1. The Power of Propensity Score Matching to Remove Confounding by Indication:  Statins and Acute Myocardial Infarction among HIV Infected and Uninfected Veterans
Elyn H. Wang, Matthew S. Freiberg, Janet P. Tate, Joyce Chang, Melissa Skanderson, Amy Justice for the Veterans Aging Cohort Study (VACS)

2. Veterans Aging Cohort Study (VACS)
Required:
Optional based on project funding:

3. Virtual Cohort (VC) SUBJECTS: 47,805 HIV infected; 99,061 uninfected
All individuals with HIV diagnoses
Age, race/ethnicity, region 2:1 matched controls
Last updated: September 2012
Next update: September 2014
SITES: All VA sites
BASELINE: 1998 (15 years of follow up)
HIV infected veterans at initiation of HIV care
Controls selected and followed in same calendar year

4. HIV and Acute Myocardial Infarction (AMI)
Prior studies show an increased risk of AMI in HIV+
- Abnormal lipids
- Chronic Inflammation
Statins have been shown to decrease CHD events
- LDL reduction
- Cholesterol-independent effects
Do HIV+ individuals experience similar effects from
statin therapy compared to HIV- individuals?

5. Methods Study population
Veterans Aging Cohort Study Virtual Cohort (VACS-VC)
from April 1, 2003 to December 31, 2009
Exclusions:
- Prior or Existing CVD
- Statin Use Within 365 Days
Data Sources
- Pharmacy Fill/Refill (Time-Updated)
Patient had to be on statins for at least 30 days
- Medical Charts and Lab Results

6. Study Populationa HIV – HIV+ Non-User Statin User AMI Events (%)
233 (0.8)
275 (1.1)
217 (1.1)
146 (1.7)
Mean Age (SD)
47.3 (9.3)
49.4 (8.5)
47.2 (9.4)
49.9 (9.0)
Male (%)
96.8
97.7
97.3
97.1
Race (%)
White
35.4
38.3
34.2
44.2
Black
50.2
47.6
51.7
41.6
Other
14.4
14.1
Alcohol (%)
15.3
12.5
15.8
11.0
Cocaine (%)
8.5
6.9
12.9
8.4
CD4 <200
31.6
21.8
VL ≥500
60.8
44.4
aStatistically significant difference between Non-user and Statin user for all characteristics

7. Other Covariates HCV Hgb <14 TG ≥150 LDL ≥130 BMI ≥30 Smoking HTN
eGFR <60
Diabetes
HDL <40

8. Results AMI Risk and Statin Use
Stratified Model HR (95% CI) p-value
HIV (-) Statin Use ( )
HIV (+) Statin Use ( )
Interaction Between Statin Use and HIV Status
(p =0.16)

9. The Next Step Propensity Score Analysis
Step 1: Estimate each patient’s propensity score
(probability of receiving treatment) based on
multiple variables
Step 2: Match each treated patient with a non-
treated patient who has a similar propensity score
Treated Patients
Not Treated

10. HIV Age Sex Race Hypertension Diabetes LDL HDL
Triglycerides
BMI
eGFR
Hgb
HCV
Smoking
Alcohol
Cocaine
Propensity Score Modeling Patient characteristics and statin initiation
Sex and Cocaine were not significant predictors

11. c = 0.804

15. Results After Matching AMI Risk and Statin Use
Matched Model HR (95% CI) p-value
HIV (-) Statin Use ( )
HIV (+) Statin Use ( )
Interaction Between Statin Use and HIV Status
(p =0.64)

17. National VACS Project Team 2011

18. Acknowledgements Consortium PI : AC Justice*
Scientific Collaborator (NIAAA): K Bryant
Affiliated PIs: S Braithwaite, K Crothers*, R Dubrow *, DA Fiellin*, M Freiberg*, V LoRe*
Participating VA Medical Centers: Atlanta (D. Rimland*, V Marconi), Baltimore (M Sajadi, R Titanji), Bronx (S Brown, Y Ponomarenko), Dallas (R Bedimo), Houston (M Rodriguez-Barradas, N Masozera), Los Angeles (M Goetz, D Leaf), Manhattan-Brooklyn (M Simberkoff, D Blumenthal, H Leaf, J Leung), Pittsburgh (A Butt, K Kraemer, M Freiberg, E Hoffman), and Washington DC (C Gibert, R Peck)
Core and Workgroup Chairs: C Brandt, J Edelman, N Gandhi, J Lim, K McGinnis, KA Oursler, C Parikh, J Tate, E Wang, J Womack
Staff: H Bathulapalli, T Bohan, J Ciarleglio, A Consorte, P Cunningham, L Erickson, C Frank, K Gordon, J Huston, F Kidwai-Khan, G Koerbel, F Levin, L Piscitelli, C Rogina, S Shahrir, M Skanderson
Major Collaborators: VA Public Health Strategic Healthcare Group, VA Pharmacy Benefits Management, Massachusetts Veterans Epidemiology Research and Information Center (MAVERIC), Yale Center for Interdisciplinary Research on AIDS (CIRA), Center for Health Equity Research and Promotion (CHERP), ART-CC, NA-ACCORD, HIV-Causal
Cross Cohort Collaborators: Richard Moore (NA-ACCORD), Jonathan Sterne (ART-CC), Brian Agan (DoD)
Major Funding by: National Institutes of Health: AHRQ (R01-HS018372), NIAAA (U24-AA020794, U01-AA020790, U01-AA020795, U01-AA020799, U24-AA022001, U24 AA022007), NHLBI (R01-HL095136; R01-HL090342) , NIAID (U01-A ), NIMH (P30-MH062294), NIDA (R01DA035616), NCI (R01 CA173754) and the Veterans Health Administration Office of Research and Development (VA REA , VA IRR Merit Award) and Office of Academic Affiliations (Medical Informatics Fellowship)
*Indicates individual is also the Chair of a Core or Workgroup

19. Acknowledgements Continued
COMpAAAS/Veterans Aging Cohort Study, a CHAART Cooperative Agreement, supported by the National Institutes of Health: National Institute on Alcohol Abuse and Alcoholism (U24-AA020794, U01-AA020790, U01-AA020795, U01-AA020799) and in kind by the US Department of Veterans Affairs.  In addition to grant support from NIAAA, we gratefully acknowledge the scientific contributions of Dr. Kendall Bryant, our scientific collaborator.
QR Codes
QR Code for VACS QR Code for VACS QR Code for VACS
Homepage INDEX CALCULATOR INDEX CALCULATOR-
MOBILE APP