Hello I need an article critique

and Brian G. Feagan Ross D. Feldman, Guang Y. Zou, Margaret K. Vandervoort, Cindy J. Wong, Sigrid A.E. Nelson

Randomized, Controlled Trial A Simplified Approach to the Treatment of Uncomplicated Hypertension: A Cluster

Print ISSN: 0194-911X. Online ISSN: 1524-4563 Copyright © 2009 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Hypertension doi: 10.1161/HYPERTENSIONAHA.108.123455 2009;53:646-653; originally published online February 23, 2009;Hypertension.

http://hyper.ahajournals.org/content/53/4/646 World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://hyper.ahajournals.org//subscriptions/

is online at: Hypertension Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information about Office. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialHypertensionin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

Clinical Trial

A Simplified Approach to the Treatment of Uncomplicated Hypertension

A Cluster Randomized, Controlled Trial

Ross D. Feldman, Guang Y. Zou, Margaret K. Vandervoort, Cindy J. Wong, Sigrid A.E. Nelson, Brian G. Feagan

Abstract—Notwithstanding the availability of antihypertensive drugs and practice guidelines, blood pressure control remains suboptimal. The complexity of current treatment guidelines may contribute to this problem. To determine whether a simplified treatment algorithm is more effective than guideline-based management, we studied 45 family practices in southwestern Ontario, Canada, using a cluster randomization trial comparing the simplified treatment algorithm with the Canadian Hypertension Education Program guidelines. The simplified treatment algorithm consisted of the following: (1) initial therapy with a low-dose angiotensin-converting enzyme inhibitor/diuretic or angiotensin receptor blocker/diuretic combination; (2) up-titration of combination therapy to the highest dose; (3) addition of a calcium channel blocker and up-titration; and (4) addition of a non—first-line antihypertensive agent. The proportion of patients treated to target blood pressure (systolic blood pressure �140 mm Hg and diastolic blood pressure �90 mm Hg for patients without diabetes mellitus or systolic blood pressure �130 mm Hg and diastolic blood pressure �80 mm Hg for diabetic patients) at 6 months was analyzed at the practice level. The proportion of patients achieving target was significantly higher in the intervention group (64.7% versus 52.7%; absolute difference: 12.0%; 95% CI: 1.5% to 22.4%; P�0.026). Multivariate analysis of patient-level data showed that assignment to the intervention arm increased the chance of reaching the target by 20% (P�0.028), when adjusted for other covariates. In conclusion, the Simplified Treatment Intervention to Control Hypertension Study indicates that a simplified antihypertensive algorithm using initial low-dose fixed-dose combination therapy is superior to guideline-based practice for the management of hypertension. (Hypertension. 2009;53:646-653.)

Key Words: hypertension management � randomized, controlled trial � fixed-dose combination therapy � cluster randomization � hypertension

Treatment of hypertension remains suboptimal despite thedevelopment of novel therapies and the widespread imple- mentation of education programs. Although multiple barriers exist to achieving better control of blood pressure, these mainly consist of patient- and practitioner-centered factors.

From the patient perspective, poor adherence to antihyper- tensive regimens is a significant component of the “treatment gap.”1–3 An important part of this problem is current prescrib- ing practices. Unfortunately, most patients require �2 med- ications to achieve optimal control,4 and these multidrug regimens are associated with lower adherence.5,6 This has been suggested to be the most important contribution to inadequate blood pressure control.7 Furthermore, switching medications, a strategy featured prominently in most national treatment guidelines, has also been linked to poor adherence.8

Thus, the use of simpler, more effective drug regimens might improve blood pressure control.

From a practitioner perspective, a number of behavioral factors have been associated with poor blood pressure con- trol. One of the most important factors is “therapeutic inertia,” whereby practitioners fail to appropriately escalate the intensity of therapy despite the presence of poorly controlled hypertension.9 We speculate that the increasingly complex treatment regimens currently advocated by experts, national guidelines, and the pharmaceutical industry might contribute to this undesirable behavior.

Given these issues, a simple, step-care– based algorithm for the pharmacological management of hypertension (Simpli- fied Treatment Intervention to Control Hypertension [STITCH]) was developed. This algorithm features the initial

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz. Received September 22, 2008; first decision October 8, 2008; revision accepted January 28, 2009. From the Robarts Clinical Trials (R.D.F., G.Y.Z., M.K.V., C.J.W., S.A.E.N., B.G.F.), Robarts Research Institute, London, Ontario, Canada; Schulich

School of Medicine and Dentistry, Department of Medicine (R.D.F., B.G.F.), Department of Physiology and Pharmacology (R.D.F.), and Department of Epidemiology and Biostatistics (G.Y.Z., B.G.F.), University of Western Ontario, London, Ontario, Canada.

This trial has been registered at www.clinicaltrials.gov (identifier NCT00129909). Correspondence to Ross D. Feldman, Robarts Research Institute, 100 Perth Dr, London, Ontario N6A 5K8, Canada. E-mail [email protected] © 2009 American Heart Association, Inc.

Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.108.123455

646 by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

use of a low dose of diuretic/angiotensin-converting enzyme inhibitor or diuretic/angiotensin receptor blocker fixed-dose combination. The rationale for this approach is that low-dose combinations of antihypertensive drugs are more effective than low-dose monotherapy and that the incidence of adverse effects with low-dose combination therapy is similar to that of placebo.10,11 Moreover, half-standard doses of diuretic/an- giotensin converting enzyme inhibitor, diuretic/angiotensin receptor blocker combinations are less likely to disturb glucose or potassium homeostasis.10 From the patient per- spective, use of fixed-dose combinations may improve adher- ence.12 From the physician perspective, this approach reduces the need for switching drugs, and treatment can be easily escalated. These properties may reduce therapeutic inertia and increase blood pressure control rates. However, the use of polytherapy with regard to antihypertensive efficiency has been questioned.13

Based on these considerations, we conducted a cluster- randomized, controlled trial that compared a simplified algo- rithm for the treatment of hypertension with conventional guideline-based care.

Methods Study Practices and Randomization Procedure Our hypothesis was that use of a simplified algorithm for the treatment of hypertension would result in better practice-level blood pressure control than conventional guideline-based care. The goal was to determine the effectiveness of the treatment algorithm within the family practice setting. Because this intervention was directed at the family practitioner and it would not be implementable at an

individual patient level, cluster design was necessarily adopted. In addition, this cluster approach minimizes the risk of treatment contamination at the patient level. Therefore practitioners, rather than patients, were randomly assigned in a 1:1 ratio to the STITCH- care treatment algorithm or to continue with usual management according to Canadian Hypertension Education Program (CHEP) guidelines. The randomization schedule was computer generated and stratified by the year of graduation of the family physician (before 1984 or 1984 and later). Because of the nature of the intervention, participants were aware of their treatment assignment.

The study was conducted at 45 (single-physician) family practices in southwestern Ontario, Canada, between February 2005 and January 2007. Initially 950 practices were approached, and 93 physicians who expressed an interest in participation were randomly assigned to either STITCH care or to care based on the CHEP recommendations (guideline care). To avoid contamination, separate protocols were sent to practitioners assigned to STITCH care and guideline care. From these practices, eligibility to continue was based on practitioner ability to identify �50 patients with uncon- trolled hypertension through chart review. Thirty five of the original 93 practices did not identify a sufficient number of patients and were not further evaluated. The 58 practices able to identify 50 potential subjects (and willing to continue) received their assigned interven- tion. Those practices that were able to identify and enroll �30 patients with confirmed, uncontrolled hypertension were analyzed. Because the study was analyzed at the cluster level, this was predetermined as the minimum number of subjects required to give a precise estimate of practice-level hypertension control. Participant flow is depicted in Figure 1.

Patients At each practice, data were collected from all of the eligible patients who were willing to participate and give written informed consent to a maximum of 50 patients. Patients were enrolled consecutively as they arrived to their physician’s practice. Eligible patients were men

3 Practices Declined Participation

93 Randomized

47 Assigned to Receive STITCH-care 25 Completed Screening

10 Declined participation 2 Withdrew 10 Ineligible (unable to identify 50 patients)

46 Assigned to Receive GUIDELINE-care 33 Completed Screening

2 Declined participation 1 Withdrew 10 Ineligible (unable to identify 50 patients)

1 Withdrew after enrollment

96 Practices Assessed for Eligibility

30 Practices enrolled patients 3 Practices ineligible (unable to enrol 30 patients) 27 Practices analyzed

24 Practices enrolled patients 6 Practices ineligible (unable to enrol 30 patients) 18 Practices analyzed

3 Withdrew after enrollment

SCREENING

INTERVENTION

ANALYSIS

Figure 1. Consort diagram of STITCH study.

Feldman et al Fixed-Dose Combination Therapy for Hypertension 647

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

and women with uncontrolled hypertension who were �18 years of age. Both untreated and partially treated patients were eligible. Uncontrolled hypertension was defined as systolic blood pressure �140 mm Hg or diastolic blood pressure �90 mm Hg for patients without diabetes mellitus or systolic blood pressure �130 mm Hg or diastolic blood pressure �80 mm Hg for patients with diabetes mellitus. Patients with ischemic heart disease, atrial fibrillation, peripheral vascular disease, stroke, stage 4 or 5 chronic kidney disease (ie, those with an estimated glomerular filtration rate of �30 mL/min per 1.73 m2) or those participating in other hypertension studies were not eligible.

For those partially treated patients enrolled in STITCH-care practices, physicians were not encouraged to discontinue the pa- tient’s baseline medications but were encouraged to make subse-

quent drug additions using the STITCH algorithm, focusing on the use of fixed-dose combinations. Patients were not excluded for reported intolerance to any individual drug recommended in the STITCH algorithm.

Interventions The STITCH algorithm consisted of 4 steps: (1) initial therapy with a half tablet of a low-dose angiotensin-converting enzyme inhibitor/ diuretic or angiotensin receptor blocker/diuretic combination; (2) up-titration of combination therapy successively to the highest dose supplied; (3) addition of a calcium channel blocker and up-titration; and (4) addition of one of the non–first-line antihypertensive agents. Practitioners assigned to the STITCH-care group were educated on

Initial therapy with a low dose ACE/diuretic or

ARB/diuretic combination

Continue with current therapy

Continue with current therapy

Up-titration of

combination therapy successively to the

highest dose

IS BLOOD PRESSURE CONTROLLED?

Yes

Yes

No

Continue with current therapy

Add calcium

channel blocker and up-titrate

No

Add an α-blocker, ß-blocker or

spironolactone

oN seY

Figure 2. STITCH-care algorithm.

Table 1. Recommended STITCH-Care Fixed-Dose Combinations

Variable Starting Dose Tablet* Initial Dose Equivalent† Maximum Dose Tablet‡

Angiotensin-converting enzyme inhibitor/diuretic

Accuretic, Pfizer 10/12.5 mg quinapril HCT 0.50 20/25 mg

Coversyl Plus, Servier 4/1.25 mg perindopril erbumine-indapamide 0.75 4/1.25 mg

Inhibace Plus, Roche 5/12.5 mg cilazapril HCT 0.75 5/12.5 mg

Zestoretic, AstraZeneca 10/12.5 mg lisinopril-HCT 0.50 20/25 mg

Prinzide, Merck Frosst 10/12.5 mg lisinopril-HCT 0.50 20/25 mg

Vaseretic, Merck Frosst 5/12.5 mg enalapril HCT 0.50 10/25 mg

Angiotensin II receptor blocker/diuretic

Atacand Plus, AstraZeneca 16/12.5 mg candesartan HCT 0.75 16/12.5 mg

Avalide, Sanofi-Aventis 150/12.5 mg irbesartan-HCT 0.75 300/25 mg

Diovan-HCT, Novartis 80/12.5 mg valsartan-HCT 0.50 160/25 mg

Hyzaar, Merck Frosst 50/12.5 mg losartan HCT 0.75 100/25 mg

Micardis Plus, Boehringer Ingelheim 80/12.5 mg telmisartan-HCT 0.75 80/12.5 mg

HCT indicates hydrochlorothiazide. *STITCH-care initial dose is half of the supplied starting dose tablet. †This was a dose equivalent of a half tablet of the starting dose tablet relative to a defined “standard daily dose,” as published previously.15

‡The maximum dose was supplied as a single tablet.

648 Hypertension April 2009

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

the use of the treatment algorithm by an implementation team consisting of the principal investigator and study personnel from the coordinating center. Simplified aids to facilitate implementation were supplied (including treatment algorithm cards and a listing of applicable low-dose, fixed-dose combinations (Figure 2 and Table 1). The principal investigator (R.D.F.) followed up with each physician by telephone on 1 occasion, 4 to 6 weeks after the initiation visit. Practitioners assigned to the guideline care group were not exposed to the STITCH algorithm. At these practices, the CHEP guidelines (that were current at the time of the study) were

reviewed with the physician, and they were provided educational materials for the management of hypertension.

The CHEP guidelines are similar in breadth and content to the Seventh Joint National Committee on Prevention, Detection, Evalu- ation, and Treatment of High Blood Pressure recommendations.14 However, in the CHEP guidelines current at the time of the study, fixed-dose combination drugs were not recommended as first-line treatment even for those with stage 2 hypertension (as they are in the Seventh Joint National Committee on Prevention, Detection, Evalu- ation, and Treatment of High Blood Pressure and in the 2008 CHEP recommendations).

For both treatment arms, the principal investigator was available to the family practitioners for consultation. Our intent was to provide equal access to expert advice and educational material to both treatment groups.

Outcome Measures The primary outcome measure was the proportion of patients achieving target blood pressure at the practice level, as determined at a 6-month follow-up assessment. A successful treatment was defined as achieving a systolic blood pressure �140 mm Hg and diastolic blood pressure �90 mm Hg for patients without diabetes mellitus or systolic blood pressure �130 mm Hg and diastolic blood pressure �80 mm Hg for patients with diabetes mellitus. Secondary efficacy outcomes were the change in systolic blood pressure and diastolic blood pressure. Blood pressure was measured based on the average of 5 readings, determined using a standardized and precalibrated automated blood pressure device (BpTRU, VSM Med Tech). Practices were educated in the use of the machine before initiation of the study, and detailed instructions for taking the average blood pressure measure- ment were attached to each machine to ensure consistent measurement across practices. All of the antihypertensive medications received by the patient during the study period were recorded on drug logs.

Assessment of total daily antihypertensive doses at the 6 month visit was based on review of drug logs. Standard dose units for hypertension treatment were assessed using our previously published approach and were generally defined as twice the recommended starting dose.15 For example, a 12.5-mg dose of hydrochlorothiazide was assessed as 0.5 standard units and a 25-mg dose assessed as 1.0 standard unit.

Table 2. Baseline Characteristics of the Practices and Patients

Practice/Patient Characteristics Guideline Care STITCH Care

Practices 27 18

Median cluster size (range) 46 (25 to 50) 46 (26 to 50)

Proportion of physicians graduated before 1984, n (%)

11 (40.7) 8 (44.4)

Mean recruitment duration by cluster, d

140 169

Urban location, n (%) 25 (92.6) 17 (94.4)

Men, n (%) 20 (74.1) 15 (83.3)

Patients

No. of patients 1246 802

Recruitment duration, d 503 531

Age, mean (range), y 60.9 (18.6 to 93.0) 61.9 (20.4 to 92.8)

Women, % 53 56

Diabetic, % 15.9 15.1

Baseline SBP, mean (SD), mm Hg

153.4 (14.9) 155.1 (13.7)

Baseline DBP, mean (SD), mm Hg

87.7 (10.9) 88.1 (10.9)

SBP indicates systolic blood pressure; DBP, diastolic blood pressure.

Table 3. Outcomes by Treatment Group

Variable

Practice ID

1 2 3 4 5 6 7 8 9 10 11 12 13

Guideline care

Cluster size 25 46 50 26 48 50 48 48 33 50 48 50 49

At target BP, % 0.04 0.57 0.62 0.46 0.65 0.80 0.42 0.10 0.64 0.66 0.65 0.56 0.37

Baseline SBP, mean 164.08 153.43 154.78 161.96 149.48 150.48 153.65 157.42 155.94 152.98 148.02 149.70 157.59

SBP mean 148.12 134.39 134.90 144.00 132.23 128.04 140.46 150.52 129.21 134.28 133.88 135.58 139.88

SBP difference �15.96 �19.04 �19.88 �17.96 �17.25 �22.44 �13.19 �6.90 �26.73 �18.70 �14.15 �14.12 �17.71

Baseline DBP, mean 93.84 85.39 89.38 88.31 85.38 87.18 84.48 98.54 87.24 87.12 83.15 83.88 87.33

DBP mean 84.92 77.07 79.22 79.38 85.69 75.30 78.04 92.83 75.30 79.66 80.38 77.28 78.35

DBP difference �8.92 �8.33 �10.16 �8.92 0.31 �11.88 �6.44 �5.71 �11.94 �7.46 �2.77 �6.60 �8.98

STITCH care

Cluster size 45 50 50 33 48 50 49 38 49 42 49 46 50

At target BP, % 0.78 0.44 0.56 0.52 0.42 0.88 0.57 0.66 0.92 0.67 0.63 0.65 0.68

Baseline SBP, mean 152.16 158.64 157.60 159.42 155.40 153.16 151.53 155.26 158.69 160.88 157.73 150.17 155.12

SBP mean 129.00 141.62 133.98 138.79 141.71 126.76 133.73 129.13 127.31 133.48 134.76 132.20 128.82

SBP difference �23.16 �17.02 �23.62 �20.64 �13.69 �26.40 �17.80 �26.13 �31.39 �27.40 �22.98 �17.98 �26.30

Baseline DBP, mean 82.09 88.80 91.76 92.39 89.06 92.38 88.00 84.95 92.57 89.07 88.02 83.15 89.22

DBP mean 73.82 88.02 80.06 81.48 82.48 80.00 80.76 72.37 71.49 81.05 79.31 74.37 76.64

DBP difference �8.27 �8.78 �11.70 �10.91 �6.58 �12.38 �7.24 �12.58 �21.08 �8.02 �8.71 �8.78 �12.58

Feldman et al Fixed-Dose Combination Therapy for Hypertension 649

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

Practitioner Satisfaction Survey On the completion of the study, participating physicians were sent a questionnaire assessing their satisfaction with their approach to the management of hypertension. The questionnaire was composed of the following 4 questions: (1) how effective is your treatment approach in managing your patients with hypertension?; (2) how feasible do you think it is to sustain the treatment approach in your practice setting?; (3) how likely would you be to recommend your treatment approach to a colleague?; and (4) overall, how satisfied are you with your approach to hypertension management? The questions were scored on a Likert scale, where 0 indicated the most negative response and 10 indicated the most positive response.

Statistical Methods Descriptive statistics were used to assess the demographic charac- teristics of the patients and physicians. Because treatment strategies to improve the management of hypertension were perceived and implemented at the level of the practice, statistical inferences for the primary outcome were based on the proportion of patients achieving target blood pressure at the level of the practice rather than at the patient level. The primary analysis compared the proportion of patients achieving target blood pressure between the 2 treatment groups using a 2-sample t test.16 A similar approach was used to compare the other secondary outcomes. The primary outcome was also analyzed using a weighted t test.17

We also evaluated patient-level data. Factors associated with successful treatment to target blood pressure were evaluated using a modified Poisson regression approach that yields risk ratios adjusted for clustering.18,19 Provider satisfaction data were analyzed using a 2-sample t test. All of the analyses were based on the intention-to- treat principle. Statistical tests were 2 sided and were performed at the 0.05 level of significance.

Sample Size Based on data obtained from a preliminary study of the effects of home blood pressure monitoring on blood pressure control,15 the interpractice SD of the proportion of patients achieving target blood pressure was estimated to be 17%. Therefore, randomization of 23 practices per group provided 80% power to detect a difference of

15% in the proportion of patients achieving target blood pressure between the treatment groups.

Ethics Approval and Informed Consent The institutional review board at the University of Western Ontario approved the protocol. All of the patients gave written informed consent.

Results A total of 45 practices were successful in identifying �50 patients with uncontrolled hypertension and enrolling �30 patients with confirmed uncontrolled hypertension. Of these, 18 practices had been assigned to STITCH care and 27 had been assigned to guideline care. The physician and patient demographics were similar in the 2 treatment groups (Table 2).

A total of 2111 patients from these 45 practices were evaluated. One patient recruited to STITCH care and 6 patients from the guideline-care arm were ineligible because their base- line blood pressure was already at target. The analyses were, thus, based on data collected from 2104 patients with uncon- trolled hypertension, of whom 49% were already being treated at the time of their baseline assessment. Through the duration of the study, 23 patients in the STITCH-care arm and 33 patients in the guideline-care arm withdrew from the study.

Proportion of Patients Achieving Target Blood Pressure Patients treated at practices that were randomized to STITCH care were more likely to reach target blood pressure (Table 3). At practices assigned to STITCH care, the mean success rate of achieving target blood pressure at 6 months was 64.7% compared with 52.7% for those assigned to guideline care (between-group difference: 12.1%; 95% CI: 1.5% to 22.4%; P�0.026). An analysis that weighted the t statistic for the

Table 3. Outcomes by Treatment Group (Continued)

Practice ID

14 15 16 17 18 19 20 21 22 23 24 25 26 27

50 50 47 50 50 47 50 44 50 50 50 44 44 49

0.60 0.44 0.66 0.72 0.80 0.45 0.38 0.55 0.52 0.76 0.46 0.57 0.57 0.24

148.96 158.27 159.34 160.48 152.52 158.21 153.70 149.98 152.24 140.58 152.12 156.98 146.52 151.37

134.72 136.78 131.96 131.98 128.12 140.77 136.90 136.86 136.00 125.98 137.38 134.39 135.64 142.04

�14.24 �21.14 �27.38 �28.50 �24.40 �17.45 �16.80 �13.11 �16.24 �14.60 �14.74 �22.59 �10.89 �9.33

82.22 88.60 92.85 91.14 83.54 85.02 91.66 87.41 84.74 86.76 89.64 81.82 95.75 89.24

74.00 80.42 83.64 79.12 72.90 76.98 84.12 81.00 76.96 73.98 81.58 74.61 81.09 85.24

�8.22 �8.18 �9.21 �12.02 �10.64 �8.04 �7.54 �6.41 �7.78 �12.78 �8.06 �7.20 �14.66 �4.00

50 43 26 49 35

0.50 0.63 0.85 0.61 0.69

151.24 147.79 156.19 153.63 160.63

130.92 130.05 124.81 134.14 131.60

�20.51 �17.74 �31.38 �19.49 �29.03

88.53 89.53 81.12 86.14 84.14

78.62 80.37 68.46 78.10 73.63

�9.81 �9.16 �12.65 �8.04 �10.51

650 Hypertension April 2009

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

number of patients in a practice also showed significant and similar results to the unweighted analysis.

Changes in Systolic and Diastolic Blood Pressures After 6 months of treatment, practices assigned to STITCH care had a mean (� SD) reduction of systolic blood pressure of 22.6�4.9 mm Hg compared with 17.5�5.2 mm Hg for practices assigned to guideline care (absolute difference: 5.2 mm Hg; 95% CI: 2.0 to 8.5 mm Hg; P�0.002; Table 3). The STITCH-care practices also demonstrated a greater reduction in mean diastolic blood pressure than those as- signed to guideline care (10.4�3.3 versus 8.2�3.1 mm Hg; absolute difference: 2.2 mm Hg; 95% CI: 0.2 to 4.2 mm Hg; P�0.03). Seven serious adverse events were reported (3 in guideline care and 4 in STITCH care), including 3 deaths (all in the guideline-care group) and 3 episodes of drug intoler- ance (all in the STITCH-care group).

Exposure to Antihypertensive Drugs Drug logs were used to assess adherence with the treatment algorithms. In the guideline-care group, 15% of patients were exposed to fixed-dose combination drugs. In contrast, in the STITCH-care group, exposure to fixed-dose combination drugs was significantly higher (85% versus 15%; P�0.001). Paralleling these findings, the number of individual drugs

prescribed (with a fixed-dose combination counting as 2 individual drugs) was significantly higher in the STITCH- care practices compared with those assigned to guideline care (2.3 versus 1.9 drugs; absolute difference: 0.4 drugs; 95% CI: 0.2 to 0.6; P�0.001). These findings are consistent with the greater use of fixed-dose combinations in patients assigned to the STITCH-care arm (see Table 4).

Although more aggressive dosing might have been facili- tated by the STITCH algorithm, the number of standard doses prescribed was similar in the 2 groups. In fact, the number of standard doses of antihypertensive drug prescribed tended to be higher in the guideline care arm: (1.7 versus 2.0 standard doses; absolute difference: �0.3; 95% CI: �0.5 to 0.01; P�0.06). However, it should be noted that the proportion of patients who were up-titrated during the study was higher in the STITCH-care group (82.6% versus 69.6%; between- group difference: 13.0%; 95% CI: 3.8% to 22.3%; P�0.007).

Predictors of Achieving Blood Pressure Targets As depicted in Table 5, patient gender, patient age group, and physician graduation before 1984 were not predictors of blood pressure control at 6 months. The multivariate model identified 2 factors that were independently associated with successful treatment to target blood pressure. Assignment to the STITCH-care arm increased the probability of achieving this

Table 4. Antihypertensive Medication Use at Baseline and Month 6

Hypertensive Medications

Baseline, % 6 Mo, %

STITCH Care (n�802) Guideline Care (n�1246) STITCH Care (n�802) Guideline Care (n�1246)

�-Adrenergic antagonists 0.1 1.2 0.6 1.5

�- and �-Adrenergic antagonist 0.1 0.2 0.2 0.6

Angiotensin II receptor antagonists 6.4 8.3 2.0 18.3

Angiotensin-converting enzyme inhibitors

13.8 20.8 5.0 32.6

�-Adrenergic antagonists 9.6 11.2 12.2 19.9

Calcium channel blockers 13.8 16.9 29.6 30.3

Centrally acting antiadrenergic agents 0 0.1 0.2 0.3

Diuretics 18.7 21.7 9.1 36.9

Angiotensin II receptor antagonist/diuretic combinations

7.5 7.6 43.5 14.4

Angiotensin-converting enzyme inhibitor/diuretic combinations

3.7 1.5 39.3 2.0

�-Adrenergic antagonist/diuretic combination

0 0.2 0 0.4

Table 5. Predictors of Achieving Blood Pressure Target

Determinant

Univariate Analyses Multivariate Model

Risk Ratio 95% CI P*† Risk Ratio 95% CI P*†

STITCH-care 1.2 1.02 to 1.40 0.03 1.2 1.0 to 1.4 0.03

Age (per 10 year increase) 1.0 0.97 to 1.06 0.48

Female 1.0 0.86 to 1.03 0.22

Not diabetic 2.5 2.02 to 3.05 �0.001 2.5 2.0 to 3.1 �0.001

Physician before 1984 1.1 0.88 to 1.26 0.63

*The analysis was conducted using a modified Poisson regression model that evaluated patient-level data. †P values were derived by adjustment for clustering in the model.

Feldman et al Fixed-Dose Combination Therapy for Hypertension 651

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

goal by 20% (risk ratio: 1.20; 95% CI: 1.02 to 1.40; P�0.03). Patients without diabetes mellitus had more than double the probability of reaching their target blood pressure independent of the treatment assignment. This is consistent with previous evidence of lower rates of blood pressure control in patients with diabetes mellitus (primarily ascribed to the lower targets for this subgroup of hypertensive patients).20

Provider Satisfaction The physicians in the STITCH-care arm were somewhat more satisfied with their approach to managing hypertension than physicians in the guideline-care arm (8.9 versus 8.2; absolute difference: 0.7; 95% CI: 0.4 to 1.4; P�0.04). The physicians in the STITCH-care arm also tended to feel that they were more effective in managing patients with hyper- tension (8.70 versus 8.00; absolute difference: 0.70; 95% CI: 0.02 to 1.40; P�0.06) and were more willing to recommend their approach to other physicians (8.70 versus 7.70; absolute difference: 1.00; 95% CI: 0.02 to 2.10; P�0.05). There was no difference between the treatment group arms in their perspective of the feasibility of sustaining their current treatment approach (8.20 versus 7.90; absolute difference: 0.30; 95% CI: 0.20 to 1.10; P�0.14).

Discussion Identification of new strategies for improving blood pressure control in a family practice setting is a healthcare priority. Our study demonstrates that a simplified algorithm for the management of hypertension is implementable, changes physician-prescribing patterns, and results in better blood pressure control than conventional guideline-based care.

Notwithstanding the competing influences of both pharma- ceutical company marketing and the educational initiatives of the CHEP, our simple implementation process, consisting of an initial meeting and single follow-up telephone call by a clinical pharmacologist, changed physician behavior. This behavioral change is probably best demonstrated by the �6-fold increase in the use of fixed-dose antihypertensive combination drugs by STITCH-care physicians in compari- son with the guideline-based practices. However, most im- portantly, blood pressure was improved in practices assigned to the STITCH-care arm. Furthermore, STITCH-care physi- cians reported greater satisfaction with their management of hypertension, suggesting that STITCH care is a practicable strategy for use by community physicians.

The STITCH algorithm featured the initial use of combi- nation therapy. It is notable that several previous studies have suggested that initial therapy with specific fixed-dose combina- tions results in superior blood pressure reduction in comparison with monotherapy-based approaches. However, these studies used mandated regimens in rigidly conducted clinical trials and relatively high initial drug doses.21,22 In contrast, our trial used a “critical pathway” approach to educating physicians regarding a simplified treatment algorithm that was not dependent on the prescription of specific drug combinations. Accordingly, we believe that these findings are more widely generalizable for the management of hypertension.

Which components of STITCH were responsible for the effectiveness of the algorithm? Although this question was

not directly addressed in the study, 3 factors should be considered. First, use of the algorithm might have facilitated the use of more intensive dosing regimens by physicians (ie, overcoming therapeutic inertia). Second, use of fixed-dose combinations may have improved adherence by patients, as has been suggested in observation studies. Finally, use of combination therapy may be intrinsically more effective because of synergy between agents. However, with respect to the first possibility, our analysis of dose intensity showed that practices assigned to STITCH care did not receive more aggressive treatment, as assessed by the number of standard doses prescribed. On the other hand, patients in the STITCH- care arm were more likely to have had their antihypertensive dose up-titrated. Therefore, whether the STITCH algorithm attenuated therapeutic inertia cannot be resolved. Thus, we speculate that the effectiveness of STITCH is more likely related to better adherence by patients and/or the use of more effective drug combinations (ie, minimizing the use of less-than-additive 2- or 3-drug combinations, a phenomenon suggested to occur with polytherapy in hypertension).13

The 2008 CHEP guidelines included a new recommendation supporting the use of initial therapy with a fixed-dose combina- tion as a first-line choice. Notably, based on the 2008 CHEP recommendations, 45% of STITCH-care and 39% of guideline- care patients would have been candidates for initial fixed-dose combination therapy at baseline. However, it is important to emphasize that starting with a fixed-dose combination is only 1 of 6 first-line recommended choices for initiation of therapy in this group. We would suggest that, based on our findings, a fixed-dose combination might be the preferred initial treatment in this subgroup (versus initiation with a single agent).

Limitations Our study has some limitations. The STITCH algorithm was implemented by a single clinical pharmacologist (R.D.F.). Whether similar results could be obtained with widespread implementation is unknown. Second, the imbalance between numbers of practices initiated in STITCH-care versus guideline-care arms requires consideration. This imbalance raises the possibility that the physicians who participated in the STITCH-care arm may have been different, in terms of their behavior, from those assigned to the GUIDELINE-care arm. However, the majority of the practice “dropouts” (35 of 48 practices) occurred before site initiation (ie, before the point at which intervention sites were introduced to the treatment algorithm). Among practices that actively started screening, dropout rates were comparable. Second, we could not identify any practice or patient level factor that would support the existence of a difference between treatment arms. Third, the treatment effect size was relatively large. We do not believe it is plausible that confounding would result in a difference of this magnitude. Moreover, the trial was de- signed with a 1:1 ratio between STITCH care and guideline care, which, in general, has higher power than an unbalanced trial. The unbalanced ratio of intervention:control sites in STITCH would have been expected to reduce the power of the study to identify significant differences. This implies that our results are likely conservative.

652 Hypertension April 2009

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from

Perspectives Use of a simplified algorithm for the treatment of hypertension featuring the initial use of low-dose, fixed-drug combination drugs is implementable, accepted by family physicians, and results in improved blood pressure control. This return to step care may be an important way forward for the treatment of hypertension. We also speculate that such an approach may be broadly applicable for the treatment of other chronic diseases.

Acknowledgments Coordinating center personnel include the following individuals: M. Abboud, B. Annunziello, T. Clayton, B. Feagan, R. Feldman, L. Jensen, E. Karugu, H. Kaufman, C. Matthews, J. MacKay, K. Greenwood, S. Nelson, N. Ng, W. Piper, L. Robinson, B. Sarazin, C. Trinidad, M. Vandervoort, C. Wong, J. Xie, and G. Y. Zou.

The following family practitioners participated in the study: Ancaster: B. B. Loewith; Brantford: W. S. Tsuchida, D. C. Vincent; Burlington: R. R. Jhirad, A. Nayar; Chatham: J. Boekhoud, S. Munro; Exeter: D. Hodder; Hamilton: R. D. Arora, T. Subramanian, L. S. Zavodni, Ilder- ton: G. Nancekievill; Kitchener: K. Bedrosian, B. Huth, J. A. King, S. MacGregor, R. Nicholson, T. Szozda; London: T. Alam, E. Arcia- Bravo, E. Armogan, V. Chawla, G. Debosz-Szczerbowski, P. Dickie, P. Dzongowski, J. Horne, S. Joshi, R. Komar, A. Mawji, S. Petrlich, A. Richard, F. Wong; St Mary’s: S. Hiscock; St Thomas: B. Hertwig; Stratford: D. Crowley; Strathroy: M. Mithoowani; Thedford: L. Gibbs; Waterloo: F. Dawood, G. Pannozzo, J. Peet; Windsor: M. Bilibajkich, D. Bridgeo, M. Zajner; and Woodstock: J. Brioux.

We acknowledge the contribution of Suzanne Vandervoort, the study’s initial coordinator. We thank Beverley Jasevicius for expert secretarial assistance.

Sources of Funding The trial was supported by an independent grant from Pfizer Canada.

The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation review; or approval of the article.

Disclosures R.D.F. reports receiving research funding from Pfizer and has been a paid consultant or on advisory boards for Bayer, Boehringer- Ingleheim, Pfizer, Bristol Myers Squibb, Sanofi-Aventis, Schering- Plough, Merck, and Servier. B.G.F. reports receiving research funding unrelated to the topic of the article from Synta, Millennium, Schering Canada, Celltech, Centocor, Elan/Biogen, Berlex, Ortho- Biotech, Protein Design Labs, ISIS, Santarus, Schering Plough, Celgene, UCB Pharma, Napo Pharma, BMS, Abbott, and Otsuka. He also reports receiving consulting and lecture fees from UCB Pharma, Schering Canada, Proctor and Gamble, Elan/Biogen, Millennium, Protein Design Labs, Berlex, AstraZeneca, Celgene, Abbott, Santa- rus, GeneLogic Inc, Cerimon Pharmaceuticals, Tioga Pharmaceuti- cals, BMS, ISIS, Serono, Teva, Genentech, and CombinatoRx. No other financial disclosures were reported.

References 1. Schroeder K, Fahey TK, Ebrahim S. How can we improve adherence to

blood pressure–lowering medication in ambulatory care? Systematic review of randomized control trials. Arch Intern Med. 2004;164:722–732.

2. Mitka M. Research probes details of poor adherence in antihypertensive drug therapy. JAMA. 2007;298:2128.

3. Perreault S, Lamarre D, Blais L, Dragomir A, Berbiche D, Lalonde L, Laurier C, St-Maurice F, Collin J. Persistence with treatment in newly

treated middle-aged patients with essential hypertension. Ann Pharma- cother. 2005;39:1401–1408.

4. Cushman WC, Ford CE, Cutler JA, Margolis KL, Davis BR, Grimm RH, Black HR, Hamilton BP, Holland J, Nwachuku C, Papademetriou V, Probstfield J, Wright JT Jr, Alderman MH, Weiss RJ, Piller L, Bet- tencourt, J, Walsh SM; ALLHAT Collaborative Research Group. Success and predictors of blood pressure control in diverse North Am settings: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). J Clin Hypertens. 2002;4:393– 404.

5. Claxton AJ, Cramer J, Pierce C. A systematic review of the associations between dose regimens and medication compliance. Clin Ther. 2001;23:1296 –1310.

6. Domino FJ. Improving adherence to treatment for hypertension. Am Fam Physician. 2005;71:2089 –2090.

7. Rose AJ, Berlowitz DR, Orner MB, Kressin NR. Understanding uncon- trolled hypertension: is it the patient or the provider? J Clin Hypertens (Greenwich). 2007;9:937–943.

8. Caro JJ, Speckman JL, Salas M, Raggio G, Jackson JD. Effect of initial drug choice on persistence with antihypertensive therapy: the importance of actual practice data. CMAJ. 1999;160:41– 46.

9. Okonofua EC, Simpson KN, Jesri A, Rehman SU, Durkalski VL, Egan BM. Therapeutic inertia is an impediment to achieving the Healthy People 2010 blood pressure control goals. Hypertension. 2006;47:345–351.

10. Law MR, Wald NJ, Morris JK, Jordan RE. Value of low dose combi- nation treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ. 2003;326:1427.

11. Nash DT. Rationale for combination therapy in hypertension man- agement: focus on angiotensin receptor blockers and thiazide diuretics. South Med J. 2007;100:386 –392.

12. Taylor AA, Shoheiber O. Adherence to antihypertensive therapy with fixed-dose amlodipine besylate/benazepril HCI versus comparable component-based therapy. Congest Heart Fail. 2003;9:324 –332.

13. Campbell NR, McAlister FA, Duong-Hua M, Tu K. Polytherapy with two or more antihypertensive drugs to lower blood pressure in elderly Ontarians. Room for improvement. Can J Cardiol. 2007;23:783–787.

14. Khan NA, Hemmelgarn B, Padwal R, Larochelle P, Mahon JL, Lewanczuk RZ McAlister FA, Rabkin SW, Hill MD, Feldman RD, Schiffrin EL, Campbell NR, Logan AG, Arnold M, Moe G, Campbell TS, Milot A, Stone JA, Jones C, Leiter LA, Ogilvie RI, Herman RJ, Hamet P, Fodor G, Carruthers G, Culleton B, Burns KD, Ruzicka M, deChamplain J, Pylypchuck G, Gledhill N, Petrella R, Bouldanger JM, Trudeau L, Hegele RA, Woo V, McFarlane P, Touzy RM, Tobe SW; Canadian Hypertension Education Program. The 2007 Canadian Hypertension Edu- cation Program recommendations for the management of hypertension: part 2 - therapy. Can J Cardiol. 2007;23:539 –550.

15. Zarnke KB, Feagan BG, Mahon JL, Feldman RD. A randomized study comparing a patient-directed hypertension management strategy with usual office-based care. Am J Hypertens. 1997;10:58 – 67.

16. Donner A, Klar N. Statistical considerations in the design and analysis of community intervention trials. J Clin Epidemiol. 1996;49:435– 439.

17. Bland JM, Kerry SM. Statistical notes: weighted comparisons of means. BMJ. 1998;316:129.

18. Zou GY. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706.

19. Zou GY. One relative risk versus two odds ratios: implications for meta-analyses involving paired and unpaired binary data. Clin Trials. 2007;4:25–31.

20. Joffres MR, Hamet P, MacLean DR, L’italien GJ, Fodor G. Distribution of blood pressure and hypertension in Canada and the United States. Am J Hypertens. 2001;14:1099 –1105.

21. Mourad JJ, Waeber B, Zannad F, Laville M, Duru G, Andréjak M, Investigators of the STRATHE Trial. Comparison of different therapeutic strategies in hypertension: a low-dose combination of perindopril- indapamide versus a sequential monotherapy or a stepped-care approach. J Hypertens. 2004;22:2379 –2386.

22. Lacourcière Y, Poirier L, Lefebvre J. Expedited blood pressure control with initial angiotensin II antagonist/diuretic therapy compared with stepped-care therapy in patients with ambulatory systolic hypertension. Can J Cardiol. 2007;23:377–382.

Feldman et al Fixed-Dose Combination Therapy for Hypertension 653

by guest on September 24, 2014http://hyper.ahajournals.org/Downloaded from