Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

RAPID3 on an MDHAQ Agreement with DAS28 and CDAI Activity Categories, Scored in 5 Versus More Than 90 Seconds

RAPID3 on an MDHAQ Agreement with DAS28 and CDAI Activity Categories, Scored in 5 Versus More Than 90 Seconds - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related

RAPID3 (Routine Assessment of Patient Index
Data) on an MDHAQ (Multidimensional Health
Assessment Questionnaire): Agreement With
DAS28 (Disease Activity Score) and CDAI
(Clinical Disease Activity Index) Activity
Categories, Scored in Five Versus More Than
Ninety Seconds
Objective. To compare the Routine Assessment of Patient Index Data 3 (RAPID3) on a Multidimensional Health
Assessment Questionnaire (MDHAQ) with the Disease Activity Score (DAS28), Clinical Disease Activity Index (CDAI),
and individual core data set measures for correlations, agreement of activity levels, and time to score.
Methods. Four rheumatologists each assessed 50 patients with rheumatoid arthritis in “real-time” clinical care. Patients
completed an MDHAQ. The rheumatologist then calculated RAPID3 (physical function, pain, patient global estimate),
performed a 28-joint count, assigned a physician global estimate, and scored a CDAI, each timed by an observer.
Erythrocyte sedimentation rate (ESR) was tested on the same date, and the DAS28-ESR was computed later, again timed
by an observer. Spearman’s rank-order correlations and comparisons of patients classified as high activity, moderate
activity, low activity, and remission according to the DAS28, CDAI, and RAPID3 were computed and compared with
kappa statistics. A second study of 25 “paper patients” was also performed to compare time to score the DAS28, CDAI,
and RAPID3 on a 0–10 versus 0–30 scale. Mean and median times to score each index were computed.
Results. The 3 indices were correlated significantly, including agreement for>80% of patients for high/moderate activity.
The mean time to perform a 28-joint count was 94 seconds, and the mean times to score the DAS28, CDAI, RAPID3 on a
0–10 scale, and RAPID3 on a 0–30 scale were 114, 106, 9.6, and 4.6 seconds, respectively.
Conclusion. RAPID3 scores provide similar quantitative information to DAS28 and CDAI, while calculated on a 0–30
scale in about 5% of the time.
Quantitative assessment of patients with rheumatoid ar-
thritis (RA) has been extensively advanced over the last 3
decades (1), primarily in clinical trials and clinical re-
search. The only quantitative data collected at most usual
Supported by grants from the Arthritis Foundation, Am-
gen, Bristol-Myers Squibb, and the Jack C. Massey Founda-
Theodore Pincus, MD, Yusuf Yazici, MD: New York Uni-
versity Hospital for Joint Diseases, New York;
J. Swearingen, PhD: University of Arkansas for Medical
Sciences, Little Rock;
Martin J. Bergman, MD: Taylor Hos-
pital, Ridley Park, Pennsylvania;
C. Lee Colglazier, MD,
Arthur M. Kunath, MD: Kunath, Burte and Temming, Crest-
view Hills, Kentucky;
Alan T. Kaell, MD: Rheumatology
Associates of Long Island, Port Jefferson Station, New York;
Evan L. Siegel, MD: Arthritis & Rheumatism Associates,
Wheaton, Maryland.
Dr. Pincus has received consultant fees, speaking fees,
and/or honoraria (less than $10,000 each) from Centocor,
Inc., and UCB, Inc., and has received research grants from
Bristol-Myers Squibb, Inc., and UCB, Inc. Dr. Bergman has
received consultant fees, speaking fees, and/or honoraria
(less than $10,000 each) from Amgen, UCB, Roche/Genen-
tech, Bristol-Myers Squibb, and TAP, and (more than
$10,000) from Abbott. Dr. Kunath has received consultant
fees, speaking fees, and/or honoraria (less than $10,000
each) from Centocor and Abbott, and (more than $10,000
each) from Pfizer and Novartis. Dr. Siegel has received con-
sultant fees, speaking fees, and/or honoraria (less than
$10,000 each) from Abbott and Amgen, and owns stock
and/or holds stock options in Amgen. Dr. Yazici has re-
Arthritis Care & Research
Vol. 62, No. 2, February 2010, pp 181–189
DOI 10.1002/acr.20066
© 2010, American College of Rheumatology

care visits are laboratory tests, because other quantitative
RA core data set measures (2–4) of formal joint counts (5)
or patient questionnaire scores (6) are not included. Most
routine rheumatology care has not changed over many
years, with decisions largely based on nonquantitative im-
pressions, despite evidence of the value of quantitative
measures to recognize poor long-term outcomes (7–9) and
use of the Disease Activity Score (DAS28) to improve
outcomes in RA clinical trials (10–18).
No single measure can serve as a gold standard to assess
and monitor all of the individual patients with RA, and
pooled indices (19) such as the DAS28 (20,21) and Clinical
Disease Activity Index (CDAI) (22) are used. An index of
only the 3 patient self-report RA core data set measures,
the Routine Assessment of Patient Index Data 3 (RAPID3)
without joint counts, is correlated significantly with the
DAS28 and CDAI in clinical trials (23) and in clinical
settings (24,25). RAPID3 distinguishes active from control
treatments at levels similar to the DAS28 and CDAI in
clinical trials of leflunomide (26,27), methotrexate (26,27),
adalimumab (28), and abatacept (23). RAPID3 on a Multi-
dimensional Health Assessment Questionnaire (MDHAQ)
(see Supplementary Appendix A, available in the online
version of this article at http://www3.interscience.wiley.
com/journal/77005015/home) is scored in approximately
10 seconds (29), whereas a 28-joint count, required for the
DAS or CDAI, involves approximately 90 seconds (29).
Previous studies of RAPID3 did not include analyses of
the time to score the DAS28 and CDAI, and were per-
formed by 3 rheumatologists who had considerable expe-
rience with quantitative measurement (25,29). In the
present report, 4 additional rheumatologists who had not
participated in prior research on RA indices analyzed cor-
relations of RAPID3 with the DAS28 and CDAI, and the
time to score these indices.
Patients and rheumatologists. Patients were seen in
usual care by 4 rheumatologists (CLC, ATK, AMK, ELS) at
3 private practice settings. Each rheumatologist assessed
50 patients with RA. None of these rheumatologists had
participated in previous studies of correlations of indices
(25) or time to perform joint counts and score RAPID3 (29).
Patients signed consent for the results to be sent anony-
mously to a data center. The study was approved by local
Institutional Review Boards for the Protection of Human
MDHAQ and RAPID3 scores. The MDHAQ (30,31) (see
Supplementary Appendix A, available in the online ver-
sion of this article at http://www3.interscience.wiley.
com/journal/77005015/home) includes, on page 1, the 3
self-report RA core data set scores for physical function,
pain, and patient global estimate. A template to convert
0–3 scores for 10 individual physical function items to a
0–10 composite score is found in the column on the right
side of page 1 (see Supplementary Appendix A, available
in the online version of this article at http://www3.
interscience.wiley.com/journal/77005015/home). The pain
and global estimate visual analog scales (VAS) are in 21
numbered circles rather than a 10-cm line format (32) to
facilitate scoring. A template to convert RAPID3 scores
from 0–30 to 0–10 has been included in many versions of
the MDHAQ. Page 1 (see Supplementary Appendix A,
available in the online version of this article at http://
also includes a Rheumatoid Arthritis Disease Activity In-
dex (RADAI) self-report joint count (33). Page 2 of the
MDHAQ (see Supplementary Appendix A, available in the
online version of this article at http://www3.interscience.
wiley.com/journal/77005015/home) includes a review of
systems, recent medical history, fatigue VAS, queries
about exercise and change in status, and demographic
First study protocol: “real-time” comparisons of
DAS28, CDAI, and RAPID3 scores. The receptionist pre-
sented an MDHAQ to all patients upon registration for a
visit. Each patient completed the MDHAQ while waiting
to see the rheumatologist. The rheumatologist calculated a
RAPID3 score from the MDHAQ before or while seeing the
patient, and entered the score on the MDHAQ. An assis-
tant using a stopwatch recorded the number of seconds to
score RAPID3 on a “time to score” data sheet, and recorded
the assigned scores for individual measures and RAPID3
on a separate “measures and indices scores” data sheet.
The rheumatologist performed a standard 28 swollen
and tender joint count (34), scored on a joint count form at
a usual point in the visit. The joint count was again timed
by the assistant with a stopwatch, and recorded on the
time to score data sheet. The swollen and tender joint
count scores were entered on the measures and indices
scores data sheet. The rheumatologist assigned a physician
global estimate of status on a 0–10 scale (where 0 � best
status and 10 � poorest status) on the measures and indi-
ces scores data sheet, while the timer did not stop the
stopwatch, and then scored a CDAI (swollen joint count
[0–28 scale], tender joint count [0–28 scale], patient global
estimate [0–10 scale], physician global estimate [0–10
scale]; total CDAI: 0–76 scale) (22). The time in seconds to
score the CDAI was recorded by the assistant and entered
on the time to score data sheet, and the CDAI was recorded
on the measures and indices scores data sheet.
An erythrocyte sedimentation rate (ESR) was ordered.
At a later time, a DAS28-ESR (35) was calculated from the
swollen joint count, tender joint count, ESR, and patient
global estimate using the DAS Web site calculator (online
at: www.das-score.nl), timed by the observer and recorded
ceived consultant fees, speaking fees, and/or honoraria (less
than $10,000 each) from Bristol-Myers Squibb, Centocor,
Roche, UCB, and Celgene.
Address correspondence to Theodore Pincus, MD, New
York University Hospital for Joint Diseases, 301 East 17th
Street, Room 1608, New York, NY 10003. E-mail:
Submitted for publication April 7, 2009; accepted in re-
vised form September 23, 2009.
182 Pincus et al

on the time to score data sheet, with the DAS28-ESR re-
corded on the measures and indices scores data sheet.
Second study protocol: time to score DAS28, CDAI, and
RAPID3 as 0–30 and 0–10 scores in “paper patients.”
After the rheumatologists had enrolled 50 patients each in
the study, a second simpler study was conducted of the
time to score indices of 25 paper patients randomly se-
lected from the 200 in the real-time study. In this second
study, individual index component measures were listed
in columns on a separate page for each of the 4 indices for
25 patients in groups of 5, with a space to enter the index
score. The 4 indices were scored in order: RAPID3 on a
0–30 scale, without conversion of a 0–30 score to 0–10,
CDAI, DAS28, and RAPID3 as a 0–10 score, with conver-
sion from a 0–30 scale using a scoring template identical
to that found on the MDHAQ.
This simpler exercise included rheumatologists A, B, C,
and D, who each had studied 50 of their own patients in
the real-time study described above, as well as 2 additional
rheumatologists E and F (MJB, YY), who had participated
in previous studies (25,29). An assistant timed and re-
corded the number of seconds to score each index by the
rheumatologist, in groups of 5. The mean and median of
the total of 25 scores were then designated as the time to
score a single measure.
Statistical analyses. Demographic data, RA core data
set measures, and indices of 50 patients of each of the 4
rheumatologists were summarized and compared using
analysis of variance for continuous data and chi-square
tests for categorical data. Spearman’s rank-order correla-
tion coefficients were computed to compare individual
core data set measures, duration of disease, and DAS28,
CDAI, and RAPID3 scores.
Cross-tabulations were computed to compare the num-
ber and proportion of patients classified in the 4 DAS28
and CDAI categories of high disease activity (DAS28:�5.1,
CDAI: �22), moderate disease activity (DAS28: 3.21–5.1,
CDAI: 10.1–22), low disease activity (DAS28: 2.61–3.2,
CDAI: 2.81–10), and remission (DAS28: 0–2.6, CDAI:
0–2.8) with the 4 proposed RAPID3 categories, measured
on a 0–30 scale of high severity (12.1–30), moderate se-
verity (6.1–12.0), low severity (3.1–6.0), and remission
(0–3.0). The term “severity” rather than “activity” is used
to describe RAPID3 categories because a self-report index
may reflect joint damage as well as disease activity. Sta-
tistical significance of the level of agreement of the differ-
ent scales was evaluated using chi-square, kappa, and
weighted kappa statistics (36). Additional analyses were
performed for each index in 2 groups, high/moderate com-
pared with low/remission, because the cut points of 3.2 for
DAS28, 10 for CDAI, and 6 for RAPID3 appear to provide
critical levels for rheumatologists to recognize incomplete
responses and strongly consider a change in therapy.
The times to score each measure and index in the first
real-time protocol in 50 patients were summarized for
each rheumatologist and for the entire group. The times to
score the 25 paper patients in a controlled setting were
estimated by dividing each group of 5 observations by 5
and subsequently calculating the mean of these estimated
times by each of the 6 rheumatologists and the entire
group. All analyses were conducted using Stata statistical
software, version 9.2 (StataCorp, College Station, TX).
Patients. The mean age of the patients was 53.4 years,
the mean duration of disease was 11.6 years, and 81%
were women (Table 1). Mean values for RA core data set
measures were 4.8 swollen joints on a 28 swollen joint
count, 7.1 tender joints on a 28 tender joint count, physi-
cian global estimate of 3.5 (0–10 scale), ESR of 24.5 mm/
hour, physical function of 2.2 (on a 0–10 MDHAQ scale;
0.67 on a 0–3 scale), pain of 4.3 (0–10 scale), and patient
global estimate of 3.7 (0–10 scale). The mean DAS28 level
(0–10 scale) was 3.7, the mean CDAI (0–76 scale) was 19.2,
the mean RAPID3 on a 0–30 scale was 10.3, and the mean
RAPID3 on a 0–10 scale was 3.4.
Variation among patients of the 4 rheumatologists was
seen. Patients of rheumatologists A and D were younger
(these rheumatologists were younger) and had shorter du-
ration of disease. The patients of rheumatologist B had
significantly higher swollen and tender joint counts, ESR,
and therefore DAS28 and CDAI, than the other 3 rheuma-
tologists, although self-report questionnaire responses of
the patients did not differ significantly. Overall, the pa-
tients appear to be a typical group of patients with RA.
Correlations and severity categories of DAS28, CDAI,
and RAPID3. DAS28 versus RAPID3 scores were signifi-
cantly correlated (Spearman’s � � 0.43, range 0.36–0.61;
P � 0.001), and CDAI versus RAPID3 scores were signifi-
cantly correlated at higher levels (� � 0.61, range 0.54–
0.77; P � 0.001) (Table 2). High activity according to the
DAS28 (�5.1) was seen in 17% of the patients; 65% of
those patients also had high severity according to RAPID3
(�12), and 82% had high/moderate severity according to
RAPID3 (�6) (Table 3). Overall, 22% of the patients met
the criteria for DAS28 remission (�2.6); 41% of those
patients met RAPID3 near-remission criteria (�3), and
59% had low severity or remission (�6) (Table 3). The
kappa value was 0.16 (range 0.12–0.20) and the weighted
kappa value was 0.27 (range 0.13–0.33) for DAS28 versus
RAPID3, indicating fair agreement (Tables 2 and 3).
High activity according to the CDAI (�22) was seen in
33% of the patients; 68% of those patients had RAPID3
high severity (�12) and 88% had RAPID3 high/moderate
severity (�6) (Table 3). Only 8% of the patients met the
criteria for CDAI remission (�2.8); 75% of those patients
were also in RAPID3 remission (�3) (Table 3). The kappa
value was 0.27 (range 0.22–0.37) and the weighted kappa
value was 0.44 (range 0.26–0.58) for CDAI versus RAPID3,
again indicating fair agreement of the indices (Tables 2
and 3).
Patients were analyzed in 2 rather than 4 categories,
according to high/moderate activity or low activity/remis-
sion (Table 2), as the most important clinical level to
identify incomplete responses in patients with RA. Among
patients with high or moderate DAS28 activity (�3.2),
80% (range 61–97%) also had high or moderate RAPID3
RAPID3 Compared With DAS28 and CDAI 183

severity (�6) (Table 2). Among patients with low activity
or remission according to DAS28 (�3.2), 53% (range
48–56%) also had low activity or remission according to
RAPID3 (�6) (Table 2). The kappa was 0.34 (range 0.05–
0.48), indicating fair agreement.
Among patients with high or moderate CDAI activity
(�10), 86% (range 73–97%) had high or moderate RAPID3
severity (�6) (Table 2). Among patients with low activity
or remission according to CDAI (�10), 69% (range 59–
100%) had RAPID3 low activity/remission (�6) (Table 2).
The kappa was 0.55 (range 0.18–0.70), indicating moder-
ate agreement.
Time to score indices in real-time clinical care. The
mean time to score a 28-joint count in real-time clinical
care was 94 seconds (range 54–120 seconds) (Figure 1A).
The mean time to score a CDAI was 106 seconds (range
72–133 seconds). The mean time to score a DAS28 was 114
seconds (range 71–145 seconds). The mean time to score
RAPID3 as a 0–10 scale (including conversion from a 0–30
score) was 9.5 seconds (range 8.3–12.5 seconds) (Figure
Time to score indices in paper patients. A second,
simple study of time to score indices in paper patients
(Figure 1B) indicated a mean time to score a RAPID3 on a
0–10 scale of 9.6 seconds (range 7.5–13.4 seconds), CDAI
5.2 seconds (range 3.4–6.8 seconds), DAS28 (at the Web
site) 17.7 seconds (range 11.4–23.4 seconds), and RAPID3
on a 0–30 scale, without adjusting to a 0–10 scale, 4.6
seconds (range 3.6–5.9 seconds) (Figure 1B).
The mean time to score RAPID3 on a 0–10 scale was
identical to real-time clinical data. However, the mean
time to score a 0–30 scale was 5 seconds (range 2.0–6.2
seconds) less than on a 0–10 scale (Figure 1B). With in-
clusion of 94 seconds for a 28-joint count, the mean time to
score a DAS28 was 111.7 seconds and the mean time to
score a CDAI was 99.2 seconds, both similar to real-time
clinical data.
Quantitative measurement has advanced treatment of
many diseases, including RA, to recognize severe long-
term outcomes (7–9) and to improve outcomes in clinical
trials (10–18). However, most measures used in clinical
trials and other research, such as formal joint counts (5) or
patient questionnaires (6), have not been incorporated into
routine care to help guide clinical decisions. Most rheu-
matology visits are conducted similarly to the 1970s, with
laboratory tests as the only quantitative data.
A careful joint examination is required for a diagnosis of
RA. A formal quantitative swollen and tender joint count
and indices that include these measures, such as the
Table 1. Demographic characteristics, RA core data set measures, and clinical indices in 200 patients seen by 4
A(n� 50) B (n � 51) C (n � 50) D (n � 49)
(n � 200) P
Age, years 52.7 � 14.1 58.9 � 17.3 58.8 � 12.5 43.0 � 15.5 53.4 � 16.2 � 0.0001†
Duration, years 5.6 � 7.6 13.4 � 9.1 14.9 � 11.7 12.5 � 12.3 11.6 � 10.8 0.0001†
Education, years 13.7 � 2.4 13.8 � 2.8 13.1 � 2.6 13.9 � 2.9 13.6 � 2.7 0.4601†
Women, no. (%) 38 (76.0) 42 (82.4) 41 (82.0) 41 (83.7) 162 (81.0) 0.5940‡
Race, no. (%)
Non-Hispanic white 49 (98.0) 46 (90.2) 48 (96.0) 43 (87.7) 186 (93.0) 0.0340‡
African American 1 (2.0) 0 2 (4.0) 2 (4.1) 5 (2.5)
Hispanic 0 5 (9.8) 0 2 (4.1) 7 (3.5)
Asian 0 0 0 2 (4.1) 2 (4.1)
RA core data set measures
Physician/assessor measures
Swollen 28-joint count 2.8 � 3.8 10.1 � 6.6 3.3 � 3.7 2.9 � 2.6 4.8 � 5.4 � 0.0001†
Tender 28-joint count 7.2 � 6.9 14.6 � 7.5 4.0 � 4.5 2.5 � 3.3 7.1 � 7.5 � 0.0001†
Physician global estimate VAS 4.0 � 2.2 4.3 � 2.5 3.1 � 1.6 2.8 � 2.2 3.5 � 2.2 0.0009†
Laboratory measure
ESR, mm/hour 20.1 � 17.1 35.6 � 22.0 22.1 � 20.5 20.1 � 18.1 24.5 � 20.5 0.0001†
Patient measures
Function 2.4 � 1.9 2.4 � 1.8 2.4 � 2.0 1.7 � 1.8 2.2 � 1.9 0.1191†
Pain VAS 5.2 � 2.6 4.6 � 3.3 4.5 � 2.8 3.1 � 2.9 4.3 � 3.0 0.0036†
Patient global estimate VAS 4.2 � 2.7 4.1 � 2.9 4.0 � 2.9 2.7 � 2.7 3.7 � 2.8 0.0312†
Clinical indices
DAS28 3.4 � 1.4 5.3 � 1.0 3.2 � 1.2 3.0 � 1.0 3.7 � 1.5 � 0.0001†
CDAI 18.1 � 12.0 33.0 � 14.8 14.3 � 9.9 10.9 � 7.9 19.2 � 14.2 � 0.0001†
RAPID3 11.7 � 6.2 11.0 � 7.3 10.8 � 7.0 7.5 � 6.9 10.3 � 7.0 0.0106†
* Values are the mean � SD unless otherwise indicated. P values �0.05 were considered significant. RA � rheumatoid arthritis; VAS � visual analog
scale; ESR� erythrocyte sedimentation rate; DAS28� Disease Activity Score; CDAI� Clinical Disease Activity Index; RAPID3� Routine Assessment
of Patient Index Data 3.
† Continuous variable; calculated by analysis of variance.
‡ Discontinuous variable; calculated by chi-square test.
184 Pincus et al

DAS28 and CDAI, are the most specific measures of RA
activity (37). However, the most specific measure is not
necessarily the most sensitive to detect change. Relative
efficiencies of patient questionnaire scores to detect differ-
ences between active and control treatments in clinical
trials of methotrexate (38), leflunomide (38), anakinra (39),
Table 2. Spearman’s rank correlation coefficients, kappa, and weighted kappa coefficients for RAPID3 versus DAS28 and
CDAI, and agreement between moderate/high severity or low severity/remission categories according to RAPID3 scores versus
moderate/high activity or low activity/remission according to DAS28 and CDAI, in 200 patients seen by 4 rheumatologists*
A B C D Total
RAPID3 vs. DAS28
Spearman’s correlation coefficient 0.61 0.39 0.54 0.36 0.43
P � 0.001 0.005 � 0.001 0.010 � 0.001
Kappa 0.13 0.12 0.20 0.12 0.16
P 0.022 0.090 0.002 0.050 � 0.001
Weighted kappa 0.29 0.13 0.33 0.16 0.27
P � 0.001 0.050 � 0.001 0.43 � 0.001
Spearman’s correlation coefficient 0.66 0.54 0.77 0.70 0.61
P � 0.001 � 0.001 � 0.001 � 0.001 � 0.001
Kappa 0.37 0.22 0.25 0.22 0.27
P � 0.001 0.004 � 0.001 � 0.001 � 0.001
Weighted kappa 0.58 0.26 0.46 0.37 0.44
P � 0.001 � 0.001 � 0.001 � 0.001 � 0.001
Agreement of RAPID3/DAS28 categories
Moderate/high, no. (%) 28/29 (97) 35/49 (71) 23/25 (92) 11/18 (61) 97/121 (80)
Kappa 0.47 0.05 0.48 0.15 0.34
P � 0.001 0.257 � 0.001 0.141 � 0.001
Low/remission, no. (%) 10/21 (48) 1/2 (50) 14/25 (56) 17/31 (55) 42/79 (53)
Kappa –––– –
P –––– –
Agreement of RAPID3/CDAI categories
Moderate/high, no. (%) 32/33 (97) 36/49 (73) 28/29 (97) 17/21 (81) 113/132 (86)
Kappa 0.61 0.18 0.70 0.51 0.55
P � 0.001 0.03 � 0.001 � 0.001 � 0.001
Low/remission, no. (%) 10/17 (59) 2/2 (100) 15/21 (71) 20/28 (71) 47/68 (69)
Kappa –––– –
P –––– –
*RAPID3 � Routine Assessment of Patient Index Data 3; DAS28 � Disease Activity Score; CDAI � Clinical Disease Activity Index.
Table 3. RAPID3 scores compared with DAS28 and CDAI in 200 patients by 4
RAPID3 scores
DAS28 scores†
High activity (�5.1) 22 (64.7) 6 (17.7) 4 (11.8) 2 (5.9) 34 (17.0)
Moderate activity (3.21–5.1) 44 (50.6) 25 (28.7) 5 (5.8) 13 (14.9) 87 (43.5)
Low activity (2.61–3.2) 9 (25.7) 10 (28.6) 9 (25.7) 7 (20.0) 35 (17.5)
Remission (0–2.6) 7 (15.9) 11 (25.0) 8 (18.2) 18 (40.9) 44 (22.0)
Total 82 (41.0) 52 (26.0) 26 (13.0) 40 (20.0) 200
CDAI scores‡
High activity (�22) 45 (68.2) 13 (19.7) 4 (6.1) 4 (6.1) 66 (33.0)
Moderate activity (10.1–22.0) 31 (47.0) 24 (36.4) 5 (7.6) 6 (9.1) 66 (33.0)
Low activity (2.9–10) 6 (11.5) 15 (28.9) 13 (25.0) 18 (34.6) 52 (26.0)
Remission (0–2.8) 0 (0.0) 0 (0.0) 4 (25.0) 12 (75.0) 16 (8.0)
Total 82 (41.0) 52 (26.0) 26 (13.0) 40 (20.0) 200
* Values are the number (percentage). RAPID3 � Routine Assessment of Patient Index Data 3; DAS28 �
Disease Activity Score; CDAI � Clinical Disease Activity Index.
† DAS28 scores vs. RAPID3 scores: � � 0.16, weighted � � 0.27.
‡ CDAI scores vs. RAPID3 scores: � � 0.27, weighted � � 0.44.
RAPID3 Compared With DAS28 and CDAI 185

adalimumab (40), abatacept (41), and infliximab (42) are
similar to (often greater than) swollen and tender joint
counts or laboratory tests.
One concern of rheumatologists is that patient question-
naire scores may reflect irreversible joint damage, and
therefore be insensitive to detect changes in disease activ-
ity with treatment (43). However, the relative efficiencies
data are similar across all of the clinical trials studied for
joint count and patient questionnaire scores. These find-
ings suggest that joint counts appear to reflect a similar
level of irreversible changes as questionnaire scores, with
similar changes over short periods to document control of
The MDHAQ and RAPID3 scores do not replace either a
careful joint examination or conversation with the patient.
On the contrary, a careful joint examination is required to
interpret MDHAQ data and RAPID3 scores in clinical de-
cisions, and the MDHAQ enhances conversation with sav-
ing of time based on the RADAI self-report joint count (33)
on page 1, and a review of systems, recent medical history,
and demographic measures on page 2 (see Supplementary
Appendix A, available in the online version of this
article at http://www3.interscience.wiley.com/journal/
77005015/home) (44).
The MDHAQ was developed from the standard HAQ,
and RAPID3 on a 0–30 scale is scored in approximately 5
seconds versus 42 seconds to score a standard HAQ (29).
MDHAQ features that enhance clinical scoring include 10
rather than 20 physical function activities, scoring tem-
plates to convert raw physical function scores from 0–3 to
0–10 for RAPID3 composite scores, and VAS for pain and
global estimate comprised of 21 numbered circles rather
than a 10-cm line, to eliminate a need for a ruler.
Data in this report confirm and extend previous reports
that RAPID3 scores are significantly correlated with
DAS28 and CDAI scores in usual clinical care (25). Over-
all, 80% of the patients identified as having high or mod-
erate activity according to the DAS28 and CDAI, the level
at which rheumatologists might identify incomplete re-
sponse and strongly consider a change in therapy, had
Figure 1. Results of “real-time” scoring in the clinic and the “paper patient” scoring
exercise. A, Real-time clinic scoring: time in seconds required to score 28-joint count and 3
composite indices to assess 200 patients with rheumatoid arthritis by 4 rheumatologists (A,
B, C, and D; P � 0.0001 for all comparisons). B, Paper patient scoring exercise: average time
in seconds by all of the sites (6 rheumatologists) to calculate clinical outcome scores. CDAI�
Clinical Disease Activity Index; DAS28 � Disease Activity Score; RAPID3 � Routine
Assessment of Patient Index Data 3. Error bars show the SD. Values are expressed as mean
� SD.
186 Pincus et al

high or moderate activity according to the RAPID3. Fewer
patients had low activity/remission according to the RAP-
ID3 compared with the DAS28 and CDAI (25).
Correlations between the RAPID3 versus the DAS28 and
CDAI, and agreement of the proportions of patients with
various activity levels according to the DAS28, CDAI, and
RAPID3, were significant, although not as high in this
study as in the previous study (25). Nonetheless, kappa
levels for the 4 rheumatologists of 0.34 for the DAS28
versus RAPID3 and 0.55 for the CDAI versus RAPID3 in an
analysis of high/moderate activity versus low activity/re-
mission are in the range of comparisons of other clinical
measures for a similar construct. For example, kappa val-
ues for a normal versus abnormal ESR compared with a
normal versus abnormal C-reactive protein (CRP) level
were 0.21, 0.44, and 0.50, and a correlation of ESR and
CRP was 0.51 for patients of the 3 rheumatologists in the
previous study (ref. 25 and Pincus T: unpublished obser-
vations), which is lower than the RAPID3 compared with
the DAS28 and CDAI.
Data concerning time to score the indices are also simi-
lar to previous findings, as summarized in Figure 2, which
includes data from the previous report (29) as well as
real-time and paper patient scoring in this report. A swol-
len and tender 28-joint count was performed in 90 seconds
in the previous study and 94 seconds in the real-time
clinical study reported here. Scoring a HAQ disability
index in the previous study (29) involved 42 seconds and
was not studied here. The CDAI was scored in 106.1 sec-
onds in real time and 99.2 seconds in paper patient scoring
(94 seconds for the joint count plus 5.2 seconds for scor-
ing). The DAS28 was scored in 113.9 seconds in real time
and 111.7 seconds in paper patient scoring (94 seconds for
the joint count plus 17.7 seconds for scoring). The RAPID3
on a 0–10 scale was scored in 9.6 seconds in the previous
study, 9.5 seconds in real time in the present study, and
9.6 seconds in paper patient scoring. The RAPID3 on a
0–30 scale was scored in 4.6 seconds in paper patient
scoring in the present study.
Several limitations to this study are seen. First, only 4
rheumatologists were included, and variation in the re-
sults was seen. Nonetheless, agreement of indices for 2
rheumatologists is quite similar to or even greater than that
seen in the previous study of 3 rheumatologists (29), and
similarities in the aggregate are seen for 7 rheumatologists
whose results have been analyzed. Second, the conditions
are likely to underestimate the time required to score each
of the indices in a real-time situation, as the physicians
recognized that they were under observation. However,
comparative times appear to provide accurate relative es-
timates, and findings again appear similar from the real-
time and paper patient studies here and previous studies
(23,29). Third, if an assistant is available to perform a joint
count and score a DAS28 or CDAI, the time savings of
RAPID3 for the physician are to a large extent irrelevant.
However, assistants are unavailable to many rheumatolo-
gists, and if available, their time is valuable. Fourth, time
to score could be similar for all of the indices in an all-
electronic environment. However, the 90–94 seconds to
perform a formal joint count would remain, and no data
can be gleaned from electronic media in fewer than 5
All quantitative measures and indices of RA must be
interpreted by a caring and knowledgeable physician to
apply to an individual patient in formulating clinical de-
cisions. Neither RAPID3 nor any index should be regarded
as a substitute for a careful history and physical examina-
tion or as indicating an invariant clinical decision. Clinical
decisions are based on clinical judgment, incorporating all
types of information pertinent to individual patients.
Nonetheless, clinical judgment is better informed by avail-
ability of quantitative data, with RAPID3, as well as a
DAS28, CDAI or any measure or index, viewed as guides,
but not replacements, for clinical judgment.
The RAPID3 appears preferable to no quantitative clin-
ical data at all (other than laboratory tests), as usually seen
in contemporary rheumatology visits. All RA disease ac-
tivity measures and indices are surrogates, and limitations
are seen for the joint count (45) and DAS28 (37,46–48), as
well as for patient questionnaires and RAPID3 (45). None-
theless, physical function on a patient questionnaire, and
not a laboratory test or radiograph, provides the most
significant clinical prognostic indicator of most severe
5–10-year outcomes of RA (other than radiographic dam-
age), including work disability, costs, and mortality (49).
The MDHAQ and RAPID3 add no additional work for
the physician for comparison from one visit to another. A
receptionist, nurse clinician, or other assistant can easily
be taught to calculate RAPID3 scores using the scoring
templates on the MDHAQ, as used by the authors in this
study. A nonquantitative, careful joint examination, as
generally performed by most rheumatologists, with quan-
titative data from RAPID3 and a self-report RADAI joint
count to monitor clinical status, may be adequate for most
Figure 2. Mean time in seconds required to score 28-joint count,
Health Assessment Questionnaire disability index (HAQ-DI),
Clinical Disease Activity Index (CDAI), Disease Activity Score
(DAS28), Routine Assessment of Patient Index Data 3 (RAPID3) on
a 0–10 scale, and RAPID3 on a 0–30 scale in a previously pub-
lished study (29), in “real time” in clinical visits of 200 patients
seen by 4 rheumatologists, and in a “paper patient” scoring exer-
cise completed by 6 rheumatologists. Not all of the measures/
indices were evaluated in each study. The time required to per-
form a 28-joint count (94 seconds) is included in the time required
to score CDAI and DAS28 reported for the paper patient exercise.
Error bars show the SD. Values are expressed as mean � SD.
RAPID3 Compared With DAS28 and CDAI 187

patient care. MDHAQ/RAPID3 scores provide valid, reli-
able, feasible, and acceptable measures for standard clini-
cal care.
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors ap-
proved the final version to be submitted for publication. Dr. Pin-
cus had full access to all of the data in the study and takes
responsibility for the integrity of the data and the accuracy of the
data analysis.
Study conception and design. Pincus, Bergman, Colglazier,
Acquisition of data. Pincus, Bergman, Colglazier, Kaell, Kunath,
Siegel, Yazici.
Analysis and interpretation of data. Pincus, Swearingen, Berg-
man, Colglazier, Kaell, Yazici.
The work reported herein was an investigator-initiated study
supported in part by research grants to Dr. Pincus from Amgen
and Bristol-Myers Squibb. The funding sources played no role in
the study design, in the collection, analysis, or interpretation of
data, writing of the report, or decision to submit the manuscript
for publication. Publication was not contingent on approval by
the funding sources, and the content was not submitted to the
funding sources for approval.
1. Tugwell P, Boers M, Brooks PM, Simon L, Strand CV. Intro-
duction OMERACT 5: International Consensus Conference on
Outcome Measures in Rheumatology. J Rheumatol 2001;28:
2. Felson DT, Anderson JJ, Boers M, Bombardier C, Chernoff M,
Fried B, et al. The American College of Rheumatology prelim-
inary core set of disease activity measures for rheumatoid
arthritis clinical trials. Arthritis Rheum 1993;36:729–40.
3. Tugwell P, Boers M, and the OMERACT Committee. Proceed-
ings of the OMERACT conferences on outcome measures in
rheumatoid arthritis clinical trials: 29 April – 3 May 1992;
Maastricht, The Netherlands. J Rheumatol 1993;20:527–91.
4. Boers M, Tugwell P, Felson DT, van Riel PL, Kirwan JR,
Edmonds JP, et al. World Health Organization and Interna-
tional League of Associations for Rheumatology core end-
points for symptom modifying antirheumatic drugs in rheu-
matoid arthritis clinical trials. J Rheumatol Suppl 1994;41:
5. Pincus T, Segurado OG. Most visits of most patients with
rheumatoid arthritis to most rheumatologists do not include a
formal quantitative joint count. Ann Rheum Dis 2006;65:
6. Wolfe F, Pincus T, Thompson AK, Doyle J. The assessment of
rheumatoid arthritis and the acceptability of self-report ques-
tionnaires in clinical practice. Arthritis Rheum 2003;49:59–
7. Yelin E, Meenan R, Nevitt M, Epstein W. Work disability in
rheumatoid arthritis: effects of disease, social, and work fac-
tors. Ann Intern Med 1980;93:551–6.
8. Pincus T, Callahan LF. What is the natural history of rheuma-
toid arthritis? Rheum Dis Clin North Am 1993;19:123–51.
9. Wolfe F. The natural history of rheumatoid arthritis. J Rheu-
matol Suppl 1996;44:13–22.
10. Mottonen T, Hannonen P, Leirisalo-Repo M, Nissila M, Kau-
tiainen H, Korpela M, et al, and the FIN-RACo Trial Group.
Comparison of combination therapy with single-drug therapy
in early rheumatoid arthritis: a randomised trial. Lancet 1999;
11. Grigor C, Capell H, Stirling A, McMahon AD, Lock P, Vallance
R, et al. Effect of a treatment strategy of tight control for
rheumatoid arthritis (the TICORA study): a single-blind ran-
domised controlled trial. Lancet 2004;364:263–9.
12. Puolakka K, Kautiainen H, Mottonen T, Hannonen P, Korpela
M, Hakala M, et al, for the FIN-RACo Trial Group. Early
suppression of disease activity is essential for maintenance of
work capacity in patients with recent-onset rheumatoid
arthritis: five-year experience from the FIN-RACo trial. Arthri-
tis Rheum 2005;52:36–41.
13. Goekoop-Ruiterman YP, de Vries-Bouwstra JK, Allaart CF,
van Zeben D, Kerstens PJ, Hazes JM, et al. Clinical and radio-
graphic outcomes of four different treatment strategies in
patients with early rheumatoid arthritis (the BeSt study): a
randomized, controlled trial. Arthritis Rheum 2005;52:3381–
14. Goekoop-Ruiterman YP, de Vries-Bouwstra JK, Allaart CF,
van Zeben D, Kerstens PJ, Hazes JM, et al. Comparison of
treatment strategies in early rheumatoid arthritis: a random-
ized trial. Ann Intern Med 2007;146:406–15.
15. Verstappen SM, Jacobs JW, van der Veen MJ, Heurkens AH,
Schenk Y, ter Borg EJ, et al. Intensive treatment with metho-
trexate in early rheumatoid arthritis: aiming for remission.
Computer Assisted Management in Early Rheumatoid Arthri-
tis (CAMERA, an open-label strategy trial). Ann Rheum Dis
16. Hetland ML, Ejbjerg BJ, Horslev-Petersen K, Jacobsen S, Vest-
ergaard A, Jurik AG, et al. MRI bone oedema is the strongest
predictor of subsequent radiographic progression in early
rheumatoid arthritis: results from a 2 year randomised con-
trolled trial (CIMESTRA). Ann Rheum Dis 2009;68:384–90.
17. Hetland ML, Stengaard-Pedersen K, Junker P, Lottenburger T,
Hansen I, Andersen LS, et al. Aggressive combination therapy
with intra-articular glucocorticoid injections and conven-
tional disease-modifying anti-rheumatic drugs in early rheu-
matoid arthritis: second-year clinical and radiographic results
from the CIMESTRA study. Ann Rheum Dis 2008;67:815–22.
18. Saunders SA, Capell HA, Stirling A, Vallance R, Kincaid W,
McMahon AD, et al. Triple therapy in early active rheumatoid
arthritis: a randomized, single-blind, controlled trial compar-
ing step-up and parallel treatment strategies. Arthritis Rheum
19. Goldsmith CH, Smythe HA, Helewa A. Interpretation and
power of pooled index. J Rheumatol 1993;20:575–8.
20. Van der Heijde DM, van ‘t Hof MA, van Riel PL, Theunisse
LM, Lubberts EW, van Leeuwen MA, et al. Judging disease
activity in clinical practice in rheumatoid arthritis: first step
in the development of a disease activity score. Ann Rheum
Dis 1990;49:916–20.
21. Van der Heijde DM, van ‘t Hof M, van Riel PL, van de Putte
LB. Development of a disease activity score based on judg-
ment in clinical practice by rheumatologists. J Rheumatol
22. Aletaha D, Smolen J. The Simplified Disease Activity Index
(SDAI) and the Clinical Disease Activity Index (CDAI): a
review of their usefulness and validity in rheumatoid arthri-
tis. Clin Exp Rheumatol 2005;23:S100–8.
23. Pincus T, Bergman MJ, Yazici Y, Hines P, Raghupathi K,
Maclean R. An index of only patient-reported outcome mea-
sures, Routine Assessment of Patient Index Data 3 (RAPID3),
in two abatacept clinical trials: similar results to Disease
Activity Score (DAS28) and other RAPID indices that include
physician-reported measures. Rheumatology (Oxford) 2008;
24. Wolfe F, Michaud K, Pincus T. A composite disease activity
scale for clinical practice, observational studies and clinical
trials: the Patient Activity Scale (PAS/PAS-II). J Rheumatol
25. Pincus T, Swearingen CJ, BergmanM, Yazici Y. RAPID3 (Rou-
tine Assessment of Patient Index Data 3), a rheumatoid arthri-
tis index without formal joint counts for routine care: pro-
posed severity categories compared to DAS and CDAI
categories. J Rheumatol 2008;35:2136–47.
26. Pincus T, Strand V, Koch G, Amara I, Crawford B, Wolfe F, et
al. An index of the three core data set patient questionnaire
188 Pincus et al

measures distinguishes efficacy of active treatment from pla-
cebo as effectively as the American College of Rheumatology
20% response criteria (ACR20) or the Disease Activity Score
(DAS) in a rheumatoid arthritis clinical trial. Arthritis Rheum
27. Pincus T, Amara I, Koch GG. Continuous indices of core data
set measures in rheumatoid arthritis clinical trials: lower re-
sponses to placebo than seen with categorical responses with
the American College of Rheumatology 20% criteria. Arthritis
Rheum 2005;52:1031–6.
28. Pincus T, Chung C, Segurado OG, Amara I, Koch GG. An
index of patient self-reported outcomes (PRO Index) discrim-
inates effectively between active and control treatment in 4
clinical trials of adalimumab in rheumatoid arthritis. J Rheu-
matol 2006;33:2146–52.
29. Yazici Y, Bergman M, Pincus T. Time to score quantitative
rheumatoid arthritis measures: 28-joint count, Disease Activ-
ity Score, Health Assessment Questionnaire (HAQ), multidi-
mensional HAQ (MDHAQ), and Routine Assessment of Pa-
tient Index Data (RAPID) scores. J Rheumatol 2008;35:603–9.
30. Pincus T, Swearingen C, Wolfe F. Toward a multidimensional
Health Assessment Questionnaire (MDHAQ): assessment of
advanced activities of daily living and psychological status in
the patient-friendly health assessment questionnaire format.
Arthritis Rheum 1999;42:2220–30.
31. Pincus T, Sokka T, Kautiainen H. Further development of a
physical function scale on a multidimensional Health Assess-
ment Questionnaire for standard care of patients with rheu-
matic diseases. J Rheumatol 2005;32:1432–9.
32. Pincus T, Bergman M, Sokka T, Roth J, Swearingen C, Yazici
Y. Visual analog scales in formats other than a 10 centimeter
horizontal line to assess pain and other clinical data. J Rheu-
matol 2008;35:1550–8.
33. Stucki G, Liang MH, Stucki S, Bruhlmann P, Michel BA. A
self-administered Rheumatoid Arthritis Disease Activity In-
dex (RADAI) for epidemiologic research. Arthritis Rheum
34. Fuchs HA, Brooks RH, Callahan LF, Pincus T. A simplified
twenty-eight–joint quantitative articular index in rheumatoid
arthritis: psychometric properties and correlation with pa-
rameters of disease activity. Arthritis Rheum 1989;32:531–7.
35. Prevoo ML, van ‘t Hof MA, Kuper HH, van Leeuwen MA, van
de Putte LB, van Riel PL. Modified disease activity scores that
include twenty-eight–joint counts: development and valida-
tion in a prospective longitudinal study of patients with rheu-
matoid arthritis. Arthritis Rheum 1995;38:44–8.
36. Landis JR, Koch GG. The measurement of observer agreement
for categorical data. Biometrics 1977;33:159–74.
37. Pincus T. The DAS is the most specific measure, but a patient
questionnaire is the most informative measure to assess rheu-
matoid arthritis. J Rheumatol 2006;33:834–7.
38. Strand V, Cohen S, Crawford B, Smolen JS, Scott DL. Patient-
reported outcomes better discriminate active treatment from
placebo in randomized controlled trials in rheumatoid arthri-
tis. Rheumatology (Oxford) 2004;43:640–7.
39. Cohen SB, Strand V, Aguilar D, Ofman JJ. Patient- versus
physician-reported outcomes in rheumatoid arthritis patients
treated with recombinant interleukin-1 receptor antagonist
(anakinra) therapy. Rheumatology (Oxford) 2004;43:704–11.
40. Pincus T, Amara I, Segurado OG, Bergman M, Koch GG.
Relative efficiencies of physician/assessor global estimates
and patient questionnaire measures are similar to or greater
than joint counts to distinguish adalimumab from control
treatments in rheumatoid arthritis clinical trials. J Rheumatol
41. Wells G, Li T, Maxwell L, Maclean R, Tugwell P. Responsive-
ness of patient reported outcomes including fatigue, sleep
quality, activity limitation, and quality of life following treat-
ment with abatacept for rheumatoid arthritis. Ann Rheum Dis
42. Pincus T, Zelinger D, Bolce RJ. High/moderate versus low
activity/remission patient proportions are similar according
to DAS28 (Disease Activity Score), CDAI (Clinical Disease
Activity Index) and RAPID3 (Routine Assessment of Patient
Index Data) in ATTRACT and ASPIRE Infliximab (INFX) clin-
ical trials in patients with rheumatoid arthritis (RA) [abstract].
Ann Rheum Dis 2009;68 Suppl 3:433.
43. Aletaha D, Smolen J, Ward MM. Measuring function in rheu-
matoid arthritis: identifying reversible and irreversible com-
ponents. Arthritis Rheum 2006;54:2784–92.
44. Pincus T, Yazici Y, Bergman M. Saving time and improving
care with a multidimensional health assessment question-
naire: 10 practical considerations. J Rheumatol 2006;33:448–
45. Pincus T. Limitations of a quantitative swollen and tender
joint count to assess and monitor patients with rheumatoid
arthritis. Bull NYU Hosp Jt Dis 2008;66:216–23.
46. Wolfe F, Michaud K, Pincus T, Furst D, Keystone E. The
Disease Activity Score is not suitable as the sole criterion for
initiation and evaluation of anti–tumor necrosis factor ther-
apy in the clinic: discordance between assessment measures
and limitations in questionnaire use for regulatory purposes.
Arthritis Rheum 2005;52:3873–9.
47. Coury F, Rossat A, Tebib A, Letroublon MC, Gagnard A,
Fantino B, et al. Rheumatoid arthritis and fibromyalgia: a
frequent unrelated association complicating disease manage-
ment. J Rheumatol 2009;36:58–62.
48. Makinen H, Hannonen P. How to assess patients with rheu-
matoid arthritis and concomitant fibromyalgia? J Rheumatol
49. Pincus T, Wolfe F. Patient questionnaires for clinical research
and improved standard patient care: is it better to have 80% of
the information in 100% of patients or 100% of the informa-
tion in 5% of patients? J Rheumatol 2005;32:575–7.
RAPID3 Compared With DAS28 and CDAI 189