46 characteristics associated with medication adherence … · 2019-11-22 · 46 characteristics...
TRANSCRIPT
Characteristics Associated With Medication Adherence Among New Omalizumab Users46
James Zazzali, PhD, MPH1; Michael S. Broder, MD, MSHS2; Eunice Chang, PhD2 1Genentech Inc., South San Francisco, CA; 2Partnership for Health Analytic Research, Beverly Hills, CA
ABSTRACTOBJECTIVE: Adherence to omalizumab (OMA) therapy has not been well studied. We sought to identify characteristics associated with adherence among new OMA users. METHODS: This was a retrospective cohort analysis using a HIPAA-compliant claims database. The study identified asthma patients who were ≥12 years old, newly treated with OMA between 7/1/2004 and 6/30/2007, and enrolled for 1 year before and 1 year after the first OMA claim. Adherence was measured by medication possession ratio (MPR) (total days of medication supplied ÷ 365) with each OMA claim considered a 28-day supply. Therapy step, as defined by the EPR3 guidelines, was assigned using a published algorithm. A base linear regression model was conducted with MPR as the dependent variable and demographics, physician specialty, respiratory comorbidities, asthma control, therapy step, and medication ratio forced into the model as independent variables. Other comorbidities were considered with forward selection and retained if significant at p<.05.RESULTS: We identified 766 new OMA users; mean age was 43 (SD 14) years, and 61% (n=466) were female. Forty-two percent of patients had allergists as their usual source of care, 20% pulmonologists, 28.3% primary care physicians, and the remainder other/unspecified. The mean number of chronic conditions was 5 (SD 2). Before starting OMA, most of patients were on EPR3 Step 5 therapy (28%), followed by Step 4 (18%), Step 3 (16%), Step 2 (15%), Step 1 (6%), and Step 6 (5%). Mean MPR in the year after treatment initiation was 0.68 (SD 0.30), and 55% persisted with therapy for at least 1 year. The final model included the base variables and diseases of circulatory system (the only additional significant predictor). Significant predictors of higher MPR were: care provided by allergists or pulmonologists, chronic otitis media, 5 or 6 vs ≥7 chronic conditions, and no disease of the circulatory system.CONCLUSIONS: We found MPR for omalizumab to be higher on average than what has been reported for combination corticosteroid/long-acting beta-agonist therapy. Care by an allergist or pulmonologist was associated with greater adherence. Confounding by disease severity is possible, but the relationship was significant after controlling for therapy step. If this finding is confirmed in other studies, it supports the value of specialist care for patients with difficult to treat asthma.
CONCLUSIONS• Among new omalizumab users, we estimated the MPR to be 0.68 with 54.7%
persisting on omalizumab at one year after initiating therapy.
• Omalizumab adherence and persistence are consistent with previously published reports, although this analysis indicates slightly higher estimates.
• The number of chronic conditions and select respiratory comorbidities were significant predictors of adherence and persistence in our multivariate models.
LIMITATIONS• Administrative claims data do not indicate the dosing schedule for omalizumab;
in this analysis, each dose of omalizumab was assumed to be a 28 day supply.
• Inclusion of concomitant asthma medications as covariates is indicative only of those medications filled and not reflective of patient adherence to therapy.
BACKGROUND• Lack of adherence to prescribed treatments for asthma is a well-known problem.
Rates of nonadherence range from 30% to 70%.1-3
• Poor asthma medication adherence is associated with decreases in asthma control and increases in emergency department visits, hospitalizations, and the need for oral corticosteroids.3,4
• Omalizumab, a humanized monoclonal antibody targeting immunoglobulin E, is approved in the United States for the treatment of adults and adolescents (≥12 years) with moderate to severe persistent allergic asthma that is inadequately controlled with inhaled corticosteroids.5
• Adherence to omalizumab therapy has not been well studied. One prior analysis of asthma patients newly treated with omalizumab estimated adherence rates to be 64.6% with 54% persisting up to one year.6
OBJECTIVE• To identify characteristics associated with adherence and persistence among new
omalizumab users
METHODSStudy Design• A retrospective cohort analysis using a HIPAA-compliant administrative claims
database of 10 million covered lives representing all major regions of the United States.
• Eligible patients were ≥12 years old, diagnosed with asthma, newly treated with omalizumab between 7/1/2004 and 6/30/2007, and followed for a year (Figures 1 and 2).
RESULTS• We identified 766 new omalizumab users; average age was 43.4 years and 61%
were female. Fifty percent of patients received asthma care from an allergist, 28% a pulmonologist, and 17% a primary care physician.
• Overall, the MPR was 0.68 (SD, ± 0.30). MPR stratified by baseline measures are presented in Table 1 and results of the multivariate analysis in Table 2.
• During follow-up, 45.3% of patients discontinued treatment. Mean (SD) time to discontinuation/end of study was 256.5 (131.6) days; median was 353 days (Table 3, Figure 3).
• The risk of discontinuing omalizumab was examined across a number of variables (Table 4).
REFERENCES1. Bender B et al. Ann Allergy Asthma Immunol. 2000;
85(5):416-421.2. Bender BG et al. Ann Allergy Asthma Immunol. 2007;
98(4):322-328.3. World Health Organization. Adherence to long-term
therapies: evidence for action. Geneva, Switzerland: World Health Organization; 2003.
4. Williams LK et al. J Allergy Clin Immunol. 2004;114(6):1288-1293.
5. Xolair [prescribing information]. South San Francisco, CA: Genentech USA, Inc.; 2010.
6. Broder MS et al. Allergy Asthma Proc. 2009;30(2):148-157.7. Deyo RA et al. J Clin Epidemiol. 1992;45(6):613-619.8. National Asthma Education and Prevention Program. Expert
Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007;120(5 suppl):S94-S138.
Index Date
First claim for omalizumab in ID period
1-year period (Preindex period)
7/1/03 6/30/08
6/30/077/1/04
Identification (ID) Period
1-year period (Postindex period)
Figure 1. Study Timeline
% o
f P
atie
nts
on
Tre
atm
ent
Days to Discontinue Omalizumab
100
75
50
25
0
0 50 100 150 200 250 300 350 400
Figure 3. Days to Discontinuation of Omalizumab.
Study Outcomes• Adherence to omalizumab therapy during the postindex period, as measured by the
medication possession ratio (MPR) – Calculated as the total days of medication supplied over the year, divided by 365 – Each omalizumab claim was considered a 28-day supply • Medication persistence during the postindex period, as measured by time
to discontinuation – Calculated as the number of days on omalizumab with no gaps of more than
45 daysStatistical Analysis• To evaluate the association of baseline measures (patient demographic and clinical
characteristics) with omalizumab adherence and persistence, we conducted a linear regression model with adherence (MPR) as the continuous dependent variable and all baseline measures as independent variables and a similar logistic regression model to predict persistence at one year.
– Based on bivariate analysis results and clinical relevance, demographics (age, sex, and region), asthma physician specialty, number of chronic conditions, evidence of allergy, respiratory-specific comorbidities, therapy step, and medication ratio were first included in the linear regression model; we then used forward selection to identify significant results for other baseline measures.
3456 patients with a claim of omalizumab in ID period
(7/1/2004–6/30/2007)
n=3391
n=3377
n=2886
n=766
65 patients with age <12 years on the index date (the date of �rst omalizumab claim in ID period)
14 patients had no asthma diagnosis in the preindex or postindex period
491 previously used omalizumab
2120 patients were not continuously enrolled in the preindex and postindex periods
Figure 2. Cohort Selection for Omalizumab New Users.
ID = identification.
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Demographics
Age, y (mean 43.4; SD 14.5) 0.157
12-17 68 (8.9) 0.72 (0.28)
18-34 111 (14.5) 0.62 (0.34)
35-44 183 (23.9) 0.66 (0.30)
45-54 234 (30.5) 0.68 (0.30)
55-64 141 (18.4) 0.71 (0.27)
65+ 29 (3.8) 0.65 (0.24)
Sex 0.174
Female 466 (60.8) 0.69 (0.30)
Male 300 (39.2) 0.66 (0.30)
Region 0.056
Midwest 184 (24.0) 0.71 (0.28)
Northeast 86 (11.2) 0.62 (0.32)
South 387 (50.5) 0.66 (0.30)
West 109 (14.2) 0.70 (0.27)
Physician Specialty
Usual asthma care physiciana 0.020
Pulmonologist 214 (27.9) 0.68 (0.28)
Primary care 128 (16.7) 0.64 (0.31)
Allergists 381 (49.7) 0.69 (0.29)
Other 22 (2.9) 0.66 (0.36)
Unknown 21 (2.7) 0.48 (0.39)
General Comorbidities
No. of chronic conditions (mean 4.7; SD 2.3) 0.047
1-2 155 (20.2) 0.63 (0.33)
3-4 227 (29.6) 0.69 (0.29)
5-6 216 (28.2) 0.71 (0.27)
7+ 168 (21.9) 0.65 (0.29)
Charlson Comorbidity Index7 (mean 1.8; SD 1.7) 0.967
1 530 (69.2) 0.68 (0.30)
2 86 (11.2) 0.68 (0.29)
3+ 150 (19.6) 0.67 (0.29)
Evidence of allergy <.001
No 57 (7.4) 0.55 (0.34)
Yes 709 (92.6) 0.69 (0.29)
Respiratory-Specific Comorbidities
COPD 0.338
No 577 (75.3) 0.67 (0.31)
Yes 189 (24.7) 0.69 (0.27)
Sinusitis 0.551
No 243 (31.7) 0.67 (0.31)
Yes 523 (68.3) 0.68 (0.29)
Rhinitis <0.001
No 93 (12.1) 0.57 (0.33)
Yes 673 (87.9) 0.69 (0.29)
Tonsillitis 0.321
No 731 (95.4) 0.67 (0.30)
Yes 35 (4.6) 0.72 (0.28)
Acute upper respiratory infection 0.471
No 515 (67.2) 0.67 (0.30)
Yes 251 (32.8) 0.69 (0.28)
Conjunctivitis 0.894
No 619 (80.8) 0.68 (0.30)
Yes 147 (19.2) 0.67 (0.29)
Chronic otitis media 0.002
No 746 (97.4) 0.67 (0.30)
Yes 20 (2.6) 0.81 (0.17)
Nasal polyposis 0.073
No 663 (86.6) 0.68 (0.29)
Yes 103 (13.4) 0.63 (0.32)
Cough 0.427
No 381 (49.7) 0.67 (0.31)
Yes 385 (50.3) 0.68 (0.28)
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Asthma Control in the Preindex Period
Poor asthma controlb 0.096
No 191 (24.9) 0.64 (0.32)
Yes 575 (75.1) 0.69 (0.29)
Asthma-related inpatient hospitalization 0.340
No 657 (85.8) 0.68 (0.30)
Yes 109 (14.2) 0.65 (0.30)
Asthma-related ED visit 0.412
No 705 (92.0) 0.67 (0.30)
Yes 61 (8.0) 0.71 (0.29)
Two or more oral corticosteroid prescriptions filled 0.031
No 265 (34.6) 0.64 (0.32)
Yes 501 (65.4) 0.69 (0.28)
Six or more short-acting beta-agonist prescriptions filled 0.310
No 434 (56.7) 0.67 (0.30)
Yes 332 (43.3) 0.69 (0.29)
EPR-3 Guidelines8 Therapy Step in Preindex Period
Therapy step prior to initiating omalizumab 0.006
No asthma meds 68 (8.9) 0.56 (0.33)
Step 1 44 (5.7) 0.61 (0.32)
Step 2 112 (14.6) 0.67 (0.30)
Step 3 121 (15.8) 0.73 (0.27)
Step 4 136 (17.8) 0.69 (0.28)
Step 5 214 (27.9) 0.69 (0.29)
Step 6 37 (4.8) 0.63 (0.33)
Unclassified 34 (4.4) 0.72 (0.28)
Medication Ratioc and Other Asthma Medication Use in the Preindex Period
Medication ratio (mean 0.64; SD 0.28 among 718 patients with medication ratios) 0.003
Low ratio (<0.5) 197 (25.7) 0.64 (0.31)
High ratio (≥0.5) 521 (68.0) 0.70 (0.28)
Missing (no controller or reliever)
48 (6.3) 0.55 (0.35)
Short-acting beta-agonist inhalers 0.044
No 115 (15.0) 0.62 (0.32)
Yes 651 (85.0) 0.68 (0.29)
Anticholinergics 0.077
No 521 (68.0) 0.66 (0.30)
Yes 245 (32.0) 0.70 (0.29)
Oral corticosteroids 0.119
No 153 (20.0) 0.64 (0.32)
Yes 613 (80.0) 0.68 (0.29)
ICS/LABA combination 0.051
No 240 (31.3) 0.64 (0.31)
Yes 526 (68.7) 0.69 (0.29)
ICS alone 0.286
No 450 (58.7) 0.67 (0.31)
Yes 316 (41.3) 0.69 (0.28)
LABA alone 0.790
No 643 (83.9) 0.67 (0.30)
Yes 123 (16.1) 0.68 (0.30)
Mast cell stabilizers 0.108
No 734 (95.8) 0.67 (0.30)
Yes 32 (4.2) 0.76 (0.25)
Methylxanthines 0.810
No 670 (87.5) 0.67 (0.30)
Yes 96 (12.5) 0.68 (0.26)
Leukotriene receptor antagonists 0.078
No 221 (28.9) 0.65 (0.31)
Yes 545 (71.1) 0.69 (0.29)
Table 1. MPR of Omalizumab Stratified by Patient Demographics and Clinical Characteristics
aThe usual asthma care physician was determined by the largest plurality of office visits with evaluation and management services and asthma diagnosis. bAsthma-related inpatient hospitalization, asthma-related ED visit, ≥2 oral corticosteroid prescriptions filled, or ≥6 short-acting beta-agonist prescriptions filled.cMedication ratio is units of asthma controllers to units of controllers + units of relievers.COPD = chronic obstructive pulmonary disease; ED = emergency department; EPR-3 = National Asthma Education and Prevention Program Expert Panel Report 3; ICS = inhaled corticosteroid; LABA = long-acting beta-agonist; MPR = medication possession ratio.
Omalizumab New Users(N=766)
MPR (Range 0-1) Mean (SD) 0.68 (0.30)[Median] [0.80]
Discontinued omalizumab n (%) 347 (45.3)Days to discontinuation/end of study Mean (SD) 256.5 (131.6)
[Median] [353]Number of omalizumab claims Mean (SD) 10.1 (5.4)
[Median] [11]1 n (%) 64 (8.4)2-5 n (%) 94 (12.3)6-11 n (%) 249 (32.5)12+ n (%) 359 (46.9)
Table 3. Omalizumab Use in the Postindex Period
HR (95% CI) P ValueAge, y 12-17 vs 65+ 0.82 (0.41 - 1.67) 0.576 18-34 vs 65+ 1.15 (0.64 - 2.16) 0.655 35-44 vs 65+ 0.86 (0.50 - 1.57) 0.600 45-54 vs 65+ 0.81 (0.48 - 1.46) 0.457 55-64 vs 65+ 0.76 (0.44 - 1.38) 0.344Female vs male 0.83 (0.66 - 1.05) 0.121Region Midwest vs West 0.76 (0.52 - 1.12) 0.164 Northeast vs West 0.98 (0.63 - 1.51) 0.917 South vs West 1.16 (0.84 - 1.62) 0.380Usual asthma care physician specialty Allergist vs primary/other 0.85 (0.64 - 1.13) 0.264 Pulmonologist vs primary/other 0.84 (0.62 - 1.14) 0.273No. of chronic conditions 1-2 vs 7+ 0.89 (0.59 - 1.35) 0.596 3-4 vs 7+ 0.81 (0.58 - 1.13) 0.212 5-6 vs 7+ 0.69 (0.51 - 0.94) 0.020Evidence of allergy 0.97 (0.53 - 1.71) 0.914Therapy step No asthma medication vs Step 6 1.25 (0.64 - 2.49) 0.518 Step 1 vs Step 6 1.21 (0.61 - 2.44) 0.580 Step 2 vs Step 6 1.18 (0.69 - 2.13) 0.560 Step 3 vs Step 6 0.79 (0.46 - 1.44) 0.430 Step 4 vs Step 6 0.92 (0.54 - 1.66) 0.769 Step 5 vs Step 6 0.90 (0.54 - 1.58) 0.689 Unclassified vs Step 6 0.73 (0.33 - 1.59) 0.437Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) 1.18 (0.66 - 2.07) 0.563 Missing (no controller or reliever) vs high ratio (≥0.5) 1.36 (0.77 - 2.43) 0.289COPD 0.93 (0.71 - 1.22) 0.613Sinusitis 0.99 (0.77 - 1.27) 0.916Rhinitis 0.65 (0.40 - 1.09) 0.082Tonsillitis 0.58 (0.29 - 1.02) 0.083Acute upper respiratory infection 1.12 (0.88 - 1.42) 0.341Conjunctivitis 1.13 (0.85 - 1.49) 0.384Chronic otitis media 0.61 (0.26 - 1.24) 0.218Nasal polyposis 1.46 (1.05 - 2.00) 0.019Cough 0.95 (0.75 - 1.19) 0.627Diseases of the circulatory system 1.32 (1.00 - 1.76) 0.052
Table 4. Risk of Discontinuing Omalizumab: Adjusted Hazard Ratios and 95% Confidence Intervals
CI = confidence interval; HR = hazard ratio.
Coefficient SE P ValueIntercept 0.49 0.10 <0.001Age, y 12-17 vs 65+ 0.04 0.07 0.567 18-34 vs 65+ -0.04 0.06 0.520 35-44 vs 65+ 0.01 0.06 0.845 45-54 vs 65+ 0.04 0.06 0.483 55-64 vs 65+ 0.06 0.06 0.298Female vs male 0.03 0.02 0.268Region Midwest vs West 0.03 0.04 0.375 Northeast vs West -0.03 0.04 0.545 South vs West -0.03 0.03 0.290Usual asthma care physician specialty Allergist vs primary/other 0.04 0.03 0.178 Pulmonologist vs primary/other 0.04 0.03 0.193No. of chronic conditions 1-2 vs 7+ -0.03 0.04 0.522 3-4 vs 7+ 0.03 0.03 0.334 5-6 vs 7+ 0.07 0.03 0.028Evidence of allergy 0.06 0.06 0.351Therapy step No asthma medication vs Step 6 -0.02 0.07 0.818 Step 1 vs Step 6 -0.01 0.07 0.933 Step 2 vs Step 6 0.01 0.06 0.817 Step 3 vs Step 6 0.08 0.06 0.138 Step 4 vs Step 6 0.04 0.05 0.506 Step 5 vs Step 6 0.04 0.05 0.443 Unclassified vs Step 6 0.10 0.07 0.182Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) -0.03 0.06 0.602 Missing (no controller or reliever) vs high ratio (≥0.5)
-0.08 0.06 0.211
COPD 0.03 0.03 0.317Sinusitis 0.00 0.02 0.927Rhinitis 0.08 0.05 0.112Tonsillitis 0.07 0.05 0.178Acute upper respiratory infection 0.00 0.02 0.950Conjunctivitis -0.02 0.03 0.477Chronic otitis media 0.13 0.07 0.050Nasal polyposis -0.09 0.03 0.005Cough 0.01 0.02 0.684Diseases of the circulatory system -0.07 0.03 0.008
Table 2. Multivariate Analysis: Characteristics Associated With Omalizumab MPR
COPD = chronic obstructive pulmonary disease; MPR = medication possession ratio.
This study was supported by Genentech, Inc., South San Francisco, CA, and Novartis Pharmaceuticals Corporation, East Hanover, NJ, and third-party writing assistance for this poster was provided by Genentech, Inc., South San Francisco CA, and Novartis Pharmaceuticals, East Hanover, NJ.
Chara
cte
rist
ics
Ass
ocia
ted W
ith M
edic
ati
on A
dhere
nce A
mong N
ew
Om
alizu
mab U
sers
46
Jam
es Z
azza
li, P
hD, M
PH
1 ; M
icha
el S
. Bro
der
, MD
, MS
HS
2 ; E
unic
e C
hang
, PhD
2 1 G
enen
tech
Inc.
, Sou
th S
an F
ranc
isco
, CA
; 2 Par
tner
ship
for
Hea
lth A
naly
tic R
esea
rch,
Bev
erly
Hill
s, C
A
AB
ST
RA
CT
OB
JEC
TIV
E: A
dhe
renc
e to
om
aliz
umab
(OM
A) t
hera
py
has
not
bee
n w
ell s
tud
ied
. W
e so
ught
to
iden
tify
char
acte
ristic
s as
soci
ated
with
ad
here
nce
amon
g ne
w
OM
A u
sers
. M
ET
HO
DS
: Thi
s w
as a
ret
rosp
ectiv
e co
hort
ana
lysi
s us
ing
a H
IPA
A-c
omp
liant
cl
aim
s d
atab
ase.
The
stu
dy
iden
tified
ast
hma
pat
ient
s w
ho w
ere
≥12
year
s ol
d,
new
ly t
reat
ed w
ith O
MA
bet
wee
n 7/
1/20
04 a
nd 6
/30/
2007
, and
enr
olle
d fo
r
1 ye
ar b
efor
e an
d 1
yea
r af
ter
the
first
OM
A c
laim
. Ad
here
nce
was
mea
sure
d
by
med
icat
ion
pos
sess
ion
ratio
(MP
R) (
tota
l day
s of
med
icat
ion
sup
plie
d ÷
365
) w
ith e
ach
OM
A c
laim
con
sid
ered
a 2
8-d
ay s
upp
ly. T
hera
py
step
, as
defi
ned
by
the
EP
R3
guid
elin
es, w
as a
ssig
ned
usi
ng a
pub
lishe
d a
lgor
ithm
. A b
ase
linea
r re
gres
sion
mod
el w
as c
ond
ucte
d w
ith M
PR
as
the
dep
end
ent
varia
ble
and
d
emog
rap
hics
, phy
sici
an s
pec
ialty
, res
pira
tory
com
orb
iditi
es, a
sthm
a co
ntro
l, th
erap
y st
ep, a
nd m
edic
atio
n ra
tio fo
rced
into
the
mod
el a
s in
dep
end
ent
varia
ble
s.
Oth
er c
omor
bid
ities
wer
e co
nsid
ered
with
forw
ard
sel
ectio
n an
d r
etai
ned
if
sign
ifica
nt a
t p
<.0
5.R
ES
ULT
S: W
e id
entifi
ed 7
66 n
ew O
MA
use
rs; m
ean
age
was
43
(SD
14)
yea
rs,
and
61%
(n=
466)
wer
e fe
mal
e. F
orty
-tw
o p
erce
nt o
f pat
ient
s ha
d a
llerg
ists
as
thei
r us
ual s
ourc
e of
car
e, 2
0% p
ulm
onol
ogis
ts, 2
8.3%
prim
ary
care
phy
sici
ans,
and
th
e re
mai
nder
oth
er/u
nsp
ecifi
ed. T
he m
ean
num
ber
of c
hron
ic c
ond
ition
s w
as
5 (S
D 2
). B
efor
e st
artin
g O
MA
, mos
t of
pat
ient
s w
ere
on E
PR
3 S
tep
5 t
hera
py
(28%
), fo
llow
ed b
y S
tep
4 (1
8%),
Ste
p 3
(16%
), S
tep
2 (1
5%),
Ste
p 1
(6%
), an
d
Ste
p 6
(5%
). M
ean
MP
R in
the
yea
r af
ter
trea
tmen
t in
itiat
ion
was
0.6
8 (S
D 0
.30)
, an
d 5
5% p
ersi
sted
with
the
rap
y fo
r at
leas
t 1
year
. The
fina
l mod
el in
clud
ed t
he
bas
e va
riab
les
and
dis
ease
s of
circ
ulat
ory
syst
em (t
he o
nly
add
ition
al s
igni
fican
t p
red
icto
r). S
igni
fican
t p
red
icto
rs o
f hig
her
MP
R w
ere:
car
e p
rovi
ded
by
alle
rgis
ts
or p
ulm
onol
ogis
ts, c
hron
ic o
titis
med
ia, 5
or
6 vs
≥7
chro
nic
cond
ition
s, a
nd n
o d
isea
se o
f the
circ
ulat
ory
syst
em.
CO
NC
LUS
ION
S: W
e fo
und
MP
R fo
r om
aliz
umab
to
be
high
er o
n av
erag
e th
an
wha
t ha
s b
een
rep
orte
d fo
r co
mb
inat
ion
cort
icos
tero
id/lo
ng-a
ctin
g b
eta-
agon
ist
ther
apy.
Car
e b
y an
alle
rgis
t or
pul
mon
olog
ist
was
ass
ocia
ted
with
gre
ater
ad
here
nce.
Con
foun
din
g b
y d
isea
se s
ever
ity is
pos
sib
le, b
ut t
he r
elat
ions
hip
was
si
gnifi
cant
aft
er c
ontr
ollin
g fo
r th
erap
y st
ep. I
f thi
s fin
din
g is
con
firm
ed in
oth
er
stud
ies,
it s
upp
orts
the
val
ue o
f sp
ecia
list
care
for
pat
ient
s w
ith d
ifficu
lt to
tr
eat
asth
ma.
CO
NC
LUS
ION
S•
Am
ong
new
om
aliz
umab
use
rs, w
e es
timat
ed t
he M
PR
to
be
0.68
with
54.
7%
per
sist
ing
on o
mal
izum
ab a
t on
e ye
ar a
fter
initi
atin
g th
erap
y.
• O
mal
izum
ab a
dhe
renc
e an
d p
ersi
sten
ce a
re c
onsi
sten
t w
ith p
revi
ousl
y p
ublis
hed
re
por
ts, a
lthou
gh t
his
anal
ysis
ind
icat
es s
light
ly h
ighe
r es
timat
es.
• T
he n
umb
er o
f chr
onic
con
diti
ons
and
sel
ect
resp
irato
ry c
omor
bid
ities
wer
e si
gnifi
cant
pre
dic
tors
of a
dhe
renc
e an
d p
ersi
sten
ce in
our
mul
tivar
iate
mod
els.
LIM
ITA
TIO
NS
• A
dm
inis
trat
ive
clai
ms
dat
a d
o no
t in
dic
ate
the
dos
ing
sche
dul
e fo
r om
aliz
umab
; in
thi
s an
alys
is, e
ach
dos
e of
om
aliz
umab
was
ass
umed
to
be
a 28
day
sup
ply
.
• In
clus
ion
of c
onco
mita
nt a
sthm
a m
edic
atio
ns a
s co
varia
tes
is in
dic
ativ
e on
ly o
f th
ose
med
icat
ions
fille
d a
nd n
ot r
eflec
tive
of p
atie
nt a
dhe
renc
e to
the
rap
y.
BA
CK
GR
OU
ND
• L
ack
of a
dhe
renc
e to
pre
scrib
ed t
reat
men
ts fo
r as
thm
a is
a w
ell-
know
n p
rob
lem
. R
ates
of n
onad
here
nce
rang
e fr
om 3
0% t
o 70
%.1-
3
• P
oor
asth
ma
med
icat
ion
adhe
renc
e is
ass
ocia
ted
with
dec
reas
es in
ast
hma
cont
rol
and
incr
ease
s in
em
erge
ncy
dep
artm
ent
visi
ts, h
osp
italiz
atio
ns, a
nd t
he n
eed
for
or
al c
ortic
oste
roid
s.3,
4
• O
mal
izum
ab, a
hum
aniz
ed m
onoc
lona
l ant
ibod
y ta
rget
ing
imm
unog
lob
ulin
E,
is a
pp
rove
d in
the
Uni
ted
Sta
tes
for
the
trea
tmen
t of
ad
ults
and
ad
oles
cent
s
(≥12
yea
rs) w
ith m
oder
ate
to s
ever
e p
ersi
sten
t al
lerg
ic a
sthm
a th
at is
inad
equa
tely
co
ntro
lled
with
inha
led
cor
ticos
tero
ids.
5
• A
dhe
renc
e to
om
aliz
umab
the
rap
y ha
s no
t b
een
wel
l stu
die
d. O
ne p
rior
anal
ysis
of
asth
ma
pat
ient
s ne
wly
tre
ated
with
om
aliz
umab
est
imat
ed a
dhe
renc
e ra
tes
to b
e 64
.6%
with
54%
per
sist
ing
up t
o on
e ye
ar.6
OB
JEC
TIV
E•
To
iden
tify
char
acte
ristic
s as
soci
ated
with
ad
here
nce
and
per
sist
ence
am
ong
new
om
aliz
umab
use
rs
ME
TH
OD
SS
tud
y D
esig
n•
A r
etro
spec
tive
coho
rt a
naly
sis
usin
g a
HIP
AA
-com
plia
nt a
dm
inis
trat
ive
clai
ms
dat
abas
e of
10
mill
ion
cove
red
live
s re
pre
sent
ing
all m
ajor
reg
ions
of t
he
Uni
ted
Sta
tes.
• E
ligib
le p
atie
nts
wer
e ≥1
2 ye
ars
old
, dia
gnos
ed w
ith a
sthm
a, n
ewly
tre
ated
w
ith o
mal
izum
ab b
etw
een
7/1/
2004
and
6/3
0/20
07, a
nd fo
llow
ed fo
r a
year
(F
igur
es 1
and
2).
RE
SU
LTS
• W
e id
entifi
ed 7
66 n
ew o
mal
izum
ab u
sers
; ave
rage
age
was
43.
4 ye
ars
and
61%
w
ere
fem
ale.
Fift
y p
erce
nt o
f pat
ient
s re
ceiv
ed a
sthm
a ca
re fr
om a
n al
lerg
ist,
28%
a
pul
mon
olog
ist,
and
17%
a p
rimar
y ca
re p
hysi
cian
. •
Ove
rall,
the
MP
R w
as 0
.68
(SD
, ± 0
.30)
. MP
R s
trat
ified
by
bas
elin
e m
easu
res
are
pre
sent
ed in
Tab
le 1
and
res
ults
of t
he m
ultiv
aria
te a
naly
sis
in T
able
2.
• D
urin
g fo
llow
-up
, 45.
3% o
f pat
ient
s d
isco
ntin
ued
tre
atm
ent.
Mea
n (S
D) t
ime
to d
isco
ntin
uatio
n/en
d o
f stu
dy
was
256
.5 (1
31.6
) day
s; m
edia
n w
as 3
53 d
ays
(T
able
3, F
igur
e 3)
.•
The
ris
k of
dis
cont
inui
ng o
mal
izum
ab w
as e
xam
ined
acr
oss
a nu
mb
er o
f var
iab
les
(Tab
le 4
).
RE
FER
EN
CE
S1.
Ben
der
B e
t al
. Ann
Alle
rgy
Ast
hma
Imm
unol
. 200
0;85
(5):4
16-4
21.
2. B
end
er B
G e
t al
. Ann
Alle
rgy
Ast
hma
Imm
unol
. 200
7;98
(4):3
22-3
28.
3. W
orld
Hea
lth O
rgan
izat
ion.
Ad
here
nce
to lo
ng-t
erm
th
erap
ies:
evi
den
ce fo
r ac
tion.
Gen
eva,
Sw
itzer
land
: W
orld
Hea
lth O
rgan
izat
ion;
200
3.4.
Will
iam
s LK
et
al. J
Alle
rgy
Clin
Imm
unol
. 200
4;11
4(6)
:12
88-1
293.
5. X
olai
r [p
resc
ribin
g in
form
atio
n]. S
outh
San
Fra
ncis
co, C
A:
Gen
ente
ch U
SA
, Inc
.; 20
10.
6. B
rod
er M
S e
t al
. Alle
rgy
Ast
hma
Pro
c. 2
009;
30(2
):148
-157
.7.
Dey
o R
A e
t al
. J C
lin E
pid
emio
l. 19
92;4
5(6)
:613
-619
.8.
Nat
iona
l Ast
hma
Ed
ucat
ion
and
Pre
vent
ion
Pro
gram
. Exp
ert
Pan
el R
epor
t 3
(EP
R-3
): G
uid
elin
es fo
r th
e D
iagn
osis
and
M
anag
emen
t of
Ast
hma-
Sum
mar
y R
epor
t 20
07. J
Alle
rgy
Clin
Imm
unol
. 200
7;12
0(5
sup
pl):
S94
-S13
8.
In
dex
Dat
eFi
rst
clai
m f
or
om
aliz
umab
in ID
per
iod
1-ye
ar p
erio
d (P
rein
dex
per
iod
)
7/1/
036/
30/0
8
6/30
/07
7/1/
04
Iden
tifi
cati
on
(ID) P
erio
d
1-ye
ar p
erio
d (P
ost
ind
ex p
erio
d)
Fig
ure
1. S
tud
y Ti
mel
ine
% of Patients on Treatment
Day
s to
Dis
cont
inue
Om
aliz
umab
100 75 50 25 0
050
100
150
200
250
300
350
400
Fig
ure
3. D
ays
to D
isco
ntin
uatio
n of
Om
aliz
umab
.
Stu
dy
Out
com
es•
Ad
here
nce
to o
mal
izum
ab t
hera
py
dur
ing
the
pos
tind
ex p
erio
d, a
s m
easu
red
by
the
med
icat
ion
pos
sess
ion
ratio
(MP
R)
– C
alcu
late
d a
s th
e to
tal d
ays
of m
edic
atio
n su
pp
lied
ove
r th
e ye
ar, d
ivid
ed b
y 36
5
–
Eac
h om
aliz
umab
cla
im w
as c
onsi
der
ed a
28-
day
sup
ply
•
Med
icat
ion
per
sist
ence
dur
ing
the
pos
tind
ex p
erio
d, a
s m
easu
red
by
time
to d
isco
ntin
uatio
n
–
Cal
cula
ted
as
the
num
ber
of d
ays
on o
mal
izum
ab w
ith n
o ga
ps
of m
ore
than
45
day
sS
tati
stic
al A
naly
sis
• T
o ev
alua
te t
he a
ssoc
iatio
n of
bas
elin
e m
easu
res
(pat
ient
dem
ogra
phi
c an
d c
linic
al
char
acte
ristic
s) w
ith o
mal
izum
ab a
dhe
renc
e an
d p
ersi
sten
ce, w
e co
nduc
ted
a li
near
re
gres
sion
mod
el w
ith a
dhe
renc
e (M
PR
) as
the
cont
inuo
us d
epen
den
t va
riab
le a
nd
all b
asel
ine
mea
sure
s as
ind
epen
den
t va
riab
les
and
a s
imila
r lo
gist
ic r
egre
ssio
n m
odel
to
pre
dic
t p
ersi
sten
ce a
t on
e ye
ar.
– B
ased
on
biv
aria
te a
naly
sis
resu
lts a
nd c
linic
al r
elev
ance
, dem
ogra
phi
cs (a
ge,
sex,
and
reg
ion)
, ast
hma
phy
sici
an s
pec
ialty
, num
ber
of c
hron
ic c
ond
ition
s,
evid
ence
of a
llerg
y, r
esp
irato
ry-s
pec
ific
com
orb
iditi
es, t
hera
py
step
, and
m
edic
atio
n ra
tio w
ere
first
incl
uded
in t
he li
near
reg
ress
ion
mod
el; w
e th
en u
sed
fo
rwar
d s
elec
tion
to id
entif
y si
gnifi
cant
res
ults
for
othe
r b
asel
ine
mea
sure
s.
3456
pat
ient
s w
ith a
cla
im
of o
mal
izum
ab in
ID p
erio
d(7
/1/2
004–
6/30
/200
7)
n=33
91
n=33
77
n=28
86
n=76
6
65 p
atie
nts
with
age
<12
yea
rs o
n th
e in
dex
dat
e (th
e d
ate
of �
rst
omal
izum
ab c
laim
in ID
per
iod
)
14 p
atie
nts
had
no
asth
ma
dia
gnos
is
in t
he p
rein
dex
or
pos
tind
ex p
erio
d
491
pre
viou
sly
used
om
aliz
umab
2120
pat
ient
s w
ere
not
cont
inuo
usly
enr
olle
d
in t
he p
rein
dex
and
pos
tind
ex p
erio
ds
Fig
ure
2. C
ohor
t S
elec
tion
for
Om
aliz
umab
New
Use
rs.
ID =
iden
tifica
tion.
N(%
)M
PR
M
ean
(SD
)P
Val
ue
All
766
(100
.0)
0.68
(0.3
0)
Dem
og
rap
hics
Ag
e, y
(mea
n 43
.4; S
D 1
4.5)
0.15
7
12-
1768
(8.9
)0.
72 (0
.28)
18-
3411
1(1
4.5)
0.62
(0.3
4)
35-
4418
3(2
3.9)
0.66
(0.3
0)
45-
5423
4(3
0.5)
0.68
(0.3
0)
55-
6414
1(1
8.4)
0.71
(0.2
7)
65+
29(3
.8)
0.65
(0.2
4)
Sex
0.17
4
Fem
ale
466
(60.
8)0.
69 (0
.30)
Mal
e30
0(3
9.2)
0.66
(0.3
0)
Reg
ion
0.05
6
Mid
wes
t18
4(2
4.0)
0.71
(0.2
8)
Nor
thea
st86
(11.
2)0.
62 (0
.32)
Sou
th38
7(5
0.5)
0.66
(0.3
0)
Wes
t10
9(1
4.2)
0.70
(0.2
7)
Phy
sici
an S
pec
ialt
y
Usu
al a
sthm
a ca
re p
hysi
cian
a0.
020
Pul
mon
olog
ist
214
(27.
9)0.
68 (0
.28)
Prim
ary
care
128
(16.
7)0.
64 (0
.31)
Alle
rgis
ts38
1(4
9.7)
0.69
(0.2
9)
Oth
er22
(2.9
)0.
66 (0
.36)
Unk
now
n21
(2.7
)0.
48 (0
.39)
Gen
eral
Co
mo
rbid
itie
s
No
. of
chro
nic
cond
itio
ns (m
ean
4.7;
SD
2.3
)0.
047
1-2
155
(20.
2)0.
63 (0
.33)
3-4
227
(29.
6)0.
69 (0
.29)
5-6
216
(28.
2)0.
71 (0
.27)
7+
168
(21.
9)0.
65 (0
.29)
Cha
rlso
n C
om
orb
idit
y In
dex
7 (m
ean
1.8;
SD
1.7
)0.
967
153
0(6
9.2)
0.68
(0.3
0)
286
(11.
2)0.
68 (0
.29)
3+
150
(19.
6)0.
67 (0
.29)
Evi
den
ce o
f al
lerg
y<
.001
No
57(7
.4)
0.55
(0.3
4)
Yes
709
(92.
6)0.
69 (0
.29)
Res
pir
ato
ry-S
pec
ific
Co
mo
rbid
itie
s
CO
PD
0.33
8
No
577
(75.
3)0.
67 (0
.31)
Yes
189
(24.
7)0.
69 (0
.27)
Sin
usit
is0.
551
No
243
(31.
7)0.
67 (0
.31)
Yes
523
(68.
3)0.
68 (0
.29)
Rhi
niti
s<
0.00
1
No
93(1
2.1)
0.57
(0.3
3)
Yes
673
(87.
9)0.
69 (0
.29)
To
nsill
itis
0.32
1
No
731
(95.
4)0.
67 (0
.30)
Yes
35(4
.6)
0.72
(0.2
8)
Acu
te u
pp
er r
esp
irat
ory
infe
ctio
n0.
471
No
515
(67.
2)0.
67 (0
.30)
Yes
251
(32.
8)0.
69 (0
.28)
Co
njun
ctiv
itis
0.89
4
No
619
(80.
8)0.
68 (0
.30)
Yes
147
(19.
2)0.
67 (0
.29)
Chr
oni
c o
titi
s m
edia
0.00
2
No
746
(97.
4)0.
67 (0
.30)
Yes
20(2
.6)
0.81
(0.1
7)
Nas
al p
oly
po
sis
0.07
3
No
663
(86.
6)0.
68 (0
.29)
Yes
103
(13.
4)0.
63 (0
.32)
Co
ugh
0.42
7
No
381
(49.
7)0.
67 (0
.31)
Yes
385
(50.
3)0.
68 (0
.28)
N(%
)M
PR
M
ean
(SD
)P
Val
ue
All
766
(100
.0)
0.68
(0.3
0)
Ast
hma
Co
ntro
l in
the
Pre
ind
ex P
erio
d
Po
or
asth
ma
cont
rolb
0.09
6
No
191
(24.
9)0.
64 (0
.32)
Yes
575
(75.
1)0.
69 (0
.29)
Ast
hma-
rela
ted
inp
atie
nt h
osp
ital
izat
ion
0.34
0
No
657
(85.
8)0.
68 (0
.30)
Yes
109
(14.
2)0.
65 (0
.30)
Ast
hma-
rela
ted
ED
vis
it0.
412
No
705
(92.
0)0.
67 (0
.30)
Yes
61(8
.0)
0.71
(0.2
9)
Tw
o o
r m
ore
ora
l co
rtic
ost
ero
id p
resc
rip
tio
ns f
illed
0.03
1
No
265
(34.
6)0.
64 (0
.32)
Yes
501
(65.
4)0.
69 (0
.28)
Six
or
mo
re s
hort
-act
ing
bet
a-ag
oni
st p
resc
rip
tio
ns f
illed
0.
310
No
434
(56.
7)0.
67 (0
.30)
Yes
332
(43.
3)0.
69 (0
.29)
EP
R-3
Gui
del
ines
8 T
hera
py
Ste
p in
Pre
ind
ex P
erio
d
The
rap
y st
ep p
rio
r to
init
iati
ng o
mal
izum
ab0.
006
No
asth
ma
med
s68
(8.9
)0.
56 (0
.33)
Ste
p 1
44(5
.7)
0.61
(0.3
2)
Ste
p 2
112
(14.
6)0.
67 (0
.30)
Ste
p 3
121
(15.
8)0.
73 (0
.27)
Ste
p 4
136
(17.
8)0.
69 (0
.28)
Ste
p 5
214
(27.
9)0.
69 (0
.29)
Ste
p 6
37(4
.8)
0.63
(0.3
3)
Unc
lass
ified
34(4
.4)
0.72
(0.2
8)
Med
icat
ion
Rat
ioc
and
Oth
er A
sthm
a M
edic
atio
n U
se in
the
Pre
ind
ex P
erio
d
Med
icat
ion
rati
o (m
ean
0.64
; SD
0.2
8 am
ong
718
pat
ient
s w
ith m
edic
atio
n ra
tios)
0.00
3
Low
rat
io (<
0.5)
197
(25.
7)0.
64 (0
.31)
Hig
h ra
tio (≥
0.5)
521
(68.
0)0.
70 (0
.28)
Mis
sing
(n
o co
ntro
ller
or
rel
ieve
r)48
(6.3
)0.
55 (0
.35)
Sho
rt-a
ctin
g b
eta-
ago
nist
inha
lers
0.04
4
No
115
(15.
0)0.
62 (0
.32)
Yes
651
(85.
0)0.
68 (0
.29)
Ant
icho
liner
gic
s0.
077
No
521
(68.
0)0.
66 (0
.30)
Yes
245
(32.
0)0.
70 (0
.29)
Ora
l co
rtic
ost
ero
ids
0.11
9
No
153
(20.
0)0.
64 (0
.32)
Yes
613
(80.
0)0.
68 (0
.29)
ICS
/LA
BA
co
mb
inat
ion
0.05
1
No
240
(31.
3)0.
64 (0
.31)
Yes
526
(68.
7)0.
69 (0
.29)
ICS
alo
ne
0.28
6
No
450
(58.
7)0.
67 (0
.31)
Yes
316
(41.
3)0.
69 (0
.28)
LAB
A a
lone
0.79
0
No
643
(83.
9)0.
67 (0
.30)
Yes
123
(16.
1)0.
68 (0
.30)
Mas
t ce
ll st
abili
zers
0.10
8
No
734
(95.
8)0.
67 (0
.30)
Yes
32(4
.2)
0.76
(0.2
5)
Met
hylx
anth
ines
0.81
0
No
670
(87.
5)0.
67 (0
.30)
Yes
96(1
2.5)
0.68
(0.2
6)
Leuk
otr
iene
rec
epto
r an
tag
oni
sts
0.07
8
No
221
(28.
9)0.
65 (0
.31)
Yes
545
(71.
1)0.
69 (0
.29)
Tab
le 1
. MP
R o
f Om
aliz
umab
Str
atifi
ed b
y P
atie
nt D
emog
rap
hics
and
Clin
ical
Cha
ract
eris
tics
a The
usu
al a
sthm
a ca
re p
hysi
cian
was
det
erm
ined
by
the
larg
est
plu
ralit
y of
offi
ce v
isits
with
eva
luat
ion
and
m
anag
emen
t se
rvic
es a
nd a
sthm
a d
iagn
osis
. bA
sthm
a-re
late
d in
pat
ient
hos
pita
lizat
ion,
ast
hma-
rela
ted
ED
vis
it, ≥
2 or
al c
ortic
oste
roid
pre
scrip
tions
fille
d,
or ≥
6 sh
ort-
actin
g b
eta-
agon
ist
pre
scrip
tions
fille
d.
c Med
icat
ion
ratio
is u
nits
of a
sthm
a co
ntro
llers
to
units
of c
ontr
olle
rs +
uni
ts o
f rel
ieve
rs.
CO
PD
= c
hron
ic o
bst
ruct
ive
pul
mon
ary
dis
ease
; ED
= e
mer
genc
y d
epar
tmen
t; E
PR
-3 =
Nat
iona
l Ast
hma
Ed
ucat
ion
and
Pre
vent
ion
Pro
gram
Exp
ert
Pan
el R
epor
t 3;
ICS
= in
hale
d c
ortic
oste
roid
; LA
BA
= lo
ng-a
ctin
g b
eta-
agon
ist;
MP
R =
med
icat
ion
pos
sess
ion
ratio
.
Om
aliz
umab
New
Use
rs(N
=76
6)M
PR
(Ran
ge 0
-1)
Mea
n (S
D)
0.68
(0.3
0)[M
edia
n][0
.80]
Dis
cont
inue
d o
mal
izum
ab
n (%
)34
7 (4
5.3)
Day
s to
dis
cont
inua
tion/
end
of s
tud
yM
ean
(SD
)25
6.5
(131
.6)
[Med
ian]
[353
]N
umb
er o
f om
aliz
umab
cla
ims
Mea
n (S
D)
10.1
(5.4
)[M
edia
n][1
1]1
n (%
)64
(8.4
)2-
5n
(%)
94 (1
2.3)
6-11
n (%
)24
9 (3
2.5)
12+
n (%
)35
9 (4
6.9)
Tab
le 3
. Om
aliz
umab
Use
in t
he P
ostin
dex
Per
iod
HR
(
95%
CI)
P V
alue
Ag
e, y
12-
17 v
s 65
+0.
82(0
.41
- 1.
67)
0.57
6 1
8-34
vs
65+
1.15
(0.6
4 -
2.16
)0.
655
35-
44 v
s 65
+0.
86(0
.50
- 1.
57)
0.60
0 4
5-54
vs
65+
0.81
(0.4
8 -
1.46
)0.
457
55-
64 v
s 65
+0.
76(0
.44
- 1.
38)
0.34
4Fe
mal
e vs
mal
e0.
83(0
.66
- 1.
05)
0.12
1R
egio
n M
idw
est
vs W
est
0.76
(0.5
2 -
1.12
)0.
164
Nor
thea
st v
s W
est
0.98
(0.6
3 -
1.51
)0.
917
Sou
th v
s W
est
1.16
(0.8
4 -
1.62
)0.
380
Usu
al a
sthm
a ca
re p
hysi
cian
sp
ecia
lty
Alle
rgis
t vs
prim
ary/
othe
r0.
85(0
.64
- 1.
13)
0.26
4 P
ulm
onol
ogis
t vs
prim
ary/
othe
r0.
84(0
.62
- 1.
14)
0.27
3N
o. o
f ch
roni
c co
ndit
ions
1-2
vs
7+0.
89(0
.59
- 1.
35)
0.59
6 3
-4 v
s 7+
0.81
(0.5
8 -
1.13
)0.
212
5-6
vs
7+0.
69(0
.51
- 0.
94)
0.02
0E
vid
ence
of
alle
rgy
0.97
(0.5
3 -
1.71
)0.
914
The
rap
y st
ep N
o as
thm
a m
edic
atio
n vs
Ste
p 6
1.25
(0.6
4 -
2.49
)0.
518
Ste
p 1
vs
Ste
p 6
1.21
(0.6
1 -
2.44
)0.
580
Ste
p 2
vs
Ste
p 6
1.18
(0.6
9 -
2.13
)0.
560
Ste
p 3
vs
Ste
p 6
0.79
(0.4
6 -
1.44
)0.
430
Ste
p 4
vs
Ste
p 6
0.92
(0.5
4 -
1.66
)0.
769
Ste
p 5
vs
Ste
p 6
0.90
(0.5
4 -
1.58
)0.
689
Unc
lass
ified
vs
Ste
p 6
0.73
(0.3
3 -
1.59
)0.
437
Med
icat
ion
rati
o L
ow r
atio
(<0.
5) v
s hi
gh r
atio
(≥0.
5)1.
18(0
.66
- 2.
07)
0.56
3 M
issi
ng (n
o co
ntro
ller
or r
elie
ver)
vs
high
rat
io (≥
0.5)
1.36
(0.7
7 -
2.43
)0.
289
CO
PD
0.93
(0.7
1 -
1.22
)0.
613
Sin
usit
is0.
99(0
.77
- 1.
27)
0.91
6R
hini
tis
0.65
(0.4
0 -
1.09
)0.
082
To
nsill
itis
0.58
(0.2
9 -
1.02
)0.
083
Acu
te u
pp
er r
esp
irat
ory
infe
ctio
n1.
12(0
.88
- 1.
42)
0.34
1C
onj
unct
ivit
is1.
13(0
.85
- 1.
49)
0.38
4C
hro
nic
oti
tis
med
ia0.
61(0
.26
- 1.
24)
0.21
8N
asal
po
lyp
osi
s1.
46(1
.05
- 2.
00)
0.01
9C
oug
h0.
95(0
.75
- 1.
19)
0.62
7D
isea
ses
of
the
circ
ulat
ory
sys
tem
1.32
(1.0
0 -
1.76
)0.
052
Tab
le 4
. Ris
k of
Dis
cont
inui
ng O
mal
izum
ab: A
dju
sted
Haz
ard
Rat
ios
and
95
% C
onfid
ence
Inte
rval
s
CI =
con
fiden
ce in
terv
al; H
R =
haz
ard
rat
io.
Co
effic
ient
SE
P V
alue
Inte
rcep
t0.
490.
10<
0.00
1A
ge,
y 1
2-17
vs
65+
0.04
0.07
0.56
7 1
8-34
vs
65+
-0.0
40.
060.
520
35-
44 v
s 65
+0.
010.
060.
845
45-
54 v
s 65
+0.
040.
060.
483
55-
64 v
s 65
+0.
060.
060.
298
Fem
ale
vs m
ale
0.03
0.02
0.26
8R
egio
n M
idw
est
vs W
est
0.03
0.04
0.37
5 N
orth
east
vs
Wes
t-0
.03
0.04
0.54
5 S
outh
vs
Wes
t-0
.03
0.03
0.29
0U
sual
ast
hma
care
phy
sici
an s
pec
ialt
y A
llerg
ist
vs p
rimar
y/ot
her
0.04
0.03
0.17
8 P
ulm
onol
ogis
t vs
prim
ary/
othe
r0.
040.
030.
193
No
. of
chro
nic
cond
itio
ns 1
-2 v
s 7+
-0.0
30.
040.
522
3-4
vs
7+0.
030.
030.
334
5-6
vs
7+0.
070.
030.
028
Evi
den
ce o
f al
lerg
y0.
060.
060.
351
The
rap
y st
ep N
o as
thm
a m
edic
atio
n vs
Ste
p 6
-0.0
20.
070.
818
Ste
p 1
vs
Ste
p 6
-0.0
10.
070.
933
Ste
p 2
vs
Ste
p 6
0.01
0.06
0.81
7 S
tep
3 v
s S
tep
60.
080.
060.
138
Ste
p 4
vs
Ste
p 6
0.04
0.05
0.50
6 S
tep
5 v
s S
tep
60.
040.
050.
443
Unc
lass
ified
vs
Ste
p 6
0.10
0.07
0.18
2M
edic
atio
n ra
tio
Low
rat
io (<
0.5)
vs
high
rat
io (≥
0.5)
-0.0
30.
060.
602
Mis
sing
(no
cont
rolle
r or
rel
ieve
r)
vs
high
rat
io (≥
0.5)
-0.0
80.
060.
211
CO
PD
0.03
0.03
0.31
7S
inus
itis
0.00
0.02
0.92
7R
hini
tis
0.08
0.05
0.11
2T
ons
illit
is0.
070.
050.
178
Acu
te u
pp
er r
esp
irat
ory
infe
ctio
n0.
000.
020.
950
Co
njun
ctiv
itis
-0.0
20.
030.
477
Chr
oni
c o
titi
s m
edia
0.13
0.07
0.05
0N
asal
po
lyp
osi
s-0
.09
0.03
0.00
5C
oug
h0.
010.
020.
684
Dis
ease
s o
f th
e ci
rcul
ato
ry s
yste
m-0
.07
0.03
0.00
8
Tab
le 2
. Mul
tivar
iate
Ana
lysi
s: C
hara
cter
istic
s A
ssoc
iate
d W
ith
Om
aliz
umab
MP
R
CO
PD
= c
hron
ic o
bst
ruct
ive
pul
mon
ary
dis
ease
; MP
R =
med
icat
ion
pos
sess
ion
ratio
.
This
stu
dy
was
sup
por
ted
by
Gen
ente
ch, I
nc.,
Sou
th S
an F
ranc
isco
, CA
, and
Nov
artis
P
harm
aceu
tical
s C
orp
orat
ion,
Eas
t H
anov
er, N
J, a
nd t
hird
-par
ty w
ritin
g as
sist
ance
for
this
pos
ter
was
pro
vid
ed b
y G
enen
tech
, Inc
., S
outh
San
Fra
ncis
co C
A, a
nd N
ovar
tis
Pha
rmac
eutic
als,
Eas
t H
anov
er, N
J.
Characteristics Associated With Medication Adherence Among New Omalizumab Users46
James Zazzali, PhD, MPH1; Michael S. Broder, MD, MSHS2; Eunice Chang, PhD2 1Genentech Inc., South San Francisco, CA; 2Partnership for Health Analytic Research, Beverly Hills, CA
ABSTRACTOBJECTIVE: Adherence to omalizumab (OMA) therapy has not been well studied. We sought to identify characteristics associated with adherence among new OMA users. METHODS: This was a retrospective cohort analysis using a HIPAA-compliant claims database. The study identified asthma patients who were ≥12 years old, newly treated with OMA between 7/1/2004 and 6/30/2007, and enrolled for 1 year before and 1 year after the first OMA claim. Adherence was measured by medication possession ratio (MPR) (total days of medication supplied ÷ 365) with each OMA claim considered a 28-day supply. Therapy step, as defined by the EPR3 guidelines, was assigned using a published algorithm. A base linear regression model was conducted with MPR as the dependent variable and demographics, physician specialty, respiratory comorbidities, asthma control, therapy step, and medication ratio forced into the model as independent variables. Other comorbidities were considered with forward selection and retained if significant at p<.05.RESULTS: We identified 766 new OMA users; mean age was 43 (SD 14) years, and 61% (n=466) were female. Forty-two percent of patients had allergists as their usual source of care, 20% pulmonologists, 28.3% primary care physicians, and the remainder other/unspecified. The mean number of chronic conditions was 5 (SD 2). Before starting OMA, most of patients were on EPR3 Step 5 therapy (28%), followed by Step 4 (18%), Step 3 (16%), Step 2 (15%), Step 1 (6%), and Step 6 (5%). Mean MPR in the year after treatment initiation was 0.68 (SD 0.30), and 55% persisted with therapy for at least 1 year. The final model included the base variables and diseases of circulatory system (the only additional significant predictor). Significant predictors of higher MPR were: care provided by allergists or pulmonologists, chronic otitis media, 5 or 6 vs ≥7 chronic conditions, and no disease of the circulatory system.CONCLUSIONS: We found MPR for omalizumab to be higher on average than what has been reported for combination corticosteroid/long-acting beta-agonist therapy. Care by an allergist or pulmonologist was associated with greater adherence. Confounding by disease severity is possible, but the relationship was significant after controlling for therapy step. If this finding is confirmed in other studies, it supports the value of specialist care for patients with difficult to treat asthma.
CONCLUSIONS• Among new omalizumab users, we estimated the MPR to be 0.68 with 54.7%
persisting on omalizumab at one year after initiating therapy.
• Omalizumab adherence and persistence are consistent with previously published reports, although this analysis indicates slightly higher estimates.
• The number of chronic conditions and select respiratory comorbidities were significant predictors of adherence and persistence in our multivariate models.
LIMITATIONS• Administrative claims data do not indicate the dosing schedule for omalizumab;
in this analysis, each dose of omalizumab was assumed to be a 28 day supply.
• Inclusion of concomitant asthma medications as covariates is indicative only of those medications filled and not reflective of patient adherence to therapy.
BACKGROUND• Lack of adherence to prescribed treatments for asthma is a well-known problem.
Rates of nonadherence range from 30% to 70%.1-3
• Poor asthma medication adherence is associated with decreases in asthma control and increases in emergency department visits, hospitalizations, and the need for oral corticosteroids.3,4
• Omalizumab, a humanized monoclonal antibody targeting immunoglobulin E, is approved in the United States for the treatment of adults and adolescents (≥12 years) with moderate to severe persistent allergic asthma that is inadequately controlled with inhaled corticosteroids.5
• Adherence to omalizumab therapy has not been well studied. One prior analysis of asthma patients newly treated with omalizumab estimated adherence rates to be 64.6% with 54% persisting up to one year.6
OBJECTIVE• To identify characteristics associated with adherence and persistence among new
omalizumab users
METHODSStudy Design• A retrospective cohort analysis using a HIPAA-compliant administrative claims
database of 10 million covered lives representing all major regions of the United States.
• Eligible patients were ≥12 years old, diagnosed with asthma, newly treated with omalizumab between 7/1/2004 and 6/30/2007, and followed for a year (Figures 1 and 2).
RESULTS• We identified 766 new omalizumab users; average age was 43.4 years and 61%
were female. Fifty percent of patients received asthma care from an allergist, 28% a pulmonologist, and 17% a primary care physician.
• Overall, the MPR was 0.68 (SD, ± 0.30). MPR stratified by baseline measures are presented in Table 1 and results of the multivariate analysis in Table 2.
• During follow-up, 45.3% of patients discontinued treatment. Mean (SD) time to discontinuation/end of study was 256.5 (131.6) days; median was 353 days (Table 3, Figure 3).
• The risk of discontinuing omalizumab was examined across a number of variables (Table 4).
REFERENCES1. Bender B et al. Ann Allergy Asthma Immunol. 2000;
85(5):416-421.2. Bender BG et al. Ann Allergy Asthma Immunol. 2007;
98(4):322-328.3. World Health Organization. Adherence to long-term
therapies: evidence for action. Geneva, Switzerland: World Health Organization; 2003.
4. Williams LK et al. J Allergy Clin Immunol. 2004;114(6):1288-1293.
5. Xolair [prescribing information]. South San Francisco, CA: Genentech USA, Inc.; 2010.
6. Broder MS et al. Allergy Asthma Proc. 2009;30(2):148-157.7. Deyo RA et al. J Clin Epidemiol. 1992;45(6):613-619.8. National Asthma Education and Prevention Program. Expert
Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007;120(5 suppl):S94-S138.
Index Date
First claim for omalizumab in ID period
1-year period (Preindex period)
7/1/03 6/30/08
6/30/077/1/04
Identification (ID) Period
1-year period (Postindex period)
Figure 1. Study Timeline
% o
f P
atie
nts
on
Tre
atm
ent
Days to Discontinue Omalizumab
100
75
50
25
0
0 50 100 150 200 250 300 350 400
Figure 3. Days to Discontinuation of Omalizumab.
Study Outcomes• Adherence to omalizumab therapy during the postindex period, as measured by the
medication possession ratio (MPR) – Calculated as the total days of medication supplied over the year, divided by 365 – Each omalizumab claim was considered a 28-day supply • Medication persistence during the postindex period, as measured by time
to discontinuation – Calculated as the number of days on omalizumab with no gaps of more than
45 daysStatistical Analysis• To evaluate the association of baseline measures (patient demographic and clinical
characteristics) with omalizumab adherence and persistence, we conducted a linear regression model with adherence (MPR) as the continuous dependent variable and all baseline measures as independent variables and a similar logistic regression model to predict persistence at one year.
– Based on bivariate analysis results and clinical relevance, demographics (age, sex, and region), asthma physician specialty, number of chronic conditions, evidence of allergy, respiratory-specific comorbidities, therapy step, and medication ratio were first included in the linear regression model; we then used forward selection to identify significant results for other baseline measures.
3456 patients with a claim of omalizumab in ID period
(7/1/2004–6/30/2007)
n=3391
n=3377
n=2886
n=766
65 patients with age <12 years on the index date (the date of �rst omalizumab claim in ID period)
14 patients had no asthma diagnosis in the preindex or postindex period
491 previously used omalizumab
2120 patients were not continuously enrolled in the preindex and postindex periods
Figure 2. Cohort Selection for Omalizumab New Users.
ID = identification.
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Demographics
Age, y (mean 43.4; SD 14.5) 0.157
12-17 68 (8.9) 0.72 (0.28)
18-34 111 (14.5) 0.62 (0.34)
35-44 183 (23.9) 0.66 (0.30)
45-54 234 (30.5) 0.68 (0.30)
55-64 141 (18.4) 0.71 (0.27)
65+ 29 (3.8) 0.65 (0.24)
Sex 0.174
Female 466 (60.8) 0.69 (0.30)
Male 300 (39.2) 0.66 (0.30)
Region 0.056
Midwest 184 (24.0) 0.71 (0.28)
Northeast 86 (11.2) 0.62 (0.32)
South 387 (50.5) 0.66 (0.30)
West 109 (14.2) 0.70 (0.27)
Physician Specialty
Usual asthma care physiciana 0.020
Pulmonologist 214 (27.9) 0.68 (0.28)
Primary care 128 (16.7) 0.64 (0.31)
Allergists 381 (49.7) 0.69 (0.29)
Other 22 (2.9) 0.66 (0.36)
Unknown 21 (2.7) 0.48 (0.39)
General Comorbidities
No. of chronic conditions (mean 4.7; SD 2.3) 0.047
1-2 155 (20.2) 0.63 (0.33)
3-4 227 (29.6) 0.69 (0.29)
5-6 216 (28.2) 0.71 (0.27)
7+ 168 (21.9) 0.65 (0.29)
Charlson Comorbidity Index7 (mean 1.8; SD 1.7) 0.967
1 530 (69.2) 0.68 (0.30)
2 86 (11.2) 0.68 (0.29)
3+ 150 (19.6) 0.67 (0.29)
Evidence of allergy <.001
No 57 (7.4) 0.55 (0.34)
Yes 709 (92.6) 0.69 (0.29)
Respiratory-Specific Comorbidities
COPD 0.338
No 577 (75.3) 0.67 (0.31)
Yes 189 (24.7) 0.69 (0.27)
Sinusitis 0.551
No 243 (31.7) 0.67 (0.31)
Yes 523 (68.3) 0.68 (0.29)
Rhinitis <0.001
No 93 (12.1) 0.57 (0.33)
Yes 673 (87.9) 0.69 (0.29)
Tonsillitis 0.321
No 731 (95.4) 0.67 (0.30)
Yes 35 (4.6) 0.72 (0.28)
Acute upper respiratory infection 0.471
No 515 (67.2) 0.67 (0.30)
Yes 251 (32.8) 0.69 (0.28)
Conjunctivitis 0.894
No 619 (80.8) 0.68 (0.30)
Yes 147 (19.2) 0.67 (0.29)
Chronic otitis media 0.002
No 746 (97.4) 0.67 (0.30)
Yes 20 (2.6) 0.81 (0.17)
Nasal polyposis 0.073
No 663 (86.6) 0.68 (0.29)
Yes 103 (13.4) 0.63 (0.32)
Cough 0.427
No 381 (49.7) 0.67 (0.31)
Yes 385 (50.3) 0.68 (0.28)
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Asthma Control in the Preindex Period
Poor asthma controlb 0.096
No 191 (24.9) 0.64 (0.32)
Yes 575 (75.1) 0.69 (0.29)
Asthma-related inpatient hospitalization 0.340
No 657 (85.8) 0.68 (0.30)
Yes 109 (14.2) 0.65 (0.30)
Asthma-related ED visit 0.412
No 705 (92.0) 0.67 (0.30)
Yes 61 (8.0) 0.71 (0.29)
Two or more oral corticosteroid prescriptions filled 0.031
No 265 (34.6) 0.64 (0.32)
Yes 501 (65.4) 0.69 (0.28)
Six or more short-acting beta-agonist prescriptions filled 0.310
No 434 (56.7) 0.67 (0.30)
Yes 332 (43.3) 0.69 (0.29)
EPR-3 Guidelines8 Therapy Step in Preindex Period
Therapy step prior to initiating omalizumab 0.006
No asthma meds 68 (8.9) 0.56 (0.33)
Step 1 44 (5.7) 0.61 (0.32)
Step 2 112 (14.6) 0.67 (0.30)
Step 3 121 (15.8) 0.73 (0.27)
Step 4 136 (17.8) 0.69 (0.28)
Step 5 214 (27.9) 0.69 (0.29)
Step 6 37 (4.8) 0.63 (0.33)
Unclassified 34 (4.4) 0.72 (0.28)
Medication Ratioc and Other Asthma Medication Use in the Preindex Period
Medication ratio (mean 0.64; SD 0.28 among 718 patients with medication ratios) 0.003
Low ratio (<0.5) 197 (25.7) 0.64 (0.31)
High ratio (≥0.5) 521 (68.0) 0.70 (0.28)
Missing (no controller or reliever)
48 (6.3) 0.55 (0.35)
Short-acting beta-agonist inhalers 0.044
No 115 (15.0) 0.62 (0.32)
Yes 651 (85.0) 0.68 (0.29)
Anticholinergics 0.077
No 521 (68.0) 0.66 (0.30)
Yes 245 (32.0) 0.70 (0.29)
Oral corticosteroids 0.119
No 153 (20.0) 0.64 (0.32)
Yes 613 (80.0) 0.68 (0.29)
ICS/LABA combination 0.051
No 240 (31.3) 0.64 (0.31)
Yes 526 (68.7) 0.69 (0.29)
ICS alone 0.286
No 450 (58.7) 0.67 (0.31)
Yes 316 (41.3) 0.69 (0.28)
LABA alone 0.790
No 643 (83.9) 0.67 (0.30)
Yes 123 (16.1) 0.68 (0.30)
Mast cell stabilizers 0.108
No 734 (95.8) 0.67 (0.30)
Yes 32 (4.2) 0.76 (0.25)
Methylxanthines 0.810
No 670 (87.5) 0.67 (0.30)
Yes 96 (12.5) 0.68 (0.26)
Leukotriene receptor antagonists 0.078
No 221 (28.9) 0.65 (0.31)
Yes 545 (71.1) 0.69 (0.29)
Table 1. MPR of Omalizumab Stratified by Patient Demographics and Clinical Characteristics
aThe usual asthma care physician was determined by the largest plurality of office visits with evaluation and management services and asthma diagnosis. bAsthma-related inpatient hospitalization, asthma-related ED visit, ≥2 oral corticosteroid prescriptions filled, or ≥6 short-acting beta-agonist prescriptions filled.cMedication ratio is units of asthma controllers to units of controllers + units of relievers.COPD = chronic obstructive pulmonary disease; ED = emergency department; EPR-3 = National Asthma Education and Prevention Program Expert Panel Report 3; ICS = inhaled corticosteroid; LABA = long-acting beta-agonist; MPR = medication possession ratio.
Omalizumab New Users(N=766)
MPR (Range 0-1) Mean (SD) 0.68 (0.30)[Median] [0.80]
Discontinued omalizumab n (%) 347 (45.3)Days to discontinuation/end of study Mean (SD) 256.5 (131.6)
[Median] [353]Number of omalizumab claims Mean (SD) 10.1 (5.4)
[Median] [11]1 n (%) 64 (8.4)2-5 n (%) 94 (12.3)6-11 n (%) 249 (32.5)12+ n (%) 359 (46.9)
Table 3. Omalizumab Use in the Postindex Period
HR (95% CI) P ValueAge, y 12-17 vs 65+ 0.82 (0.41 - 1.67) 0.576 18-34 vs 65+ 1.15 (0.64 - 2.16) 0.655 35-44 vs 65+ 0.86 (0.50 - 1.57) 0.600 45-54 vs 65+ 0.81 (0.48 - 1.46) 0.457 55-64 vs 65+ 0.76 (0.44 - 1.38) 0.344Female vs male 0.83 (0.66 - 1.05) 0.121Region Midwest vs West 0.76 (0.52 - 1.12) 0.164 Northeast vs West 0.98 (0.63 - 1.51) 0.917 South vs West 1.16 (0.84 - 1.62) 0.380Usual asthma care physician specialty Allergist vs primary/other 0.85 (0.64 - 1.13) 0.264 Pulmonologist vs primary/other 0.84 (0.62 - 1.14) 0.273No. of chronic conditions 1-2 vs 7+ 0.89 (0.59 - 1.35) 0.596 3-4 vs 7+ 0.81 (0.58 - 1.13) 0.212 5-6 vs 7+ 0.69 (0.51 - 0.94) 0.020Evidence of allergy 0.97 (0.53 - 1.71) 0.914Therapy step No asthma medication vs Step 6 1.25 (0.64 - 2.49) 0.518 Step 1 vs Step 6 1.21 (0.61 - 2.44) 0.580 Step 2 vs Step 6 1.18 (0.69 - 2.13) 0.560 Step 3 vs Step 6 0.79 (0.46 - 1.44) 0.430 Step 4 vs Step 6 0.92 (0.54 - 1.66) 0.769 Step 5 vs Step 6 0.90 (0.54 - 1.58) 0.689 Unclassified vs Step 6 0.73 (0.33 - 1.59) 0.437Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) 1.18 (0.66 - 2.07) 0.563 Missing (no controller or reliever) vs high ratio (≥0.5) 1.36 (0.77 - 2.43) 0.289COPD 0.93 (0.71 - 1.22) 0.613Sinusitis 0.99 (0.77 - 1.27) 0.916Rhinitis 0.65 (0.40 - 1.09) 0.082Tonsillitis 0.58 (0.29 - 1.02) 0.083Acute upper respiratory infection 1.12 (0.88 - 1.42) 0.341Conjunctivitis 1.13 (0.85 - 1.49) 0.384Chronic otitis media 0.61 (0.26 - 1.24) 0.218Nasal polyposis 1.46 (1.05 - 2.00) 0.019Cough 0.95 (0.75 - 1.19) 0.627Diseases of the circulatory system 1.32 (1.00 - 1.76) 0.052
Table 4. Risk of Discontinuing Omalizumab: Adjusted Hazard Ratios and 95% Confidence Intervals
CI = confidence interval; HR = hazard ratio.
Coefficient SE P ValueIntercept 0.49 0.10 <0.001Age, y 12-17 vs 65+ 0.04 0.07 0.567 18-34 vs 65+ -0.04 0.06 0.520 35-44 vs 65+ 0.01 0.06 0.845 45-54 vs 65+ 0.04 0.06 0.483 55-64 vs 65+ 0.06 0.06 0.298Female vs male 0.03 0.02 0.268Region Midwest vs West 0.03 0.04 0.375 Northeast vs West -0.03 0.04 0.545 South vs West -0.03 0.03 0.290Usual asthma care physician specialty Allergist vs primary/other 0.04 0.03 0.178 Pulmonologist vs primary/other 0.04 0.03 0.193No. of chronic conditions 1-2 vs 7+ -0.03 0.04 0.522 3-4 vs 7+ 0.03 0.03 0.334 5-6 vs 7+ 0.07 0.03 0.028Evidence of allergy 0.06 0.06 0.351Therapy step No asthma medication vs Step 6 -0.02 0.07 0.818 Step 1 vs Step 6 -0.01 0.07 0.933 Step 2 vs Step 6 0.01 0.06 0.817 Step 3 vs Step 6 0.08 0.06 0.138 Step 4 vs Step 6 0.04 0.05 0.506 Step 5 vs Step 6 0.04 0.05 0.443 Unclassified vs Step 6 0.10 0.07 0.182Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) -0.03 0.06 0.602 Missing (no controller or reliever) vs high ratio (≥0.5)
-0.08 0.06 0.211
COPD 0.03 0.03 0.317Sinusitis 0.00 0.02 0.927Rhinitis 0.08 0.05 0.112Tonsillitis 0.07 0.05 0.178Acute upper respiratory infection 0.00 0.02 0.950Conjunctivitis -0.02 0.03 0.477Chronic otitis media 0.13 0.07 0.050Nasal polyposis -0.09 0.03 0.005Cough 0.01 0.02 0.684Diseases of the circulatory system -0.07 0.03 0.008
Table 2. Multivariate Analysis: Characteristics Associated With Omalizumab MPR
COPD = chronic obstructive pulmonary disease; MPR = medication possession ratio.
This study was supported by Genentech, Inc., South San Francisco, CA, and Novartis Pharmaceuticals Corporation, East Hanover, NJ, and third-party writing assistance for this poster was provided by Genentech, Inc., South San Francisco CA, and Novartis Pharmaceuticals, East Hanover, NJ.
Characteristics Associated With Medication Adherence Among New Omalizumab Users46
James Zazzali, PhD, MPH1; Michael S. Broder, MD, MSHS2; Eunice Chang, PhD2 1Genentech Inc., South San Francisco, CA; 2Partnership for Health Analytic Research, Beverly Hills, CA
ABSTRACTOBJECTIVE: Adherence to omalizumab (OMA) therapy has not been well studied. We sought to identify characteristics associated with adherence among new OMA users. METHODS: This was a retrospective cohort analysis using a HIPAA-compliant claims database. The study identified asthma patients who were ≥12 years old, newly treated with OMA between 7/1/2004 and 6/30/2007, and enrolled for 1 year before and 1 year after the first OMA claim. Adherence was measured by medication possession ratio (MPR) (total days of medication supplied ÷ 365) with each OMA claim considered a 28-day supply. Therapy step, as defined by the EPR3 guidelines, was assigned using a published algorithm. A base linear regression model was conducted with MPR as the dependent variable and demographics, physician specialty, respiratory comorbidities, asthma control, therapy step, and medication ratio forced into the model as independent variables. Other comorbidities were considered with forward selection and retained if significant at p<.05.RESULTS: We identified 766 new OMA users; mean age was 43 (SD 14) years, and 61% (n=466) were female. Forty-two percent of patients had allergists as their usual source of care, 20% pulmonologists, 28.3% primary care physicians, and the remainder other/unspecified. The mean number of chronic conditions was 5 (SD 2). Before starting OMA, most of patients were on EPR3 Step 5 therapy (28%), followed by Step 4 (18%), Step 3 (16%), Step 2 (15%), Step 1 (6%), and Step 6 (5%). Mean MPR in the year after treatment initiation was 0.68 (SD 0.30), and 55% persisted with therapy for at least 1 year. The final model included the base variables and diseases of circulatory system (the only additional significant predictor). Significant predictors of higher MPR were: care provided by allergists or pulmonologists, chronic otitis media, 5 or 6 vs ≥7 chronic conditions, and no disease of the circulatory system.CONCLUSIONS: We found MPR for omalizumab to be higher on average than what has been reported for combination corticosteroid/long-acting beta-agonist therapy. Care by an allergist or pulmonologist was associated with greater adherence. Confounding by disease severity is possible, but the relationship was significant after controlling for therapy step. If this finding is confirmed in other studies, it supports the value of specialist care for patients with difficult to treat asthma.
CONCLUSIONS• Among new omalizumab users, we estimated the MPR to be 0.68 with 54.7%
persisting on omalizumab at one year after initiating therapy.
• Omalizumab adherence and persistence are consistent with previously published reports, although this analysis indicates slightly higher estimates.
• The number of chronic conditions and select respiratory comorbidities were significant predictors of adherence and persistence in our multivariate models.
LIMITATIONS• Administrative claims data do not indicate the dosing schedule for omalizumab;
in this analysis, each dose of omalizumab was assumed to be a 28 day supply.
• Inclusion of concomitant asthma medications as covariates is indicative only of those medications filled and not reflective of patient adherence to therapy.
BACKGROUND• Lack of adherence to prescribed treatments for asthma is a well-known problem.
Rates of nonadherence range from 30% to 70%.1-3
• Poor asthma medication adherence is associated with decreases in asthma control and increases in emergency department visits, hospitalizations, and the need for oral corticosteroids.3,4
• Omalizumab, a humanized monoclonal antibody targeting immunoglobulin E, is approved in the United States for the treatment of adults and adolescents (≥12 years) with moderate to severe persistent allergic asthma that is inadequately controlled with inhaled corticosteroids.5
• Adherence to omalizumab therapy has not been well studied. One prior analysis of asthma patients newly treated with omalizumab estimated adherence rates to be 64.6% with 54% persisting up to one year.6
OBJECTIVE• To identify characteristics associated with adherence and persistence among new
omalizumab users
METHODSStudy Design• A retrospective cohort analysis using a HIPAA-compliant administrative claims
database of 10 million covered lives representing all major regions of the United States.
• Eligible patients were ≥12 years old, diagnosed with asthma, newly treated with omalizumab between 7/1/2004 and 6/30/2007, and followed for a year (Figures 1 and 2).
RESULTS• We identified 766 new omalizumab users; average age was 43.4 years and 61%
were female. Fifty percent of patients received asthma care from an allergist, 28% a pulmonologist, and 17% a primary care physician.
• Overall, the MPR was 0.68 (SD, ± 0.30). MPR stratified by baseline measures are presented in Table 1 and results of the multivariate analysis in Table 2.
• During follow-up, 45.3% of patients discontinued treatment. Mean (SD) time to discontinuation/end of study was 256.5 (131.6) days; median was 353 days (Table 3, Figure 3).
• The risk of discontinuing omalizumab was examined across a number of variables (Table 4).
REFERENCES1. Bender B et al. Ann Allergy Asthma Immunol. 2000;
85(5):416-421.2. Bender BG et al. Ann Allergy Asthma Immunol. 2007;
98(4):322-328.3. World Health Organization. Adherence to long-term
therapies: evidence for action. Geneva, Switzerland: World Health Organization; 2003.
4. Williams LK et al. J Allergy Clin Immunol. 2004;114(6):1288-1293.
5. Xolair [prescribing information]. South San Francisco, CA: Genentech USA, Inc.; 2010.
6. Broder MS et al. Allergy Asthma Proc. 2009;30(2):148-157.7. Deyo RA et al. J Clin Epidemiol. 1992;45(6):613-619.8. National Asthma Education and Prevention Program. Expert
Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007;120(5 suppl):S94-S138.
Index Date
First claim for omalizumab in ID period
1-year period (Preindex period)
7/1/03 6/30/08
6/30/077/1/04
Identification (ID) Period
1-year period (Postindex period)
Figure 1. Study Timeline
% o
f P
atie
nts
on
Tre
atm
ent
Days to Discontinue Omalizumab
100
75
50
25
0
0 50 100 150 200 250 300 350 400
Figure 3. Days to Discontinuation of Omalizumab.
Study Outcomes• Adherence to omalizumab therapy during the postindex period, as measured by the
medication possession ratio (MPR) – Calculated as the total days of medication supplied over the year, divided by 365 – Each omalizumab claim was considered a 28-day supply • Medication persistence during the postindex period, as measured by time
to discontinuation – Calculated as the number of days on omalizumab with no gaps of more than
45 daysStatistical Analysis• To evaluate the association of baseline measures (patient demographic and clinical
characteristics) with omalizumab adherence and persistence, we conducted a linear regression model with adherence (MPR) as the continuous dependent variable and all baseline measures as independent variables and a similar logistic regression model to predict persistence at one year.
– Based on bivariate analysis results and clinical relevance, demographics (age, sex, and region), asthma physician specialty, number of chronic conditions, evidence of allergy, respiratory-specific comorbidities, therapy step, and medication ratio were first included in the linear regression model; we then used forward selection to identify significant results for other baseline measures.
3456 patients with a claim of omalizumab in ID period
(7/1/2004–6/30/2007)
n=3391
n=3377
n=2886
n=766
65 patients with age <12 years on the index date (the date of �rst omalizumab claim in ID period)
14 patients had no asthma diagnosis in the preindex or postindex period
491 previously used omalizumab
2120 patients were not continuously enrolled in the preindex and postindex periods
Figure 2. Cohort Selection for Omalizumab New Users.
ID = identification.
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Demographics
Age, y (mean 43.4; SD 14.5) 0.157
12-17 68 (8.9) 0.72 (0.28)
18-34 111 (14.5) 0.62 (0.34)
35-44 183 (23.9) 0.66 (0.30)
45-54 234 (30.5) 0.68 (0.30)
55-64 141 (18.4) 0.71 (0.27)
65+ 29 (3.8) 0.65 (0.24)
Sex 0.174
Female 466 (60.8) 0.69 (0.30)
Male 300 (39.2) 0.66 (0.30)
Region 0.056
Midwest 184 (24.0) 0.71 (0.28)
Northeast 86 (11.2) 0.62 (0.32)
South 387 (50.5) 0.66 (0.30)
West 109 (14.2) 0.70 (0.27)
Physician Specialty
Usual asthma care physiciana 0.020
Pulmonologist 214 (27.9) 0.68 (0.28)
Primary care 128 (16.7) 0.64 (0.31)
Allergists 381 (49.7) 0.69 (0.29)
Other 22 (2.9) 0.66 (0.36)
Unknown 21 (2.7) 0.48 (0.39)
General Comorbidities
No. of chronic conditions (mean 4.7; SD 2.3) 0.047
1-2 155 (20.2) 0.63 (0.33)
3-4 227 (29.6) 0.69 (0.29)
5-6 216 (28.2) 0.71 (0.27)
7+ 168 (21.9) 0.65 (0.29)
Charlson Comorbidity Index7 (mean 1.8; SD 1.7) 0.967
1 530 (69.2) 0.68 (0.30)
2 86 (11.2) 0.68 (0.29)
3+ 150 (19.6) 0.67 (0.29)
Evidence of allergy <.001
No 57 (7.4) 0.55 (0.34)
Yes 709 (92.6) 0.69 (0.29)
Respiratory-Specific Comorbidities
COPD 0.338
No 577 (75.3) 0.67 (0.31)
Yes 189 (24.7) 0.69 (0.27)
Sinusitis 0.551
No 243 (31.7) 0.67 (0.31)
Yes 523 (68.3) 0.68 (0.29)
Rhinitis <0.001
No 93 (12.1) 0.57 (0.33)
Yes 673 (87.9) 0.69 (0.29)
Tonsillitis 0.321
No 731 (95.4) 0.67 (0.30)
Yes 35 (4.6) 0.72 (0.28)
Acute upper respiratory infection 0.471
No 515 (67.2) 0.67 (0.30)
Yes 251 (32.8) 0.69 (0.28)
Conjunctivitis 0.894
No 619 (80.8) 0.68 (0.30)
Yes 147 (19.2) 0.67 (0.29)
Chronic otitis media 0.002
No 746 (97.4) 0.67 (0.30)
Yes 20 (2.6) 0.81 (0.17)
Nasal polyposis 0.073
No 663 (86.6) 0.68 (0.29)
Yes 103 (13.4) 0.63 (0.32)
Cough 0.427
No 381 (49.7) 0.67 (0.31)
Yes 385 (50.3) 0.68 (0.28)
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Asthma Control in the Preindex Period
Poor asthma controlb 0.096
No 191 (24.9) 0.64 (0.32)
Yes 575 (75.1) 0.69 (0.29)
Asthma-related inpatient hospitalization 0.340
No 657 (85.8) 0.68 (0.30)
Yes 109 (14.2) 0.65 (0.30)
Asthma-related ED visit 0.412
No 705 (92.0) 0.67 (0.30)
Yes 61 (8.0) 0.71 (0.29)
Two or more oral corticosteroid prescriptions filled 0.031
No 265 (34.6) 0.64 (0.32)
Yes 501 (65.4) 0.69 (0.28)
Six or more short-acting beta-agonist prescriptions filled 0.310
No 434 (56.7) 0.67 (0.30)
Yes 332 (43.3) 0.69 (0.29)
EPR-3 Guidelines8 Therapy Step in Preindex Period
Therapy step prior to initiating omalizumab 0.006
No asthma meds 68 (8.9) 0.56 (0.33)
Step 1 44 (5.7) 0.61 (0.32)
Step 2 112 (14.6) 0.67 (0.30)
Step 3 121 (15.8) 0.73 (0.27)
Step 4 136 (17.8) 0.69 (0.28)
Step 5 214 (27.9) 0.69 (0.29)
Step 6 37 (4.8) 0.63 (0.33)
Unclassified 34 (4.4) 0.72 (0.28)
Medication Ratioc and Other Asthma Medication Use in the Preindex Period
Medication ratio (mean 0.64; SD 0.28 among 718 patients with medication ratios) 0.003
Low ratio (<0.5) 197 (25.7) 0.64 (0.31)
High ratio (≥0.5) 521 (68.0) 0.70 (0.28)
Missing (no controller or reliever)
48 (6.3) 0.55 (0.35)
Short-acting beta-agonist inhalers 0.044
No 115 (15.0) 0.62 (0.32)
Yes 651 (85.0) 0.68 (0.29)
Anticholinergics 0.077
No 521 (68.0) 0.66 (0.30)
Yes 245 (32.0) 0.70 (0.29)
Oral corticosteroids 0.119
No 153 (20.0) 0.64 (0.32)
Yes 613 (80.0) 0.68 (0.29)
ICS/LABA combination 0.051
No 240 (31.3) 0.64 (0.31)
Yes 526 (68.7) 0.69 (0.29)
ICS alone 0.286
No 450 (58.7) 0.67 (0.31)
Yes 316 (41.3) 0.69 (0.28)
LABA alone 0.790
No 643 (83.9) 0.67 (0.30)
Yes 123 (16.1) 0.68 (0.30)
Mast cell stabilizers 0.108
No 734 (95.8) 0.67 (0.30)
Yes 32 (4.2) 0.76 (0.25)
Methylxanthines 0.810
No 670 (87.5) 0.67 (0.30)
Yes 96 (12.5) 0.68 (0.26)
Leukotriene receptor antagonists 0.078
No 221 (28.9) 0.65 (0.31)
Yes 545 (71.1) 0.69 (0.29)
Table 1. MPR of Omalizumab Stratified by Patient Demographics and Clinical Characteristics
aThe usual asthma care physician was determined by the largest plurality of office visits with evaluation and management services and asthma diagnosis. bAsthma-related inpatient hospitalization, asthma-related ED visit, ≥2 oral corticosteroid prescriptions filled, or ≥6 short-acting beta-agonist prescriptions filled.cMedication ratio is units of asthma controllers to units of controllers + units of relievers.COPD = chronic obstructive pulmonary disease; ED = emergency department; EPR-3 = National Asthma Education and Prevention Program Expert Panel Report 3; ICS = inhaled corticosteroid; LABA = long-acting beta-agonist; MPR = medication possession ratio.
Omalizumab New Users(N=766)
MPR (Range 0-1) Mean (SD) 0.68 (0.30)[Median] [0.80]
Discontinued omalizumab n (%) 347 (45.3)Days to discontinuation/end of study Mean (SD) 256.5 (131.6)
[Median] [353]Number of omalizumab claims Mean (SD) 10.1 (5.4)
[Median] [11]1 n (%) 64 (8.4)2-5 n (%) 94 (12.3)6-11 n (%) 249 (32.5)12+ n (%) 359 (46.9)
Table 3. Omalizumab Use in the Postindex Period
HR (95% CI) P ValueAge, y 12-17 vs 65+ 0.82 (0.41 - 1.67) 0.576 18-34 vs 65+ 1.15 (0.64 - 2.16) 0.655 35-44 vs 65+ 0.86 (0.50 - 1.57) 0.600 45-54 vs 65+ 0.81 (0.48 - 1.46) 0.457 55-64 vs 65+ 0.76 (0.44 - 1.38) 0.344Female vs male 0.83 (0.66 - 1.05) 0.121Region Midwest vs West 0.76 (0.52 - 1.12) 0.164 Northeast vs West 0.98 (0.63 - 1.51) 0.917 South vs West 1.16 (0.84 - 1.62) 0.380Usual asthma care physician specialty Allergist vs primary/other 0.85 (0.64 - 1.13) 0.264 Pulmonologist vs primary/other 0.84 (0.62 - 1.14) 0.273No. of chronic conditions 1-2 vs 7+ 0.89 (0.59 - 1.35) 0.596 3-4 vs 7+ 0.81 (0.58 - 1.13) 0.212 5-6 vs 7+ 0.69 (0.51 - 0.94) 0.020Evidence of allergy 0.97 (0.53 - 1.71) 0.914Therapy step No asthma medication vs Step 6 1.25 (0.64 - 2.49) 0.518 Step 1 vs Step 6 1.21 (0.61 - 2.44) 0.580 Step 2 vs Step 6 1.18 (0.69 - 2.13) 0.560 Step 3 vs Step 6 0.79 (0.46 - 1.44) 0.430 Step 4 vs Step 6 0.92 (0.54 - 1.66) 0.769 Step 5 vs Step 6 0.90 (0.54 - 1.58) 0.689 Unclassified vs Step 6 0.73 (0.33 - 1.59) 0.437Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) 1.18 (0.66 - 2.07) 0.563 Missing (no controller or reliever) vs high ratio (≥0.5) 1.36 (0.77 - 2.43) 0.289COPD 0.93 (0.71 - 1.22) 0.613Sinusitis 0.99 (0.77 - 1.27) 0.916Rhinitis 0.65 (0.40 - 1.09) 0.082Tonsillitis 0.58 (0.29 - 1.02) 0.083Acute upper respiratory infection 1.12 (0.88 - 1.42) 0.341Conjunctivitis 1.13 (0.85 - 1.49) 0.384Chronic otitis media 0.61 (0.26 - 1.24) 0.218Nasal polyposis 1.46 (1.05 - 2.00) 0.019Cough 0.95 (0.75 - 1.19) 0.627Diseases of the circulatory system 1.32 (1.00 - 1.76) 0.052
Table 4. Risk of Discontinuing Omalizumab: Adjusted Hazard Ratios and 95% Confidence Intervals
CI = confidence interval; HR = hazard ratio.
Coefficient SE P ValueIntercept 0.49 0.10 <0.001Age, y 12-17 vs 65+ 0.04 0.07 0.567 18-34 vs 65+ -0.04 0.06 0.520 35-44 vs 65+ 0.01 0.06 0.845 45-54 vs 65+ 0.04 0.06 0.483 55-64 vs 65+ 0.06 0.06 0.298Female vs male 0.03 0.02 0.268Region Midwest vs West 0.03 0.04 0.375 Northeast vs West -0.03 0.04 0.545 South vs West -0.03 0.03 0.290Usual asthma care physician specialty Allergist vs primary/other 0.04 0.03 0.178 Pulmonologist vs primary/other 0.04 0.03 0.193No. of chronic conditions 1-2 vs 7+ -0.03 0.04 0.522 3-4 vs 7+ 0.03 0.03 0.334 5-6 vs 7+ 0.07 0.03 0.028Evidence of allergy 0.06 0.06 0.351Therapy step No asthma medication vs Step 6 -0.02 0.07 0.818 Step 1 vs Step 6 -0.01 0.07 0.933 Step 2 vs Step 6 0.01 0.06 0.817 Step 3 vs Step 6 0.08 0.06 0.138 Step 4 vs Step 6 0.04 0.05 0.506 Step 5 vs Step 6 0.04 0.05 0.443 Unclassified vs Step 6 0.10 0.07 0.182Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) -0.03 0.06 0.602 Missing (no controller or reliever) vs high ratio (≥0.5)
-0.08 0.06 0.211
COPD 0.03 0.03 0.317Sinusitis 0.00 0.02 0.927Rhinitis 0.08 0.05 0.112Tonsillitis 0.07 0.05 0.178Acute upper respiratory infection 0.00 0.02 0.950Conjunctivitis -0.02 0.03 0.477Chronic otitis media 0.13 0.07 0.050Nasal polyposis -0.09 0.03 0.005Cough 0.01 0.02 0.684Diseases of the circulatory system -0.07 0.03 0.008
Table 2. Multivariate Analysis: Characteristics Associated With Omalizumab MPR
COPD = chronic obstructive pulmonary disease; MPR = medication possession ratio.
This study was supported by Genentech, Inc., South San Francisco, CA, and Novartis Pharmaceuticals Corporation, East Hanover, NJ, and third-party writing assistance for this poster was provided by Genentech, Inc., South San Francisco CA, and Novartis Pharmaceuticals, East Hanover, NJ.
Characteristics Associated With Medication Adherence Among New Omalizumab Users46
James Zazzali, PhD, MPH1; Michael S. Broder, MD, MSHS2; Eunice Chang, PhD2 1Genentech Inc., South San Francisco, CA; 2Partnership for Health Analytic Research, Beverly Hills, CA
ABSTRACTOBJECTIVE: Adherence to omalizumab (OMA) therapy has not been well studied. We sought to identify characteristics associated with adherence among new OMA users. METHODS: This was a retrospective cohort analysis using a HIPAA-compliant claims database. The study identified asthma patients who were ≥12 years old, newly treated with OMA between 7/1/2004 and 6/30/2007, and enrolled for 1 year before and 1 year after the first OMA claim. Adherence was measured by medication possession ratio (MPR) (total days of medication supplied ÷ 365) with each OMA claim considered a 28-day supply. Therapy step, as defined by the EPR3 guidelines, was assigned using a published algorithm. A base linear regression model was conducted with MPR as the dependent variable and demographics, physician specialty, respiratory comorbidities, asthma control, therapy step, and medication ratio forced into the model as independent variables. Other comorbidities were considered with forward selection and retained if significant at p<.05.RESULTS: We identified 766 new OMA users; mean age was 43 (SD 14) years, and 61% (n=466) were female. Forty-two percent of patients had allergists as their usual source of care, 20% pulmonologists, 28.3% primary care physicians, and the remainder other/unspecified. The mean number of chronic conditions was 5 (SD 2). Before starting OMA, most of patients were on EPR3 Step 5 therapy (28%), followed by Step 4 (18%), Step 3 (16%), Step 2 (15%), Step 1 (6%), and Step 6 (5%). Mean MPR in the year after treatment initiation was 0.68 (SD 0.30), and 55% persisted with therapy for at least 1 year. The final model included the base variables and diseases of circulatory system (the only additional significant predictor). Significant predictors of higher MPR were: care provided by allergists or pulmonologists, chronic otitis media, 5 or 6 vs ≥7 chronic conditions, and no disease of the circulatory system.CONCLUSIONS: We found MPR for omalizumab to be higher on average than what has been reported for combination corticosteroid/long-acting beta-agonist therapy. Care by an allergist or pulmonologist was associated with greater adherence. Confounding by disease severity is possible, but the relationship was significant after controlling for therapy step. If this finding is confirmed in other studies, it supports the value of specialist care for patients with difficult to treat asthma.
CONCLUSIONS• Among new omalizumab users, we estimated the MPR to be 0.68 with 54.7%
persisting on omalizumab at one year after initiating therapy.
• Omalizumab adherence and persistence are consistent with previously published reports, although this analysis indicates slightly higher estimates.
• The number of chronic conditions and select respiratory comorbidities were significant predictors of adherence and persistence in our multivariate models.
LIMITATIONS• Administrative claims data do not indicate the dosing schedule for omalizumab;
in this analysis, each dose of omalizumab was assumed to be a 28 day supply.
• Inclusion of concomitant asthma medications as covariates is indicative only of those medications filled and not reflective of patient adherence to therapy.
BACKGROUND• Lack of adherence to prescribed treatments for asthma is a well-known problem.
Rates of nonadherence range from 30% to 70%.1-3
• Poor asthma medication adherence is associated with decreases in asthma control and increases in emergency department visits, hospitalizations, and the need for oral corticosteroids.3,4
• Omalizumab, a humanized monoclonal antibody targeting immunoglobulin E, is approved in the United States for the treatment of adults and adolescents (≥12 years) with moderate to severe persistent allergic asthma that is inadequately controlled with inhaled corticosteroids.5
• Adherence to omalizumab therapy has not been well studied. One prior analysis of asthma patients newly treated with omalizumab estimated adherence rates to be 64.6% with 54% persisting up to one year.6
OBJECTIVE• To identify characteristics associated with adherence and persistence among new
omalizumab users
METHODSStudy Design• A retrospective cohort analysis using a HIPAA-compliant administrative claims
database of 10 million covered lives representing all major regions of the United States.
• Eligible patients were ≥12 years old, diagnosed with asthma, newly treated with omalizumab between 7/1/2004 and 6/30/2007, and followed for a year (Figures 1 and 2).
RESULTS• We identified 766 new omalizumab users; average age was 43.4 years and 61%
were female. Fifty percent of patients received asthma care from an allergist, 28% a pulmonologist, and 17% a primary care physician.
• Overall, the MPR was 0.68 (SD, ± 0.30). MPR stratified by baseline measures are presented in Table 1 and results of the multivariate analysis in Table 2.
• During follow-up, 45.3% of patients discontinued treatment. Mean (SD) time to discontinuation/end of study was 256.5 (131.6) days; median was 353 days (Table 3, Figure 3).
• The risk of discontinuing omalizumab was examined across a number of variables (Table 4).
REFERENCES1. Bender B et al. Ann Allergy Asthma Immunol. 2000;
85(5):416-421.2. Bender BG et al. Ann Allergy Asthma Immunol. 2007;
98(4):322-328.3. World Health Organization. Adherence to long-term
therapies: evidence for action. Geneva, Switzerland: World Health Organization; 2003.
4. Williams LK et al. J Allergy Clin Immunol. 2004;114(6):1288-1293.
5. Xolair [prescribing information]. South San Francisco, CA: Genentech USA, Inc.; 2010.
6. Broder MS et al. Allergy Asthma Proc. 2009;30(2):148-157.7. Deyo RA et al. J Clin Epidemiol. 1992;45(6):613-619.8. National Asthma Education and Prevention Program. Expert
Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma-Summary Report 2007. J Allergy Clin Immunol. 2007;120(5 suppl):S94-S138.
Index Date
First claim for omalizumab in ID period
1-year period (Preindex period)
7/1/03 6/30/08
6/30/077/1/04
Identification (ID) Period
1-year period (Postindex period)
Figure 1. Study Timeline
% o
f P
atie
nts
on
Tre
atm
ent
Days to Discontinue Omalizumab
100
75
50
25
0
0 50 100 150 200 250 300 350 400
Figure 3. Days to Discontinuation of Omalizumab.
Study Outcomes• Adherence to omalizumab therapy during the postindex period, as measured by the
medication possession ratio (MPR) – Calculated as the total days of medication supplied over the year, divided by 365 – Each omalizumab claim was considered a 28-day supply • Medication persistence during the postindex period, as measured by time
to discontinuation – Calculated as the number of days on omalizumab with no gaps of more than
45 daysStatistical Analysis• To evaluate the association of baseline measures (patient demographic and clinical
characteristics) with omalizumab adherence and persistence, we conducted a linear regression model with adherence (MPR) as the continuous dependent variable and all baseline measures as independent variables and a similar logistic regression model to predict persistence at one year.
– Based on bivariate analysis results and clinical relevance, demographics (age, sex, and region), asthma physician specialty, number of chronic conditions, evidence of allergy, respiratory-specific comorbidities, therapy step, and medication ratio were first included in the linear regression model; we then used forward selection to identify significant results for other baseline measures.
3456 patients with a claim of omalizumab in ID period
(7/1/2004–6/30/2007)
n=3391
n=3377
n=2886
n=766
65 patients with age <12 years on the index date (the date of �rst omalizumab claim in ID period)
14 patients had no asthma diagnosis in the preindex or postindex period
491 previously used omalizumab
2120 patients were not continuously enrolled in the preindex and postindex periods
Figure 2. Cohort Selection for Omalizumab New Users.
ID = identification.
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Demographics
Age, y (mean 43.4; SD 14.5) 0.157
12-17 68 (8.9) 0.72 (0.28)
18-34 111 (14.5) 0.62 (0.34)
35-44 183 (23.9) 0.66 (0.30)
45-54 234 (30.5) 0.68 (0.30)
55-64 141 (18.4) 0.71 (0.27)
65+ 29 (3.8) 0.65 (0.24)
Sex 0.174
Female 466 (60.8) 0.69 (0.30)
Male 300 (39.2) 0.66 (0.30)
Region 0.056
Midwest 184 (24.0) 0.71 (0.28)
Northeast 86 (11.2) 0.62 (0.32)
South 387 (50.5) 0.66 (0.30)
West 109 (14.2) 0.70 (0.27)
Physician Specialty
Usual asthma care physiciana 0.020
Pulmonologist 214 (27.9) 0.68 (0.28)
Primary care 128 (16.7) 0.64 (0.31)
Allergists 381 (49.7) 0.69 (0.29)
Other 22 (2.9) 0.66 (0.36)
Unknown 21 (2.7) 0.48 (0.39)
General Comorbidities
No. of chronic conditions (mean 4.7; SD 2.3) 0.047
1-2 155 (20.2) 0.63 (0.33)
3-4 227 (29.6) 0.69 (0.29)
5-6 216 (28.2) 0.71 (0.27)
7+ 168 (21.9) 0.65 (0.29)
Charlson Comorbidity Index7 (mean 1.8; SD 1.7) 0.967
1 530 (69.2) 0.68 (0.30)
2 86 (11.2) 0.68 (0.29)
3+ 150 (19.6) 0.67 (0.29)
Evidence of allergy <.001
No 57 (7.4) 0.55 (0.34)
Yes 709 (92.6) 0.69 (0.29)
Respiratory-Specific Comorbidities
COPD 0.338
No 577 (75.3) 0.67 (0.31)
Yes 189 (24.7) 0.69 (0.27)
Sinusitis 0.551
No 243 (31.7) 0.67 (0.31)
Yes 523 (68.3) 0.68 (0.29)
Rhinitis <0.001
No 93 (12.1) 0.57 (0.33)
Yes 673 (87.9) 0.69 (0.29)
Tonsillitis 0.321
No 731 (95.4) 0.67 (0.30)
Yes 35 (4.6) 0.72 (0.28)
Acute upper respiratory infection 0.471
No 515 (67.2) 0.67 (0.30)
Yes 251 (32.8) 0.69 (0.28)
Conjunctivitis 0.894
No 619 (80.8) 0.68 (0.30)
Yes 147 (19.2) 0.67 (0.29)
Chronic otitis media 0.002
No 746 (97.4) 0.67 (0.30)
Yes 20 (2.6) 0.81 (0.17)
Nasal polyposis 0.073
No 663 (86.6) 0.68 (0.29)
Yes 103 (13.4) 0.63 (0.32)
Cough 0.427
No 381 (49.7) 0.67 (0.31)
Yes 385 (50.3) 0.68 (0.28)
N (%)MPR
Mean (SD) P Value
All 766 (100.0) 0.68 (0.30)
Asthma Control in the Preindex Period
Poor asthma controlb 0.096
No 191 (24.9) 0.64 (0.32)
Yes 575 (75.1) 0.69 (0.29)
Asthma-related inpatient hospitalization 0.340
No 657 (85.8) 0.68 (0.30)
Yes 109 (14.2) 0.65 (0.30)
Asthma-related ED visit 0.412
No 705 (92.0) 0.67 (0.30)
Yes 61 (8.0) 0.71 (0.29)
Two or more oral corticosteroid prescriptions filled 0.031
No 265 (34.6) 0.64 (0.32)
Yes 501 (65.4) 0.69 (0.28)
Six or more short-acting beta-agonist prescriptions filled 0.310
No 434 (56.7) 0.67 (0.30)
Yes 332 (43.3) 0.69 (0.29)
EPR-3 Guidelines8 Therapy Step in Preindex Period
Therapy step prior to initiating omalizumab 0.006
No asthma meds 68 (8.9) 0.56 (0.33)
Step 1 44 (5.7) 0.61 (0.32)
Step 2 112 (14.6) 0.67 (0.30)
Step 3 121 (15.8) 0.73 (0.27)
Step 4 136 (17.8) 0.69 (0.28)
Step 5 214 (27.9) 0.69 (0.29)
Step 6 37 (4.8) 0.63 (0.33)
Unclassified 34 (4.4) 0.72 (0.28)
Medication Ratioc and Other Asthma Medication Use in the Preindex Period
Medication ratio (mean 0.64; SD 0.28 among 718 patients with medication ratios) 0.003
Low ratio (<0.5) 197 (25.7) 0.64 (0.31)
High ratio (≥0.5) 521 (68.0) 0.70 (0.28)
Missing (no controller or reliever)
48 (6.3) 0.55 (0.35)
Short-acting beta-agonist inhalers 0.044
No 115 (15.0) 0.62 (0.32)
Yes 651 (85.0) 0.68 (0.29)
Anticholinergics 0.077
No 521 (68.0) 0.66 (0.30)
Yes 245 (32.0) 0.70 (0.29)
Oral corticosteroids 0.119
No 153 (20.0) 0.64 (0.32)
Yes 613 (80.0) 0.68 (0.29)
ICS/LABA combination 0.051
No 240 (31.3) 0.64 (0.31)
Yes 526 (68.7) 0.69 (0.29)
ICS alone 0.286
No 450 (58.7) 0.67 (0.31)
Yes 316 (41.3) 0.69 (0.28)
LABA alone 0.790
No 643 (83.9) 0.67 (0.30)
Yes 123 (16.1) 0.68 (0.30)
Mast cell stabilizers 0.108
No 734 (95.8) 0.67 (0.30)
Yes 32 (4.2) 0.76 (0.25)
Methylxanthines 0.810
No 670 (87.5) 0.67 (0.30)
Yes 96 (12.5) 0.68 (0.26)
Leukotriene receptor antagonists 0.078
No 221 (28.9) 0.65 (0.31)
Yes 545 (71.1) 0.69 (0.29)
Table 1. MPR of Omalizumab Stratified by Patient Demographics and Clinical Characteristics
aThe usual asthma care physician was determined by the largest plurality of office visits with evaluation and management services and asthma diagnosis. bAsthma-related inpatient hospitalization, asthma-related ED visit, ≥2 oral corticosteroid prescriptions filled, or ≥6 short-acting beta-agonist prescriptions filled.cMedication ratio is units of asthma controllers to units of controllers + units of relievers.COPD = chronic obstructive pulmonary disease; ED = emergency department; EPR-3 = National Asthma Education and Prevention Program Expert Panel Report 3; ICS = inhaled corticosteroid; LABA = long-acting beta-agonist; MPR = medication possession ratio.
Omalizumab New Users(N=766)
MPR (Range 0-1) Mean (SD) 0.68 (0.30)[Median] [0.80]
Discontinued omalizumab n (%) 347 (45.3)Days to discontinuation/end of study Mean (SD) 256.5 (131.6)
[Median] [353]Number of omalizumab claims Mean (SD) 10.1 (5.4)
[Median] [11]1 n (%) 64 (8.4)2-5 n (%) 94 (12.3)6-11 n (%) 249 (32.5)12+ n (%) 359 (46.9)
Table 3. Omalizumab Use in the Postindex Period
HR (95% CI) P ValueAge, y 12-17 vs 65+ 0.82 (0.41 - 1.67) 0.576 18-34 vs 65+ 1.15 (0.64 - 2.16) 0.655 35-44 vs 65+ 0.86 (0.50 - 1.57) 0.600 45-54 vs 65+ 0.81 (0.48 - 1.46) 0.457 55-64 vs 65+ 0.76 (0.44 - 1.38) 0.344Female vs male 0.83 (0.66 - 1.05) 0.121Region Midwest vs West 0.76 (0.52 - 1.12) 0.164 Northeast vs West 0.98 (0.63 - 1.51) 0.917 South vs West 1.16 (0.84 - 1.62) 0.380Usual asthma care physician specialty Allergist vs primary/other 0.85 (0.64 - 1.13) 0.264 Pulmonologist vs primary/other 0.84 (0.62 - 1.14) 0.273No. of chronic conditions 1-2 vs 7+ 0.89 (0.59 - 1.35) 0.596 3-4 vs 7+ 0.81 (0.58 - 1.13) 0.212 5-6 vs 7+ 0.69 (0.51 - 0.94) 0.020Evidence of allergy 0.97 (0.53 - 1.71) 0.914Therapy step No asthma medication vs Step 6 1.25 (0.64 - 2.49) 0.518 Step 1 vs Step 6 1.21 (0.61 - 2.44) 0.580 Step 2 vs Step 6 1.18 (0.69 - 2.13) 0.560 Step 3 vs Step 6 0.79 (0.46 - 1.44) 0.430 Step 4 vs Step 6 0.92 (0.54 - 1.66) 0.769 Step 5 vs Step 6 0.90 (0.54 - 1.58) 0.689 Unclassified vs Step 6 0.73 (0.33 - 1.59) 0.437Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) 1.18 (0.66 - 2.07) 0.563 Missing (no controller or reliever) vs high ratio (≥0.5) 1.36 (0.77 - 2.43) 0.289COPD 0.93 (0.71 - 1.22) 0.613Sinusitis 0.99 (0.77 - 1.27) 0.916Rhinitis 0.65 (0.40 - 1.09) 0.082Tonsillitis 0.58 (0.29 - 1.02) 0.083Acute upper respiratory infection 1.12 (0.88 - 1.42) 0.341Conjunctivitis 1.13 (0.85 - 1.49) 0.384Chronic otitis media 0.61 (0.26 - 1.24) 0.218Nasal polyposis 1.46 (1.05 - 2.00) 0.019Cough 0.95 (0.75 - 1.19) 0.627Diseases of the circulatory system 1.32 (1.00 - 1.76) 0.052
Table 4. Risk of Discontinuing Omalizumab: Adjusted Hazard Ratios and 95% Confidence Intervals
CI = confidence interval; HR = hazard ratio.
Coefficient SE P ValueIntercept 0.49 0.10 <0.001Age, y 12-17 vs 65+ 0.04 0.07 0.567 18-34 vs 65+ -0.04 0.06 0.520 35-44 vs 65+ 0.01 0.06 0.845 45-54 vs 65+ 0.04 0.06 0.483 55-64 vs 65+ 0.06 0.06 0.298Female vs male 0.03 0.02 0.268Region Midwest vs West 0.03 0.04 0.375 Northeast vs West -0.03 0.04 0.545 South vs West -0.03 0.03 0.290Usual asthma care physician specialty Allergist vs primary/other 0.04 0.03 0.178 Pulmonologist vs primary/other 0.04 0.03 0.193No. of chronic conditions 1-2 vs 7+ -0.03 0.04 0.522 3-4 vs 7+ 0.03 0.03 0.334 5-6 vs 7+ 0.07 0.03 0.028Evidence of allergy 0.06 0.06 0.351Therapy step No asthma medication vs Step 6 -0.02 0.07 0.818 Step 1 vs Step 6 -0.01 0.07 0.933 Step 2 vs Step 6 0.01 0.06 0.817 Step 3 vs Step 6 0.08 0.06 0.138 Step 4 vs Step 6 0.04 0.05 0.506 Step 5 vs Step 6 0.04 0.05 0.443 Unclassified vs Step 6 0.10 0.07 0.182Medication ratio Low ratio (<0.5) vs high ratio (≥0.5) -0.03 0.06 0.602 Missing (no controller or reliever) vs high ratio (≥0.5)
-0.08 0.06 0.211
COPD 0.03 0.03 0.317Sinusitis 0.00 0.02 0.927Rhinitis 0.08 0.05 0.112Tonsillitis 0.07 0.05 0.178Acute upper respiratory infection 0.00 0.02 0.950Conjunctivitis -0.02 0.03 0.477Chronic otitis media 0.13 0.07 0.050Nasal polyposis -0.09 0.03 0.005Cough 0.01 0.02 0.684Diseases of the circulatory system -0.07 0.03 0.008
Table 2. Multivariate Analysis: Characteristics Associated With Omalizumab MPR
COPD = chronic obstructive pulmonary disease; MPR = medication possession ratio.
This study was supported by Genentech, Inc., South San Francisco, CA, and Novartis Pharmaceuticals Corporation, East Hanover, NJ, and third-party writing assistance for this poster was provided by Genentech, Inc., South San Francisco CA, and Novartis Pharmaceuticals, East Hanover, NJ.