Tools for MS Research: Databases and Outcomes Assessment Instruments
John Booss, MD; Moderator
Timothy Vollmer, MD
Joseph Guarnaccio, MD
Nicholas LaRocca, PhD
Leigh Anderson, MD
Kimberly Schafer, MS
Michael Rosen, PhD
How can VA staff harness the vast amount of data available in a system as large as VA? Three different registries in varying stages of development and which collect data from slightly varying angles were presented.
Registries are a way to `stir the pot', by forcing clinicians to record clinical information in a uniform manner. Registries can provide the information required to influence public policy so that it appropriately responds to specific populations.
If data is properly collected from a certain population, useful information can be obtained. However, how are clinicians using the registry programs; do the registries accurately identify and group special populations and their unique needs? For instance, within VA there is concern for the MS population and the SCI population. Each requires the system to respond in a certain way, but neither population is identified uniquely.
NARCOMS (North American Research Consortium on MS) is the research arm of the Consortium of MS Centers, a multi-disciplinary organization of professionals and researchers specializing in the care of persons with MS. The NARCOMS patient registry is a recruitment database developed to provide a resource for linking investigators and patients with required profiles in an expedited manner in order to work together towards finding the cure for MS. The registry is voluntary, confidential, for research purposes only, and free of cost to patients. In developing the database, NARCOMS found that patients with MS were strong advocates for research, wished to have a great deal of information about current studies, and were extremely cooperative about collaborating on research projects.
It should be emphasized that the NARCOMS patient registry is a recruitment database, rather than a population based sample. It may not reflect the actual patient population, but it does provide profiles on patients interested in participating in clinical trials. Having said that, future plans for the registry include developing it as an outcome database that can be used for benchmarking by various MS clinical programs around the country.
Data gathered includes basic demographics such as marital status, living situations, educational level, and employment status. A minimal attempt is made to identify impact of MS on patient's ability to be employed. There is also some information on third party healthcare insurance sources. Military service and its relationship to care within the VA healthcare system has just begun to be gathered, with over 3,000 individuals enrolled since last year and approximately 100 more joining each week. There is some information about family history and genetic association with this disease in terms of twins and family associated disease.
Diagnostic evaluations are more difficult to collect. It is not intuitively obvious for a patient to identify the type of MS they have. Therefore, queries about relapse and attack frequency are used. It is, however, difficult to tell from this information if the patient is relapse-remitting or secondary progressive. Some information is gathered on medication and therapies, both historical and present. The Extended Disability Status Scale is used to quantitatively describe the patient's clinical status: information about seven domains of performance is collected: mobility, hand function, cognitive functions, fatigue, bowel, bladder, and sensory functions. The scale asks how these symptoms impact the patient in terms of their daily activities.
NARCOMS Findings
Approximately 10,000 people have self enrolled, with data provided for half of the enrollees verified. The current enrolled population includes half veterans (note: 627 remain in military) and half non veterans. Within the veteran population, 90% are male. The non-veteran population follows the national profile of one third male/two thirds female ratio.
The average age of diagnosis does not vary between the veteran/non-veteran population, however, the current age among veterans is higher than the average population of MS patients. Relapses appear to occur at similar rates throughout all populations. Veterans with MS are significantly less likely to be employed outside of the home.
Finally, of significance, 27% of the non-veteran population is using one of three FDA approved immunological therapies as compared to 3% of the veteran population.
The Spinal Cord Dysfunction Registry got its start around 1993 as a project that was funded by Health Services Research and Development and the Paralyzed Veterans of America. Originally geared towards Spinal Cord Injury, it was expanded to include MS patients within a year of its origin. The registry is continuously under development each VA medical center assessing and suggesting improvements on a regular basis.
The Registry has two purposes. First, it is a clinical tool used for demographic information, nature and level of dysfunction, medications, and evaluations. Second, it serves as a national database of information for use by VA healthcare policymakers, as well as researchers. Ideally this sort of database would assist providers in identifying patients who might respond to new therapies. Also, it allows for functional status to be entered at each visit, enabling a series of functional measurements to be taken over a period of time.
The Registry allows for reports to be drawn on a national basis or at a local medical center basis. Statistical information at individual sites can also be gathered.
The National MS Society's (NMSS) Longitudinal Database is still under development. It was begun with the goal of prospectively following a large sample of people with MS over a long period of time. Establishing such a cohort will be unlike any in the past, at least in the United States. The Society plans to look at both cross-sectional and longitudinal questions—cross-sectional early in the process and longitudinal later. Most of the research questions need to be addressed using longitudinal data. The Society is interested in a variety of outcomes, including clinical, economic, and socioeconomic. While clinical outcomes are the most familiar to researchers, the impact of the disease on the financial status of individuals and their quality of life has been overlooked in the past. The Society wants to look at a variety of predictors that may be related to outcomes and the way in which various intervening variables may affect the relationship between predictors and outcomes.
The database is being developed with the assistance of a multidisciplinary panel which includes experts in the area of MS, as well as a number of people who specialize in key areas such as health economics, neuro-epidemiology, and spinal cord injury clinical care, but who have not necessarily yet applied this expertise to MS. A contractor, APT Associates of Cambridge, MA has been selected to actually implement the project, largely because of their experience with previous statistical MS surveys and with large databases.
Because MS has a prevalence of one in a thousand, it is nearly impossible to do geographically based sampling. Initially the database will have 1,876 patients. These will be divided and recruited through a number of different sources. There will be 375 patients each from three types of healthcare providers: primary care physicians, neurologists, and MS specialists. Another 375 incident cases will be recruited and distributed among these three groups as well. Patients who are in long-term care facilities and patients who are not currently being followed by physicians are also planned for recruitment. It should be noted that the latter group is rarely included in such studies.
The American Medical Association Physician master file will be utilized for the sampling plan. (Samples of primary care physicians and non-MS specialists will be constructed to identify MS specialists, a grouping that requires a rigorous definition that will hold up in terms of scientific validity.) Each physician who agrees to participate will help to recruit three patients so that the burden on any given provider will be relatively small.
All patients will be administered a baseline questionnaire, an annual follow-up, a baseline clinical exam, an annual clinical exam, and a semi-annual phone interview. The phone interview will be particularly geared toward socioeconomic data. In addition to this basic core data set, baseline blood samples on 400 newly diagnosed patients will be collected along with annual blood samples. The serum will be separated and preserved for immunologic studies. Three hundred patients, stratified by severity, will have DNA typing and baseline. Also, some 200 patients will have baseline and annual MRIs.
The prospective cohort will be research oriented, national in scope, and as broadly representative as possible for comprehensive data. It should not, however, be considered the same as a population-based cohort.
The MS Society ultimately wants to enlarge its investigator base, so that a lot more pooling of information on MS from a variety of registries and databases will add to the original identified research questions. The goal of a common core data set is a valuable one. This would justify the effort and care that would be involved in trying to link people between databases, even though the bioethical issue of increasingly restricted regulations in terms of medical information is, and will continue to be, a significant issue.
The question of patient feedback in all three registries was raised. NARCOMS has perhaps one of the most developed areas for incorporating patient input, but Dr. Vollmer suggested that even more could be done. Current methods include focus groups and write in comments. It was suggested that online cohort would be valuable for more real-time communication.
Go to TOP.
|
|
|