1. Challenges of a Large Pragmatic Trial: the ALLHAT Experience
Duke Industry Statistics Symposium
Are Pragmatic Trials Ready for Prime Time?
September 7, 2017
Challenges of a Large Pragmatic Trial: the ALLHAT Experience
Barry R. Davis, MD, PhD
University of Texas School of Public Health
Houston, TX

2. Outline Description of ALLHAT PRECIS-2 for ALLHAT Expected Challenges
Unexpected Challenges
Summary
I will tell you about the designs of ALLHAT and SHEP and some of the trials’ infrastructures, as the approaches to some of the important trial activities can vary depending on those things. In particular, I’ll talk about training and certification and IRB issues, and various things related to the ongoing conduct of the trial such as communications, endpoints, protocol and operational changes and personnel turnover.

3. Health and Human Services National Heart, Lung, and Blood Institute
U.S. Department of
Health and Human Services
National Institutes
of Health
National Heart, Lung, and Blood Institute
The Antihypertensive and Lipid- Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)
Treatment and complications among the million Americans with hypertension are estimated to cost the nation $37 billion annually, with antihypertensive drug costs alone accounting for an estimated $15.5 billion/year. There is conclusive evidence that antihypertensive drug therapy can substantially reduce the risk of hypertension-related morbidity and mortality. However, the optimal choice for initial pharmacotherapy of hypertension is uncertain.
Earlier clinical trials documented the benefit of lowering blood pressure (BP) using primarily thiazide diuretics or beta-blockers. After these studies, several newer classes of antihypertensive agents, i.e. angiotensin-converting-enzyme inhibitors (ACEIs), calcium channel blockers (CCBs), alpha-adrenergic blockers, and more recently angiotensin receptor blockers (ARBs) became available. Over the past decade, major placebo-controlled trials have documented that ACEIs and CCBs reduce cardiovascular events in hypertensive individuals. However, their relative value compared with older, less expensive agents remains unclear. There has been considerable uncertainty regarding effects of some classes of antihypertensive drugs on risk of coronary heart disease (CHD). The relative benefit of various agents in high-risk subgroups such as older, diabetic, and Black hypertensive persons also needed to be established.
JAMA. 2002;288:

4. Pragmatic Study Characteristics
Enroll broadly representative and heterogeneous patient populations
Conducted in typical clinical and health care settings
Large in size
To discern hypothesized differences in effectiveness
To evaluate differences in patient subgroups
As simple as possible in design and conduct
Maximal inclusion, minimal exclusion criteria
Minimally intrusive on routine clinical practice in terms of procedures, data collection
Examples: ALLHAT1 , MI-FREE2
1ALLHAT Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs. diuretic. JAMA 2002; 288:2981–97;
2Choudhry, N.K., et al. Full coverage for preventive medications after myocardial infarction. N Engl J Med Dec 1;365(22):2088–97;

6. Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)
Practice-based, randomized, multi-center trial
Antihypertensive component
42,418 high-risk hypertensives 55+ years
Whether newer agents reduce incidence of CHD compared to a diuretic
Blinded, no placebo

7. Plus: Lipid-lowering component -
1/4 of antihypertensive trial cohort (10,336)
moderately hypercholesterolemic patients
does reduction of LDL cholesterol reduce all-cause mortality?
unblinded (pravastatin vs “usual care”)

8. Secondary Objectives All-cause mortality Stroke Combined CHD
Combined CVD
Cancer
Major subgroups
Age 65+, women, African-Americans, diabetics

9. Randomized Design of ALLHAT
Amlodipine
Chlorthalidone
Doxazosin
Lisinopril
High-risk hypertensive patients
Consent / Randomize
Eligible for lipid-lowering
Not eligible for lipid-lowering
Here you can see a diagram of the randomization scheme for ALLHAT. High risk hypertensives were evaluated according to the eligibility criteria, and consented and randomized to one of four antihypertensive medications. Among those randomized into the hypertension part of the program, further evaluation revealed those who were additionally eligible for the cholesterol-lowering part of the trial, and those were consented and randomized to pravastatin or “usual care”. Average follow-up was about 5 years.
Consent / Randomize
Pravastatin Usual care
Follow until death or end of study (4-8 years)

10. Infrastructure
623 clinical sites in the U.S., Canada, Puerto Rico, and U.S. Virgin Islands
42,418 participants in the antihypertensive component
10,355 participants in the lipid component
Randomization began 2/94
Follow-up ended 3/02
The infrastructure of ALLHAT is critical as a background to all of the things that I’ll mention later. There were 623 clinical sited in the US, Canada, Puerto Rico and the US Virgin Islands. 42,418 participants were randomized into the antihypertensive component and 10,355 into the cholesterol-lowering component. Randomization began in February 1994 and ended 4 years later for the hypertension component in January 1998; randomization to the cholesterol component ended 4 months later. Follow-up ended in March 2002.

11. 33
CANADA
Quebec
Ontario
Washington
Oregon
California
MEXICO
Nevada
Idaho
Montana
North Dakota
South Dakota
Wyoming
Utah
Colorado
Arizona
New Mexico
Texas
Oklahoma
Kansas
Nebraska
Minnesota
Iowa
Wisconsin
Michigan
Illinois
Missouri
Indiana
Arkansas
Louisiana
Mississippi
Alabama
Georgia
Tennessee
Kentucky
Ohio WV
Virginia
N. Carolina
S. Carolina
Pennsylvania DE NJ MD
New York VT NH CT MA RI
Maine
New Brunswick
Nova Scotia
Prince Edward Island
Island of New Foundland
Puerto Rico (Area Enlarged)
St. Croix, Virgin Islands
(Area Enlarged)
The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial
Florida
-623 Clinical Centers
-USA, Canada, Caribbean
-Diverse practice settings
-Primary sponsor: NIH-NHLBI
-Concurrent support from the VA
-Assistance from pharmaceutical companies but no role in scientific conduct of trial
Program Office,
NHLBI
This map shows the ALLHAT sites in the United States, Canada, Puerto Rico, and the US Virgin Islands.
CTC, UT
The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA. 2002;288:

12. Infrastructure (continued)
9 regions
VA sites, Canadian sites
7 geographically distributed
sites per region
1-2 Regional MD’s
2-4 Regional Study Coordinators
Regional teams were, at the beginning of ALLHAT, a relatively new concept for NHLBI. Because of the many communication challenges with so many clinical sites across such a wide geographic area, the sites were separated into 9 regions to facilitate communication and monitoring. The VA sites comprised their own region regardless of geographic area, and the Canadian sites comprised their own region. The remaining sites were assigned to regions based on geographic location. Each region included between sites and, depending on the number of sites, 1 or 2 regional physicians and 2-4 Regional Study Coordinators.

13. Role of the Regional Teams
Initial regulatory paperwork
Training
Regular communication with sites
Address questions from site staff
Site monitoring visits
Steering Committee, subcommittees
Writing groups
Assist CTC with forms, MOO, procedures
Participate in planning for Investigators’ Meetings
Site recruitment
The Regional teams have a larger role in ALLHAT compared to what we usually think of as a role for a clinical trials monitor. Since this was a fairly new concept, their roles were expanded and modified over time as ALLHAT faced new challenges, and better and more efficient ways were found to work with the regions to accomplish some goal such as site recruitment,
READ LIST
You can see that this is a truly collaborative effort.

14. PRECIS-2 Study population Eligibility 5 Recruitment 4 Location
Setting 5
Organization 5
Methods
Primary Outcome 5
Follow-up 4
Primary Analysis 5
Intervention delivery
Flexibility 4
Adherence 4

16. Expected Challenges
Design and analysis
Study protocol, forms, manual of operations
Randomization for two trials
Promote study awareness among physicians & patients at national level

17. Expected Challenges (Continued)
Site recruitment and IRB approval
Acquisition and distribution of study drugs
Large meeting coordination and travel
Regional administration & communications
Financial management

18. Expected Challenges (Continued)
Minority recruitment goals
Support center selection and oversight
QC, monitoring, and reports
Assist clinical sites in recruitment and adherence
Steering Committee, Subcommittees, DSMB
Publications and presentations

19. Eligibility Criteria for Antihypertensive Trial
Inclusion
Aged > 55 years
BP eligibility -  140/90 mm Hg but  180/110 mm Hg at two visits
1+ major CVD risk factors - MI or stroke (> 6 months),history of revascularization, documented ASCVD, diabetes,, HDL < 35 mg/dL, LVH (ECG or echo), current cigarette smoking
Exclusion
MI, stroke within 6 months, Symptomatic CHF or ejection fraction < 35%, creatinine  2 mg/dL, requiring diuretics, CCB, ACEI, or alpha blockers for reasons other than hypertension
ALLHAT included men and women aged > 55 years. Blood pressure eligibility was determined at two visits, and was based on the patient’s antihypertensive treatment status at the first screening visit.
Patients with untreated systolic and/or diastolic hypertension ( 140/90 mm Hg but  180/110 mm Hg at two visits) were eligible.
Patients with treated hypertension were eligible if:
BP was ≤ 160/100 mm Hg on 1-2 antihypertensive drugs at Visit 1 and
≤ 180/110 mm Hg at Visit 2, when medication may have been partially withdrawn
No washout period was required in ALLHAT.

20. Unexpected Challenges
Recruitment
Early discontinuation of doxazosin arm
Cancellation of closeout training
Follow-up specialists
Dissemination

21. Unexpected Challenge 1 - Recruitment
Go from 270 to 400 sites (to 600)
Add cigarette smoking as risk factor
Lower age cutoff from 60 to 55
Extend recruitment period for 1 year
Increase $ for randomization
Pressel et al. Controlled Clinical Trials 22:674–686 (2001)
Wright et al. Controlled Clinical Trials 22: (2001)

22. Unexpected Challenge 1 - Recruitment
6. Add Puerto Rico
7. Add field personnel program
8. Extend recruitment period for 6 months to end of 1/98
9. Add 9th Region – Canada
10. Publicity – radio, print, direct mail
Pressel et al. Controlled Clinical Trials 22:674–686 (2001)
Wright et al. Controlled Clinical Trials 22: (2001)

23. Unexpected Challenge 2 - Early Discontinuation of Trial Arm
1/24/00 – NHLBI Director accepts recommendation to drop arm; futility of primary endpoint & statistical, clinical significance of major secondary endpoint
1/28/00 – 1st Transition Committee meeting
2/3/00 – Steering Committee and Regions informed
2/11/00 – Closeout materials finalized
Pressel et al. Controlled Clinical Trials 22:29–41 (2001)

24. Unexpected Challenge 2 - Early Discontinuation of Trial Arm
2/16-18/00 – Transition kits to sites
2/18/00 – Paper submitted to JAMA Express
3/8/00 – NHLBI press release
3/15 – ACC presentation
4/5 – Paper appears on WWW; in print on 4/19
Pressel et al. Controlled Clinical Trials 22:29–41 (2001)

25. Unexpected Challenge 3 - Cancellation of Closeout Training Meeting (9/14-15/01) for Investigators
Major investigators meeting in Kansas City to train sites on the closeout procedure
Closeout visits to start 10/1/01
Events of Sept. 11 resulted in cancellation of meeting (over 1000 people)
Crisis recovery – What do we do about training?

26. Unexpected Challenge 3 - Cancellation of Closeout Training Meeting (9/14-15/01) for Investigators
CTC, NHLBI, 9 Regions formulated plan in KC
Send out training kits, series of conference calls, site visits, local training meetings, one-on-one calls
Effort begun immediately
61% trained by 9/30/01, 97% by 10/31/01
All sites trained by November 2001

27. Unexpected Challenge 4 - Follow-up Specialists to Locate Participants
Losses to follow-up large problem in long-term practice-based large trial – many sites had little experience
Need database search and follow-up specialist
Follow-up specialist protocol
Started in 4/2000
Pressel et al. SCT 2002, 2003

28. Unexpected Challenge 4 - Follow-up Specialists to Locate Participants
One, then two, then three F/U specialists
Incorporating searches into closeout
10/1/01 – 3/29/02
Pressel et al. SCT 2002, 2003

29. Unexpected Challenge 5 - Dissemination
Dissemination of results is part of every trial
Usually just papers and presentations
Usually more done if industry-sponsored
Dissemination to be integral part of ALLHAT
Pressel et al. SCT 2005

30. Unexpected Challenge 5 - Dissemination
Planning started with Dissemination Committee
Beyond publication of primary results
Press conference, training spokespersons, major meetings, work with Publications Committee, work with NHBEP, press kits, web site, lay journals, etc.
Bartholomew et al. Clinical Trials, 2009
Stafford et al., Archives of Internal Med, 2010

31. Summary
A large pragmatic trial can present several expected and unexpected challenges
ALLHAT challenges mostly related to:
600+ practice-based sites
No specific funding except for forms completed
Competing priorities
Offset by:
Relatively simple protocol
Few forms
Few data items collected per visit
In summary, I would say that the ALLHAT challenges were mostly related to the busy practice-based setting of most sites, the lack of specific funding for staff, and competing priorities (those being the busy practice itself and other trials). However, these were somewhat offset by the relatively simple protocol, few form types, and few data items collected at each visit. These last factors were a good match for the clinics in this study. If our protocol would have been more complicated in any way, it would have been nearly impossible to accomplish successfully in busy practices with lots of other protocols with little specific study funding.

32. Lessons Learned
Experienced team is a must
Anticipate problems
Rely on others inside & outside trial
Involve the investigators and coordinators
Open communications, flexibility