Causal Effect of Ambulatory Specialty Care on Mortality Following Myocardial Infarction: A Comparison of Propensity Score and Instrumental Variable Analyses

  • Mary Beth Landrum
  • John Z. Ayanian
Article

Abstract

The quality and outcomes of care provided by primary care physicians and specialists are increasingly important issues in health policy research. Estimating the effect of specialty care on patient outcomes however is complicated by the observational nature of the studies. Patients treated by specialists are often different in terms of observed and unobserved characteristics that can bias estimates of specialty effects. We illustrate and compare two different analytic approaches, propensity scores and instrumental variables, to infer the causal effect of cardiology care in the ambulatory setting on 18-month mortality among 5467 elderly patients who survived at least 3 months after being hospitalized for a myocardial infarction in New York state during 1994 and 1995. Using both approaches we found reductions in 18-month mortality associated with ambulatory cardiology care. However, reasonable deviations from the assumptions underlying each method led to estimated differences in mortality ranging from a 6% absolute reduction in mortality to a 2% increase among patients who received cardiology care. Choosing an analytic strategy depends on both available data and the policy question of interest. We believe that comparative analyses such as this one, with extensive assessment of the assumptions underlying each method, can provide valuable insights into important policy questions reliant on the analysis of observational data.

causal inference propensity scores instrumental variables acute myocardial infarction specialty care 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. J. D. Angrist, G. W. Imbens and D. B. Rubin, “Identification of causal effects using instrumental variables,” Journal of the American Statistical Association, 434, pp. 444-454, 1996.Google Scholar
  2. J. D. Angrist and G. W. Imbens, “Two-stage least squares estimation of average causal effects in models with variable treatment intensity,” Journal of the American Statistical Association, 90, pp. 431-442, 1995.Google Scholar
  3. J. Z. Ayanian, E. Guadagnoli, B. J. McNeil and P. D. Cleary, “Treatment and outcomes of acute myocardial infarction among patients of cardiologists and generalist physicians,” Archives of Internal Medicine, 1997.Google Scholar
  4. S. J. Borowsky, R. L. Kravitz and M. Laouri, et al., “Effect of physician specialty on use of necessary coronary angiography,” Journal of the American College of Cardiology, 26, pp. 1484-1491, 1995.Google Scholar
  5. A. F. Connors, T. Speroff and N. V. Dawson. et al., “The effectiveness of right heart catheterization in the initial care of critically ill patients,” Journal of the American Medical Association, 276, pp. 889-897, 1996.Google Scholar
  6. R. B. D'Agostino, “Tutorial in biostatistics: propensity score methods for bias reduction in the comparisons of a treatment to a non-randomized control group,” Statistics in Medicine, 17, pp. 2265-2281, 1998.Google Scholar
  7. M. E. Glickman and S. T. Normand, “The derivation of a latent threshold instrumental variables model,” Statistica Sinica, 10, pp. 517-544, 2000.Google Scholar
  8. C. D. Frances, M. G. Shlipak, H. Noguchi, P. Heidenreich and M. McClellan, “Does physician specialty affect the survival of elderly patients with myocardial infarction?” Health Services Research, 35, pp. 1093-1116, 2000.Google Scholar
  9. K. M. Harris and D. K. Remler, “Who is the marginal patient? understanding instrumental variables estimates of treatment effects,” Health Services Research, 33, pp. 1337-1360, 1998.Google Scholar
  10. K. Hirano, G. W. Imbens, D. B. Rubin and X. H. Zhou, “Assessing the effect of an influenza vaccine in an encouragement design,” Biostatistics, 1, pp. 69-88, 2000.Google Scholar
  11. P. Holland, “Statistics and causal inference,” Journal of the American Statistical Association, 81, pp. 945-970, 1986.Google Scholar
  12. G. W. Imbens and J. D. Angrist, “Identification and estimation of local average treatment effects,” Econometrica, 62, pp. 467-475, 1994.Google Scholar
  13. J. Johnston and J. DiNardo. Econometric Methods, McGraw-Hill: New York, 1997.Google Scholar
  14. J. G. Jollis, E. R. DeLong and E. D. Peterson, et al., “Outcome of acute myocardial infarction according to the specialty of the admitting physician,” New England Journal of Medicine, 335, pp. 1880-1887, 1996.Google Scholar
  15. T. A. Marciniak, E. F. Ellerbeck and M. J. Radford, et al., “Improving the quality of care for medicare patients with acute myocardial infarction: results from the cooperative cardiovascular project,” Journal of the American Medical Association, 279, pp. 1351-1359, 1998.Google Scholar
  16. M. McClellan, B. J. McNeil and J. P. Newhouse, “Does more intensive treatment of acute myocardial infarction in the elderly reduce mortality?” Journal of the American Medical Association, 272, pp. 859-866, 1994.Google Scholar
  17. I. S. Nash, D. B. Nash, V. Fuster, “Do cardiologists do it better?” Journal of the American College of Cardiology, 29, pp. 274-278, 1997.Google Scholar
  18. C. A. Polanczyk, L. E. Rohde, L. Goldman, E. F. Cook, E. J. Thomas, E. R. Marcantonio, C. M. Mangione and T. H. Lee, “Right heart catheterization and cardiac complications in patients undergoing noncardiac surgery: an observational study,” Journal of the American Medical Aociation, 286, pp. 309-314, 2001.Google Scholar
  19. P. R. Rosenbaum and D. B. Rubin, “The central role of the propensity score in observational studies for causal effects,” Biometrika, 70, pp. 41-55, 1983a.Google Scholar
  20. P. R. Rosenbaum and D. B. Rubin, “Assessing sensitivity to an unobserved binary covariate in an observational study with binary outcome,” Journal of the Royal Statistical Society, Series B, 45, pp. 212-218, 1983b.Google Scholar
  21. P. R. Rosenbaum and D. B. Rubin, “Reducing bias in observational studies using subclassification on the propensity score,” Journal of the American Statistical Association, 79, pp. 516-524, 1984.Google Scholar
  22. P. R. Rosenbaum and D. B. Rubin, “Constructing a control group using multivariate matched sampling methods that incorporate the propensity score,” American Statistician, 39, pp. 33-38, 1985a.Google Scholar
  23. P. R. Rosenbaum and D. B. Rubin, “The bias due to incomplete matching,” Biometrics, 41, pp. 103-166, 1985b.Google Scholar
  24. D. B. Rubin, “Estimating causal effects of treatments in randomized and nonrandomized studies,” Journal of Educational Psychology, 66, pp. 688-701, 1974.Google Scholar
  25. D. B. Rubin, “Bayesian inference for causal effects,” The Annals of Statistics, 6, pp. 34-58, 1978.Google Scholar
  26. D. B. Rubin, “Using Multivariate Matched Sampling and Regression Adjustment to control bias in observational studies,” Journal of the American Statistical Association, 74, pp. 318-328, 1979.Google Scholar
  27. D. R. Thiemann, J. Coresh, W. J. Oetgen and N. R. Powe, “The association between hospital volume and survival after acute myocardial infarction in elderly patients.” New England Journal of Medicine, 21, pp. 1640-1648, 1999.Google Scholar

Copyright information

© Kluwer Academic Publishers 2001

Authors and Affiliations

  • Mary Beth Landrum
    • 1
  • John Z. Ayanian
    • 1
    • 2
  1. 1.Department of Health Care PolicyHarvard Medical SchoolBostonUSA
  2. 2.Department of Medicine, Division of General MedicineBrigham and Women's Hospital and Harvard Medical SchoolUSA

Personalised recommendations