Author + information
- Arthur J. Moss, MD∗ (, )
- Executive Committees of the Multicenter Research Group (MRG),
- Multicenter Automatic Defibrillator Implantation Trial (MADIT)
- ↵∗Address correspondence to:
Dr. Arthur J. Moss, University of Rochester Medical Center, 265 Crittenden Boulevard, CU 420653, Rochester, New York 14642-0653.
The Heart Research Follow-up Program at the University of Rochester Medical Center has an ongoing record of successfully carrying out large-scale clinical cardiovascular studies and randomized clinical trials for >35 years, and nearly all of the original executive committee members continue to be actively involved and still serve in leadership positions in the studies. Three large-scale clinical studies were published by the Multicenter Research Group (MRG): the Multicenter Post-infarction Program in 1983 (1), the diltiazem trial in 1988 (2), and risk factors for myocardial ischemia in 1993 (3). Four MADIT (Multicenter Automatic Defibrillator Implantation Trials) have been published between 1996 and 2012 (4–7), and in 2016 we are starting our fifth MADIT study, MADIT-Diabetes, using a subcutaneous implantable cardioverter-defibrillator.
Before embarking on each major clinical trial, we defined the questions that the study was intended to answer. In testing a new therapy, we utilized randomized, placebo-controlled studies. We always conducted these studies at multiple sites, with a single central coordinating center managing the data.
Our studies consistently used time-to-event outcomes. The simplest of these is death, but unless such studies are very large, there will rarely be enough deaths to provide meaningful comparisons between treatment groups. We generally used time to death or time to a cardiac event such as heart failure, whichever occurred first. This approach increased the number of events for statistical analysis. However, it ignored information on repeated events or on deaths that may follow a nonfatal event.
In designing our clinical trials, we pre-specified the number of patients needed to enroll to provide a definitive answer to the scientific question at issue. For a study to be considered “positive” in terms of providing evidence to assist in securing approval of a new treatment by the U.S. Food and Drug Administration, convincing evidence of effectiveness requires a 2-sided alpha p value ≤0.05, or equivalently that a 95% confidence interval for the true parameter value being estimated (treatment difference, risk difference, or hazard ratio) not include the null value of the parameter.
The MADIT investigations were conducted using fully sequential studies of the effectiveness of implantable defibrillators with truncated triangular designs, which allow very early stopping in case of an adverse effect or strongly positive effect of treatment. More extensive follow-up was used to establish moderately sized beneficial effects.
The data safety monitoring board (DSMB) was responsible for the safety of study subjects and for the efficacy of study procedures, including the timely and reliable acquisition and management of study information. Thus, the DSMB protected the interests of trial participants but also the interests of the sponsor and the larger community in ensuring a scientifically valid outcome. The perspective of the DSMB was unique because it alone had access to all facts pertaining to the study as they accrued throughout the course of the trial.
DSMB members were intellectually and financially independent of trial investigators. The study sponsor provided them with adequate reimbursement for their services. This arrangement, which guaranteed data confidentiality and lack of conflict of interest, was stipulated in a contract between members and the sponsoring institution that was put in force before the trial began. In addition to reimbursement and expenses, DSMB members received full liability insurance for any action they took while serving in this capacity.
The principal investigator had the ultimate responsibility of overseeing all aspects of the trial. The principal investigator was responsible for ensuring that the appropriate university infrastructure for the trial was provided by the sponsoring institution. The University of Rochester was the sponsoring academic institution for all of the trials (1–7). This included the institutional review board and the research grants and contract office. The institutional review board ensured protection of study subjects and compliance with all institutional, state, and federal regulations regarding human research. The research grant and contracts office had responsibility for negotiating and overseeing contractual and financial aspects of the study.
Contractual arrangements with enrolling centers within and outside of the United States were challenging, as unique institutional requirements frequently required negotiations on a center-by-center basis. Centers participating in the trial needed the essential personnel, including a center principal investigator and a study coordinator. Each enrollment center had to demonstrate the capacity to oversee enrollment, patient safety, and accurate data collection.
We maximized the clinical studies in the following ways:
1. Our adult cardiovascular populations were age 21 to >80 years. We included as large a proportion of females as possible as well as those with relevant comorbidities, various races, and a spectrum of socioeconomic factors.
2. Enrollment and exclusion criteria were clearly described in writing.
3. Before the onset of the study or trial, each enrolling center provided evidence that it had the ability to provide the required patient enrollment and follow-up.
4. Copies of original source documents were retained, and these were signed and dated by a member of the enrolling site team. All source documentation that was provided to the data management center had personal health identifiers redacted and labeled with the subject study identifiers.
5. We established a potential adjustment for extending the trial (decision by DSMB) if enrollment and/or the number of primary events was less than anticipated over time.
6. We generally utilized a fixed budget, but in certain situations, a supplementary budget was negotiated with the National Institutes of Health or industry—once again, this was handled by the DSMB because they were the only ones who knew all the enrollment data, efficacy and safety endpoints, and approximately when the trial would be completed.
7. A specific publication policy was established before the study or trial ended identifying the authors for the primary and the high-level secondary analyses (mostly the executive committee members). Subsequent tertiary substudies were performed by enrolling center cardiologists and their cardiology fellows, with numerous publications derived with this approach.
8. Each clinical trial was ended by the DSMB when 1 of the pre-specified efficacy or safety endpoints was met. On occasion, we extended further analysis of a study for several years after the primary endpoint was met when trying to evaluate a mortality endpoint over a more extended period of time.
In summary, our cardiovascular studies and trials were eminently successful in advancing the science of medical therapeutics. The principal investigator was the same person for 35 years, and this individual effectively involved all of the executive committee members in the important research decisions. We also established effective academic-industrial collaboration with investigator-driven studies that included unimpeded access to all data by the executive committee at the conclusion of each study. We effectively collaborated with industry while still achieving scientific independence. A most important benefit and one valued by all was that this professional involvement during the 35 research years led to lifelong friendships among the investigators.
The Executive Committee members thank Seah Nisam for his advice and counsel over the years in the various clinical studies that our group carried out.
Executive Committees: Multicenter Research Group: Jesia Benhorin, MD, Thomas Bigger, MD, Poul Eric Block-Thomsen, MD, PhD, Monty Bodenheimer, MD, Mary Brown, MS, Robert Case, MD (deceased), Edward M. Dwyer, MD, Shirley Eberly, MS, Joseph Fleiss, PhD (deceased), Charles Francis, MD, John Gillespie, MD, Robert Goldstein, MD, Henry Greenberg, MD, Mark Haigney, MD, Helmut Klein, MD, Ronald Krone, MD, Valentina Kutyifa, MD, PhD, Edgar Lichstein, MD, Emanuela Locati, MD, PhD, Frank Marcus, MD, Arthur J. Moss, MD, David Oakes, PhD, Charles Odoroff, PhD (deceased), Daniel Ryan, MD, Wojciech Zareba, MD, PhD. MADIT (Multicenter Automatic Defibrillator Implantation Trials): Christopher Beck, PhD, Mary Brown, MS, David Cannom, MD, James Daubert, MD, Mark Estes, MD, Ilan Goldenberg, MD, Henry Greenberg, MD, Jack Hall, PhD (deceased), Stephen Hammes, MD, PhD, David Huang, MD, Helmut Klein, MD, Reinoud Knops, MD, PhD, Michael Kosiborod, MD, Valentina Kutyifa, MD, PhD, Arthur J. Moss, MD, David Oakes, PhD, Jeanne Poole, MD, Claudio Schuger, MD, Jagmeet Singh, MD, PhD, Scott Solomon, MD, David Wilber, MD, Wojciech Zareba, MD, PhD.
- American College of Cardiology Foundation