|Home | About | Journals | Submit | Contact Us | Français|
To describe and communicate data collected in the CHEER Infrastructure proof of concept study to facilitate understanding of the potential capabilities of practice-based research networks, and to present pilot data for development of future research initiatives.
Prospective observational study of CHEER infrastructure operational capacity using a convenience sample of all patients presenting to the practices with tinnitus, dizziness, or a combination of these symptoms.
The CHEER Network of community and academic practice sites.
The data collection exercise collected demographic, clinical, treatment, and health-related quality of life surveys on tinnitus, dizziness, and migraine disorders. Descriptive analysis of the data is presented.
Of the sites in the CHEER network, 73% (16/22) successfully enrolled subjects; a total of 1532 patients were enrolled in 8 months. Tinnitus alone, dizziness alone, and both occurred in 28%, 34%, and 29%, respectively. Patients complaining of tinnitus and dizziness had lower quality of life than those sufferers with one disorder. Migraine was associated with 27% of patients. The most frequent diagnoses for patients with tinnitus and dizziness were Ménière’s disease (34%), vertiginous migraine (18%), and benign paroxysmal positional vertigo (16%).
Descriptive data on patients with common disorders can be rapidly collected within the framework of a practice-based research network. Our data provide valuable pilot information on the targeted disorders, providing a baseline for development of future epidemiological data and clinical trials.
CHEER (Creating Healthcare Excellence through Education and Research), a practice-based research network focused on hearing and communication disorders, is based on a collaborative model of academic and community partnerships in otolaryngology. The research backbone of the network is tripartite. The Duke Clinical Research Institute (DCRI) provides clinical research infrastructure including network development and education, data management, and regulatory support. CHEER accesses thought leadership and ties to regional community sites through established collaborations with four academic medical center co-principal investigators. The third component is the American Academy of Otolaryngology, Head and Neck Surgery Foundation (AAO-HNSF) that provides physicians and ancillary health care providers a contextual reference for network involvement within the specialty.
Evaluation of the functionality and efficiency of the CHEER network infrastructure was accomplished through the Otology Data Collection (ODC) project through which we collected data on patients presenting with tinnitus and/or dizziness. The project was designed to evaluate processes essential to data collection in a multicenter study as many of the participating sites are community sites and new to research. These common otologic complaints were selected for the study since they fall within the mission of the National Institute on Deafness and Other Communication Disorders, have disease-specific measures, and evidence-based treatment literature is limited. A second goal of the project was to develop databases to further characterize the symptoms of patients presenting with tinnitus and dizziness in order to inform future research in these areas.
Evaluation and treatment of tinnitus and dizziness contribute significant costs to the health care system and are inadequately understood due to the heterogeneity of causes and clinical presentations. According to the American Tinnitus Association, 50 million Americans experience tinnitus and 12 million experience symptoms that are severe enough to seek medical attention. Two million patients are so seriously debilitated that they cannot function on a normal day-to-day basis. The financial impact of tinnitus is significant. Disability awards were received by 289,159 veterans for their tinnitus totaling $345,495,552 in annual compensation (2004).1 This is just one example of the cost to society, and does not reflect other quality of life impacts. Dizziness and vertigo are frequent complaints of patients seen in otolaryngology offices, and tools for evaluation and treatment are often inadequate. In studies by Neuhauser et al, dizziness and vertigo present a 1-year prevalence of 20% and a 1-year incidence of 3% in a general population.2,3 The impact of dizziness on cost and quality of life is considerable.4 One-fourth of patients with chronic dizziness reduce work hours, change occupations, or stop working as a result of their dizziness, and half indicate a substantial reduction in their efficiency at work.4
Two prior reports on the CHEER network discuss in more detail its development, structure, and progress.5,6 This third report on the CHEER network summarizes the data collected through the ODC Project. Methodologically, the structure of the study was driven by the project’s primary intent of testing the network’s ability to deploy a project. In this report, we describe the data that were the byproduct of the research methodology, providing an innovative and potentially compelling picture of the future of comparative effectiveness research, particularly where precise disease definitions are still under debate.
In January 2010, two CHEER multidisciplinary expert panels were convened to identify evidence gaps and research priorities within tinnitus and dizziness. The ODC Project was proposed by the CHEER Hub PIs and was vetted further at these meetings, accessing multidisciplinary input for consensus on the final set of study questionnaires and surveys. The resulting survey set included patient surveys including, the Tinnitus Handicap Inventory (THI),7 the Dizziness Handicap Inventory (DHI),8 and the Migraine Assessment Tool,9 and background questions and demographics. The physician portion included associated clinical information and diagnosis.
To protect patient privacy, study surveys were given a unique identifier number and stripped of patient identifiers. All research data were stored in a central database, accessible to all participating sites for online entry in REDCap, a secure research data system. The Duke University institutional review board (IRB) was approved for exemption and extended to CHEER community sites. CHEER sites participating in the study requiring an independent IRB review (n=5) were provided administrative support.
A convenience sample of patients presenting for visits for either tinnitus and/or dizziness participated in the project. No compensation was provided to the patient or the participating CHEER sites. Eligible patients had to be ≥18 years of age; literate and able to read and speak English, or able to participate with the aid of a medical interpreter; possess sufficient mental capacity to comply with study requirements; and present to clinic with symptoms of tinnitus/dizziness.
The project was launched in February 2010 with rolling activation of sites (timing based on their readiness given IRB needs and/or internal process setup) beginning in March 2010 and continued through November 2010. Data were extracted from the REDCap database and analyzed using summary SAS reports (SAS Institute, Inc., Cary, NC).
Of the active trained sites in the CHEER network, 16 of 22 or 73%, across 13 states participated in the ODC project (see Acknowledgments). Nine sites (56%) were private practices and 7 were academic. A total of 1532 patients were enrolled with average and median enrollment of 95.8 and 64.5 patients per site, respectively.
Patient demographics are provided in Table 1; 16% of the patients indicated a race other than Caucasian.
Thirty-four percent of patients indicated their main concern was dizziness, 28% indicated tinnitus, 29% indicated both tinnitus & dizziness, and the remaining 9% provided no response to this question but completed other sections of the survey packet. This is important to note as patient surveys were not removed from the analysis if they skipped questions. Appropriate denominators for the various subsections of the report are noted in corresponding tables.
Data are stratified by the patient-reported main concern to highlight different patient profiles. Patients with primarily tinnitus or dizziness alone may look very different from patients with both tinnitus & dizziness, and their underlying demographics and/or contributing factors may vary as well as their responsiveness to particular treatments (Table 1).
Table 2a details the profile of tinnitus patient respondents. The majority of tinnitus & dizziness (76%, n=336) and tinnitus (68%, n=287) patients responded that their tinnitus began more than 6 months ago. Approximately one-third of patients indicated extreme bother from tinnitus, but could carry out activities of daily living while another one-third indicated that their tinnitus is only a problem on certain days. Of tinnitus and dizziness patients, 15% indicated that their tinnitus was so extreme that it affected their life greatly as compared with 7% of tinnitus patients.
Data collected on strategies tried by the patients could help future subtyping therapeutic studies. For example, for tinnitus & dizziness patients, the top three strategies that patients indicated worked ‘sometimes’ or ‘always’ (and had denominators of patients greater than 20) were background sound (56/63 or 89%), anti-anxiety medicine (45/52 or 87%), and other medications (65/74 or 87%). Using the same criteria, the top three strategies for tinnitus patients are background sound (57/67 or 85%), amplification (15/24 or 63%), and anti-anxiety medicine (20/37 or 54%). On average, patients indicated they had tried 2.25 strategies with a maximum of 10 and a median of two strategies tried.
The average THI (high score worse, 0–100) score for tinnitus & dizziness patients was 36.5 as compared with 28.6 for tinnitus patients, both a grade of 2 (mild severity). For tinnitus & dizziness patients, 22% fell into grades 4 (severe) and 5 (catastrophic) as compared with 12% of tinnitus patients. The most frequently reported problem/item on the THI for both tinnitus & dizziness and tinnitus patients was “I feel I have no control over my tinnitus” while the least frequently reported problem/item was “I feel desperate because of my tinnitus.”
Provider data (Table 2b) collected on tinnitus patients included the provider’s determination of whether the tinnitus was subjective or objective and audiogram information. For the tinnitus and dizziness patients in the study, tinnitus was determined to be objective in 3% of the patients versus 6% of the tinnitus patients. When the audiogram was performed, it showed normal hearing in both ears one-fifth of the time.
Table 3a provides the results of the survey detailing the profile of dizziness patient respondents. The majority of tinnitus & dizziness (85%, n=351) and dizziness (82%, n=407) patients responded that their dizziness was severe enough to warrant further investigation of their experience. The specific screening question was worded as follows:
I have experienced a feeling of severe spinning (either myself or the room spinning, may be associated with nausea and vomiting). Defined as vertigo, this sensation is severe enough that I have to stop what I am doing. [If NO, stop this questionnaire]
Patients responding to the dizziness portions of the questionnaire reported that dizziness had an impact on their quality of life and daily living. Thirty-nine percent of dizziness and 26% of tinnitus and dizziness patients had more than 10 vertigo episodes in the last month, and 23% of tinnitus and dizziness and 24% of dizziness patients missed more than 10 days of work in that same timeframe. Tinnitus and dizziness patients had associated symptoms more often than dizziness patients; however, the top associated symptoms differed by group. The top two associated symptoms for tinnitus & dizziness were tinnitus in predominantly one ear (54%, n=207) and fullness or pressure in one ear only (47%), whereas the top two associated symptoms for dizziness patients were headache or head pain (32%) and light sensitivity (25%). Neck pain, headaches previously diagnosed as migraines, and family history of migraine was present a third of the time or more in both groups.
The average DHI8 (higher scores worse, 0–100) score for tinnitus & dizziness patients was 41.7 compared with 40.0 for dizziness patients, both a grade of 2 (moderate severity). For tinnitus and dizziness patients, 24% were grade 3 (severe) compared with 18% of dizziness patients. The most frequently reported problem/item on the DHI for both tinnitus & dizziness and dizziness patients was “quick head movements increase problem” while the least frequently reported problem/item was “afraid to stay home alone.” For the subgroup of dizzy patient respondents that filled out the MAT,9 27% of both groups scored as having a diagnosis of migraine.
Physician data (Table 3b) collected on dizziness patients included the provider’s determination of ‘most likely diagnosis’ and audiogram and electronystagmography (ENG) information. The top three ‘most likely diagnoses’ for the tinnitus & dizziness patients were definite Ménière’s disease (38%, n=103), vertiginous migraine (18%), and benign paroxysmal positional vertigo (BPPV) (16%), whereas for dizziness patients, the top three were BPPV (35%), vertiginous migraine (24%), and vestibular neuronitis (24%). When the audiogram was performed, it was reported as “normal hearing” in both ears 20% of the time in tinnitus & dizziness patients and 30% in dizziness patients. When performed, the ENG showed normal caloric function in both ears most of the time and at least a 25% weakness in the clinically affected ear 35% of the time in tinnitus & dizziness patients and 25% in dizziness patients.
The primary objective of the ODC project was to test the research infrastructure through deployment of a descriptive, epidemiologic study on tinnitus and dizziness. The project, now successfully executed and completed, provides the needed proof of concept of the functionality and processes initially adopted in the CHEER community-based research network. The data described here are an important byproduct of testing the operational metrics of CHEER. Notwithstanding that the study was not structured for the purpose of reaching a target enrollment or hypothesis testing, the success of this project and the resulting data further advance CHEER’s mission of becoming the national resource for practice-based research in hearing and communicative sciences.
Additional credibility for the data itself is apparent when comparing results with those seen in the literature. In a review of epidemiologic studies published on tinnitus from 1993–2003,10 findings across the literature include: age and sex were not consistently associated with prevalence of tinnitus; hearing loss was consistently associated with tinnitus; bilateral tinnitus was more common than unilateral with left-sided tinnitus typically reported more than right; and clinical risk factors associated with tinnitus included depression. In our data, there was a close to equal split in sex for tinnitus patients and the average age was 54. Our data indicate that normal hearing in both ears only was noted one-fifth of the time, in concert with the findings in the review. Similarly, we also found that bilateral tinnitus was more common (44%) than unilateral and that left-sided tinnitus was reported more often than right-sided (22% vs. 20%). Thirty-eight patients (9%) indicated using antidepressants as a treatment strategy, potentially lending to an association between tinnitus and depression. The review also pointed out the need for larger studies and the need for a more diverse patient population in terms of race and ethnicity. Our patient demographics show some diversity, with 14% of tinnitus patients indicating a race other than Caucasian. The patient demographics for the diseases in this study are consistent with the overall patient demographics at the participating sites. However, for future studies, the incorporation of a more diverse patient population through focused site recruitment is in progress for the CHEER network.
Neuhauser et al2,3 reported dizziness and vertigo prevalence higher in women than men in the general population. Our study supports this statistic; 71% of our dizziness patients were female, compared with the overall pool of respondents where 61% were female. Bronstein et al4 reported the considerable impact of dizziness on social and work life. We found this to be true as 39% of dizziness patients reported having more than 10 episodes of vertigo, and 24% had missed more than 10 days of work in the past month. Bronstein found that half of the patients affected by dizziness feel a substantial drop in their efficiency at work, so much so that 25% may give up or change their work as a result. A 2010 report by Hegemann and Palla11 reviewed recent advances in the field of neurotology, focusing on vestibular tests. They found that while there have been improvements, treatment options are still limited. The ability of the CHEER network to recruit large volumes of patients in a timely fashion across multiple sites could provide the necessary resource to explore treatment options further.
Given that the primary goal of the ODC project was to implement and test the processes adopted by the CHEER infrastructure, the data presented here are both limited and robust. Conceptually, we can begin to see and describe disease patterns of tinnitus and dizziness that should be considered in the design of future research efforts. Since this data collection exercise was admittedly not hypothesis-driven by the diseases, we have intentionally avoided database mining and statistical manipulations for correlations or conclusions that might be over interpreted. We only describe the data in this report and propose interesting observations that might lead curious clinician-scientists to collaboratively explore practice-based research networks like CHEER.
The power and expediency of the CHEER network in recruiting ample patient sizes across a diverse set of practices and geographic locations is real. In the ODC study, we enrolled 1532 (427 tinnitus; 527 dizziness; 443 both tinnitus & dizziness) patients in 8 months across 16 sites. In contrast to this metric, the development of the THI was based on recruitment of 84 patients in the item development phase and 66 patients in the validation phase across 2 sites.7 Similarly, in the development of the DHI,8 the final version of the tool was administered to 106 consecutive patients at one site to demonstrate internal consistency reliability. In a review of randomized controlled trials in tinnitus and dizziness, the vast majority included less than 100 patients and, regardless of sample size, most are single site studies. For comparative effectiveness research, studies aimed at subtyping, profiling, and stratifying patients are a necessary precursor. Additionally, research translation and generalizability is facilitated by involving community sites and diverse populations in research.
The data are also valuable in serving as the foundation for developing research questions and protocols. Questions and future research opportunities could include:
Further work in patient profiling and subcategorization needs to be a priority to guide development and conduct of future rigorous treatment trials. This is particularly true in the setting of high public health impact disease or disease symptoms that still have unclear etiologies and likely represent a collection of disorders. We must explore the texture of disease and disease presentations so that treatments are effective and acceptable to health care providers and patients.
The data here are robust in their descriptive and exploratory value, and what they predict can be accomplished in the future with adequate resources and through collaboration of academic medical centers with community partnerships. We believe that engagement and success on this grassroots level are essential to future research efforts and key steps to translating research into practice.
We acknowledge all of the CHEER sites and especially the initiative and motivation of the 16 CHEER sites that participated in this study for no compensation but the greater good: Duke University Medical Center (NC); Charlotte Eye, Ear, Nose, & Throat Associates (NC); Low Country ENT (SC); Medical University of South Carolina (SC); Philadelphia Ear, Nose and Throat Associates (PA); Coastal Ear, Nose & Throat (NJ); Ear, Nose & Throat Surgeons of Western New England (MA); University of Michigan (MI); Ear, Nose & Throat Associates, PC (IN); New York Ear and Eye Infirmary (NY); Ear Institute of Chicago (IL); Puget Sound Hearing & Balance (WA); University of Texas Southwestern Medical Center (TX); University of Texas Medical Branch (TX); Oregon Health & Science University (OR); Commonwealth Ear, Nose & Throat (KY).
We also would like to offer a special acknowledgement to Maureen Hannley, PhD. Dr Hannley's intellectual and creative contributions to the development of CHEER have directly contributed to the success of this network.
Funding: Grant number: 5R33DC008632-04; NIDCD R21/R33 Phased Infrastructure Grant for Patient-Oriented Research
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.