Asthma Control Test

Primary Care Respiratory Journal (2009);18(1): 41-49
ORIGINAL RESEARCH
The Asthma Control Test
TM
(ACT) as a predictor of GINA
guideline-defined asthma control: analysis of a
multinational cross-sectional survey
*Mike Thomas
a
, Stephen Kay
b
, James Pike
b
, Angela Williams
c
,
Jacqueline R Carranza Rosenzweig
c
, Elizabeth V Hillyer
d
, David Price
a
a
C entre of A cadem ic Prim ary C are, U niversity of A berdeen, Scotland, U K
b
A delphi Real W orld Products, Bollington, U K
c
G lobal H ealth O utcom es, G laxoSm ithKline R& D , Research Triangle Park, N orth C arolina, U SA
d
Respiratory Research Ltd, N orw ich, U K
Subm itted 28th M ay 2008; revised version received 18th Septem ber 2008; further revisions received 23rd O ctober 2008;
accepted 28th N ovem ber 2008; online 24th February 2009
Abstract
Aims: To evaluate w hether the A sthm a C ontrol Test
TM
(A C T) score is predictive of G lobal Initiative for A sthm a (G IN A ) guideline-defined
classification levels of asthm a control. The A C T is a validated, 5-item , patient-com pleted m easure of asthm a control w ith a recall period
of four w eeks.
Methods: C ross-sectional survey com paring A C T score and G IN A classification of asthm a control am ong 2949 patients attending prim ary
care physicians and specialists in France, G erm any, Italy, Spain, the U K, and the U SA .
Results: The area under the receiver operating characteristics curve for A C T score predicting G IN A control w as 0.84 (95% C I 0.820.85).
A n A C T score of <19 (not w ell-controlled asthm a) correctly predicted G IN A -defined partly controlled/uncontrolled asthm a 94% of the
tim e, w hile an A C T score of >20 predicted G IN A -defined controlled asthm a 51% of the tim e, w ith kappa statistic of 0.42, representing
m oderate agreem ent.
Conclusions: A n A C T score <19 is useful for identifying patients w ith poorly controlled asthm a as defined by G IN A .
2009 G eneral Practice A irw ays G roup. A ll rights reserved.
M Thom as, et al. Prim Care Resp J 2009; 18(1): 41-49.
doi:10.4104/pcrj.2009.00010
Keywords asthma, Asthma Control Test, control, GINA, exacerbations, multinational survey, patient-completed outcome assessment,
Europe, United States
* Corresponding author: D r M ike Thom as, C entre of A cadem ic Prim ary C are, U niversity of A berdeen, Scotland, A B25 2AY, U K.
Tel: +44 (0)1285 760671 Fax: +44 (0)1224 550683 E-m ail: m ikethom as@ doctors.org.uk
41
PRIMARY CARE RESPIRATORY JOURNAL
w w w .thepcrj.org
doi:10.4104/pcrj.2009.00010
Introduction
The 2006 update to the G lobal Initiative for A sthm a (G IN A )
guideline em phasises the im portance of evaluating asthm a
control, rather than asthm a severity, in order to guide asthm a
m anagem ent decisions. C lassification of disease severity is a
static m easure that, w hilst useful in initiating treatm ent, is less
helpful in guiding subsequent treatm ent.
1,2
G IN A guidelines
suggest that classification of asthm a control m ore directly
reflects the effectiveness of therapeutic interventions and
thus it m ay be m ore useful clinically. C urrent guidelines define
asthm a control as: no lim itations of activities; no nocturnal
sym ptom s; m inim al or no daytim e sym ptom s; m inim al or no
need for rescue therapy; norm al lung function; and no
exacerbations.
3-5
G uideline-defined asthm a control can be attained and
m aintained for the m ajority of patients eligible to participate
in a controlled trial setting,
6,7
but it is frequently not achieved
in real w orld practice.
8-11
Poorly controlled asthm a accounts
for a disproportionate share of the costs of asthm a and
represents a heavy socioeconom ic burden.
12-15
H ow ever, m any
patients w orldw ide have sub-optim ally controlled asthm a,
The full version of this paper, w ith A ppendices and online
Tables and Figures, is available online at w w w .thepcrj.org
Copyright GPIAG - reproduction prohibited
http://www.thepcrj.org
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
M Thomas et al.
42
and they are often not aw are that better control can be
achieved; m oreover, physicians m ay overestim ate levels of
control or the extent of im provem ent achieved w ith therapy,
often because of inadequate assessm ent.
16,17
In addition,
selection of asthm a control criteria not consistent w ith
current asthm a guidelines m ay ham per assessm ent.
18
The A sthm a C ontrol Test
TM
(A C T) w as developed as a
screening tool to address the need for a sim ple, rapidly-
com pleted assessm ent tool in clinical practice. The A C T is a
validated, patient-com pleted m easure of asthm a control
com prising five questions that assess activity lim itation, shortness
of breath, night-tim e sym ptom s, use of rescue m edication, and
patient overall rating of asthm a control over the previous four
w eeks (see A ppendix 1 online).
19-21
The questions are scored from
1 (w orst) to 5 (best), and the A C T score is the sum of the
responses, giving a m axim um best score of 25. A n A C T score of
19 dem onstrated the highest area under the receiver operating
characteristic (RO C ) curve, and thus a score of >20 is the optim al
cut-off point defining w ell-controlled asthm a over the previous
four w eeks
19-22
although, as N athan et al
19
describe, the cut-off
point can be chosen according to application.
It is im portant in clinical practice to identify patients
w hose asthm a is not w ell-controlled, since these patients
require review of their therapy as w ell as assessm ent of risk
factors for poor asthm a control.
23
The A C T w as designed for
use in daily practice as a supplem entary m easure to the
physicians assessm ent and/or lung function testing, but it has
not been validated as a predictor of G IN A -defined asthm a
control. The objective of this m ultinational cross-sectional
survey w as to evaluate w hether the A C T can predict G IN A -
defined asthm a control, w ith particular em phasis on the
binary split betw een G IN A -defined partly
controlled/uncontrolledasthm a versus controlledasthm a.
Methods
Disease Specific Programmes
The D isease Specific Program m es (D SP) are large,
m ultinational observational studies of clinical practice that are
conducted every 12 to 18 m onths by the A delphi G roup in
the U SA and in five European countries (France, G erm any,
Italy, Spain, and the U K).
24
D esigned to survey patients and
physicians on their perceptions of treatm ent effectiveness,
sym ptom s, and im pact of com m on chronic diseases, each
D SP is specific to a disease area and includes questionnaires
com pleted by up to 1000 physicians and 12,000 patients.
D escriptions of D SP m ethods for studying asthm a and allergic
rhinitis have been published.
25,26
The respiratory program m e w as initiated in 2000. H ere w e
sum m arise the m ethods specific to patients w ith asthm a
included in the sixth w ave (Respiratory D SP VI),
conducted in the first quarter of 2007.
Respiratory DSP VI survey participants and
procedures
Physicians recruited for the Respiratory D SP VI num bered 120
in France, G erm any, Italy, and Spain (50 prim ary care
practitioners, 50 pulm onologists or country equivalent, and
20 allergists in each country), 100 in the U K (50 prim ary care
practitioners and 50 chest specialists), and 180 in the U S (75
prim ary care practitioners, 75 pulm onologists, and 30
allergists). Larger sam ples of physicians w ere recruited from
m ore densely populated areas. Physicians w ere asked to
collect inform ation for their next six consecutive patients >12
years of age w ith physician-diagnosed asthm a, irrespective of
the reason for the consultation. The physician form s,
com pleted after the consultation w ith no direct input from
the patient, included patient dem ographic data, disease
sym ptom s and severity, diagnostic and treatm ent history, and
health resource utilisation m easures.
Patients included in the survey by their physicians could
then be invited to com plete a survey form im m ediately after
the consultation. C om pletion of the survey form w as not
obligatory, and physicians did not see or influence patient
responses. The study protocol follow ed ethical procedures
including verbal inform ed consent of all physicians and
patients for anonym ous and aggregated reporting of research
findings based on the questionnaires em ployed.
The patient survey included questions on asthm a
sym ptom s and severity, the im pact of asthm a and any
m easures taken by the patient to control their asthm a, as w ell
as satisfaction and com pliance w ith m edication, expectations
of therapy, and health resource utilisation. Physicians w ere
provided w ith an honorarium for study com pletion. Patient
com pensation depended on local country regulations (som e
w ere not com pensated, som e received vouchers, som e
received paym ent).
Outcome measures
Included in the Respiratory DSP VI survey w ere the ACT tool and
variables to identify asthm a control level as per G INA classification
(other survey data not reported here). The ACT w as included as
part of the patient-com pleted form , and G INA-defined asthm a
control w as determ ined prim arily from patient responses.
Translations of the A C T w ere validated. The translations of
the Respiratory D SP VI survey w ere m ade using forw ard-
backw ard translation and w ere not validated. Specific cognitive
debriefing w as not done before initiating the survey; how ever,
the form s w ere piloted to see how they w ere com pleted, given
that patients w ould be com pleting them alone.
The G IN A classification and our survey definitions of G IN A
asthm a control are sum m arised in Table 1, w ith differences
betw een the tw o noted (all relate to tim e fram e). Q uestions
relating to the first four item s in the G IN A classification
daytim e sym ptom s, lim itations of activities, nocturnal
w w w .thepcrj.org
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
The ACT as a predictor of GINA-defined asthma control
43
sym ptom s, and need for rescue m edication w ere included
on both physician and patient form s w ith reference to the
previous four w eeks, as in the A C T (online A ppendices 1, 2,
and 3). Q uestioning on the fifth item in the G IN A
classification lung function w as included on the physician
form w ith reference to the previous 12 m onths.
Q uestioning about the sixth item in the G IN A control
definition asthm a exacerbations w as w ith reference to the
preceding 12 m onths on both physician and patient form s;
how ever, the tim ing of the exacerbation w as enquired about
on (and thus data derived from ) only the physician-com pleted
form s (see A ppendices 2 and 3 online). O f note, the G IN A
definitions for partly controlled and uncontrolled categories
are not m utually exclusive based on the exacerbation item
because the tim e period for an exacerbation is not specified
(see Table 1).
3
W e therefore elected to specify an exacerbation
in the preceding seven days as defining uncontrolled asthm a,
and an exacerbation w ithin the preceding year (but not in the
previous seven days) to define partly controlled asthm a in
term s of the G IN A classification.
Analysis
Physicians and their patients w ere linked by assigned study
num bers, and data included in the analyses w ere anonym ised
and restricted to those from m atched and fully com pleted
patient and physician form s. The prim ary analyses w ere
perform ed including all patients as w ell as by country.
The analyses evaluated the relationship betw een A C T
scores and G IN A -defined asthm a control, taking the G IN A
classification as the trueclassification and the A C T score as
the predictorclassification. O ur prim ary analyses evaluated
the relationship betw een A C T scores and G IN A -defined partly
controlled/uncontrolled versus controlled asthm a. In
w w w .thepcrj.org
Levels of A sthm a C ontrol
C haracteristic C ontrolled Partly controlled U ncontrolled
(A ll of the follow ing) (A ny m easure present
in any week)
D aytim e sym ptom s N one (tw ice or less/w eek) M ore than tw ice/w eek
Lim itations of activities N one A ny Three or m ore
N octurnal sym ptom s/aw akening N one A ny features of partly
N eed for reliever/rescue treatm ent N one (tw ice or less/w eek) M ore than tw ice/w eek controlled asthm a
Lung function (PEF or FEV
1
) N orm al <80% predicted or personal present in any week
best (if know n)
Exacerbations N one O ne or m ore/year* O ne in any week
FEV
1
= forced expiratory volum e in 1 second; PEF = peak expiratory flow .
*A ny exacerbation should prom pt review of m aintenance treatm ent to ensure that it is adequate.
By definition, an exacerbation in any w eek m akes that an uncontrolled asthm a w eek.
Lung function is not a reliable test for children 5 years and younger.
Table 1A. Levels of asthma control according to the Global Initiative for Asthma (GINA).
3
Reprinted with permission.
Levels of A sthm a C ontrol
C haracteristic C ontrolled Partly controlled U ncontrolled
(A ll of the follow ing) (A ny m easure present
in previous 4 weeks )
D aytim e sym ptom s* N one (tw ice or less/w eek) M ore than tw ice/w eek
Lim itations of activities* N one A ny Three or m ore
N octurnal sym ptom s/aw akening* N one A ny features of partly
N eed for reliever/rescue treatm ent* N one (tw ice or less/w eek) M ore than tw ice/w eek controlled asthm a
present in any w eek
of the last 4 weeks
Exacerbations N one O ne or m ore in prior year O ne or m ore in the
previous 7 days
*D ata derived from patient-com pleted form s
D ata derived from physician-com pleted form s
Text in bold italics signifies differences from G IN A definitions in Table 1A

Table 1B. Working definition of GINA-defined asthma control used in the survey.
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
M Thomas et al.
44
sensitivity analyses, w e exam ined the effect of including
exacerbations for the G IN A definitions of uncontrolled and
partly controlled asthm a; thus, w e re-ran the analyses: firstly
w ith all exacerbation data taken out; and secondly w ith
exacerbations defining partly controlled asthm a taken out
(i.e., exacerbations >7 days to 1 year) but leaving them in for
defining uncontrolled asthm a. A n additional sensitivity
analysis, w ith all exacerbations rem oved, focused on
uncontrolled asthm a (thus, uncontrolled versus partly
controlled/controlled asthm a). W e calculated the sensitivity,
specificity, positive predictive value (predictive value of a
positive test), negative predictive value (predictive value of a
negative test), and percentage of patients correctly classified
overall using each A C T score as the cut-off point for G IN A -
defined controlled asthm a and, in the sensitivity analysis, the
cut-off point for G IN A -defined uncontrolled asthm a.
For the prim ary analysis, w e plotted RO C curves (sensitivity
versus 1 specificity) for the full range of A C T cut-off points.
From these w e determ ined the area under the curve (A U C ),
w ith 95% confidence intervals (C I). The A U C sum m arises the
relationship betw een the tw o m easures by incorporating
inform ation from all A C T values. If the A C T score w ere a
perfect predictor this area w ould equal 1; if it w ere no better
than random chance it w ould equal 0.5 (the straight line
draw n on the RO C curves). W e tested for differences in the
A U C betw een countries using the W ald X
2
test.
The kappa statistic w as used to m easure agreem ent
betw een an A C T cut-off point of >20 defining w ell-controlled
asthm a and the G IN A binary split of partly
controlled/uncontrolled versus controlled asthm a.
27
In
addition, w e evaluated the kappa statistic for defining
uncontrolled asthm a in the sensitivity analysis.
D escriptive statistics w ere used to sum m arise A C T scores
relative to the three G IN A classifications of asthm a control
including the exacerbation criterion. A s A C T scores w ere non-
norm ally distributed across all G IN A as w ell as G IN A partly
controlled and uncontrolled categories, w e report m edian
A C T scores w ith interquartile ranges (IQ R) for each category.
The Kruskal-W allis test w as used to test for differences in A C T
m edians am ong G IN A categories. Pairw ise com parisons w ere
m ade using the M ann-W hitney U test w ith Bonferroni
adjustm ents for m ultiple testing.
A ll statistical analysis w as carried using STATA
Version 10
(StataC orp LP, C ollege Station, Texas, U S).
Results
Survey participants
Physicians com pleted asthm a patient record form s for 3503
patients in Europe and 1080 patients in the U SA . Eighty-five
percent of these form s (3877/4583) could be m atched w ith a
corresponding patient self-com pleted form . A total of 2949
of the 3877 (76% ) patient form s, corresponding to 64% of
physician form s, had com plete responses for the A C T- and
G IN A -related questions and w ere included in the analyses.
Patient num bers by country are sum m arised in Table 2.
Patients ranged in age from 12 to 93 years (m ean [SD ], 42
[17]), and 56% (1657/2941) w ere fem ale (a few dem ographic
data w ere m issing). M ost patients w ere w hite (2598/2930 or
89% ); patients of H ispanic (102, 3% ) and A frican-A m erican
descent (98, 3% ) w ere equally represented, w ith sm aller
num bers w ho described them selves as A fro-C aribbean (62,
2% ), A sian (55, 2% ), and other (15, <1% ). Tw elve percent
w ere current sm okers, 24% w ere ex-sm okers, and 64% had
never sm oked.
Respiratory DSP VI Survey results
Lung function data w ere available from only 539 of 2949
(18% ) m atched physician-com pleted form s and thus w ere
not included in the m ain analyses. H ow ever, analyses
incorporating lung function data gave results very sim ilar to
those obtained w hen the data w ere excluded, both for the
overall study population as w ell as for the subgroup w ith lung
function data (online Tables 13, see w w w .thepcrj.org).
GINA partly controlled/uncontrolled versus controlled
asthma as defined by ACT score
W ith a cut-off point of >20 for the A C T score defining w ell-
controlled asthm a, and a binary split for G IN A classification
(partly controlled/uncontrolled versus controlled asthm a), an
A C T score of <19 (not w ell-controlled asthm a) correctly
predicted G IN A -defined partly controlled/uncontrolled
asthm a 93.9% of the tim e (Table 3). A n A C T score of >20
predicted G IN A -defined controlled asthm a 51.3% of the
tim e. Table 4 sum m arises the positive and negative predictive
values, as w ell as sensitivity and specificity, of the A C T score
cut point of >20 for all patients and by country.
The area under the RO C curve for the A C T score
predicting the G IN A control classification w as 0.84 (95% C I
w w w .thepcrj.org
PRF Matched PSC Matched PSC
for analysis*
(n=4583) (n=3877) (n=2949)
France 741 732 (99% ) 697 (94% )
G erm any 720 710 (99% ) 495 (69% )
Italy 720 702 (98% ) 495 (69% )
Spain 726 573 (79% ) 395 (54% )
U nited Kingdom 596 200 (34% ) 154 (26% )
U nited States 1080 960 (89% ) 713 (66% )
*N o. form s w ith corresponding PRF as w ell as com plete responses for inclusion
in the analyses.
Percentages are relative to num ber of PRFs for that country.

Table 2. Number of physician-completed patient record
forms (PRF) and patient self-completed (PSC) responses
available for analysis, by country.
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
45
0.820.85) for all countries com bined. This A U C can be
interpreted as the probability that the A C T score for a
random ly selected patient w ith partly controlled/uncontrolled
asthm a w ill be less than that for a random ly selected patient
w ith controlled asthm a. A m ong individual countries, the area
under the RO C curve varied significantly (W ald X
2
p=0.0048),
from 0.80 (Italy) to 0.89 (U K) see Figure 1.
U sing the cut-off point of >20 for A C T w ell-controlled
asthm a, the kappa level of agreem ent for the entire patient
population w as 0.42. The kappa statistic is a m eans of
m easuring agreem ent beyond chance betw een tw o sets of
observations using categorical data and is interpreted as
follow s: 0.811.0, alm ost perfect; 0.610.80, substantial;
0.410.60, m oderate; 0.210.40, fair; and 0.00.20, slight;
and <0, poor agreem ent.
27
Individual country kappa scores
(0.290.52) represented fair to m oderate agreem ent (Table
5).
Sensitivity analyses
In sensitivity analyses, w hen w e rem oved all exacerbations as
a criterion, m ore patients w ere defined as having G IN A -
controlled asthm a. The positive predictive value of the A C T
rem ained sim ilar w hile the negative predictive value
im proved: an A C T score of <19 correctly predicted G IN A -
defined partly controlled/uncontrolled asthm a 88.4% of the
tim e, and an A C T score of >20 predicted G IN A -defined
controlled asthm a 71.6% of the tim e (online Table 4). The
area under the RO C curve im proved to 0.89 and the kappa
statistic im proved substantially to 0.58. W hen w e rem oved
only exacerbations occurring >7 days to 1 year previously, the
results w ere virtually identical (sam e kappa and area under
RO C curve to tw o digits; data not show n).
O nline Table 5 and Figures 1 and 2 report the results of
sensitivity analyses using the binary split of G IN A uncontrolled
versus partly controlled/controlled asthm a and rem oving
exacerbations as a criterion. A s com pared w ith the prim ary
analyses, the positive predictive value fell substantially
(33.6% ) w hile the negative predictive value of an A C T score
>20 rose to alm ost 99% , i.e. correctly predicting G IN A -
defined partly controlled/controlled asthm a 99% of the tim e.
A t the cut-off point of A C T >20, the overall kappa statistic
w as 0.35, w hile the m axim um kappa values for predicting
uncontrolled asthm a occurred at A C T scores of <14 (kappa
0.55) and <15 (kappa 0.53).
GINA control categories and ACT scores
The distributions of the three G IN A control categories relative
to A C T scores are depicted in Figure 2 for all patients.
The m edian (IQ R) A C T scores for G IN A categories w ere 14
(1117), 20 (1722), and 23 (2125) for uncontrolled, partly
controlled, and controlled asthm a, respectively (p<0.001).
The m edian A C T scores for each G IN A category varied
significantly w ithin each country (p<0.001), w ith significant
differences (p<0.001) for all pairw ise com parisons; m edian
A C T scores w ere num erically sim ilar for any given G IN A
category am ong countries (see Table 6).
Table 7 sum m arises the num bers of patients m eeting each
G IN A criterion for all patients w ho did not have controlled
asthm a by the G IN A definition but w ho had an A C T score
>20 (w ell-controlled asthm a). A total of 56 patients w ere
classified as having G IN A uncontrolled asthm a w ith an A C T
score >20, including 36 (64% ) w ith an exacerbation in the
previous seven days, and 21 (38% ) w ith three or m ore
sym ptom s in any one w eek of the preceding four w eeks (one
w w w .thepcrj.org
Positive Negative
Correctly predictive predictive
Sensitivity Specificity classified value value
>5 0 100 32 -- 32
>6 0 100 32 100 32
>7 1 100 32 100 32
>8 2 100 33 100 32
>9 3 100 34 100 32
>10 6 100 36 100 33
>11 8 100 37 100 34
>12 11 100 39 100 34
>13 15 100 42 100 35
>14 18 100 44 99 36
>15 23 99 47 99 37
>16 28 99 51 98 39
>17 34 98 55 97 41
>18 42 97 59 97 44
>19 51 95 65 96 47
>20 60 92 70 94 51
>21 70 84 75 91 57
>22 79 74 78 87 63
>23 87 59 78 82 67
>24 91 42 76 77 69
>25 94 27 73 74 68
A ll data are percentages.
For each A C T score cut-off point, the:
* Sensitivity is defined as the percentage of patients w ith G IN A -
defined partly controlled/uncontrolled asthm a w ho w ere
identified by the A C T as belonging to this group.
Specificity is defined as the percentage of patients w ith G IN A -
defined controlled asthm a w ho w ere identified by A C T as
belonging to this group.
Positive predictive value, or the predictive value of a positive
test, is the percentage of patients w hom A C T predicts correctly
to have partly controlled/uncontrolled asthm a based on the
G IN A classification.
** N egative predictive value, or the predictive value of a negative
to have controlled asthm a based on the G IN A classification.
Table 3. Performance of the ACT score at different cut-off
points in predicting GINA categories of asthma control
(controlled versus partly controlled/ uncontrolled) for all
patients (n=2949)
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
patient m et both criteria). O f the 756 patients w ho w ere
classified as having G IN A partly controlled asthm a w ith an
A C T score >20, 322 (43% ) had had >1 exacerbation in the
preceding 12 m onths and no other issues, w hile 176 (23% )
had had no exacerbation and just one issue during the
previous four w eeks causing them to m eet the criteria for
G IN A partly controlled asthm a.
M Thomas et al.
46
w w w .thepcrj.org
Positive Negative
G lobal (n=2949) 60 92 70 94 51
France (n=697) 58 93 72 93 58
G erm any (n=495) 59 93 73 93 60
Italy (n=495) 56 85 63 93 38
Spain (n=395) 55 97 65 98 41
U K (n=154) 73 88 77 94 56
U S (n=713) 64 91 72 94 51
Table 4. Specificity, sensitivity, positive and negative
predictive values for ACT score cut-off point of >20 for
well-controlled asthma and GINA binary split of
controlled vs partly controlled/uncontrolled asthma,
globally and by country
ACT >20 ACT <20 Kappa
Country GINA classification N (%) N (%) statistic
A ll C ontrolled 856 (92) 78 (8) 0.42
Partly/uncontrolled 812 (40) 1203 (60)
France C ontrolled 250 (93) 20 (7) 0.46
G erm any C ontrolled 184 (93) 14 (7) 0.48
Italy C ontrolled 100 (85) 17 (15) 0.29
Spain C ontrolled 95 (97) 3 (3) 0.36
U K C ontrolled 38 (88) 5 (12) 0.52
U S C ontrolled 189 (91) 19 (9) 0.45
Table 5. Kappa level of agreement at ACT score cut-off
point of >20 for well-controlled asthma and GINA binary
split of controlled vs partly controlled/uncontrolled
asthma
N ACT median* (IQR) ACT range
GINA controlled
A ll 934 23 (2125) 1125
France 270 23 (2225) 1425
G erm any 198 23 (2124) 1525
Italy 117 23 (2124) 1325
Spain 98 23 (2224) 1825
U K 43 23 (2124) 1725
U S 208 23 (2225) 1125
GINA partly controlled
A ll 1439 20 (1722) 525
France 323 20 (1722) 625
G erm any 208 20 (1822) 925
Italy 298 20 (1821) 925
Spain 237 20 (1722) 925
U K 66 19 (1621) 525
U S 307 20 (1722) 825
GINA uncontrolled
A ll 576 14 (1117) 525
France 104 13 (1016) 524
G erm any 89 14 (1116) 623
Italy 80 15 (1318) 824
Spain 60 15 (1317.5) 525
U K 45 12 (1016) 523
U S 198 14 (1117) 525
*p<0.001 (Kruskal-W allis) for all betw een-group com parisons for each country;
p<0.001 (M ann-W hitney w ith Bonferroni adjustm ent) for all pairw ise
com parisons for each country.
Table 6. GINA classifications and corresponding median
ACT scores for all patients and by country
GINA GINA
partly controlled uncontrolled
(n=756) (n=56)
D aytim e sym ptom s, >2/w k 123 (16% ) 23 (41% )
Lim itations of activities, any 155 (21% ) 23 (41% )
N octurnal sym ptom s, any 242 (32% ) 35 (63% )
N eed for rescue inhaler, >2/w k 86 (11% ) 23 (41% )
>3 of above issues in 1 w eek of prior 4 N /A 21 (38% )
A sthm a exacerbation
499 (66% )* 36 (64% )*

N /A = not applicable
*Patients could m eet >1 criterion.
Exacerbation recorded in prior 7 days for G IN A uncontrolled asthm a and at any

tim e in >7 days and prior 12 m onths for G IN A partly controlled asthm a.
Table 7. Patients who did not have controlled asthma by
the GINA definition but who had an ACT score >20 (well-
controlled asthma): numbers meeting each GINA criterion
for partly controlled or uncontrolled asthma*
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
47
Discussion
In this m ultinational survey, the A C T w as useful in predicting
G IN A -defined asthm a control categories and w as particularly
useful in confirm ing patients w hose asthm a w as not
controlled according to the G IN A classification. W e found
that an A C T score of <19 correctly predicted G IN A partly
controlledor uncontrolledasthm a 94% of the tim e overall
and >93% of the tim e in each country. The area under the
RO C curve, the single m easure incorporating the m ost
inform ation on the relationship betw een A C T predicting
G IN A , w as quite adequate at 0.84 and of a sim ilar m agnitude
to prior findings for the A C T as w ell as other case
identification instrum ents for asthm a, allergy, and chronic
obstructive pulm onary disease.
28-30
A n A C T score >20 predicted G IN A -defined controlled
asthm a 51% of the tim e, and the kappa statistic (0.42)
suggested a m oderate agreem ent using the cut-off point of
>20 for w ell controlledasthm a. This is largely because
substantial num bers of patients w ith an A C T score >20 had
G IN A partly controlled, and a few G IN A uncontrolled,
asthm a. M any of the discrepancies could be explained by
either the tim ing of exacerbations or by variability in item
content and grading betw een the A C T and G IN A definitions.
O verall, 66% of patients w ith G IN A partly controlled asthm a
and an A C T score >20 failed to m eet the G IN A definition of
controlled asthm a either because of an old exacerbation or
isolated sym ptom s, w ith nocturnal sym ptom s reported in
isolation m ost com m only. The A C T lists specific sym ptom s (for
exam ple, shortness of breath, night-tim e w heezing), w hile
the G IN A categories are less specific (nam ely, daytim e or
night-tim e sym ptom s), perhaps capturing sym ptom s outside
the A C T definition. M oreover, som e discrepancies likely arose
from faulty questionnaire com pletion, not uncom m on in a
large survey of this nature.
There is no gold standard for m easuring asthm a control;
even the G IN A classification is described as a w orking
schem e based on current opinion [that] has not been
validated.
3
The G IN A classification does not include a
tim efram e and thus can be used to define long- or short-term
control. The fact that exacerbations and any sym ptom s during
the preceding year are captured, provides a long-term picture
of asthm a control that can be useful in determ ining optim al
asthm a therapy.
The A C T w as developed and validated, using a criterion
m easure of control (specialistsrating of asthm a control after
spirom etry), to serve as a screening tool for assessing short-
term asthm a control over the preceding four w eeks.
19,20
Patientsrecall of asthm a sym ptom s decreases over tim e and
thus a short recall period (24 w eeks) is recom m ended for
patient-reported sym ptom history.
4
The A C T is rapidly
com pleted by patients, and the dichotom ous scoring system
is convenient for busy clinicians. H ow ever, reliance on a single
questionnaire could result in the potential for over- or under-
treatm ent. N o questionnaire is a perfect replacem ent for a
thorough m edical history and clinical judgem ent.
Results of sensitivity analyses support the robustness of
our prim ary analyses and help to characterise further the
relationship betw een the A C T and G IN A control
classifications. W hen w e rem oved exacerbations as a
criterion, m ore patients w ere defined as having controlled
asthm a by the G IN A definition, and the negative predictive
value of the A C T score im proved substantially to 72% , w ith
w w w .thepcrj.org
Figure 1. Receiver operating characteristics (ROC) curves
for the Asthma Control Test score predicting the Global
Initiative for Asthma (GINA) control classification for the
six countries included in the survey. The legend shows
area under the ROC curve (95% CI) for each country.
Figure 2. Distribution of GINA control categories relative
to ACT scores for outpatients with asthma (n=2949) in
five European countries and the US.
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
M Thomas et al.
48
kappa im proved to 0.58. For predicting uncontrolled asthm a,
the m axim um kappa values w ere at A C T scores of <14 (0.55)
and <15 (0.53). O f note, an A C T score of <15 w as identified
previously by Schatz and cow orkers
20
as the optim al cut-off
point defining uncontrolled asthm a.
N onetheless, the use of prior exacerbations in identifying
at-risk patients is supported by results of enquiries into
asthm a deaths
31
and by the G IN A guidelines, w hich include
"no exacerbations" as a criterion for asthm a control.
3
A sthm a
control is a com plex concept, and as no single m easure
encom passes the full com plexity of asthm a, com posite
outcom e m easures that include m easures of current sym ptom
control and disease im pact as w ell as future risk are
advocated.
32
In the clinic (or consulting room ) it m ay be difficult to
evaluate exacerbations, and the relevance of a prior
exacerbation w ith regard to asthm a control m ay be difficult to
judge; for exam ple, the relevance w ill vary according to
w hether there is a seasonality to exacerbations or w hether
treatm ent had been changed since an exacerbation.
N evertheless, to ignore exacerbations is to m iss an im portant
facet of the disease. O ur study w as a real w orld study and
thus reflective of real w orld hurdles that clinicians face in
interpreting G IN A guidelines. Indeed, applying any guideline
criterion can be difficult in a real w orld setting.
A survey of this nature has several lim itations, including
the possibility of patient selection bias, since the participants
represent a convenience sam ple and m ay not be
representative of the overall population of patients w ith
asthm a although they m ay be representative of the
consulting population in w hom these instrum ents are often
used. W e recruited larger num bers of physicians in densely
populated areas, and physicians w ere asked to invite
consecutive patients to participate, w ith the goal of collecting
cross-sectional data from an unselected real w orld sam ple. In
m ost countries, over tw o-thirds of patients returned
com pleted surveys that could be m atched to their physicians
form s. The tw o exceptions w ere Spain and the U K, w here
only 54% and 26% of patients, respectively, com pleted their
surveys, raising the possibility of non-response bias for those
countries. Poor U K response rates appear to have resulted
from a com bination of physicians not asking patients to
com plete the surveys, com bined w ith low response rates from
patients. A nother lim itation of this survey is that the quality of
data relied on provision of accurate data from physicians and
patients. In addition, som e of the survey questions w ere not
directly equivalent to the G IN A criteria; for exam ple, the
question on the patient-com pleted form for exercise
lim itation ("[has] your asthm a stopped you taking part in
every day activities?") could be interpreted differently from
the w ording used in the actual G IN A table ("lim itations of
activities"). M oreover, the Respiratory D SP VI questionnaire
w as not validated. Finally, w e cannot rule out recall bias on
the part of participating patients.
A strength of this survey is its inclusion of a large
m ultinational patient population from both prim ary and
specialist care. The greater percentage of w om en than m en
(56% vs. 44% ) in our survey m irrors the greater prevalence of
asthm a am ong adult w om en w orldw ide.
1
M oreover, our
findings confirm the prevalence of uncontrolled asthm a
am ong consulting patients,
9,10
w ith only 934/2949 (32% ) of
patients categorised as having G IN A controlled asthm a and
1668/2949 (57% ) having an A C T score of >20, the cut-off
point for w ell-controlled asthm a.
In conclusion, w e found that an A C T score <19 is useful
to identify patients w ith poorly controlled asthm a for w hom
a full clinical review is needed. Further studies are needed to
evaluate the benefits of the A C T over tim e in a real w orld
setting. The A C T is easily and rapidly com pleted by patients
and can serve as a useful tool in the clinic to assess asthm a
control, ideally in conjunction w ith a com plete m edical history
and lung function testing.
D ata from this study w ere presented as a late breaker abstract at
the International Prim ary C are Respiratory G roup conference,
28-31 M ay 2008, Seville, Spain.
Conflicts of interest
Elizabeth V H illyer has received freelance w riting assignm ents from M erck and
A erocrine. She has no shares in any pharm aceutical com pany.
Professor D avid Price has consultant arrangem ents w ith: A ltana, Boehringer-
Ingelheim , G laxoSm ithKline, Ivax and Pfizer. H e or his team have received grants
and research support for research in respiratory disease from the follow ing
organisations: U K N ational H ealth Service, A ltana Pharm a, A straZeneca, Boehringer-
Ingelheim , G laxoSm ithKline, Ivax, M erck Sharpe and D ohm e, N ovartis, Pfizer, and
Schering Plough. Professor Price has also spoken for: A ltana Pharm a, Boehringer-
Ingelheim , G laxoSm ithKline, M erck, Sharpe and D ohm e and Pfizer.
A ngela E. W illiam s and Jacqueline R C arranza Rosenzw eig are em ployees of
G laxoSm ithKline.
N either M ike Thom as nor any m em ber of his close fam ily has any shares in
pharm aceutical com panies. In the last 3 years he has received fees for acting as a
consultant form M SD , Schering and G SK, has received speakers honoraria for
w w w .thepcrj.org
a) Difficulties
Low patient yield in U K
b) Alternative methodologies
A reverse approach, in w hich the G IN A classification, w hich
has not been validated, is tested against the A C T as the gold
standard could have been of interest
c) New questions arising
N eed to test A C T in a real w orld clinical setting, adjusting
therapy according to score, to evaluate outcom es
d) Lessons for clinical practice A C T score <19 useful
clinically in identifying patients w ith poorly controlled
asthm a
Discussion Summary
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
49
speaking at sponsored m eetings from the follow ing com panies m arketing
respiratory and allergy products: A straZeneca, Boehringer Inglehiem , G SK, M SD ,
Schering-Plough, Teva. H e has received honoraria for attending advisory panels
w ith; A ltana, A stra Zeneca, BI, G SK, M SD , M erck Respiratory, Schering-Plough,
Teva. H e has received sponsorship to attend international scientific m eetings from :
G SK, M SD , A stra Zeneca. H e has received funding for research projects from : G SK,
M SD , A stra Zeneca. H e holds a research fellow ship from A sthm a U K.
Funding
Funding support w as provided by G lobal H ealth O utcom es, G laxoSm ithKline R& D
Building 11, 1st Floor, Stockley Park W est, U xbridge, M iddlesex U B11 1BU , U K.
Acknowledgement
The authors w ould like to acknow ledge the help of Victoria H iggins in these
analyses.
References
1. G lobal Initiative for A sthm a. G IN A report, G lobal Strategy for A sthm a
M anagem ent and Prevention 2006. Available at: http://w w w .ginasthm a.org/
[A ccessed: 20 M ay 2008].
2. H um bert M , H olgate S, Boulet LP, Bousquet J. A sthm a control or severity: that
is the question. A llergy 2007;62(2):95-101.
3. G lobal Initiative for A sthm a (G IN A ). G lobal Strategy for A sthm a M anagem ent
and Prevention 2007. Available at: http://w w w .ginasthm a.org/ [A ccessed: 20
M ay 2008].
4. N ational A sthm a Education and Prevention Program . Expert Panel Report 3:
G uidelines for the D iagnosis and M anagem ent of A sthm a. Full report 2007.
Available at: http://w w w .nhlbi.nih.gov/guidelines/asthm a/ [A ccessed: 20 M ay
2008].
5. British Thoracic Society, Scottish Intercollegiate G uidelines N etw ork. British
G uideline on the M anagem ent of A sthm a. A vailable at:
http://w w w .sign.ac.uk/guidelines/fulltext/101/index.htm l [A ccessed: 20 M ay
2008].
6. Batem an ED , Boushey H A , Bousquet J, et al. C an guideline-defined asthm a
control be achieved? The G aining O ptim al A sthm a C ontroL study. Am J Respir
Crit Care Med 2004;170(8):836-44.doi:10.1164/rccm .200401-033O C .
7. Batem an ED , Bousquet J, Braunstein G L. Is overall asthm a control being
achieved? A hypothesis-generating study. Eur Respir J 2001;17(4):589-95.
8. C azzoletti L, M arcon A , Janson C , et al. A sthm a control in Europe: a real-w orld
evaluation based on an international population-based study. J Allergy Clin
Immunol 2007;120(6):1360-7. doi:10.1016/j.jaci.2007.09.019.
9. C hapm an KR, Boulet LP, Rea RM , Franssen E. Suboptim al asthm a control:
prevalence, detection and consequences in general practice. Eur Respir J
2008;31(2):320-5. doi:10.1183/09031936.00039707.
10. Partridge M R, van der M olen T, M yrseth SE, Busse W W . A ttitudes and actions
of asthm a patients on regular m aintenance therapy: the IN SPIRE study. BMC
Pulm Med 2006;6:13. doi:10.1186/1471-2466-6-13.
11. Rabe KF, A dachi M , Lai C K, et al. W orldw ide severity and control of asthm a in
children and adults: the global asthm a insights and reality surveys. J Allergy Clin
Immunol 2004;114(1):40-7. doi:10.1016/j.jaci.2004.04.042.
12. A ccordini S, C orsico A , C erveri I, et al. The socio-econom ic burden of asthm a
is substantial in Europe. Allergy2008;63(1):116-24.
13. Peters SP, Ferguson G , D eniz Y, Reisner C . U ncontrolled asthm a: a review of the
prevalence, disease burden and options for treatm ent. Respir Med
2006;100(7):1139-51. doi:10.1016/j.rm ed.2006.03.031.
14. Van G anse E, Laforest L, Pietri G , et al. Persistent asthm a: disease control,
resource utilisation and direct costs. Eur Respir J 2002;20(2):260-7.
15. A ccordini S, Bugiani M , A rossa W , et al. Poor control increases the econom ic
cost of asthm a. A m ulticentre population-based study. Int Arch Allergy
Immunol 2006;141(2):189-98. doi:10.1159/000094898.
16. Juniper EF, C hauhan A , N eville E, et al. C linicians tend to overestim ate
im provem ents in asthm a control: an unexpected observation. Prim Care Respir
J 2004;13(4):181-4. doi:10.1016/j.pcrj.2004.04.003.
17. Price D , Ryan D , Pearce L, et al. The burden of paediatric asthm a is higher than
health professionals think: results from the A sthm a In Real Life (A IR) study. Prim
Care Resp J 2002;11(2):30-3.
18. Boulet LP, Phillips R, O 'Byrne P, Becker A . Evaluation of asthm a control by
physicians and patients: com parison w ith current guidelines. Can Respir J
2002;9(6):417-23.
19. N athan RA , Sorkness C A , Kosinski M , et al. D evelopm ent of the asthm a control
test: a survey for assessing asthm a control. J Allergy Clin Immunol
2004;113(1):59-65. doi:10.1016/j.jaci.2003.09.008.
20. Schatz M , Sorkness C A , Li JT, et al. A sthm a C ontrol Test: reliability, validity, and
responsiveness in patients not previously follow ed by asthm a specialists. J
Allergy Clin Immunol 2006;117(3):549-56. doi:10.1016/j.jaci.2006.01.011.
21. Schatz M , M osen D M , Kosinski M , et al. Validity of the A sthm a C ontrol Test
com pleted at hom e. Am J Manag Care 2007;13(12):661-7.
22. Vega JM , Badia X, Badiola C , et al. Validation of the Spanish version of the
A sthm a C ontrol Test (A C T). J Asthma 2007;44(10):867-72.
doi:10.1080/02770900701752615.
23. H orne R, Price D , C leland J, et al. C an asthm a control be im proved by
understanding the patient's perspective? BMC Pulm Med 2007;7:8.
doi:10.1186/1471-2466-7-8.
24. A delphi G roup Products. D isease Specific Program m es. A vailable at:
http://w w w .adelphigroup.com /com panies/com pany_group_products.asp
[A ccessed: 20 M ay 2008].
25. Barnes N C , W illiam s A E. U nscheduled healthcare resource use am ong asthm a
patients receiving low -dose inhaled corticosteroids m aintenance treatm ent. Int
J Clin Pract 2005;59(9):1017-24. doi:10.1111/j.1742-1241.2005.00615.x
26. H iggins V, Kay S, Sm all M . Physician and patient survey of allergic rhinitis:
m ethodology. Allergy 2007;62 Suppl 85:6-8. doi:10.1111/j.1398-
9995.2007.01547.x.
27. Landis JR, Koch G G . The m easurem ent of observer agreem ent for categorical
data. Biometrics 1977;33(1):159-74. doi:10.2307/2529310.
28. Price D B, Tinkelm an D G , N ordyke RJ, Isonaka S, H albert RJ. Scoring system and
clinical application of C O PD diagnostic questionnaires. Chest
2006;129(6):1531-9. doi:10.1378/chest.129.6.1531.
29. W allenstein G V, C arranza-Rosenzw eig J, Kosinski M , Blaisdell-G ross B, G ajria K,
Jhingran P. A psychom etric com parison of three patient-based m easures of
asthm a control. Curr Med Res Opin 2007;23(2):369-77.
doi:10.1185/030079906X167426.
30. W eintraub JM , Sparrow D , W eiss ST. Receiver operating characteristics curve
analysis of cutaneous skin test reactions to predict hay fever and asthm a
sym ptom s in the N orm ative A ging Study. Allergy 2001;56(3):243-6.
doi:10.1034/j.1398-9995.2001.056003243.x.
31. H arrison B, Stephenson P, M ohan G , N asser S. A n ongoing C onfidential Enquiry
into asthm a deaths in the Eastern Region of the U K, 2001-2003. Prim Care
Resp J 2005;14(6):303-13. doi:10.1016/j.pcrj.2005.08.004.
32. Taylor D R, Batem an ED , Boulet LP et al. A new perspective on concepts of
asthm a severity and control. Eur Respir J 2008;32(3):545-54.
doi:10.1183/09031936.00155307.
w w w .thepcrj.org
Available online at http://www.thepcrj.org
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
Asthma ControlTest
TM
The following test can help people with asthma (12 years or older) assess their asthma control. Please circle the
appropriate score for each question. There are FIVE questions in total. Please answer the questions as honestly as
possible.
35a. D uring the past 4 weeks, how often did your asthmaprevent you from getting as m uch done at w ork, school or hom e?
A ll M ost of Som e of A little of N one of
the tim e the tim e the tim e the tim e the tim e
35b. D uring the past 4 weeks, how often have you had shortness of breath?
M ore than O nce 3-6 tim es 1-2 tim es N ot
once a day a day a w eek a w eek at all
35c. D uring the past 4 weeks, how often did your asthma sym ptom s (w heezing, coughing, shortness of breath, chest
tightness or pain) w ake you up at night or earlier than usual in the m orning?
4 or m ore nights 2 to 3 nights O nce O nce N ot
a w eek a w eek a w eek or tw ice at all
35d. D uring the past 4 weeks, how often have you used your rescue inhaler (such as salbutam ol)?
3 or m ore 1 or 2 tim es 2 or 3 tim es O nce a w eek N ot
tim es a day a day a w eek or less at all
35e. H ow w ould you rate your asthmacontrol during the past 4 weeks?
N ot Poorly Som ew hat W ell C om pletely
controlled controlled controlled controlled controlled
M Thomas et al.
000
w w w .thepcrj.org
Appendices and Online Tables and Figures
Appendix 1. Asthma Control Test
TM
as included in the Respiratory DSP VI survey
1 2 3 4 5
1 2 3 4 5
1 2 3 4 5
1 2 3 4 5
1 2 3 4 5
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
000
w w w .thepcrj.org
Appendix 2. Questions on physician-completed survey form used to derive GINA control classification in the
Respiratory DSP VI survey
D SYMPTOMS -continued
3. Thinking about this patient over the past 4 weeks, how often has this patient experienced any of the following?
(please tick one box for each characteristic)
A B
a) Daytime symptoms Tw ice a w eek or less or M ore than tw ice a w eek
b) Nocturnal symptoms N one of the tim e or O ne or m ore tim es
or awakening
c) Need for rescue treatment Tw ice a w eek or less or M ore than tw ice a w eek
d) Limitation of activities N one of the tim e or O ne or m ore tim es
4. If this patient has three or more ticks in column B above, have 3 or more of these occured in any one week
during the last 4 weeks?
Yes No
C DIAGNOSTICHISTORY
1a. Please tick if the following spirometry tests have been measured in this patient in the last 12 months
1b. When was this patients last spirometry test?
1c. What was the most recent level recorded in the patient (if known)?
1d. What was the most recent percentage predicted value recorded in this patient (if known)?
1e. Was this most recent test performed pre or post bronchodilator?
FEV
1
(Forced Expiratory Volum e in 1 second)
FVC (Forced Vital C apacity)
PEFR (Peak Expiratory Flow rate)
FEV1/FVC
(a) Test
measured in last
12 months?
Yes N o D k
1 2 3
(b) Last test?
last 4-12 O ver
3 m ths 12 m ths
m ths ago ago
1 2 3
Results
(c) level
recorded
(litres)
(litres)
(litres)
%
(d) %
predicted
(if known)
%
%
%
n/a
(e) Pre or post
bronchodilator
Pre Post D k
1 2 3
3a. How many exacerbations has this patient suffered inthe last 12 months?
3b. How long ago did this patient suffer their last exacerbation? days ago weeks ago months ago
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
M Thomas et al.
000
w w w .thepcrj.org
Appendix 3. Questions on patient-completed form used to derive GINA control classification in the Respiratory
DSP VI survey
8a. In the last 4 weeks, have you experienced any of the following in the list below? (please tick one box for each
point listed)
Column (A) Column (B)
a) D aytim e sym ptom s Tw ice a w eek or less or M ore than tw ice a w eek
b) N ight tim e sym ptom s or w aking
up earlier in the m orning than usual
c) Your asthm a has stopped you
taking part in every-day activities
d) The need to take your rescue
inhaler such as salbutam ol
8b. If you have 3 or more ticks in column (B), did these happen in any one week during the last 4 weeks?
Yes N o
N one of the tim e or O ne or m ore tim es
N one of the tim e or O ne or m ore tim es
Tw ice a w eek or less or M ore than tw ice a w eek
17. Have you ever had times in the last 12 months when you had to seek further medical help for your increased
asthma symptoms?
Yes (Go to question 18) N o (Go to question 21)
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
000
w w w .thepcrj.org
Positive Negative
>5 0 100 30 -- 30
>6 0 100 30 100 30
>7 1 100 30 100 30
>8 2 100 31 100 30
>9 3 100 32 100 31
>10 6 100 34 100 31
>11 8 100 35 100 32
>12 11 100 37 100 32
>13 14 100 40 100 33
>14 18 100 42 99 34
>15 22 99 45 99 35
>16 28 99 49 98 37
>17 34 98 53 97 39
>18 41 97 58 97 41
>19 49 95 63 96 44
>20 58 92 68 94 49
>21 69 86 74 92 54
>22 79 75 78 88 60
>23 86 60 78 83 65
>24 91 43 76 79 66
>25 94 28 74 75 66
A rea under RO C = 0.84 (95% C I 0.820.85)
For each A C T score cut point, the:
* Sensitivity is defined as the percentage of patients w ith G IN A -
defined partly controlled/uncontrolled asthm a w ho w ere
identified by the A C T as belonging to this group.
Specificity is defined as the percentage of patients w ith G IN A -
defined controlled asthm a w ho w ere identified by A C T as
belonging to this group.
Positive predictive value, or the predictive value of a positive
to have partly controlled/uncontrolled asthm a based on the
G IN A classification.
** N egative predictive value, or the predictive value of a negative
to have controlled asthm a based on the G IN A classification.
Online Table 1. Performance of the ACT score at different
cut-off points in predicting GINA categories of asthma
control (controlled versus partly controlled/uncontrolled)
for all patients (n=2949), with lung function criterion
included for those patients who had lung function data
(n=539)
Positive Negative
>5 (no 5 responses)
>6 0 100 17 -- 17
>7 0 100 17 100 17
>8 2 100 19 100 18
>9 3 100 20 100 18
>10 6 100 22 100 18
>11 9 100 24 100 19
>12 12 100 27 100 19
>13 17 100 32 100 20
>14 21 100 35 100 21
>15 26 100 39 100 22
>16 30 100 42 100 23
>17 38 100 48 100 25
>18 44 100 54 100 27
>19 51 98 59 99 30
>20 59 96 65 98 32
>21 68 87 71 96 36
>22 76 81 77 95 41
>23 83 66 80 92 45
>24 89 51 82 90 48
>25 94 29 83 86 51
A rea under RO C = 0.85 (95% C I 0.820.89)
for patients with lung function data (n=539) - lung
function criterion applied in GINA definition
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
M Thomas et al.
000
w w w .thepcrj.org
Positive Negative
>5 0 100 46 -- 46
>6 1 100 46 100 46
>7 1 100 46 100 46
>8 3 100 47 100 46
>9 4 100 48 100 47
>10 7 100 49 98 47
>11 10 100 51 98 48
>12 13 100 53 97 49
>13 18 99 55 98 50
>14 23 99 58 98 52
>15 28 99 61 97 54
>16 35 98 63 96 56
>17 42 97 67 94 58
>18 51 96 71 94 62
>19 61 94 76 92 67
>20 71 89 79 88 72
>21 81 80 81 83 78
>22 89 69 80 78 84
>23 95 55 77 72 90
>24 98 39 71 66 93
>25 99 26 65 61 93
A rea under RO C = 0.85 (95% C I 0.810.88)
for patients with lung function data (n=539)lung
function criterion not applied in GINA definition
Positive Negative
>5 0 100 46 -- 46
>6 1 100 46 100 46
>7 1 100 46 100 46
>8 3 100 47 100 46
>9 4 100 48 100 47
>10 7 100 49 98 47
>11 10 100 51 98 48
>12 13 100 53 97 49
>13 18 99 55 98 50
>14 23 99 58 98 52
>15 28 99 61 97 54
>16 35 98 63 96 56
>17 42 97 67 94 58
>18 51 96 71 94 62
>19 61 94 76 92 67
>20 71 89 79 88 72
>21 81 80 81 83 78
>22 89 69 80 78 84
>23 95 55 77 72 90
>24 98 39 71 66 93
>25 99 26 65 61 93
A rea under RO C curve = 0.89 (95% C I 0.88 0.90)
for all patients (n=2949)exacerbation criterion removed
from GINA definition
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d
000
w w w .thepcrj.org
Positive Negative
>5 0 100 85 85
>6 2 100 85 89 85
>7 3 100 85 82 85
>8 8 100 86 85 86
>9 12 99 86 78 86
>10 18 99 86 73 87
>11 27 98 87 73 88
>12 33 97 87 69 89
>13 45 96 88 68 91
>14 53 95 88 65 92
>15 63 93 88 61 93
>16 69 89 86 54 94
>17 78 86 84 49 96
>18 86 81 81 45 97
>19 92 74 77 39 98
>20 95 66 70 34 99
>21 98 55 62 28 99
>22 98 44 52 24 99
>23 99 33 43 21 100
>24 100 23 34 19 100
>25 100 15 28 18 100
A rea under RO C curve = 0.91 (95% C I 0.890.92)
control with the binary split of uncontrolled versus partly
controlled/controlled (exacerbation criterion removed
from GINA definition) for all patients (n=2949)
Online Figure 1. Receiver operating characteristics (ROC)
curves for the Asthma Control Test score predicting the
Global Initiative for Asthma (GINA) control classification
using GINA binary split of uncontrolled vs partly
controlled/controlled asthma and the exacerbation
criterion removed from GINA definitionfor the six
countries included in the survey. The legend shows area
under the ROC curve (95% CI) for each country.
Online Figure 2. Distribution of GINA control categories
relative to ACT scores for outpatients with asthma
(n=2949) in five European countries and the US -
exacerbation criterion removed from GINA definition.
C
o
p
y
r
i
g
h
t

G
e
n
e
r
a
l

P
r
a
c
t
i
c
e

A
i
r
w
a
y
s

G
r
o
u
p

R
e
p
r
o
d
u
c
t
i
o
n

p
r
o
h
i
b
i
t
e
d

Asthma Control Test

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Asthma Control Test

Uploaded by

Copyright:

Available Formats

Primary Care Respiratory Journal (2009);18(1): 41-49

By definition, an exacerbation in any w eek m akes that an uncontrolled asthm a w eek.

D ata derived from physician-com pleted form s

Text in bold italics signifies differences from G IN A definitions in Table 1A

Percentages are relative to num ber of PRFs for that country.

499 (66% )* 36 (64% )*

Exacerbation recorded in prior 7 days for G IN A uncontrolled asthm a and at any

You might also like