of the American College of Cardiology/
tiative. The pharmacy department an ICD-9-CM code documenting a American Heart Association Task Force
on Practice Guidelines (Committee
significantly benefited from serving a diagnosis of hypertension. to Revise the 1995 Guidelines for the
central role in this project by (1) in- Another limitation of the phar- Evaluation and Management of Heart
creasing the level of expertise in hy- macy claims data was in the analysis Failure). J Am Coll Cardiol. 2001; 38:
2101-13.
pertension and its comorbidities of drug therapy usage (Table 4). 9. Godley P, Pham H, Rohack J et al. Op-
among participating pharmacists, (2) Fixed-dose combination products portunities for improving the quality of
improving collaborative relation- were identified by a single pharmacy hypertension care in a managed care set-
ting. Am J Health-Syst Pharm. 2001;
ships between pharmacy staff and claim and were analyzed as mono- 58:1728-33.
SWHP’s QI, IS, and administrative therapy rather than dual therapy. 10. The state of managed care quality 2000.

Am J Health-Syst Pharm—Vol 60 Mar 15, 2003 563

 REPORTS Hypertension care

Washington, DC: National Committee 15. Ruilope LM, Coca A. The role of combi- for 316,099 white men. Arch Intern Med.
for Quality Assurance; 2000. nation therapy in the treatment of hyper- 1992; 152:56-64.
11. Yokoyama KK, Doan QD, Godley PJ et al. tension. Blood Press Suppl. 1998; 1:22-6. 20. UK Prospective Diabetes Study Group.
Effect of physician profiles and academic 16. Hilleman DE, Ryschon KL, Mohiuddin Tight blood pressure control and risk of
detailing on cost and utilization of selec- SM et al. Fixed-dose combination vs macrovascular and microvascular com-
tive serotonin reuptake inhibitors. J Man- monotherapy in hypertension: a meta- plications in type 2 diabetes: UKPDS 38.
age Care Pharm. 2002; 8:23-31. analysis evaluation. J Hum Hypertens. BMJ. 1998; 317:703-13.
12. SHEP Cooperative Research Group. Pre- 1999; 13:477-83. 21. George PB, Tobin KJ, Corpus RA et al.
vention of stroke by antihypertensive 17. Greenberg RN. Overview of patient com- Treatment of cardiac risk factors in dia-
drug treatment in older persons with iso- pliance with medication dosing: a litera- betic patients: how well do we follow the
lated systolic hypertension. Final results ture review. Clin Ther. 1984; 6:592-9. guidelines? Am Heart J. 2001; 142:857-
of the Systolic Hypertension in the Elder- 18. Hansson L, Zanchetti A, Carruthers SG et 63.
ly Program (SHEP). JAMA. 1991; 265: al. Effects of intensive blood-pressure 22. The Hypertension in Diabetes Study
3255-64. lowering and low-dose aspirin in patients Group. Hypertension in Diabetes Study
13. Neutel JM. Low-dose antihypertensive with hypertension: principal results of the (HDS): II. Increased risk of cardiovascu-
combination therapy: its rationale and Hypertension Optimal Treatment (HOT) lar complications in hypertensive type 2
role in cardiovascular risk management. randomised trial. Lancet. 1998; 351:1755- diabetic patients. J Hypertens. 1993;
Am J Hypertens. 1999; 12:73S-9S. 62. 11:319-25.
14. Epstein M, Bakris G. Newer approaches 19. Neaton JD, Wentworth D. Serum choles- 23. Stamler J, Stamler R, Neaton JD. Blood
to antihypertensive therapy. Use of fixed- terol, blood pressure, cigarette smoking, pressure, systolic and diastolic, and car-
dose combination therapy. Arch Intern and death from coronary heart disease. diovascular risks. US population data.
Med. 1996; 156:1969-78. Overall findings and differences by age Arch Intern Med. 1993; 153:598-615.

564 Am J Health-Syst Pharm—Vol 60 Mar 15, 2003