Patient Measures Are Worth the Work
Dr. Pisetsky’s foray into patient measures (“Twenty Questions” Parts 1 and 2, Feb. 2007, p. 6 and March 2007, p.6) along with Drs. Smolen and Aletaha’s article (“Target Remission,” March 2007, p. 1) are of considerable interest. Many of us in clinical practice are wrestling with the questions: “How are my patients doing?” and “Does my patient need better/more aggressive treatment?” You chose to go first with a calculator method, whereas Drs. Smolen and Aletaha have moved more to simple addition.
My first attempt at patient measures was to use a Disease Activity Score (DAS) calculator. Like you, I found having a number helped me get a better idea of how my patient was doing, but to do this I needed to get the patient’s global response. Essentially, this requires the patient to fill out some form of a questionnaire, but my attitudes toward questionnaires were somewhat skeptical, to say the least. I had done clinical trials and had worked with the Health Assessment Questionnaire (HAQ) Disability Index, but had found it cumbersome. Not only did I need to review the two pages, I also needed to correct the domain scores to account for assistive devices. When that was done, it had to be added up and divided by eight. As a solo-practice rheumatologist, spending that amount of time on each patient would have put me out of practice.
At a meeting, I had the good fortune to talk with Ted Pincus, MD, who was kind enough to send me his adaptation of the HAQ, the multidimensional HAQ (MDHAQ), which he had developed to address HAQ’s limitations. It includes only activities that can be performed by all patients (eliminating questions like the one that perplexed you: Can you shampoo your hair?) and allows for a quick, visual review of the results. I would recommend that you consider using the MDHAQ to see if some of your concerns about questionnaires are addressed.
Using the MDHAQ, Dr. Pincus, Yusuf Yazici, MD, and I are developing of a user-friendly index, RAPID (rheumatology assessment of patient index data), a one-page questionnaire completed by patients while they sit in the waiting room prior to the actual office visit. Our work, presented at the 2006 ACR meeting, demonstrated that this form can be reviewed and scored in under 10 seconds. It is given to every patient—regardless of the diagnosis—at every visit. This simplifies my front-desk’s life because they don’t have to pick and choose (and explain) who does or does not get the questionnaire.
It was interesting to read that you administered the HAQ orally to the patient. This defeats the purpose of the questionnaire. Besides the possibility that the answers may be inaccurate (patients often like to “please” their doctors and may give answers they think you want to hear), it also makes HAQ use unrealistic for the majority of clinicians, who rightly feel that time is a valuable commodity. The questions included in the RAPID are simple and, in the vast majority of cases, require no assistance by anyone to complete. They include a physical function scale consisting of 10 activities, along with a patient global score (needed if you want to calculate DAS, Simplified Disease Activity Index [SDAI], or Clinical Disease Activity Index [CDAI]), along with a patient pain Visual Analogue Scale (VAS). For further simplicity, the usual 10-cm VAS has been changed to a 21-numbered circle format that eliminates the need to use a ruler or to have standardized photocopying.
These three scores should look familiar to all rheumatologists, as they are the patient reported outcomes (PRO) measures used to calculate ACR scores. Once the actual office visit begins, I quickly score the RAPID and use the information to get an idea of how the patient is doing.
The time needed for me to get the scores and record the RAPID is 10 seconds. Compare this with the 90 seconds it takes to do a formal 28-joint count and the additional 14 seconds to calculate the DAS using the calculator on the DAS Web site. (These data were also presented at the 2006 ACR Annual Meeting.) And that assumes that the lab tests are available and that they were performed correctly. What’s just as important, the resultant score is an excellent measure of disease activity.
The total RAPID score is 0–30 or 0–10, the mean of the three scores. It gives comparable results to the SDAI and CDAI discussed by Drs. Smolen and Aletaha, with similar correlation to DAS. Our goal is to reach a score of 6 on a 0–30 scale or a mean score of 2, indicating low disease activity, or 3 on a 0–30 scale or a mean score of 1, indicating near remission. The major difference is that the RAPID does not include a formal joint count. This is an important distinction, as we have reduced the time required, but more importantly, we have a scale that can be done by everyone. Many studies have shown that, while virtually all rheumatologists perform a careful joint exam on every patient, very few do a formal joint count. While a formal joint count is essential in clinical trials, it has to be asked: Does it make sense to develop practice-based criteria on procedures that are not done in practice?
There is one additional benefit of using patient responses for monitoring and assessing patients. It does something novel and unique in modern medicine. It actually asks the patient to tell how he or she is doing. One has to ask: Whose perception is more accurate—the patient’s or the physician’s?
Like any other measure, RAPID is a clinical tool that helps rheumatologists make clinical decisions. It does not replace the physician, nor does it supersede common sense. An elevated erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) will have a strong influence on DAS/SDAI, but one would never use that value alone in the decision-making process without first asking the patient about a recent infection or examining the patient for signs of disease activity. If that same patient had no tender or swollen joints but was feeling poorly (patient global=50) and had an ESR of 50 due to recent pneumonia, he or she would have a DAS28 of 4.4. No one would change a disease modifying antirheumatic drug (DMARD) or biologic based on that alone.
Technology is wonderful and a lot of fun, but you don’t need to be married to it (this from a solo practitioner with five computers in his office). Your “fancy” hand-held device didn’t make your life any easier; it made you spend more time to get the information. At the risk of being crude, for the average practicing rheumatologist, time is money. The paper questionnaire, at less than three cents per copy, is easy to read and easy to score. For those of us who are using computers, the results can be scanned into the EMR or quickly entered into the record manually.
As Drs. Smolen and Aletaha correctly point out, our treatment target should be remission. Having a tool that is easily completed, rapidly scored, and clinically useful will make this goal more likely to be embraced by all. The RAPID provides most if not all of these features.
Marty Bergman, MD
Arthritis and Rheumatology, Ridley Park, Pa.
Give PROs Another Go
I really enjoyed Dr. Pisetsky’s story about the HAQ in The Rheumatologist (“Twenty Questions”). It is so wonderful that one of the top-10 U.S. scientist-rheumatologists does not hesitate to reveal all his fatal mistakes in using the HAQ in the clinic. However, I could not resist pointing out some matters that such a prominent rheumatologist apparently never learned, with concern regarding what might be expected from an ordinary clinical rheumatologist.
- Deadly sin 1: You invaded and ruined the “Patient-reported outcomes” and turned them into “Pisetsky-reported outcomes;”
- Deadly sin 2: This patient will probably never see any value in PROs because of his traumatic experience the first time; and
- Deadly sin 3: Misunderstanding the scale of the HAQ: “As it turned out, the man’s HAQ was actually low, his perception of his functioning in fact worse than the reality.” Please note: low HAQ is good; high HAQ is bad!
My dear friend Dr. Pisetsky, you really messed up with your patient….Try again with the correct method and then write a new story, delighted of the value of PROs!
I did not reach the ceiling in learning about PROs and my experience is pale compared with the great geniuses Drs. Pincus and James Fries, MD; however, I would like to share with you what I have seen.
An interpreter or investigator between a patient and a piece of paper or a silvery little box is a major confounding thing. First, most patients want to please their doc and “Pisetsky (or Sokka)-reported outcomes” tend to be significantly better than real PROs. Second, the doc is always busy, such as in your example case, and a patient is forced to provide responses under pressure while you are “ready to push the button.” In fact, PROs save time for the clinician if administered correctly: the patient completes the form while waiting to see a health professional who starts the visit by “eyeballing” the questionnaire to provide an immediate overview of patient’s health status and concerns at that time, ready to be recognized and discussed.
“People are different and outcome measures need to capture all of this wonderful variety,” wrote Dr. Pisetsky. PROs such as the HAQ/MDHAQ, pain, global assessment, fatigue, morning stiffness, self-report joint counts, and so on, are presented as scores, but levels of the scores don’t automatically lead to certain clinical decisions, just as we don’t treat CRP or ESR levels, but use the information in developing a treatment plan. Being open to widen a traditional bio-medical model of treating diseases helps me understand that in, chronic conditions, PROs—if administered correctly—may be part of an important process to help communication between a patient and a health professional and may lead to a better patient adherence to medications and improved long-term outcomes.
My dear friend Dr. Pisetsky, you really messed up with your patient. Maybe you should try again with the correct method and then write a new story, delighted of the value of PROs!
Tuulikki Sokka, MD, PhD
Jyväskylä Central Hospital, Finland