Learning Objective
The learning objective of the five-hour workshop is understand why and how to implement a registry study.

What Are Registry Studies?
Registry studies are observational studies in which the events which happen to subjects with a specific disease or condition are recorded without predefined treatment. This is in opposition to interventional studies in which subjects with a specific disease or condition are treated per an approved protocol and data are collected about the treatment. We say that registry studies are 'data driven' as opposed to interventional studies, which are 'protocol driven'. This whitepaper and the accompanying workshop are based on " Registries for Evaluating Patient Outcomes: A User's Guide" developed by the Agency for Healthcare Research and Quality.

Why Registries?
Registry studies have an advantage over controlled studies (very few device studies are randomized, so I won't use the phrase 'randomized controlled trial' or RCT) in that they are designed to assess effectiveness, the performance of a device in everyday practice. Interventional or controlled studies are designed to assess efficacy, the performance of a device in a controlled setting of subjects without co-morbidities or confounding conditions. Registry studies are especially desirable in populations where interventional studies are unethical, where large numbers of subjects are needed, or when post-approval studies are required by FDA.

There are many reasons for device manufacturers to conduct registry studies. Here are a few:
[1] To study the effectiveness of a device in real-life medical practice.
[2] To learn how treatment with your device compares with other treatments in a population with, say, certain co-morbidities or demographics.
[3] To study the nature, severity, and frequency of occurrence of adverse events and adverse device effects.
[4] To gain information for new indications or claims.
[5] To gain information to support reimbursement applications.
[6] To gain marketing intelligence.
[7] To obtain competitive intelligence (how effective is the competition?)
[8] To build site relationships.
[9] To provide publishing opportunities in populations where interventional studies may be unethical.
[10] To conduct required post-approval studies.

Regulations and Funding Sources
Surprising as it may seem, the regulations for registry studies are far more complex than for controlled studies. That's partly because lots of groups conduct registry studies besides device manufacturers; for example, government health agencies, privately funded non-profit agencies, insurance providers, hospital quality assurance functions, national disease registries, and even the National Death Index. Of course state and local regulations and the federal HIPAA regulations apply to every registry study, but a good way to sort out other regulatory requirements is to consider the study's funding source. There are three major sources of funding for registry studies.

Sponsor funded
If the registry is funded by a device company and the data will ultimately be used to support a marketing application or product claim, it is a sponsored study. Then the investigational device exemption regulations, 21 CFR Part 812, apply to the study, unless the device is exempt under Part 812.2(c) or is required by FDA under Section 522 of the Act. If subject to Part 812 you treat the study like any controlled, interventional study; if exempt from Part 812 only state and local requirements and HIPAA apply; if subject to Section 522 there is little guidance.

Federally funded
If the registry is funded by a federal agency, such as a registry of sexually transmitted diseases or the Women's Health Initiative or the Framingham Study on cardiovascular disease in men, the regulations from 45 CFR Part 46, i.e., the Common Rule, apply. For more information on the Common Rule you can look at the OHRP website.

Privately funded
But if the study is privately funded, say by the Easter Seals Foundation, there are no federal regulations other than HIPPA rules. Device manufacturers occasionally take advantage of this fact and conduct a registry outside of Part 812 or Part 46. As a manufacturer you might choose to do this if, say, you were gathering effectiveness information about a competitor. But you cannot ever use the data later to support a marketing claim, instead you would have to repeat the study.

Planning
Planning a registry study is pretty much the same as planning a controlled study. You will need a protocol, a monitoring plan, a database management plan, data analysis plan (including plans for handling missing data), informed consent, HIPAA authorization or waiver, IRB approval, case report forms, investigator agreements, financial agreements, and all the other documentation associated with a clinical study. On an interesting note, some registry studies collect data about medical practices. The investigator is also a subject and must sign an informed consent.

The most important step in planning is to be clear and in consensus about your purpose. Possibly the most common reason for poor results in a registry study is a purpose that evolves as the study progresses. Whether the purpose is to gather competitive intelligence, obtain reimbursement data, or to do a cost analysis, you should get top management to sign off on the purpose, tape it to your wall, and look at it every day. Then, when the team asks why your study doesn't meet the xyz purpose, you can say that it was only intended to meet the abc purpose.

Other important planning issues include identifying the key questions (hypotheses), the core data, the target population, and the study's scope. Unlike most interventional studies, data for a registry study might come from multiple sources including literature, other existing databases such as Medicare reimbursement databases, or existing medical records. Obtaining informed consent, and whether or not new informed consent is needed, for existing data sources may be a difficult issue to resolve.

Prospective and Retrospective
I am often asked about the difference between prospective and retrospective studies and whether registry studies are one or the other. Prospective studies are designed to gather data about events that haven't happened yet; they look forward in time. Retrospective studies are designed to gather data about events that have already happened; they look backward in time. Registry studies can be either prospective, retrospective, or both. If you engage an investigator to abstract data from existing medical records, those data are retrospective. But you might also want the investigator to follow those patients into the future, requesting specific data be obtained at future follow-up visits; these data are prospective.

Registry Designs
There are four basic designs for registry studies:
[1] cohort studies, in which you follow a group of patients with a common characteristic to see if they develop a particular endpoint or outcome. For example, we might follow all men who have stents to see if US Presidents live longer than other men.
[2] case-control studies, where we follow patients (cases) with a particular outcome and patients without the outcome but who match the case demographics (controls) and look for prognostic factors. For example, we might match men with stents to men without stents to see if we can predict why some men are successful stent survivors and others are not.
[3] case-series studies, in which we take observations on a consecutive series of "n" patients. We might measure the frequency of adverse events or missed follow-up visits.
[4] a case study, or observations taken from one patient.

For the most part device manufacturers will sponsor either case-control studies or case-series studies. Case-control studies are useful for obtaining data that compares one device or treatment to another, while case-series studies are useful for obtaining data about the frequency of an event. Cohort studies are usually too large for a single sponsor to afford, a case study is too small to have much value except for an unusual subject.

Cost, Monitoring, and Timing of Registries
As for the cost and timing of a registry study, don't expect any bargains. They cost about the same as interventional studies and require the same level of monitoring to assure data integrity. Enrollment should proceed more quickly because the selection criteria for subjects should be more inclusive; more patients should be eligible for the study.

A quiz, of course
The five-hour workshop is following by a half-hour online quiz, taken on your own time within the following two weeks. The quiz is designed to test your learning of the concepts discussed in the lecture and your ability to apply those concepts to real-life issues.

Course level
The workshop is an intermediate level course. Participants are expected to have a basic knowledge of clinical research and device regulations.

You will receive
[x] PowerPoint slides.
[x] Expert and experienced speaker.
[x] A copy of "Registries for Evaluating Patient Outcomes: A User's Guide" from AHRQ.
[x] Information on how to access Medicare reimbursement databases.
[x] An article on "Treatment of Missing Data" by David P Howell.
[x] An example informed consent from a hospital registry study.
[x] Chance for Q&A.
[x] A graded quiz with immediate test results.
[x] Certificate of Attendance and 0.55 CEUs.

Who should attend
[x] Device manufacturers looking for alternatives to controlled, interventional studies.
[x] Device manufacturers conducting Section 522 Post-Market Surveillance Studies.
[x] Clinical professionals involved in implementing registry studies.
[x] Regulatory professionals involved in reporting registry studies.
[x] Investigators who are conducting registry studies.
[x] Company managers and decision-makers.

Instructor
Dr. Nancy J Stark is President of Clinical Device Group, a consulting and contracting firm for medical device pre-approval issues. You can read about Dr. Stark at www.nancyjstark.com and about Clinical Device Group at www.clinicaldevice.com.

Date, time, registration
This five-hour event will be presented on 16 March 2010 at 11:00 am Central time and will be available OnDemand a few days later. Event materials will be distributed the day before the e-conference. To register please click here.

System requirements
[x] Personal computer.
[x] Internet Access.
[x] Telephone.