EDITORIAL - It's Good to Feel Better But It's Better to Feel Good

[Guest guest]

Journal of Rheumatology

January 2005

Editorial

--------------------------------------------------------------------------------

It's Good to Feel Better

But It's Better to Feel Good

MAXIME DOUGADOS, MD,

Professor of Rheumatology,

Faculté de Médecine René Descartes, Paris V,

Assistance Publique, Hôpitaux de Pais,

Hôpital Cochin,

27 rue du Faubourg Saint Jacques,

75014 Paris, France.

Address reprint requests to Dr. Dougados.

--------------------------------------------------------------------------------

A lot of effort has been made by the scientific rheumatology community to

standardize the clinical evaluation of musculoskeletal disorders using a

rational and scientific approach. The first step was to make a list of

domains of interest (for most musculoskeletal disorders, the domains

considered most important from a patient's perspective were pain and

functional impairment, and from a doctor's perspective, inflammation and

structural damage). The second step was to propose tools allowing evaluation

of each domain [for example, a 0-100 mm visual analog scale (VAS) for

evaluation of the domain " pain " ]. Such tools are usually continuous

variables and have intrinsic excellent metrological properties in terms of

validity, reliability, and discriminant capacity.

Because of these characteristics, these tools have been largely used in

clinical trials for evaluation of treatment modalities. The results of

therapeutic trials using them are usually expressed at a group level as mean

changes [for example, pain (VAS) decreased from 62 ± 12 to 34 ± 15 and from

63 ± 13 to 58 ± 12 in the active and placebo groups, respectively; p <

0.001]. The difficulty for the clinician is interpretation (in terms of

clinical relevance) of the results obtained and presented. In order to

facilitate his or her critical appraisal of presented results, the clinician

strongly favors presentation of results not at a group level but at an

individual level, as a percentage of improved patients in each study group.

To be able to present the results at an individual level while using a

continuous variable, it is mandatory to determine a cutoff value of change

in this variable above which a patient can be classified as " improved. "

The methodology allowing definition of such a cutoff is not easy. Concerning

symptomatic variables, one has to keep in mind the definition given by

Jaeschke, et al: " the smallest difference in score in the domain of interest

which patients perceive as beneficial and which could mandate... a change in

the patient's management " 1. If one can easily consider a longitudinal study

with a baseline and a final visit to calculate the " difference " in the

variable, the choice of the gold standard at final visit allowing us to

classify a patient as improved is more debatable. Based on the above

definition, this gold standard could be either the decision to change the

symptomatic therapy at final visit, or the overall patient opinion. This

methodological question was debated during the OMERACT 2000 meeting in

Toulouse and the conclusion strongly favored the patient's perspective2.

Another difficulty also arises from the definition of " smallest difference "

and " beneficial. "

In this issue of The Journal, Pavy, et al report results of a study on

ankylosing spondylitis aimed at proposing a cutoff for symptomatic outcome

variables frequently used in ankylosing spondylitis (i.e., the Bath indices)

based on the definition of a minimum clinically important difference3. In

their study, the authors consider as a gold standard the overall patient

opinion at end of study using a 15 point scale; for the statistical

analysis, they converted the 15 point scale into a dichotomous variable:

improved yes/no, where " good deal better, " " great deal better, " or " very

great deal better " was considered as " improved. " Several remarks can be made

concerning this approach:

The difference in score that patients perceive as beneficial is probably

different from the difference in the score patients perceive as

deterioration. The value of the minimum clinically important deterioration

is usually of lower magnitude than that of minimum clinically important

improvement4.

Since the objective is to detect the " smallest " difference, one can argue

that determination of the cutoff should not be extrapolated from the score

obtained in the whole group of patients who perceive improvement (whatever

the magnitude, from a " little " to " very great deal " ); it should be derived

from the score obtained in the subgroup of patients who perceive a

clinically relevant improvement, even of moderate magnitude.

Several studies have shown that such a cutoff is usually closely related

to the baseline score5,6. Patients with a high score at baseline need a

broad level of change to consider themselves clinically improved.

Probably the most important point is to clearly define the concept under

evaluation. In other words, is the main objective of the patient to be in

better condition (concept of improvement) or in good condition (concept of

state)? Based on results obtained in osteoarthritis evaluating these 2

concepts (minimum clinically important improvement and patient acceptable

symptom state) it appeared that the minimum clinically important improvement

values for the symptomatic outcome variables (e.g., pain, patient global

assessment, functional impairment) were closely related to the baseline

scores but not the patient-acceptable symptom-state values6,7. In other

words, whatever the level of symptoms at baseline, the main objective for

the patient seems to be to reach a state they consider acceptable. It is

obvious that the best condition is the absence of symptoms. This concept

(absence of any symptom) is one of remission. For many musculoskeletal

disorders and especially for specific domains such as pain and/or fatigue,

this objective is very difficult to achieve. Therefore, alongside this

concept of remission, the concept of low disease activity state is emerging.

For symptomatic outcome variables, the concept of patient-acceptable

symptom-state is similar to the one of low disease activity state. When

considering a disease as an entity (e.g., rheumatoid arthritis) the

objective is to merge both the patient's and the doctor's perspectives. For

this purpose, as we have seen, domains other than subjective symptoms have

to be included in the definition of low disease activity state, for example,

inflammation (e.g., C-reactive protein, number of sites of synovitis in

rheumatoid arthritis). This concept was discussed during the OMERACT 2004

Monterey meeting and the results should appear soon in this journal8.

In conclusion, concepts of improvement and state are both important to

consider. They enable us to present results of clinical research studies

and, in particular, therapeutic trials at an individual level. This

presentation is more meaningful for the clinician, especially since it

allows the calculation of the number needed to treat. In this respect, the

data presented in this issue by Pavy, et al are important to consider. These

results should be compared across different sets of patients for a specific

outcome variable (e.g., Bath Ankylosing Spondylitis Functional Index for

evaluation of functional impairment in ankylosing spondylitis), as well as

across diseases for a common outcome variable (e.g., patient's overall

assessment, pain, etc.). The data obtained in different studies will, I

hope, be discussed during international meetings such as OMERACT; such

discussions should result in proposals of endorsed cutoffs for the concepts

of improvement and state, for outcome variables to be used in evaluation of

the most frequent musculoskeletal disorders.

REFERENCES

Search PubMed for:

1. Jaeschke R, Singer J, Guyatt G. Measurement of health status.

Ascertaining the minimal clinically important difference. Control Clin

Trials 1989;10:407-15. [MEDLINE]

2. Wells G, J, Beaton D. Minimal clinically important difference

module: summary, recommendations and research agenda. J Rheumatol

2001;28:452-5. [MEDLINE]

3. Pavy S, Brophy S, Calin A. Establishment of the minimum clinically

important difference for the Bath Ankylosing Spondylitis Disease Index: a

prospective study. J Rheumatol 2005;32:80-5.

4. Schwartz AL, Meek PM, Nail LM, et al. Measurement of fatigue determining

minimally important clinical difference. J Clin Epidemiol 2002;55:239-44.

[MEDLINE]

5. Riddle DL, Stratford PW, Binkley M. Sensitivity to change of the

Roland- Back Pain Questionnaire: part 2. Phys Ther 1998;78:1197-207.

[MEDLINE]

6. Tubach F, Ravaud P, Baron G, et al. Evaluation of clinically relevant

changes in patient-reported outcomes in knee and hip osteoarthritis: The

minimal clinically important improvement [E-pub]. Ann Rheum Dis 2004;

Internet. [cited November 8, 2004]. Available from: //www.annrheumdis.com

7. Tubach F, Ravaud P, Baron G, et al. Evaluation of clinically relevant

states in patient-reported outcomes in knee and hip osteoarthritis: The

patient acceptable symptom state [E-pub]. Ann Rheum Dis 2004; Internet.

[cited November 8, 2004]. Available from: //www.annrheumdis.com

8. Wells G, Boers M, Shea B, et al. Low disease activity state for

rheumatoid arthritis: a preliminary definition. J Rheumatol 2005;32:in

press.

http://www.jrheum.com/subscribers/05/01/1.html

Not an MD

I'll tell you where to go!

Mayo Clinic in Rochester

http://www.mayoclinic.org/rochester

s Hopkins Medicine

http://www.hopkinsmedicine.org