Stephen A. Paget, MD: My name is Dr. Stephen Paget, and I'm the Physician-in-Chief at Hospital for Special Surgery. I'm pleased today to introduce to you Dr. Robert Marx, who is an Assistant Professor of Surgery and Orthopaedic Surgery and Assistant Professor at the School of Public Health at the Weill Medical College of Cornell University. He is the Director of the Center for Clinical Outcomes at Hospital for Special Surgery, and has a particular interest in defining the outcomes of various types of procedures, how to assess them, and what kind of clinical trials can best be done to define appropriate therapies for patients.
Dr. Marx, tell me a little bit about what clinical research is.
Robert Marx, MD: Clinical research is research that focuses on patients - studying patients who are undergoing treatment to define the optimal method of treatment for a certain patient with a given condition.
Stephen A. Paget, MD: We know what a basic scientist is - bench scientists who are in a laboratory and who deal with cells and test tubes and proteins. How do clinical researchers get their education and what arena do they work in?
Robert Marx, MD: Clinical researchers often are physicians who specialize in clinical research, with either a master's or PhD in clinical epidemiology or outcomes research. Also, clinical research may be a master's- or PhD-trained individual who is a non-physician.
In terms of the comparison between basic science and clinical research, the ultimate methodology is the same, which is the scientific method -- proposing the hypothesis and then testing that hypothesis.
The problem in clinical research is that we can't control the environment the way a basic scientist can. Basic scientists, for example, study animals like rats, which can be controlled in cages with their environment totally predetermined, whereas clinical researchers study patients. So, for example, we can't have a group of patients smoke and a group of patients not smoke and then determine the difference, because we can't interfere with peoples lives to that extent. Therefore, clinical research has a lot of different methods and strategies to control the environment as best possible, or at least adjust the things that we can't control.
Stephen A. Paget, MD: Should patients be frightened to participate in clinical research trials?
Robert Marx, MD: Not at all. In fact, no trial would be deemed unethical or unsafe, because they all have to be improved by the Institutional Review Board at the institution. The Institutional Review Board (IRB) is a group of physicians and non-physicians, lay people, who review the trials to make sure that they are safe and ethical for patients. So any patient enrolled in a clinical trial is at least getting safe medical care.
Stephen A. Paget, MD: You mentioned before the term "outcomes research." What does that mean?
Robert Marx, MD: "Outcomes" typically refers to how the patient does as a result of the treatment, what their final outcome is. I guess that this term has evolved in the last 10 or 15 years because we are now more interested in what really matters to patients. Traditionally, we have examined things like radiographs or range of motion of a joint, which may not actually matter to the patient. What they want to be able to do is walk or perhaps run or lift things, and that may or may not be related to more traditional measures of impairment, such as a radiograph. We have now progressed to measure outcome with quality of life measures and health-related quality of life measures. We really ask patients how they are doing and measure that in a more detailed fashion that is more relevant to the patients.
Stephen A. Paget, MD: So, the patient has diabetes. They follow their blood sugar and if you do a study on diabetes, you could follow the blood sugar. If somebody has cancer, you obviously follow various parameters. What tools or instruments are available to you in outcomes research that are as sensitive as the ones I just mentioned?
Robert Marx, MD: There are many tools that have been developed to measure what we call health-related quality of life in patients, and these really complement the measures you spoke of, such as blood sugar in a diabetic patient. These are typically questionnaires that ask patients about how they are doing, how they feel, what they are able to do, whether they are having pain. Essentially these instruments are broken down into disease-specific or condition-specific and generic. Generic measures ask them about their entire health or mental health, their physical health, their emotional health, and the disease or condition measures ask them more specifically how they are doing. For example, with respect to their knee problem, we ask if can they walk, can they climb stairs, can they sleep comfortably at night. And these two types of measures are complementary because the generic health status measures allow comparisons of cross-conditions and the disease- or condition-specific measures are more related to that particular condition, and therefore we can measure change in a more detailed and accurate fashion.
Stephen A. Paget, MD: You've been involved in various types of research projects. Could you give us some idea of what kinds of clinical research you've done, what kind of specific trials, and then also some specific examples of the knee research you have been involved in?
Robert Marx, MD: There are many different kinds of clinical research or outcomes research and many different architectures of research methodology. To give you a couple of examples, one study we did, was looking at whether shoulder dislocations cause arthritis in the shoulder. That is currently not believed to be the case, and the way to study that is to look at someone who has a shoulder dislocation and wait 40 years to see if they develop arthritis. Now that is clearly impractical and very difficult to do.
Instead, we employed a methodology called a "case control study" where we found people with arthritis and people without arthritis and looked back to see if they have shoulder dislocations, which can be done in a relatively short period of time. We actually found that the people with arthritis had a higher incidence of dislocations -- linking dislocation to arthritis really for the first time, something that has not previously been demonstrated. That does not prove causation however. That is similar to the way smoking was linked to lung cancer in case control studies. People with lung cancer and people without lung cancer were asked whether they had a history of smoking.
When that is done over and over, there is biologic possibility that causation can then be inferred, as in the case of smoking. We are not quite there yet with dislocations and shoulder arthritis.
Another methodology in clinical research is a prospective randomized clinical trial, where a group of patients who are eligible for two therapies are randomly assigned to one treatment or the other. This can only be in a case where we are not sure which treatment is better and we believe them to be equal. When that is possible and that is the case, it is then possible to randomly allocate the patients to one treatment arm or the other, where they are receiving care that is believed to be equally appropriate. The value of the randomization is that the groups are then equal for known and unknown prognostic variables.
That is to say, the groups are essentially equal for all things we know about and even things we don't know about, and even things we don't know about that may affect the outcome. And what that does is it brings us as close as possible to what we discussed before, which is basic science, where we are controlling the bias as much as possible.
In terms of ongoing study of knee problems, what we are doing is a prospective cohort study. A cohort is a group of people with a certain problem, and we are studying these patients in a prospective fashion to determine how they do after surgery to repair cartilage in the knee. This is currently an evolving field, as we try to repair cartilage and prevent arthritis down the road, and there are different surgical techniques to repair cartilage in the knee. And we are currently studying these in a prospective fashion, using patient-based outcome measures to see how people are doing and tracking them over time into the future.
Stephen A. Paget, MD: What kind of questions do you ask of them to define what their quality of life is and how the procedure has affected them?
Robert Marx, MD: What we are using for this research is established, validated questionnaires. We use, for example, the Activities of Daily Living (ADL) in knee outcome questionnaire, which was developed in Pittsburgh. It is a 17-question multiple-choice format asking people about their ability to walk, get out of a chair, and climb stairs. And we also use the SF-36, or Short Form 36, which is a 36-question scale asking people a gamut of questions about their general health. This is a widely used questionnaire across all fields of medicine. It allows us to compare the health of people with knee problems to those with kidney failure and those with heart disease.
Stephen A. Paget, MD: Patient satisfaction is important, isn't it?
Robert Marx, MD: Patient satisfaction is critical, and is not always related to the patient outcome. There are many factors that can contribute to satisfaction. It is quite a complex issue.
Stephen A. Paget, MD: How do you feed back to the doctors the information you gain from these outcome studies?
Robert Marx, MD: What we are doing now, for example, with the cartilage study that we just talked about, we collect our data, we tabulate and analyze it and submit it to meetings of physicians from across the world. Then we publish our research to communicate the results of our findings more broadly and, with respect on a more patient-specific level, to the physicians treating those patients. They have access to all of our information if they want to determine how the patient's quality of life has changed with respect to their treatment.
Stephen A. Paget, MD: Many physicians treat patients for various problems. There are many people in this hospital and others that treat knee problems. Is there a variation from physician to physician with regard to outcome, and how do you address that with the physician? Is there feedback to them to let them know where they fit in the universe?
Robert Marx, MD: I think that from physician to physician there is certainly a lot of variation in terms of how physicians treat a given patient. I think that is one of the challenges in clinical research -- to provide evidence to physicians so that they can base their treatment on sound scientific data, to reduce the variation and treatment for a given patient, and hopefully improve their outcome down the road.
Stephen A. Paget, MD: So the general concept is to constantly reset your level of activity depending upon the outcomes that you use.
Robert Marx, MD: That is exactly correct. We want to determine how patients are feeling, how they are doing after treatment and, based on that information, provide feedback to physicians and the medical community at large to reevaluate our treatments constantly and hopefully improve them.
Stephen A. Paget, MD: Are institutions like Hospital for Special Surgery very supportive of this type of research?
Fortunately for me, Hospital for Special Surgery is extremely supportive and very interested in promoting health outcomes research to improve patient's health. So certainly HSS is very supportive and, in the general medical community at large, support is definitely increasing for this, and we hope to be leaders in this field.
Thank you very much.
Dr. Marx was interviewed by Dr. Stephen A. Paget, Physician-in-Chief, Hospital for Special Surgery