Three Essays on Physician Prescribing Behavior

Download Report

Transcript Three Essays on Physician Prescribing Behavior

Three Essays on
Physician Prescribing
Behavior
Brian K. Chen
Orals Examination
December 4, 2006
Haas School of Business
University of California at Berkeley
Motivation
Prescription drug expenditures: the fastest growing component of health care
expenditures
 1990s:
Prescription drug expenditures grew by 5% to 23%
annually in most industrialized countries



United States: Fastest growing component of $1.9 trillion health care
industry
In 2004: Third largest component in the US health care expenditures
at 9%, following hospital (31%) and physician (22%) services
In 2002: Drug expenditures up by 15.3%, outstripping expenditures
in hospital (9.5%) and physician (7.7%) services
 By
2001: US$607 billion spent on prescription drugs
worldwide

In nominal terms, top 20 GDP in the world
 New
drugs explain up to 40% of annual drug expenditures
growth
Dissertation Outline and Research Questions

Chapter 1: What are the determinants in the
adoption decision of new drugs?
 Do
physician, patient, and hospital characteristics matter in
the likelihood/rate of adoption of new drugs?

Chapter 2: What is the health outcome impact of
new drugs?
 Do

new drugs lead to better health outcomes?
Chapter 3: If high costs to patient affect drug use, do
physicians take patient costs into consideration?
Why are these questions important?

Numerous implications
 Who gets new drugs? Who prescribes new drugs?
 Theoretical interest – consistent early adopters?
 Policy interest
 As first stage analysis for second essay
 Are new drugs worth their cost?
 If yes, what are the cost savings? How to encourage appropriate use
of new drugs
 If no, what are the additional costs compared to older, just-aseffective drugs?
 If
financial burden prevents access to new drugs, do
physicians take this into consideration?


If only marginal improvement, physicians should prescribe older
drugs to the financially burdened
If substantial improvement, implications for drug copayment policies
Contribution

Chapter 1:
 Very
little known about physician adoption of new drugs
 But I need: theoretical framework?

Chapter 2:
 No
strong empirical evidence on the effectiveness of new
drugs versus older drugs that corrects for selection bias
 But I need: INSTRUMENT for treatment
Background

Quick statistics
 Land
Area: 13,823 square miles
 Population (2006): 23,000,000
 2005 GDP: $U.S. 611.5 billion ($U.S. 326.5 billion)
 2005 Per Capita GDP: $U.S. 26,700 ($14,200)

Health Care in Taiwan:
 2003:
$U.S. 11 billion
 National Health Insurance, virtually 100% coverage
 5.7 hospital beds per 1,000 people, 1.4 physicians
trained in Western medicine for every 1,000
Salient Features of Taiwan’s Health Care System

Closed System
 Physicians
are employees
Freedom of Choice
 Lack of system of referrals
 Commingling of diagnostic and dispensing
services

◄Chapter 1►
Adoption/Diffusion of New Therapeutic Agents
Literature Review: Adoption of New Drugs

“Epidemic” studies
 Menzel
(1955), Coleman (1957), Peay (1988), Denig
(1991) Nair (2006) (related)

“Firm Heterogeneity” studies
 Steffesen

(1999), Tamblyn (2003), Dybahl (2005)
Bayesian model of adoption
 Coscelli
(2004)
Motivation to Prescribe

Firm heterogeneity model: there exist
characteristics that predict adoption of new
technology
 What
these characteristics are remains an open
empirical question
 Are these characteristics constant across new drugs?
Patient Demand
 Marketing Activities

Conceptual Framework: Predictions

Physician characteristics:






Patient characteristics





Prime age  greater adoption
Gender  unclear, general view is male  greater adoption
Past practice volume  greater adoption?
Type of doctor  family practice  less adoption
*Past use of drugs manufactured by same company  greater adoption
Age  unclear; depends on drug
Gender  unclear
*Higher Education  greater adoption
Condition severity  unclear
Hospital characteristics
Academic  greater adoption
Urban  greater adoption
 Family practice  less adoption at hospitals



*Drug characteristics

New drug-action mechanism?  slower adoption for agents with new mechanism
Top Prescription Drugs in Taiwan
by Sales, 2004
Top Drugs by Sales in Taiwan, 2004
Rank
Drug Name
Indication
Claims ($New Taiwan Dollar)*
1
AMLODIPINE (Norvasc)
Hypertension/Angina
2,717,461,581
2
VALSARTAN (Diovan)
Hypertension
1,438,903,860
3
ATORVASTATIN (Lipitor)
Cholesterol
1,279,879,641
4
FELODIPINE (Plendil/Lexxel)
Hypertension
1,174,234,843
5
ROSIGLITAZONE (Avandia)
Diabetes (Type 2)
1,027,488,496
6
CLOPIDOGREL (Plavix)
Heart Failure/Stroke
1,011,987,948
7
LOSARTAN (Cozaar)
Hypertension
1,003,209,815
8
GLICLAZIDE
Diabetes (Type 2)
1,001,107,515
9
METFORMIN
Diabetes (Type 2)
983,970,085
10
NIFEDIPINE (Adalat/Procardia)
Hypertension
973,298,312
11
FACTOR VIII
Anemia (Test)
891,448,733
12
ENALAPRIL (Vaseretic/Vasotec)
Hypertension/Heart Failure
852,860,619
13
ZOLPIDEM (Ambien)
Insomnia
848,801,879
14
CEFAZOLIN (Ancef)
Infection (Bacterial)
791,589,797
15
CIPROFLOXACIN (Ciloxan)
Infection (Bacterial, of the eye)
774,191,733
16
ALBUMIN
Liver/Kidney Disease (Test)
734,444,611
17
GLIMEPIRIDE (Amaryl)
Diabetes (Type 2)
713,329,867
18
IRBESARTAN (Avapro)
Hypertension
707,404,461
19
CELECOXIB (Celebrex)
Arthritis
674,028,771
20
SIMVASTATIN (Zocor)
Cholesterol
667,612,759
21
CARVEDILOL (Coreg)
Heart Failure/Hypertension
665,092,124
22
RISPERIDONE (Risperdal)
Schizophrenia
641,028,107
23
LOVASTATIN (Advicor)
Cholesterol
622,792,048
24
ATENOLOL
Hypertension
600,122,674
25
DOXAZOSIN (Cardura)
Enlarged Prostate
590,823,465
Source: National Health Insurance Bureau, Taiwan
*1 U.S. Dollar = 33 New Taiwan Dollars in 2004
Growth Rate
13.7%
21.3%
44.5%
20.1%
29.9%
62.1%
11.0%
9.2%
11.6%
2.4%
6.9%
8.0%
34.7%
-5.3%
10.6%
7.5%
36.0%
54.5%
36.1%
13.2%
18.5%
21.8%
46.5%
5.4%
19.8%
Top ICD-9-CM codes in Taiwan
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
ICD9CM/ACODE
A312
4659
460
4660
A311
4019
463
4619
A320
462
A233
A429
A322
A346
A420
4650
5589
A314
A310
6929
25000
7890
4871
A323
A239
7061
7804
4658
A269
4640
Disease Name
Acute sinitis
Upper respiratory infection
Common cold
Acute bronchitis
Hypertension, essential
Acute tonsillitis
Acute sinitis
Acute bronchitis
Pharyngitis
Acute conjuntivitis
Atopic dermatitis
Influenza
Constipation
Cellulitis and abscess
Frequent colds
Acute gastroenteritis
Chronic rhinitis
Acute tonsillitis
Contact dermatitis
Diabetes mellitus
Abdominal pain
Influenza
Asthma
Diabetes mellitus with retinopathy
Acne vulgaris
Dizziness and giddiness
Frequent colds
Hypertension, essential
Acute laryngitis
Frequency
1348973
461.9
1291130
335047
293805
235797
213079
200034
199661
190653
466
157934
148104
372
141388
706.1
140634
487.1
139419
564
137914
682.9
125392 460 or 465, 465.0, 4
122106
120790
472
119881
463
119615
111389
93431
89837
81885
80829 250.50+362.01
79410
78436
78127 460 or 465, 465.0, 4
76532
401.9
75676
Drugs introduced between 1997-2004

Atorvastatin (Lipitor) (19114/2097)





Rosiglitazone (Avandia) (15281/1052)




Date of introduction: November 1, 2000
Therapeutic class: statins
Indication: to lower cholesterol and thereby reduce cardiovascular disease.
With 2005 sales of US$12.2 billion under the brand name Lipitor, it is the largest
selling drug in the world
Date of introduction: March 1, 2001
Therapeutic Class: thiazolidinedione
Indication: Anti-diabetic drug (Diabetes Type II)
Clopidogrel (Plavix) (7378/728)
Date of introduction: January 1, 2001
Therapeutic Class: Antiplatelet agent
 Indication: is a potent oral antiplatelet agent often used in the treatment of
coronary artery disease, peripheral vascular disease, and cerebrovascular disease.
 In 2005 it was the world's second highest selling pharmaceutical with sales of
US$5.9 billion


Other new drugs
 Celecoxib (Celebrex)
 Arthritis/Pain (April 1, 2001) (but: side effects) (15574/3952)
 Esomeprazole (Nexium)
 Heartburn/Acid Reflux (January 1, 2002) (4250)
 Olanzapine (Zyprexa)
 Schizophrenia/Bipolar (February 1, 1999) (5284)
 Venlafaxine (Effexor)
 Antidepressant ( October 1, 2000) (2296)
 Montelukast (Singulair)
 Asthma (July 1, 2001) (2489)
 Quetiapine (Seroquel)
 Schizophrenia/Bipolar (April 1, 2000) (2795)
Disease Code Combinations
only < 1% of visits have no ICD9 code
Hypertension
1
0
0
0
1
1
1
0
0
0
1
1
1
0
HHD*
0
0
1
0
0
0
1
0
1
1
0
1
1
1
Cholesterol
0
1
0
0
0
1
0
1
0
1
1
0
1
1
Diabetes
0
0
0
1
1
0
0
1
1
0
1
1
0
1
Lipitor Takers** Population***
1719
26956
1400
10316
1046
12739
1661
44792
781
5614
829
3756
58
720
837
4098
401
2554
482
2004
402
1499
19
70
15
46
191
645
Percentage
0.063770589
0.135711516
0.082110056
0.037082515
0.139116494
0.220713525
0.080555556
0.204245974
0.157008614
0.240518962
0.268178786
0.271428571
0.326086957
0.296124031
*HHD = Hypertensive Heart Disease
**A patient is considered a Lipitor-taker if he or she has been prescribed Lipitor at least once
***"Population" is the number of unique individuals with at least one visit with the relevant disease code
combinations
These figures are drawn from the master data file. Because an individual may have different diagnosis
code combinations from visit to visit, the 14 combinations are not mutually exclusive
Description of Data

Panel Data
 Eight
years of complete medical claims data for a random
selection of 200,000 individuals from Taiwan’s population
of 23 million
 HOSB, PER, DOC and ID files
 The age, gender, and expenditures of the randomly
selected individuals do not differ significantly from the
population

Time Series (Random Subsamples)
 Outpatient Expenditures
 Inpatient
Expenditures
 Prescription Drugs at Contracted Pharmacies (complete)
Summary Statistics - Hypertension
Variable
FEE_YM
HOSP_ID
ID
ID_BMDY
ID_SEX
FUNC_MDY
FUNC_TYP
CASE_TYP
PRSN_ID
DRUG_DAY
num_drug
ICD9_1
ICD9_2
ICD9_3
DRUG_AMT
drug_cpy
PART_NO
PART_AMT
T_AMT
per_lipitor_mo
per_doc_lip_mo
lip_pat_vst
per_lipitor
d_acadhosp
urban
age
PRSN_SEX
prsn_age
prsn_exper
prsn_tenure
prac_vol
dayssince
hosp_adms_hy
avg_medamt_hy
avg_partamt_hy
avg_stay_hy
Obs
Unique
Mean
Min
310805
49
514.367
491
310805
4846 .
.
310805
22594 .
.
310805
12997 -7605.51
-24029
310780
3 .
.
310805
1462
15670.3
14975
308578
41 .
.
310805
25 .
.
310805
13400 .
.
310805
56 24.17292
1
310805
26 4.212822
1
310805
1851 .
.
256681
1947 .
.
185677
1813 .
.
310805
6986
1142.13
0
310805
11 90.38072
0
310805
42 .
.
310805
80 162.9943
0
310805
8516 1505.093
16
310805
49 0.019745
0
310805
49
0.03059
0
310805
2
0.01887
0
310805
211 0.047465
0
310805
2 0.036322
0
310805
2 0.489632
0
310805
99 63.75954
3
308207
4 .
.
307688
80 44.70343
-798
308178
55 6.250401
-828
279917
18 2.245155
-2
225871
92 4780.519
500
301660
2127 68.72651
1
310805
12 0.181358
0
310805
9715
5551.92
0
310805
5126 338.1398
0
310805
315 0.991601
0
Max
Label
539 Claims Date
.
Hospital ID
.
Patient ID
14636 Patient Birthdate
.
Patient Sex
16436 Office Visit Date
.
Department Visited
.
Reimbursement Type
.
Physician ID
90 Prescription Duration
27 Number of Drugs
.
Diagnosis 1
.
Diagnosis 2
.
Diagnosis 3
75787 Drug Amount
200 Patient Drug Copay Amount
.
Patient Copay Code
930 Patient Total Copay
77037 Total Claims Amount
0.039856 Percent of patient-visits with Lipitor by month
0.062314 Percent of physicians who prescribed Lipitor by month
1 Equals 1 if Lipitor received
1 Percentage of total visits with Lipitor (by Patient)
1 Equals 1 if academic hospital
1 Equals 1 if urban
107 Patient Age
.
Physician Sex
92 Physician Age
68 Physician Experience (Years since certification)
26 Physician Tenure (Years at work place)
69500 Physician Practice Volume
2869 Days between visits
11 Number of Hospital Admissions in last 6 months
1214610 Average Hospital Exp. In last 6 months
225604 Average Inpatient Copay in last 6 months
84 Average length of stay in last 6 months
Empirical Strategy –
Likelihood of adoption

Probit/Logit Model
Pr  Adoption  1    Pati    Physi , j    Hosp i ,k  ui , j ,k
 Pat:
Patient Characteristics: age, gender, past number of
visits, ER visits, hospitalizations, multiple conditions?
 Phys: Physician Characteristics: age, gender, experience,
tenure, past prescription pattern
 Hosp: Hospital Characteristics: Academic, urban, family
practice
 Endogenous variable? Omitted variables (Neglected
heterogeneity)? New diagnoses?
Empirical Strategy –
Duration to Adoption

Right-Censored Duration Model
 Continuous
or Discrete Time-Scaling
 Nonparametric or parametric functional form?:


Weibull (increasing)
 t   t  1 ,   1

Log-logistic  t    u   ,   1
1
 Effect


of Covariates (same as previous slide)
Proportional Hazard (+coeff  - hazard / +duration)
Accelerated Lifetime Hazard (1 unit +coeff  % +duration)
 Other

1  u 
issues
Multiple spells? Time-varying covariates? (move from one hospital
to another?) Unobserved Heterogeneity?
Diffusion pattern
Lipitor (for Hypertensive Patients Only)
Preliminary Results – Panel Data
Likelihood of Lipitor Adoption
Dep var: adoption
family practice
public
clinic
academic hospital
urban
patient age
patient sex
serious
poor
d_hlc
d_hl
physician sex
physician age
physician experience
physician tenure
practice volume
ER visits
hospital admissions
average stay
_cons
OLS/Random Effects
R.E. Logit (Panel), Odds Ratio R.E. Logit (Panel), Marginal Effects
R.E. Probit (Panel), Marginal Effects
Coef.
Std. Err.
OR
Std. Err.
dy/dx
Std. Err. X
dy/dx
Std. Err.
X
-0.0023039**
0.0012944 0.7885027***
0.0615047 -0.2376194***
0.078 0.150923 -0.1277919***
0.04393 0.150923
-0.0053071***
0.0019605 0.5598564***
0.0337863 -0.580075***
0.06035 0.263539 -0.2406292***
0.03407 0.263539
-0.014053***
0.0014696 0.0932925***
0.0087427 -2.372015***
0.09371
0.3185 -1.046141***
0.04508
0.3185
0.0152431***
0.00592 1.758961***
0.1570815 0.5647234***
0.0893 0.040609 0.2879713***
0.05537 0.040609
0.0033508***
0.0014845 1.509043***
0.0816024 0.4114759***
0.05408 0.483731 0.1503279***
0.03058 0.483731
0.0001778***
0.0000483 1.001481
0.0020495 0.0014795
0.00205
63.7609 0.0015028
0.00121
63.7609
0.003106***
0.0013404 1.355327***
0.0667762 0.3040428***
0.04927 0.525462 0.1387352***
0.02902 0.525462
-0.0077691***
0.0022525 0.2406417***
0.0594353 -1.424446***
0.24699 0.027459 -0.5539457***
0.11864 0.027459
-0.0019786
0.0058783 0.3819593***
0.1185521 -0.9624411***
0.31038 0.010389 -0.3743522***
0.18029 0.010389
0.0643002**
0.0379062 4.696839***
2.227083 1.54689***
0.47417 0.000545 0.8631665***
0.2681 0.000545
0.0792764***
0.0056185 11.75206***
0.6187012 2.464028***
0.05265 0.069652 1.262681***
0.03184 0.069652
0.0091677***
0.0025417 1.862283***
0.1276077 0.6218033***
0.06852 0.067642 0.2963631***
0.04025 0.067642
-0.000077*
0.0000494 0.9918423***
0.0015784 -0.0081912***
0.00159
44.6498 -0.0042903***
0.0009
44.6498
-0.0000333
0.0000486 0.997841
0.0024619 -0.0021613
0.00247
6.51965 -0.0016277
0.00117
6.51965
0.0019483***
0.0002169 1.142842***
0.0076847 0.1335182***
0.00672
2.19627 0.0693052***
0.00371
2.19627
0.000000258
1.86E-07 1.000017***
6.66E-06 0.0000174***
0.00001
4457.04 0.00000825***
0
4457.04
-0.0099165***
0.0012375 0.2882533***
0.0612194 -1.243916***
0.21238 0.021999 -0.5823007***
0.10103 0.021999
-0.0015808***
0.00076 0.8719133***
0.0462464 -0.1370653***
0.05304 0.185314 -0.0661082***
0.02798 0.185314
0.00000355
0.0001058 1.003719
0.0077491 0.0037122
0.00772
1.00633 0.0025798
0.00401
1.00633
0.0011855
0.0042243
N
253,088
253,143
R2
0.0389
Chi2
3746.96
Dep. Var.: Adoption = 1 if patient received Lipitor during the visit, 0 otherwise
253,143
253,143
3746.96
2845.04
Preliminary Results – “Pooled” Data
Likelihood of Lipitor Adoption
Dep Var: Adoption
family practice
public
clinic
academic hospital
urban
patient age
patient sex
serious
poor
d_hlc
d_hl
physician sex
physician age
physician experience
physician tenure
practice volume
ER visits
hospital admissions
average stay
_cons
N
Chi2
Pseudo R2
Probit: Marginal Effects
dy/dx
Std. Err.
X
-0.0056235***
0.00258
0.153412
-0.0054855***
0.00214
0.223603
-0.0343888***
0.00223
0.360295
0.0239782***
0.00719
0.035014
0.0016535
0.00223
0.470821
0.0003363***
0.00006
60.522
0.0050531***
0.00191
0.50589
-0.0024749
0.0065
0.015851
-0.0147356***
0.00546
0.007546
0.3558736***
0.16508
0.000434
0.2285449***
0.01403
0.057706
0.0182379***
0.00533
0.05928
-0.0002237**
0.00012
44.8525
-0.0001103**
0.00007
6.46414
0.003362***
0.00034
1.59079
-0.000000221
0
4385.67
-0.030924***
0.00664
0.050649
0.0003737
0.00172
0.208186
-0.0001204
0.00026
1.11786
Logit Regression
Odds Ratio
Std. Err.
0.7880558**
0.0997434
0.7577589***
0.0755574
0.1750072***
0.0250033
1.987782***
0.3049082
1.090944
0.1001267
1.014505***
0.0030491
1.19617***
0.0943077
0.9247244
0.2872006
0.3689485
0.2661378
23.38733***
17.17291
14.97437***
1.290243
1.749463***
0.2256541
0.9913706***
0.0040963
0.9951993
0.0034832
1.151739***
0.0156969
0.9999912
0.0000114
0.272909***
0.0841252
1.032864
0.0846072
0.9956885
0.0122598
18,421
1392.25
0.2368
18,421
1590.20
0.2350
Dep. Var.: Adoption = 1 if patient received Lipitor during the visit, 0 otherwise
Marginal Effects
dy/dx
Std. Err.
X
-0.0044693***
0.0022 0.153412
-0.0052383***
0.00176 0.223603
-0.031211***
0.00211 0.360295
0.0191818***
0.00575 0.035014
0.0017691
0.00187 0.470821
0.0002919***
0.00006
60.522
0.0036313***
0.0016 0.50589
-0.00153
0.00585 0.015851
-0.0130552***
0.00566 0.007546
0.3098128**
0.16249 0.000434
0.1952882***
0.01361 0.057706
0.0144977***
0.00422 0.05928
-0.0001757***
0.00008 44.8525
-0.0000975
0.00007 6.46414
0.0028635***
0.00029 1.59079
-0.000000178
0 4385.67
-0.0263218***
0.00617 0.050649
0.0006554
0.00166 0.208186
-0.0000876
0.00025 1.11786
Preliminary Results – Year by Year
Likelihood of Lipitor Adoption
Dep. Var: Adoption
family practice
public
clinic
academic hospital
urban
patient age
patient sex
serious
poor
d_hlc
d_hl
physician sex
physician age
physician experience
physician tenure
practice volume
ER visits
hospital admissions
average stay
N
Chi2
Pseudo R2
2001
Odds Ratio
0.6302883
0.9075834
0.2663509***
1.020202
1.197663
1.010161
1.039622
0.4865983
23.84015***
14.47979***
1.981936**
0.9946476**
1.0144
1.119848***
0.9999812
0.3484968
1.020416
1.000033
10300
210.87
0.1758
Std. Err.
0.2275293
0.2226236
0.0929507
0.5056124
0.3095196
0.0083097
0.2103167
0.4997534
27.12236
3.028132
0.6188189
0.0024587
0.0184827
0.0469927
0.0000358
0.3548526
0.2478937
0.035679
2002
Odds Ratio
Std. Err.
0.9523267
0.1820943
0.6728642***
0.1019138
0.1620778***
0.040325
1.268227
0.307244
1.940949***
0.2974637
1.002437
0.0047581
1.065695
0.1292174
0.7187478
0.338731
0.5536279
0.5681946
5.493735
5.888289
10.62204***
1.351887
1.66949***
0.3227004
0.9938469**
0.0026193
1.0144
0.0117581
1.035121
0.0252343
0.9999842
0.0000209
0.2186974**
0.1570677
0.8712102
0.1320062
1.004319
0.0175388
2003
Odds Ratio
0.7280197*
0.6207893***
0.1715873***
1.460769**
1.597852***
1.008308**
1.229874**
0.3261428**
0.2523154
Std. Err.
0.127925
0.0809304
0.0354593
0.2864846
0.2019848
0.0040193
0.1253863
0.1676118
0.2585209
9.562978***
1.522148***
0.9698728***
1.01343
1.040202**
1.000017
0.4469373**
0.9838344
0.9936948
1.041586
0.2492969
0.0067897
0.009481
0.0198238
0.0000169
0.1725207
0.1178092
0.0161526
10690
532.87
0.1857
11121
730.05
0.1881
2004
Odds Ratio
0.9902941
0.8926806
0.1475397***
1.4422*
1.380867***
1.005365
1.390368***
0.3395853**
0.3717542
14.79736***
10.2344***
1.4698**
0.9748898***
1.002569
1.003511
1.000019
0.3189974**
1.027071
0.9904582
8272
695.80
0.2044
Std. Err.
0.1694217
0.1217703
0.0307479
0.3017895
0.163111
0.0042933
0.1503748
0.1751703
0.3803973
11.58786
1.16819
0.2791801
0.0071935
0.007345
0.0194944
0.0000171
0.1591509
0.1085529
0.01867
Discussion


Need to reconstruct data from scratch
Different types of severity
 multiplicity

of conditions, or severity of a single condition
Not surprising:
 academic,
urban providers more likely to adopt, patients with
multiple indications more likely to be given Lipitor

A little surprising?
 Female
physicians more likely to adopt (probably problem
from merged data); female patients more likely to receive

Quite surprising?
 More
serious patients less likely to be given Lipitor
Future Agenda

Better understand







What factors lead to CONSISTENT adoption?
Disease conditions
Patients’ disease progression
Drug action mechanism
Physician decision-making process
Drug sales representatives’ activities
Future Research





Random Utility Model of Prescribing Behavior?
Spillover effects
Opinion Leaders
Ethnolinguistic differences
Celebrex study: when do physicians reject new drugs?
◄Chapter 2►
Do new drugs lead to better health
outcomes?
Research Question
Do new drugs lead to better health outcomes?
 More specifically, do patients who take Lipitor,
Avandia, or Plavix experience a reduction in

 ER
visits, hospital admissions, hospital lengths of
stay (problem?), and/or medical expenditures
(compared to patients taking older drugs)?
Quote

“Too often,” says Robert Seidman, chief pharmacy
officer at health insurer WellPoint, “we're choosing the
newer, pricier drug without considering whether older
drugs would get the job done just as well”

Lipitor: $612/180 20mg tablets
Zocor: $799/180 20mg tablets  but soon generics
Mevacor: $228.31/180 20mg tablets
 www.drugstore.com prices


Literature Review

Lichtenberg (1996)


Lichtenberg (2001)


Patients who consume newer drugs experience fewer work-loss days than
patients who consume older drugs; and the former tend to have lower
non-drug expenditures, reducing total expenditures
Lichtenberg (2002)


Number of hospital bed-days declined most rapidly for those diagnoses
with the greatest change in the total number of drugs prescribed and
greatest change in the distribution of drugs (proxy for novelty)
With larger dataset, and 3 years instead of 1 year of observation,
Lichtenberg argues that a reduction in the age of drugs decreased nondrug expenditures 7.2 times as much as it increased drug expenditures.
(8.3 times for Medicare population)
Lichtenberg (2005)

Effect of the launch of new drugs: Average 1 week increase in life
expectancy in the entire population
Conceptual Framework

Empirical question: Estimation of Average
Treatment Effect
 Are
the high cost of new drugs justified based on
their health outcome impact?
 Lichtenberg studies do not address selection bias in
treatment
Atorvastatin (Lipitor):
Clinical Research

Collaborative Atorvastatin Diabetes Study (CARDS),


2,800 patients with type-2 diabetes, no history of heart disease, and
relatively-low levels of cholesterol,
Positive Health outcome:

patients who took Lipitor had a 37 percent reduction in major
cardiovascular events




which included heart attacks, stroke, chest pain that required hospitalization,
cardiac resuscitation, and coronary revascularization procedures.
48 percent fewer Lipitor treated patients experienced strokes compared
to those who received placebo
overall mortality rate for Lipitor patients was 27 percent lower than for
those on placebo.
But: Study Sponsored by Pfizer / No comparison with older
drugs / Relatively Healthy Population
Atorvastatin (Lipitor)
Clinical Research - Hypertension

LIPITOR significantly reduced the rate of coronary events






either fatal coronary heart disease (46 events in the placebo group vs 40
events in the LIPITOR group)
or nonfatal MI (108 events in the placebo group vs 60 events in the
LIPITOR group)]
relative risk reduction of 36% (based on incidences of 1.9% for
LIPITOR vs 3.0% for placebo), p=0.0005
The risk reduction was consistent regardless of age, smoking status,
obesity or presence of renal dysfunction. The effect of LIPITOR was
seen regardless of baseline LDL levels. Due to the small number of
events, results for women were inconclusive.
N = 10,305 (Anglo-Scandinavian Cardiac Outcomes Trial)
Source: www.lipitor.com
Mixed Results for Lipitor Vs. Zocor
By THERESA AGOVINO, AP Business Writer
Tuesday, November 15, 2005 06 57 PM




High doses of the cholesterol-lowering drug Lipitor were no
better at preventing major heart problems than regular doses of
rival Zocor, according to the latest study on efforts to
aggressively treat the conditions released Tuesday.
Lipitor outperformed Zocor on several fronts such as lowering
cholesterol and preventing nonfatal heart attacks. The findings
will continue to give it an advantage in the market even if generic
Zocor is less expensive, some doctors said.
But: HIGH DOSE OF LIPITOR vs. REGULAR DOSE OF
ZOCOR
What about LIPITOR vs. MEVACOR, PRAVACHOL,
LESCOL, CRESTOR
Empirical Strategy

Naïve Fixed Effects Regression
Yijt  i   j   t    treatment    y  cond _ durijty   ijt
y
Yitj  health outcome (ER visits, hospitaliz ation, hospital lengthof stay) or medical expenditur es for person i with condition j in year t
i  person fixed effects
 j  condition fixed effects (combinati on of conditions )
 t  year fixed effects
treatment  new drug usage indicator
cond _ dur  1 if condition( s) j borne by patient i started y years ago, otherwise 0.
Threats to Identification
Selection for treatment most likely not random
 Selection Bias in Treatment

 Perhaps
physicians assign nonrandom populations to
treatment
 Perhaps patients seek physicians who prescribe new
drugs (e.g., Lipitor)
Correction for Selection Bias

Instrumental Variable Approach
 Gives
internally valid causal effects for individuals whose
treatment status is manipulable by the instrument




Candidates: the combination of covariates from Chapter 2 as an
instrument for the treatment (i.e., use of new drug, such as Lipitor)
With patient’s pre-adoption status in the instruments to avoid patient
self-selection
However, may reduce statistical power
Note: we can see if patients actually self-select into treatment
 But:
instruments (predicts adoption) may also affect the
dependent variable (measures for health outcome)?
Correction for Selection Bias

Selection on Observables
 Propensity
Score Matching
 Analysis of the Effects of Unobservables?
Cost Analysis

Lipitor Costs (Taiwan NHID formulary 2004, in USD):
 $1.04

per 10 mg tablet; $1.40 per 20 mg; $1.75 per 40 mg
What are the cost savings?
 If
new drug reduces emergency and hospital services
 Savings = reduced cost in emergency and hospital services –
increased drug costs

What are the additional costs?
 If
new drug has not health outcome impact?
 Additional cost = difference in price of new and old drugs
Distribution of new Lipitor takers
“Treatment” vs. “Non-Treatment”
Variable
Obs
Mean
Std. Dev. Min
Treatment = 0 (received Lipitor < 50% of visits after first Lipitor)
Age
17770 61.99561 11.14983
Total number of visits
17770 69.03523 37.57343
number of visits after Lipitor
17770 21.39961 12.30476
Number of visits with Lipitor
17770 4.650366 3.463945
How long on Lipitor (days)
17770 190.9992 251.5393
How long since first visit (ever)
17770 2313.325 525.1061
serious
17770 0.0270681 0.1622865
Treatment = 1 (received Lipitor > 50% of visits after first Lipitor)
Age
23691 62.32628 10.75469
Total number of visits
23691 58.21945 32.66184
number of visits after Lipitor
23691 12.67751 10.39991
Number of visits with Lipitor
23691 9.914356 8.203874
How long on Lipitor (days)
23691 335.9669 314.1132
How long since first visit (ever)
23691 2196.106 650.6024
serious
23691 0.0054451 0.0735913
Max
25
6
3
1
0
49
0
198
246
68
18
1214
2872
1
3
1
1
1
0
0
0
198
192
51
46
1295
2901
1
Graphical Evidence – ER visits
No adjustment for selection bias
Graphical Evidence –
ER visits (1009 Lipitor takers)
Graphical Evidence –
Smoothed ER visits (1009 Lipitor takers)
Graphical Evidence –
ER visits (656 consistent takers)
Graphical Evidence –
Smoothed ER visits (656 consistent takers)
Graphical Evidence - Hospitalization
Graphical Evidence –
Hospitalization (1009 Lipitor takers)
Graphical Evidence –
Smoothed Hospitalization (1009 takers)
Graphical Evidence –
Hospitalization(656 consistent takers)
Graphical Evidence –
Smoothed Hospitalization(656 consistent takers)
Graphical Evidence –
Average Length of Stay
Graphical Evidence –
Average Length of Stay (1009 Lipitor takers)
Graphical Evidence –
Smoothed average lengths of stay (1009 Lipitor takers)
Graphical Evidence –
Average lengths of stay (656 consistent takers)
Graphical Evidence –
Smoothed average lengths of stay (656 consistent takers)
Graphical Evidence –
Average Hospital Expenditures
Graphical Evidence –
Average expenditures (1009 takers)
Graphical Evidence –
Smoothed average expenditures (1009 takers)
Graphical Evidence –
Average expenditures (656 consistent takers)
Graphical Evidence –
Smoothed average expenditures (656 consistent takers)
Discussion

Consistent Lipitor use does lead to better health outcomes?










Not just selection bias if consistency does improve health outcome?
But suggestive evidence that healthier patients are more likely to receive
Lipitor consistently?
But: numerous possibilities for errors while merging
2004 data consistently slightly “bizarre”
Treatment indicator very rough. 1 prescription of Lipitor over 2 visits =
50%  treatment
Need to consider consistency over a prescribed period of time: 3
months?
What did they take before Lipitor?
Need to include all indications for use of Lipitor
Need to adjust for patient heterogeneity
SELECTION ISSUES
Future Agenda

Better understand
 Drug
adoption decisions, based on chapter 1
 Quest for proper instrument
 Hypertension, Hypertensive Heart Diseases,
Diabetes, High Cholesterol
 Clinical trial results for the new drugs
Thank you for your attention
and valuable assistance
◄Chapter 3►
Do physicians consider patient out-ofpocket expenses when prescribing
drugs?
Research Question

Do physicians consider patient out-of-pocket expenses?
 In
August 1999, Taiwan implemented a modest, linear
prescription drug copayment system
 Patients with one of 97 chronic conditions can be exempt
from outpatient prescription copayments if physicians give
an “chronic illness extended prescription certificate”
 As of 2005, only 13% of eligible patient-visits receive the
extended prescription certificate

Do physicians with high practice volume only give the
extended prescription certificate? Or do patients have
to demand the certificate?
Literature Review

Three bodies of literature:
 Impact

of cost-sharing on patients’ drug utilization choice:
Soumerai et al (1987, 1991, 1994), Nelson (1984), Tamblyn (2001).
 Patient-Physician

Especially: Supplier-induced demand (SID): Rice (1983), Yip (1998)
 Physician

Principal-Agent Relationship
consideration of patient out-of-pocket expenses
Only survey studies available:
Contribution and Limitation

Contribution
 As
far as I know, first paper to investigate through nonsurvey data whether physicians consider patient out-ofpocket expenses in their prescribing behavior
 Policy implications:


Greater payment for physicians to give certificate; or greater effort to
inform patients of their financial rights
Limitation
 Copayment
is insignificant (capped at $3.33 USD until 2001,
then capped at $6.66 USD)
 Generalizability?
 Correlation Study
Conceptual Framework

Physicians generally earn greater income
through increased practice volume
 Physicians
give certificates if the already have high
practice volume
 Or patients may demand certificate: proxied by
competition and patient sophistication
 Or both
Empirical Strategy

First: Fixed Effects Regression
 Investigate
effects of copayment on number of drugs,
prescription duration, adjusted drug amount, and adjusted
drug quantity
Yit  i   j   k    copaymentit    log num _ vstit     other _ copayit   it
Yit  Prescripti on duration, number of drugs, adjusted drug amount, and adjusted drug quantity
i  Person fixed effects
 j  Therapeuti c class fixed effects
 k  Year fixed effects
log num _ vstit   logged cumulative number of visits
other _ copayit  Non - drug copayment amount
Empirical Strategy

Second: Logit/Probit Estimation
 Effects
of physician practice volume, patient
sophistication level, and market competition on the
likelihood of giving “extended prescription
certificate”
Data Files






Ambulatory Care Expenditures by Visit
Details of Ambulatory Care Orders
Inpatient Expenditures by Admission
Details of Inpatient Orders
Expenditures for Prescriptions Dispensed at
Contracted Pharmacies
Details of Prescriptions Dispensed at Contracted
Pharmacies
Ambulatory Care Expenditures by Visit
Field Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Ambulatory Care Expenditures by Admission
Variable Name
Description
FEE_YM
Month and Year of Application
APPL_TYPE
Application Type
HOSP_ID
Hospital ID
APPL_DATE
Application Date
CASE_TYPE
Case Type
SEQ_NO
Sequence Number
CURE_ITEM_NO1
Code for Chronic illnesses
CURE_ITEM_NO2
CURE_ITEM_NO3
CURE_ITEM_NO4
FUNC_TYPE
Hospital Department Visited
FUNC_DATE
Date of Hospital Visit
TREAT_END_DATE
Treatment End Date
ID_BIRTHDAY
Patient's Birthdate
ID
Patient's ID Number
CARD_SEQ_NO
Patient's NHIB Card Sequence
GAVE_KIND
Type of Reimbursement
PART_NO
Copayment Code
ACODE_ICD9_1
ICD-9-CM Code
ACODE_ICD9_2
ICD-9-CM Code 2
ACODE_ICD9_3
ICD-9-CM Code 3
ICD_OP_CODE
Surgical Code
DRUG_DAY
Prescription Duration (days)
MED_TYPE
How Prescription is Filled
PRSN_ID
Physician's ID Number
PHAR_ID
Pharmacist's ID Number
DRUG_AMT
Drug Expenditures
TREAT_AMT
Treatment Fee
TREAT_CODE
Treatment Code
DIAG_AMT
Diagnosis Fee
DSVC_NO
Drug Handling Fee Code
DSVC_AMT
Drug Handling Fee Fee
BY_PASS_CODE
DRG Code
T_AMT
Total Amount
PART_AMT
Copayment Amount
T_APPL_AMT
Total Amount Applied
ID_SEX
Patient's Gender
Note
Western Medicine, Tradition, Dentistry, etc.
Some chronic illnesses are exempt from
certain payments.
Details of Ambulatory Care Orders
Details of Ambulatory Care Orders
Field Number Variable Name Description
1
FEE_YM
Application Month and Year
2
APPL_TYPE
Application Type
3
HOSP_ID
Hospital ID
4
APPL_DATE
Application Date
5
CASE_TYPE
Case Type
6
SEQ_NO
Sequence Number
7
ORDER_TYPE Order Type
8
DRUG_NO
Drug Code
9
DRUG_USE
Drug Dosage
10
DRUG_FRE
Drug Frequency
11
UNIT_PRICE
Unit Price of Drug
12
TOTAL_QTY
Total Quantity (of Drugs)
13
TOTAL_AMT
Total Amount
Note
1. Diagnostic test; 2: Prescription drugs; 3: Special Materials
Inpatient Files

Inpatient Expenditures by Admission
 Identification
information; patient age and gender;
date of admission and release; ICD9CM codes, ICD
operation codes, DRG code, various fees, various
copay amounts

Details of Inpatient Orders
 Identification
information; drug dispensed or
services rendered
Summary Statistics, Master File
Variable
visit
FEE_YM
HOSP_ID
ID
ID_BMDY
ID_SEX
FUNC_MDY
FUNC_TYP
PRSN_ID
DRUG_NO
UNIT_PRI
TOTAL_QTY
D_SUBTOT
DRUG_DAY
num_drug
ICD9_1
ICD9_2
ICD9_3
DRUG_AMT
drug_cpy
PART_NO
PART_AMT
T_AMT
total_vst
R_HOSP_ID
R_CASE_T
DRUG_MDY
Obs
Unique
Mean
3.76E+07 1.48E+07
6739285
3.76E+07
96
14964.93
3.76E+07
24621 .
3.76E+07
191525 .
3.76E+07
31848
-477.187
3.76E+07
3 .
3.76E+07
2994
14979.37
3.76E+07
62 .
3.76E+07
46854 .
3.76E+07
31858 .
3.76E+07
5423
29.06636
3.76E+07
825
15.306
3.76E+07
4857
108.4071
3.76E+07
87
8.283142
3.76E+07
41
4.283822
3.71E+07
12407 .
1.83E+07
11348 .
9508421
8517 .
3.76E+07
15094
448.6745
3.71E+07
12
21.51059
3.73E+07
69 .
3.71E+07
308
86.68472
3.71E+07
23033
863.5745
1.48E+07
709
132.8225
5905315
9428 .
5814920
37 .
5905315
2923
15630.46
Min
1
13515
.
.
-58438
.
-15276
.
.
.
0
0
0
0
1
.
.
.
0
0
.
0
0
1
.
.
12848
Max
1.48E+07
16406
.
.
296358
.
233586
.
.
.
61192
34083
535605
99
41
.
.
.
535605
200
.
2030
535863
1531
.
.
123283
Label
Visit Number*
Claims Date
Hospital ID
Patient ID
Patient Birthdate
Patient Sex**
Office Visit Date
Department Visited
Physician ID
Drug Code
Drug Unit Price
Drug Total Quantity
Drug Subtotal
Prescription Duration
Number of Different Drugs
Diagnosis Code 1
Diagnosis Code 2
Diagnosis Code 3
Drug Amount
Patient Drug Copayment
Patient Copayment Code
Patient Copayment Amount
Total Claims Amount
Number of Total Visits
Prescribing Hospital ID
Reimbursement Type
Date of Prescription
All monetary amounts in New Taiwan Dollars. US$1 = NTD$33
*Each unique patient-visit receives a unique visit number
**A person is listed as unspecified if his/her gender is unknown. This represents only 0.15% of all patient-visits.
Master Data File includes all outpatient patient-visits with a prescription for at least one drug for all 200,000
randomly selected individuals in the data set from 1997-2004