Clinical Research/Clinical Trials in Otology: Setting the Research Agenda for Development of an Intervention
May 9–10, 2007
Bethesda Marriott Suites, Bethesda, Maryland
The NIDCD sponsored a workshop to bring together leaders in Otology and clinical trials to focus on what it would take to develop interventions in otology. The goal was to encourage cross and multidisciplinary science towards the development or establishment of evidence-based treatments for otologic conditions/disorders, and to determine, on the basis of available data regarding epidemiology and stage of science, a prioritization of otologic conditions for which intervention-oriented research may be ready. Eight speakers from various academic, industry, and federal organizations were invited to present: an introduction and overview of the available databases for epidemiologic data, the realities of what it would take to develop an intervention, whether pharmacologic or behavioral, and clinical trial methodologies specific to surgical trials and real world practical trials.
Participants were organized into three sections: Middle Ear, Inner Ear, and Vestibular. Within each group, there was a discussion of which otologic conditions were of public health and clinical significance. These discussions are captured in three Worksheets. (Worksheets for each area are available here: Middle Ear, Inner Ear, and Vestibular.) There was an initial attempt to clarify the state of evidence-based treatments available in each of these areas. If only minimal evidence was available, participants identified the stage of science regarding epidemiology, level of fundamental understanding, therapeutic targets, and preclinical research currently available. The intent was to identify future areas of research and areas poised for clinical trials research.
Brief overviews of the speaker presentations are presented below.
Ling Chin, M.D., M.P.H.
NIDCD Translational Research Branch
Using author Thomas L. Friedman's book The World Is Flat as a jumping-off point, Dr. Chin described for the group the need to move away from a vertical way of thinking for conducting research, such as through individual departments and laboratories, toward a more horizontal thinking approach. This approach would include incorporating multiple disciplines, team science, interconnected networks, and so on. According to Dr. Chin, clinical research teams of the future will likely include people who traditionally have not all been at the same table at one time, such as basic scientists and clinicians, health policy decision makers and third-party payers, community liaisons and practitioners. She outlined the need for three teams of people in the area of clinical trials: scientific/technical individuals to develop a protocol, an operational team to carry out the protocol (including data managers, study coordinators, regulatory compliance officers, and monitors), and a dissemination team to translate research findings into widespread practices, and treatment guidelines.
Development/Refingement of Interventions
Development of Intervention/Treatments: Pharmacotherapy
Paul Glue, M.D., Ph.D.
Dr. Glue discussed the process required for the development of new drugs, which could also be applied to most medical devices and behavioral treatments. The development process is slow (10-15 years), costly (greater than $1 billion for each approval), and involves an enormous amount of attrition due to very high failure rates at all stages of development. Stages include:
- Discovery, where a disease and molecular target are identified
- Exploratory/development — phase I and phase II testing
- Full development — phase III testing
- Registration with a regulatory agency
A random approach to testing approved drugs for new indications should also be based on some understanding of the applicability of the approved drug for an unlabeled indication, in order to minimize the inefficient expenditure of time, effort, and financial resources; thus the standard of the phase II/III learn-relearn cycle is still necessary (e.g., therapeutic target, dose-response, dose regimen).
In conclusion, Dr. Glue suggested that although translating preclinical knowledge into treatment is challenging, there are well-defined regulatory and statistical procedures for doing so, which require a multidisciplinary team approach.
Behavioral Intervention Development: A Stage Model Approach
Michael Pantalon, Ph.D.
Yale University School of Medicine
Dr. Pantalon discussed how to use the stage model approach for developing behavioral interventions for certain medical disorders. He discussed the accepted scientific standards for testing behavioral interventions and, using tinnitus as an example, walked participants through the process of developing a behavioral intervention. This included defining the target problem (e.g., tinnitus), the target behavior the intervention strives to change, and the target mechanism of action, be it a new intervention or an adaptation of an older one.
Dr. Pantalon then described steps in the NIH Stage Model for developing behavioral interventions:
- Stage one (early development): such as development of the intervention and early testing – for example, this is where tinnitus retraining therapy current stands
- Stage two (next steps for ninterventions that prove promising): such as larger randomized control trials
- Stage three (effectiveness studies): such as testing interventions in community settings with community providers
Dr. Pantalon stressed the importance of detailing a behavioral intervention in a manual to ensure consistency among providers and to enable researchers studying the intervention's effectiveness to measure how precisely providers follow the procedure (preservation of intervention fidelity).
The Need for Expertise-Based Randomized Controlled Trials
P.J. Devereaux, M.D., Ph.D.
Dr. Devereaux discussed the benefits of conducting expertise-based randomized controlled trials (RCTs) for surgical interventions, as opposed to conventional RCTs. In conventional RCTs, not only are the participants randomized but the doctors and the treatment they administer are randomized. In expertise-based RCTs, the participants are randomized, but the doctors are selected for their specific expertise to provide the treatment under study. The benefits of expertise-based RCTs include the reduction of expertise bias, the reduction of ethical concerns (since all patients will receive treatment from a doctor with expertise in the procedure they receive), and a decreased probability that procedures will be switched midway through the protocol. Applicability and transferability are increased as a greater number of centers and surgeons are included in the trial.
Practical Clinical Trials—Evidence of Outcomes of STAR*D (Sequenced Treatment Alternatives to Relieve Depression)
Madhukar Trivedi, M.D.
University of Texas Southwestern Medical School
Dr. Trivedi shared results from the NIMH-funded study known as the Sequenced Treatment Alternatives to Relieve Depression (STAR*D), a randomized, multi-center clinical trial to study the treatment of non-psychotic major depressive disorder (MDD). The study explored how to achieve long-term remission for patients, with the understanding that some patients respond to one treatment, while others may not. The study offered a sequential paradigm if one treatment was not beneficial to them. The first level of treatment was the antidepressant citalopram, a selective serotonin reuptake inhibitor (SSRI). Participants who did not respond well to citalopram could move to the next level of treatment, and so on, up to four levels. The researchers found that after two treatment steps, 55 percent of patients were in full remission. After the second step, however, remission was low.
Bringing Otologic Devices to the U.S. Market
Srinivas Nandkumar, Ph.D.
FDA Center for Devices and Radiological Health (CDRH), Office of Device Evaluation
Dr. Nandkumar described the process by which medical devices can move from the idea phase to the U.S. market, and how the FDA ensures the safety and effectiveness of those devices. He described the three classifications of medical devices:
- Class I, which are of simple design and low risk (includes hearing aids and otoscopes)
- Class II, which are more complex and higher risk (includes tinnitus maskers, audiometers, and bone conduction hearing aids)
- Class III, which are the most complex and the highest risk (includes cochlear implants, implantable middle ear hearing devices, and auditory brainstem implants).
Depending on the classification of the device, a regulatory submission may be required. These include the Pre-market Notification, or 510(k), which is required for Class II and non-exempt Class I devices and which must demonstrate that the device is equivalent in safety and effectiveness to a pre-existing device, and the more stringent Pre-market Application, or PMA, which is required for Class III devices, and which must show that the device is reasonably safe and effective for its intended use. Dr. Nandkumar also advised that researchers review any FDA guidance that may be available on the device as well as standards on the device published by professional organizations and recognized by the FDA, all of which are available on the CDRH Website. He also offered guidance for the design of an appropriate clinical study, including setting a clear objective, enrolling the appropriate study population, and defining unambiguous success and failure criteria.
National Center for Complementary and Alternative Medicines (NCCAM)
Partap Khalsa, DC, Ph.D., DABCO
National Center for Complementary and Alternative Medicine (NCCAM)
Dr. Khalsa described the mission and history of this NIH institute and how it might pertain to research in otology. He explained that NCCAM's mission is somewhat different from other NIH institutes, because it focuses on treatments as opposed to diseases. The treatments can be broken down into:
- Biologically based systems, such as vitamins and minerals, herbals, etc. (the study of molecules)
- Mind-body medicine, such as meditation and yoga, Tai Chi, behavioral therapies, etc.
- Manipulative and body-based systems, such as chiropractic, massage, etc.
- Energy therapies, such as Qi Gong and electromagnetic fields
- Whole medical systems, such as homeopathy and acupuncture
According to Dr. Khalsa, researchers should first demonstrate evidence of efficacy of a CAM intervention --through early phase I or IIa clinical trials, substantial case reports in the literature, or long history of use-- and then proceed to a small Phase II trial using the R21 mechanism (appropriately sized, well-designed, dose/response clinical trial). He also noted that dose response is one of the most critical aspects of early trials and that the R21 or R01 (or U01) grant mechanisms are used at NCCAM for these kinds of trials.
In Support of Clinical Research Networks
David Witsell, M.D.
Duke Clinical Research Institute
Dr. Witsell spoke about the need to develop clinical research networks as a means of addressing the many challenges involved in conducting multi-center clinical trials as well as enabling the translation of research findings into practice. He stressed the importance of the data coordinating center as part of each network to provide oversight and to ensure that many complexities—such as legal and regulatory issues-- are centrally addressed. Such coordination would enable the collaboration of centers on many types of multiple clinical research studies. He noted that both physician and network objectives should be considered, whereby physicians are principally concerned about improving patient outcome, and network objectives are largely focused on building collaboration and disseminating findings to the medical community. Dr. Witsell discussed Duke Clinical Research Institute's recently awarded NIDCD research grant for a practice-based research network, called Creating Healthcare Excellence through Education and Research (CHEER), under an R21/R33 funding mechanism.
Data Sources and Availability
Medical Visits for Ear Disorders: Data from the NAMCS and NHAMCS
Susan Schappert, M.A.
Centers for Disease Control and Prevention (CDC) National Center for Health Statistics
Ms. Schappert discussed findings from two national surveys: the National Ambulatory Medical Care Survey (NAMCS), which documents patient visits to non-federal doctors' offices, and the National Hospital Ambulatory Medical Care Survey (NHAMCS), which documents patient visits to emergency and outpatient departments in non-federal hospitals. Data from 2003-2004 show that Otitis Media and other eustachian tube disorders were one of the more frequently diagnosed medical problems in doctors' office visits, accounting for roughly 14.5 million visits. Otitis Media and eustachian tube disorders accounted for 3.5 million visits to general and family practice physicians, 7 million visits to pediatricians, and 2.6 million visits to otolaryngologists and their most common diagnosis that year.
Combining the number of ambulatory care visits to doctors' offices and emergency and outpatient departments with a primary diagnosis of ear disorders yielded these numbers: Otitis Media (18.7 million), infective Otitis Externa (2.7 million), impacted cerumen (2.6 million), otalgia (1.6 million), hearing loss (1.5 million), vertigo (1 million), tinnitus (600,000), otorrhea (287,000), and all other ear disorders (1.6 million). The total visits for ear disorders that year were approximately 30.5 million.
The surveys' scope, methodology, and availability for public use were discussed, and the link to the website is: www.cdc.gov/nchs/r&d/rdc.htm
Demand for Otology Procedures: Selected Results from the AAO-HNS 2003 Workforce Study
Susan Holzer, M.A.
American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS)
Ms. Holzer discussed findings of the AAO-HN'S 2003 Workforce Study that pertained to the topic of otology. The study analyzed 2001 data from the MEDSTAT MarketScan database and the Centers for Medicare & Medicaid Services (CMS) Medicare Utilization File to determine patient demand for otolaryngology procedures around the country and the role of otolaryngologists in relation to other specialists in conducting these procedures. MEDSTAT data were subdivided into fee for services (FFS) and managed care procedures. The study compared the number of claims filed by otolaryngologists to the total number of claims filed. High among claims for MEDSTAT (FFS, managed care) as well as Medicare were myringotomy/tympanostomy and impacted cerumen removal. The study then looked at individual otological procedures and compared the numbers performed by otolaryngologists to those performed by family practice physicians, pediatricians, dermatologists, plastic surgeons, and other specialists.
NIDCD Epidemiologic Data Re: Clinical Research in Otology
Howard Hoffman, M.A.
NIDCD Epidemiology and Biostatistics Program
Mr. Hoffman discussed three national population-based resources that provided epidemiological data on hearing loss and otology. These included the following:
- National Health Interview Survey (NHIS), an annual questionnaire that is completed by an adult in roughly 33,000 households (it also has a child component)
- National Health and Nutrition Examination Survey (NHANES), a much smaller annual survey (2000 households) that incorporates a hearing test
- Early Childhood Longitudinal Study—Birth Cohort (ECLS-B), a long-term study that follows children from the time they are infants until they are five years old, and incorporates a household interview with hearing testing.
The number of questions in the NHIS hearing module has varied over the years, from 9 questions in 1999, 6 questions in 2002 (as part of the Healthy People program), to 54 questions in 2007. Likewise, the focus of NHANES hearing exams and questions has varied over the years, with 1988–1994 data focusing on children ages 6–19, 1999–2004 data focusing on adults ages 20–69, and 2005–2006 data focusing on adults over 70 years of age. The 2005–2008 data will be focusing on adolescents and noise-induced hearing loss.
Information from NHANES can provide national estimates and age-specific prevalence of hearing loss, tinnitus, hearing exams, use of hearing protection, use of hearing aids, as well as co-morbidities and risk factors. NIDCD researchers have combined NHANES data with data from the community-based Beaver Dam Study to determine distribution and prevalence of hearing loss by age, employing a pure tone average of thresholds at various frequencies to define the term "hearing loss". Balance questions have also been incorporated into the NHIS questionnaire and NHANES 1999–2004 and those data are available to researchers. Other NIDCD–supported epidemiological studies include the hearing component of the Jackson Mississippi Heart Study, the hearing component of the Hispanic Community Health Study, and the NIH Neuroepidemiology Toolbox(part of the Blueprint for Neuroscience Research), among others.
Link to the Recent Past
NIDCD: Past, Current Leading to the Future
Amy Donahue, Ph.D.
NIDCD Hearing and Balance/Vestibular Branch
Dr. Donahue discussed NIDCD's recent efforts to expand clinical and translational research programs in hearing and balance. In April 2004, NIDCD sponsored a workshop in which NIDCD-funded researchers helped identify opportunities for and barriers to increased translational research - the transfer of scientific advances to patients. Four categories within NIDCD's research portfolio were identified: molecular diagnostic/therapeutics (i.e., drug and vaccine development); bridging basic science to clinical science, clinical studies, and introduction and emergence into clinical practice. Outcomes of that workshop included the creation of a translational research grant initiative involving R01s and R21s as well as a special announcement for Core Center Grants in translational research, the development of a translational research branch within NIDCD, and a follow-up workshop on clinical research, which took place in March 2005. The clinical research workshop brought in speakers from outside of NIH and explored initiatives that NIDCD can take to help foster clinical research. Participants delved into such topics as multi-site and single-site clinical trials, phase I, II, and III clinical trials, preliminary studies, enabling platforms, such as measurement tools and assays, and infrastructure needs. Outcomes of this workshop included the development of a developmental grant for patient-oriented research to support the creation of partnerships among researchers and interested organizations, and an R21/R33 phased infrastructure grant for patient-oriented research, to support the development of data collection and management tools that could assist in clinical research. In addition, NIDCD issued two program announcements in 2006--an R01 and an R21--requesting more patient-oriented research, both of which are active until 2009. In the coming years, the translational and the clinical research initiatives will be evaluated to revisit workshop recommendations and to determine how well current efforts are addressing those goals.