Transcript Odd Ratio - Personal

APPROPRIATENESS AND VARIATION IN DRUG UTILIZATION
ACROSS PATIENTS WITH DEPRESSION
Stacey R. Long, MS1; Rebecca L. Robinson, MS2; Stella Chang, MPH1; Stephen Able, PhD2; Onur Baser, PhD1; Amelito Torres, MBA1; Ralph W. Swindle, PhD1
1
The MEDSTAT Group; 2 US Medical Division, Outcomes Research, Eli Lilly and Company
ABSTRACT
Methodology
Introduction. Retrospective claims were assessed to determine factors associated
with meeting National Committee for Quality Assurance (NCQA) guidelines for
Antidepressant Medication Management and the impact of NCQA compliance on
expenditures.
• Study Design
– Retrospective cohort study using patient-level administrative claims data
Results. 29% of patients had optimal provider visits during the 12-week acute
treatment phase. These patients were more likely to initiate therapy on SNRIs and
“other antidepressants” (all p<.001). Patients with capitated insurance coverage,
initiating on TCAs, not seen by mental health specialists or residing in the South
were INTRODUCTION
less likely to be compliant (all p<.001). 68% and 49% met minimum NCQA
guidelines
during
the
acute
6-month
continuation
treatment
Toj fise rnvisd
viesnv
dsivhs
ifvnamand
snfeiuhbv
fd nvjn vieyf
dkjfsdfdkf dfdks fdjf
efns dvnuefuphases,
webkdfvn; xioh k djfi sernv isd
vies nvds iv hsifvna
msnfeiuhbv
vjn viey fe fnin
sdv
nuef uw
ebkdfvn;comorbid
xioh k djf iser nchronic
visd vie snvd
sivhsif vn amsnfeiuhbv
respectively.
Males,
thosefdnyounger
age,
having
conditions,
not
fdn
vjnv
ie
yfe
fnsd
vnue
fu
web
kdfvn;xioh
kdjf
ise
rnvi
s
dvi
esn
vdsi
vhs
ifvna
ies
nvd
sivhs
ifvn
amsnfeiu
hbv
fd
nvjnvi
seen by non-mental health specialists or residing in the South were less likely to beeyf
efn sdv n ue fuw ebkdfvn;xioh kdjfisernv is dvie snvd si vh sifv nam snf eiuhbv fd nvj nvie yfefn sdvnuef uwe bk.Toj fise
compliant
in either
treatment
(alldkjfsdfdkf
p<.001).
Compared
to SSRIs,
rnvisd viesnv
dsivhs ifvnam
snfeiuhbvphase
fd nvjn vieyf
dfdks fdjf
efns dvnuefu webkdfvn;
xioh k users
djfi sernv of
isd vies
“othernvds
antidepressants”
were
less
likely
be compliant
and
morefdn
iv hsifvna msnfeiuhbv fdn
vjn viey
fe fn sdv
nuef to
uw ebkdfvn;
xioh k djf iser
n visdSNRI
vie snvdusers
sivhsif vnwere
amsnfeiuhbv
vjnv
ie
yfe
fnsd
vnue
fu
web
kdfvn;xioh
kdjf
ise
rnvi
s
dvi
esn
vdsi
vhs
ifvna
ies
nvd
sivhs
ifvn
amsnfeiu
hbv
fd
nvjnvi
eyf efn
likely to be compliant in the acute and continuation phases; initiators on TCAs were
sdv n ue fuw ebkdfvn;xioh kdjfisernv is dvie snvd si vh sifv nam snf eiuhbv fd nvj nvie yfefn sdvnuef uwe bk.Toj fise rnvisd
moreviesnv
likely
to be compliant in the continuation phase (all p<.001). NCQA compliant
dsivhs ifvnam snfeiuhbv fd nvjn vieyf dkjfsdfdkf dfdks fdjf efns dvnuefu webkdfvn; xioh k djfi sernv isd vies nvds iv
patients
incurred
additional
expenditures
persiyear
tonvjnonhsifvna
msnfeiuhbvan
fdn vjn
sdv n ue fuw$1430
ebkdfvn;xioh
kdjfisernv is dvie snvd
vh sifv compared
nam snf eiuhbv fd
nvie yfefn
compliant
sdvnuefpatients.
uwe bk.sifv nam snf eiuhbv fd nvj nvie yfef
Conclusions. Compliance with NCQA guidelines was less than optimal and
associated with type of initiating drug, comorbidities, gender, age, and geographic
region. Improved management of these patients could result in reduced illness
burden.
INTRODUCTION
• The importance of the appropriate use of antidepressant (AD) is growing1-4 as
costs associated with mental disorders substantially increased over the past
decade.5
• Around 16% of AD costs are associated with patients that were never adequately
treated.3 Many patients (35% to 51%) receive treatment of inadequate duration
or dosage level.3-4
• The majority of depressed individuals respond to effective AD treatment.6
However, approximately 29% to 46% of depressed patients show only partial or
no response to ADs.2 Up to 70% of patients may respond to ADs but fail to
achieve complete remission.7
• Guidelines on the adequate care in AD management have been developed by
the National Committee for Quality Assurance (NCQA). Standardized
performance measures (HEDIS; Health Plan Employer Data and Information Set)
published annually by NCQA are widely used by decision makers in managed
health care plans.8
Demographic Characteristics by Initiating Treatment Class
All Initiators TCA
SSRI
(N=48,098) (n=1,766) (n=32,974)
%
Percent of All Patients
%
%
%
SNRI
(n=4,819)
Other ADs
(n=8,539)
%
%
100.00
3.7
68. 6
10.0
17.8
71.4
71.4
72.8
73.3
64.5
43.9 (11.01) 48.3 (9.43) 43.5 (11.22) 44.4 (10.59) 44.4 (10.50)
Capitated Insurance, %
46.2
43.9
46.0
47.0
47.1
Geographic Region, %
• To determine the percentage of depressed patients who receive effective
treatment for depression in the 6 months following initiation of antidepressant
pharmacotherapy, as measured by the NCQA Guidelines for Antidepressant
Medication Management
• To assess factors associated with compliance with the NCQA guidelines
West
Northeast
21.1
23.8
22.0
15.9
20.3
North Central
30.1
25.8
30.3
29.3
30.8
South
42.9
44.9
41.9
49.5
42.6
Mean (SD) CCI Scorea
Mental Health
Specialty Careb %
a
b
5.8
5.4
0.53 (1.21) 0.77 (1.46)
42.1
52.7
5.8
0.52 (1.20)
39.5
5.3
6.3
0.55 (1.18) 0.53 (1.20)
44.7
48.4
CCI=Charlson Comorbidity Index
Care by Mental Health Specialist included any billed visit/encounter with a psychiatrist,
mental health & chemical dependency treatment facility, psychologist, or psychiatric nurses
Poster presented at 10th Annual ISPOR Meeting
Mental Health Expenditures
$1,600
67.9%
70%
$1,400
60%
$1,200
49.4%
50%
$1,000
40%
$800
28.9%
30%
$600
15.0%
20%
$400
$200
10%
$0
0%
Practitioner
contacts
Acute treatment
Continuing
treatment
Optimal
Contacts
All 3 guidelines
Acute T reatment
Continuous
T reatment
All 3 Measures
Predicted Annual Overall Expenditures= $7,042;
Predicted Annual Mental Health Expenditures= $1,617
Figure 1. Odds Ratios of Initiating Treatment Class versus SSRIs for
3 Components of NCQA Guidelines
TCA
2
SNRI
Other AD
*
1
*
*
*
*
*
*
Practitioner
contacts
0
*
*
*
AcuteTreatment
Continuous
Treatment
All 3 Measures
*Significantly different from SSRI (p<.001) after controlling for covariates in logistic regression
*Individual and Overall guideline compliance significantly associated with expenditures (p<.001).
Limitations
•The results are based on retrospective analyses of claims data, although reliance on
these data has certain limitations.
– Some patients with claims meeting the study inclusion criteria might not in fact be
patients with depression who are newly initiating on antidepressants.
– Patients may receive treatment that is not submitted to their health plan for
reimbursement, and thus not included in the claims data.
– Claims data do not include clinical indicators that would allow us to assess the
appropriateness of care on an individual basis.
– Claims data lack some important mediating variables that may predict outcomes
such as race and socioeconomic status.
•The study population is comprised of privately insured patients with depression. Results
may not be generalizable to uninsured patients or those with publicly funded insurance.
CONCLUSIONS
Figure 2. Odds Ratios of Demographic & Clinical Characteristics for
3 Components of NCQA Guidelines
5
Male
MH Specialty Care
Bipolar
CCI Score>0
Capitated Health Plan
Age (5-year)
Anxiety
*
4
Mean (SD) Age
Funding provided by Eli Lilly and Company
Overall Expenditures
80%
• Outcome Measures
Compliance with the NCQA Guidelines for Antidepressant Medication
Management
• Optimal provider contacts
– 1+ contact with prescribing practitioner in 84 days post index
– 3+ total contacts with healthcare providers in 84 days post index
• Effective acute phase treatment
– Remained on antidepressant therapy during 84 days post index as
measured by days supply listed on prescription claims
• Effective continuation phase treatment
– Remained on antidepressant therapy during 180 days post index as
measured by days supply listed on prescription claims
• Statistical Analyses
Logistic regression modeling to determine factors associated with compliance
with the NCQA guidelines
• Covariates: age, gender, geographic region, insurance plan type, Charlson
Comorbidity Index score, initially prescribed antidepressant, and if the patient
received care by a mental health specialist
Figure 3. Regression-Adjusted Marginal Effects of NCQA Measures
of Compliance* on Annual Overall and
Mental Health Expenditures
90%
% meeting criteria
• Data Source
– January 2001 - March 2003 MarketScan® Commercial Claims and Encounter
Database
• A health care claims database containing the inpatient, outpatient and
outpatient prescription drug experience of several million employees and their
dependents (annually), covered under a variety of fee-for-service and capitated
health plans
• Inclusion Criteria
– Diagnosis of depression (ICD-9-CM diagnosis codes: 296.2x, 296.3x, 300.4, 309.0,
309.1, or 311.x)
– Outpatient prescription drug claim for a tricyclic antidepressant (TCA), selective
serotonin reuptake inhibitor (SSRI), venlafaxine (SNRI: serotonin norepinephrine
reuptake inhibitor) “other antidepressants” (bupropion, nefazodone, or mirtazapine)
• The date of the first such drug claim was set as the index date.
– Continuous enrollment for 6 months before and 12 months after the index date
– Age 18 to 64 at index date
 Exclusion Criteria
– Diagnosis of schizophrenia, bipolar disorder, or psychosis (ICD-9-CM diagnosis
codes: 295.xx, 296.4x - 296.7x, 296.89, 296.9x) in the 6 months preceding the index
date
– Antidepressant drug claim(s) in the 6 months preceding the index date
Female,
Study Objectives
100%
Odds Ratios
Methods. Using the MarketScan® Commercial Claims and Encounter database,
adult patients with depression and initiating on tricyclic antidepressants (TCAs),
selective serotonin reuptake inhibitors (SSRIs), venlafaxine (SNRI), and “other
antidepressants” (bupropion, nefazodone, or mirtazapine) were followed for 12
months (between 2001-Q12003). Claimants with diagnoses of schizophrenia,
bipolar, psychoses, or antidepressant use in the prior 6 months were excluded.
Factors associated with compliance (demographics, comorbidities, initiating
antidepressants) were assessed in the sample (n=48,098) using logistic regression.
Exponential conditional mean (ECM) models were used to determine the marginal
effect of compliance on overall and mental health expenditures.
Percentage of Patients Meeting 3 Components of NCQA Guidelines
3
*
*
2
*
1
*
*
Practitioner contacts
0
*
*
*
*
*
AcuteTreatment
*
*
*
*
*
*
Continuous
Treatment
*Significantly associated (p<.05) with guidelines in logistic regression
Note: Age variable in 5-year increments, all other variables are binary. Region was a significant variable in all models as
well, with individuals residing in the South less likely to be compliant on all measures.
Charlson Comorbidity Index (CCI) score >0 indicates at least one of the following acute events or chronic diseases:
Myocardial infarction, Congestive heart failure, Peripheral vascular disease, Dementia, Cerebrovascular disease, Chronic
pulmonary disease, Connective tissue disease, Ulcer, Liver disease, Hemiplegia, Diabetes Mellitus, Moderate or severe
renal disease, Any tumor, Leukemia, Lymphoma.
• Many patients received treatment for depression that did not conform to the
minimum guidelines established by NCQA.
• Various patient, treatment, and provider factors influenced compliance with the
guidelines. Knowledge of these factors could improve depression management by
identifying patients most at risk for not receiving effective treatment.
• SNRI initiators were more likely than SSRI initiators to receive effective
treatment in the acute and continuous phases, while those initiating on "other
antidepressants were less likely (p<.001). TCA and SSRI initiators did not
differ with respect to odds of meeting acute phase treatment guidelines, while
TCA users were more likely than SSRI users to meet continuous phase
guidelines (p<.001).
• Patients treated by a mental health care specialist were more likely to meet all
three NCQA treatment guidelines (p<.001).
• In the initial year following initiation of therapy, guideline compliance was associated
with $1430 in additional overall expenditures, 66% of which were coded as directly
related to treating depression (both p<.001).
• Further research on the incremental impact of guideline compliance on subsequent
year expenditures is warranted.
REFERENCES
1. Fava, M., Management of nonresponse and intolerance: switching strategies. Journal of Clinical
Psychiatry, 2000. 2: p. 10-2.
2. Loosbrock, D.L., et al., Appropriateness of prescribing practices for serotonergic antidepressants.
Psychiatric Services, 2002. 53(2): p. 179-84.
3. Weilburg, J.B., et al., Costs of antidepressant medications associated with inadequate treatment.
American Journal of Managed Care, 2004. 10(6): p. 357-65.
4. Shasha, M., et al., Serotonin reuptake inhibitors and the adequacy of antidepressant treatment.
International Journal of Psychiatry in Medicine, 1997. 27(2): p. 83-92.
5. Thorpe K.E., et al. Which medical conditions account for the rise in health care spending? Health
Affairs, 2004 W4: p. 437-445.
6. National Institute of Mental Health. Depression: a treatable illness. 2003. Bethesda, MD: National
Institute of Mental Health. NIH Publication No. 03-5299. March 2003.
7. O’Reardon J.P., Amsterdam JD. Treatment-Resistant Depression: Progress and Limitations. Psychiatry
Annals,1998; 11: p. 633-640.
8. National Committee for Quality Assurance (NCQA). HEDIS 2004. Health plan employer data &
information set. Vol. 2, Technical specifications. Washington, DC: National Committee for Quality
Assurance. 2003.