A Comparison of Statistical Modeling Strategies for Analyzing Length of Stay after CABG Surgery

  • Peter C. AustinEmail author
  • Deanna M. Rothwell
  • Jack V. Tu


Investigators in clinical research are often interested in determining the association between patient characteristics and post-operative length of stay (LOS). We examined the relative performance of seven different statistical strategies for analyzing LOS in a cohort of patients undergoing CABG surgery. We compared linear regression; linear regression with log-transformed length of stay; generalized linear models with the following distributions: Poisson, negative binomial, normal, and gamma; and semi-parametric survival models.

Nine of twenty patient characteristics were found to be significantly associated with increased LOS in all models. The models disagreed upon the statistical significance of the association between the remaining patient characteristics and increased LOS. Generalized linear models with Poisson, negative binomial, and gamma distributions, and the Cox regression model demonstrated the greatest consistency. With the exception of Cox regression, all models had similar ability to predict length of stay in the actual data. However, the generalized linear models tended to have marginally lower prediction error than the linear models. Using four measures of prediction error, Cox regression had substantially higher prediction error than the other models. Generalized linear models were best able to predict patient length of stay in Monte Carlo simulations that were performed.

Researchers should consider generalized linear models with normal, Poisson, or negative binomial distributions for predicting length of stay following CABG surgery. Post-operative length of stay is a complex phenomenon that is difficult to incorporate into a simple parametric model due to a small proportion of patients having very long lengths of stay.

length of stay CABG surgery generalized linear model predictive modeling 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Birkes, D. and Dodge, Y., Alternative methods of regression, John Wiley &; Sons, Inc., New York, NY, 1993.Google Scholar
  2. 2.
    Blough, D.K. and Ramsey, S.D., “Using generalized linear models to assess medical care costs,” Health Services &; Outcomes Research Methodology 1, 185-202, 2000.Google Scholar
  3. 3.
    Buchinsky, M., “Changes in the U.S. wage structure 1963-1987: Application of quantile regression,” Econometrica 62, 405-458, 1994.Google Scholar
  4. 4.
    Cameron, A.C. and Trivedi, P.K., Regression analysis of count data, Cambridge University Press: Cambridge, 1998.Google Scholar
  5. 5.
    Cleveland, W.S., “Robust locally weighted regression and smoothing scatterplots,” Journal of the American Statistical Association 74, 829-836, 1979.Google Scholar
  6. 6.
    Cox, D.R. and Oakes, D., Analysis of survival data. Chapman &; Hall: London, 1984.Google Scholar
  7. 7.
    Draper, B. and Luscombe, G., “Quantification of factors contributing to length of stay in an acute pschogeriatric ward,” International Journal of Geriatric Psychiatry 13, 1-7, 1998.Google Scholar
  8. 8.
    Duan, N., “Smearing estimate:Anonparametric retransformation method,” Journal of the American Statistical Association 78, 605-610, 1983.Google Scholar
  9. 9.
    Dudley, R.A., Harrell, F.E., Smith, L.R., Mark, D.B., Califf, R.M., Pryor, D.B., et al., “Comparison of analytic models for estimating the effect of clinical factors on the cost of coronary artery bypass graft surgery,” J. Clin. Epidemiol. 46, 261-271, 1993.Google Scholar
  10. 10.
    Eastwood, E.A., Hagglund, K.J., Ragnarsson, K.T., Gordon, W.A., and Marino, R.J., “Medical rehabilitation length of stay and outcomes for persons with traumatic spinal cord injury,” Archives of Physicical Medicine &; Rehabilitation 80, 1457-1463, 1999.Google Scholar
  11. 11.
    Efron, B. and Tibshirani, R.J., An Introduction to the Bootstrap. Chapman &; Hall, New York, NY, 1993.Google Scholar
  12. 12.
    Escalante, A. and Beardmore, T.D. “Predicting length of stay after hip or knee replacement for rheumatoid arthritis,” Journal of Rheumatology 24, 146-152, 1997.Google Scholar
  13. 13.
    Evans, M., Hastings, N., and Peacock, B., Statistical Distributions, 2nd edition. John Wiley &; Sons, Inc., New York, NY, 1993.Google Scholar
  14. 14.
    Ghali, W.A., Hall, R.E., Ash, A.S., and Moskowitz, M.A. “Identifying pre-and postoperative predictors of cost and length of stay for coronary artery bypass surgery,” American Journal of Medical Quality 14, 248-254, 1999.Google Scholar
  15. 15.
    Glick, H.A. and Polsky, D., “Analytic approaches for the evaluation of costs,” Hepatology 29 (Suppl. 1), 18S-22S, 1999.Google Scholar
  16. 16.
    Harvey, R.L., Roth, E.J., Heinemann, A.W., Lovell, L.L., McGuire, J.R., and Diaz, S., “Stroke rehabilitation: Clinical predictors of resource utilization,” Archives of Physical Medicine &; Rehabilitation 79, 1349-1355, 1998.Google Scholar
  17. 17.
    Howard, D., Burke, G.L., Evans, G.W., and Crouse, J.R. 3rd, Riley, W., Arnett, D., et al., “Relations of intimal-medial thickness among sites within the carotid artery as evaluated by B-mode ultrasound. ARIC investigators,” Atherosclerosis Risk in Communities. Stroke 1581-1587, 1994.Google Scholar
  18. 18.
    Huntley, D.A., Cho, D.W., Christman, J., and Csernansky, J.G., “Predicting length of stay in an acute psychiatric hospital,” Psychiatric Services 49, 1049-1053, 1998.Google Scholar
  19. 19.
    Kalbfleisch, J.D. and Prentice, R.L., The Statistical Analysis of Failure Time Data. John Wiley &; Sons, New York, NY, 1980.Google Scholar
  20. 20.
    Klein, J.P. and Moeschberger, M.L., Survival Analysis: Techniques for Censored and Truncated Data. Springer-Verlag, New York, NY, 1997.Google Scholar
  21. 21.
    Koenker, R. and Bassett, G. Jr., “Regression Quantiles,” Econometrica 46, 33-50, 1978.Google Scholar
  22. 22.
    Lambert, C.M., Hurst, N.P., Forbes, J.F., Lochhead, A., Macleod, M., and Nuki, G., “Is day care equivalent to inpatient care for active rheumatoid arthritis?” Randomised controlled clinical and economic evaluation. British Medical Journal 316, 965-969, 1998.Google Scholar
  23. 23.
    Lassere, M. and Edmonds, J., “Psychological measures: Practical issues in observational studies and clinical monitoring,” Journal of Rheumatology 24, 1004-1007, 1997.Google Scholar
  24. 24.
    Lawless, J.F., Statistical Models and Methods for Lifetime Data. John Wiley &; Sons, New York, NY, 1982.Google Scholar
  25. 25.
    Lipscomb, J., Ancukiewicz, M., Parmigiani, G., Hasselblad, V., Samsa, G., and Matchar, D.B., “Predicting the cost of illness: A comparison of alternative models applied to stroke,” Med. Decis. Making 18 (suppl), S39-S56, 1998.Google Scholar
  26. 26.
    Manning, W.G., “The logged dependent variable, heteroscedasticity, and the retransformation problem,” Journal of Health Economics 17, 283-295, 1998.Google Scholar
  27. 27.
    Manning, W.G. and Mullahy, J., “Estimating log models: To transform or not to transform,” Journal of Health Economics 20, 461-494, 2001.Google Scholar
  28. 28.
    MathSoft Inc. S-Plus, Version 5.1. Cambridge, MA, 1999.Google Scholar
  29. 29.
    MaWhinney, S., Brown, E.R., Malcolm, J., Villaneuva, C., Groves, B.M., Quaife, R.A., et al., “Identification of risk factors for increased cost, charges, and length of stay for cardiac patients,” Annals of Thoracic Surgery 70, 702-710, 2000.Google Scholar
  30. 30.
    McCullagh, P. and Nelder, J.A., Generalized Linear Models. Chapman &; Hall: London, 1983.Google Scholar
  31. 31.
    Myers, R.H., Montgomery, D.C., and Vining, G.G., Generalized Linear Models. With Applications in Engineering and the Sciences. John Wiley &; Sons: New York, NY, 2002.Google Scholar
  32. 32.
    Naylor, C.D., Rothwell, D.M., and Tu, J.V., “Austin PC and the steering committee of the cardiac care network of Ontario. Outcomes of coronary artery bypass surgery in Ontario,” in Cardiovascular health services in Ontario: An ICES atlas. (C.D. Naylor, P.M., Slaugher, eds.), Institute for Clinical Evaluative Sciences, Toronto, Canada, pp. 189-197, 1999.Google Scholar
  33. 33.
    Roddy, S.P., Estes, J.M., Kwoun, M.O., O'donnell, T.F. Jr., and Mackey, W.C., “Factors predicting prolonged length of stay after carotid endarterectomy,” Journal of Vascular Surgery 32, 550-554, 2000.Google Scholar
  34. 34.
    Rosen, A.B., Humphries, J.O., Muhlbaier, L.H., Kiefe, C.I., Kresowik, T., and Peterson, E.D., “Effect of clinical factors on length of stay after coronary artery bypass surgery: Results of the cooperative cardiovascular project,” American Heart Journal 138, 69-77, 1999.Google Scholar
  35. 35.
    SAS Institute, Inc. SAS Version 8. Cary, NC, 1999.Google Scholar
  36. 36.
    Tu, J.V. and Rothwell, D.M., Cardiac Surgery in Ontario during fiscal 1998: An update to the Steering Committee of the Cardiac Care Network (CCN) of Ontario. Institute for Clinical Evaluative Sciences: Toronto, Ontario, 2000.Google Scholar
  37. 37.
    Van Beynum, I.M., Smeitink, J.A., den Heijer, M., te Poele, M.T., and Blom, H.J., “Hyperhomocysteinemia: A risk factor for ischemic stroke in children,” Circulation 99, 2070-2072, 1999.Google Scholar
  38. 38.
    Wee, J.Y., Bagg, S.D., and Palepu, A., “The berg balance scale as a predictor of length of stay and discharge destination in an acute stroke rehabilitation setting,” Archives of Physical Medicine &; Rehabilitation 80, 448-452, 1999.Google Scholar

Copyright information

© Kluwer Academic Publishers 2002

Authors and Affiliations

  • Peter C. Austin
    • 1
    • 2
    Email author
  • Deanna M. Rothwell
    • 1
  • Jack V. Tu
    • 1
    • 2
    • 3
  1. 1.Institute for Clinical Evaluative SciencesG1 06TorontoCanada;
  2. 2.Department of Public Health SciencesUniversity of TorontoCanada
  3. 3.Clinical Epidemiology and Health Care Research Program (Sunnybrook & Women's College Site); Division of General Internal Medicine, Department of Medicine, Sunnybrook & Women's College Health Sciences CentreUniversity of TorontoCanada

Personalised recommendations