Multilevel Poisson sample selection models and alternative methods for estimating hospital effects on long-term outcomes

  • Kyle M. Kepreos
  • Wyndy L. Wiitala
  • Anne E. Sales
  • Hallie C. Prescott
  • Theodore J. Iwashyna


Hospital care focuses on improving patients’ long-term quality of life, yet hospital quality metrics typically focus on short-term processes. Attempting to understand a patient’s long-term process introduces sample selection bias since patients must survive the hospitalization in order to observe post-hospitalization outcomes. As a result, proper analysis of long-term outcomes should account for clustering, due to the hierarchical structure of hospital data, as well as sample selection bias. The objective of this paper was to evaluate random effect parameter estimation and higher-level ranking of long-term count outcomes and short-term selection processes in the presence of cluster and selection bias by comparing multilevel Poisson models, multilevel zero-inflated Poisson models, and multilevel Poisson sample selection models (MPSSMs) in a series of simulations. We simulated an outcome resembling a post-discharge Poisson count with a pre-specified selection process determining a patient’s hospitalization survival with each hospital having a unique effect on both processes. In order to clarify the methodology, we also analyzed a real-world hospital dataset involving a count outcome conditioned on the selection process of hospital survival. Across all simulations, the random effect parameter estimates were directly compared and the empirical Bayes estimates were extracted, ranked, and compared using the Spearman rank correlation. Results show that the MPSSM produces more accurate random effect parameter estimates and higher-level empirical Bayes ranks. When modeling multilevel effects on long-term count outcomes observed after a short-term selection process, higher-level effects are more reliably measured using MPSSMs.


Selection bias Multilevel Poisson models Multilevel zero-inflated Poisson models Multilevel Poisson sample selection models Empirical Bayes ranking 



This work was supported by VA HSR&D IIR 11-109. The authors’ opinions do not necessarily represent the views of the US Department of Veterans Affairs or the US Government.

Supplementary material

10742_2015_139_MOESM1_ESM.pdf (661 kb)
Supplementary material 1 (PDF 660 kb)


  1. Bärnighausen, T., Bor, J., Wandira-Kazibwe, S., Canning, D.: Correcting HIV prevalence estimates for survey nonparticipation using Heckman-type selection models. Epidemiology 22(1), 27–35 (2011). doi: 10.1097/EDE.0b013e3181ffa201 CrossRefPubMedGoogle Scholar
  2. Buntin, M.B., Zaslavsky, A.M.: Too much ado about two-part models and transformation? Comparing methods of modeling Medicare expenditures. J Health Econ. 23(3), 525–542 (2004). doi: 10.1016/j.jhealeco.2003.10.005 CrossRefPubMedGoogle Scholar
  3. Cooper, D.J., Rosenfeld, J.V., Murray, L., Arabi, Y.M., Davies, A.R., D’Urso, P., Kossmann, T., Ponsford, J., Seppelt, I., Reilly, P., Wolfe, R., DECRA Trial Investigators, Australian and New Zealand Intensive Care Society Clinical Trials Group: Decompressive craniectomy in diffuse traumatic brain injury. N. Engl. J. Med. 364(16), 1493–1502 (2011). doi: 10.1056/NEJMoa1102077 CrossRefPubMedGoogle Scholar
  4. Elixhauser, A., Steiner, C., Harris, D.R., Coffey, R.M.: Comorbidity measures for use with administrative data. Med. Care 36(1), 8–27 (1998)CrossRefPubMedGoogle Scholar
  5. Flores-Lagunes, A., Schnier, K.E.: Estimation of sample selection models with spatial dependence. J. Appl. Econom. 27(2), 173–204 (2012). doi: 10.1002/jae.1189 CrossRefGoogle Scholar
  6. Hagihara, A., Hasegawa, M., Abe, T., Nagata, T., Wakata, T., Miyazaki, S.: Prehospital epinephrine use and survival among patients with out-of-hospital cardiac arrest. JAMA 307(11), 1161–1168 (2012). doi: 10.1001/jama.2012.294 CrossRefPubMedGoogle Scholar
  7. Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning—Data Mining, Inference, and Prediction, 2nd edn. Springer, New York (2009)Google Scholar
  8. Heckman, J.J.: Dummy endogenous variables in a simultaneous equation system. Econometrica 46, 931–959 (1978)CrossRefGoogle Scholar
  9. Hollander, M., Wolfe, D.A.: Nonparametric Statistical Methods, 2nd edn. Wiley, New York (1999)Google Scholar
  10. Iwashyna, T.J., Angus, D.C.: Declining case fatality rates for severe sepsis: good data bring good news with ambiguous implications. JAMA 311(13), 1295–1297 (2014). doi: 10.1001/jama.2014.2639 CrossRefPubMedGoogle Scholar
  11. Iwashyna, T.J., Ely, E.W., Smith, D.M., et al.: Long-term cognitive impairment and functional disability among survivors of severe sepsis. J. Am. Med. Assoc. 304(16), 1787–1794 (2010). doi: 10.1001/jama.2010.1553 CrossRefGoogle Scholar
  12. Iwashyna, T.J., Odden, A., Rohde, J., Bonham, C., Kuhn, L., Malani, P., Chen, L., Flanders, S.: Identifying patients with severe sepsis using administrative claims: patient-level validation of the angus implementation of the international consensus conference definition of severe sepsis. Med. Care 52(6), e39–e43 (2014). doi: 10.1097/MLR.0b013e318268ac86 CrossRefPubMedCentralPubMedGoogle Scholar
  13. Lachenbruch, P.: Analysis of data with excess zeros. Stat. Methods Med. Res. 11, 297–302 (2002). doi: 10.1191/0962280202sm289ra CrossRefPubMedGoogle Scholar
  14. Marra, G., Radice, R.: Estimation of a regression spline sample selection model. Comput. Stat. Data Anal. 61, 158–173 (2013). doi: 10.1016/j.csda.2012.12.010 CrossRefGoogle Scholar
  15. Mikkelsen, M.E., Christie, J.D., Lanken, P.N., Biester, R.C., Thompson, B.T., Bellamy, S.L., Localio, A.R., Demissie, E., Hopkins, R.O., Angus, D.C.: The adult respiratory distress syndrome cognitive outcomes study: long-term neuropsychological function is survivors of acute lung injury. Am. J. Respir. Crit. Care Med. 185(12), 1307–1315 (2012). doi: 10.1164/rccm.201111-2025OC CrossRefPubMedCentralPubMedGoogle Scholar
  16. Miranda, A., Rabe-Hesketh, S.: Maximum likelihood estimation of endogenous switching and sample selection models for binary, ordinal, and count variables. Stata J. 6(3), 285–308 (2006)Google Scholar
  17. Molas, M., Lesaffre, E.: Hurdle models for multilevel zero-inflated data via h-likelihood. Stat. Med. 29(30), 3294–3310 (2010). doi: 10.1002/sim.3852 CrossRefPubMedGoogle Scholar
  18. Morris, C.: Parametric empirical Bayes inference: theory and applications. J. Am. Stat. Assoc. 78(381), 47–55 (1983)CrossRefGoogle Scholar
  19. Prescott, H.C., Kepreos, K.M., Wiitala, W.L., Iwashyna, T.J.: Temporal changes in the influence of hospitals and regional healthcare networks on severe sepsis mortality. Crit. Care Med. (2015). doi: 10.1097/CCM.0000000000000970 Google Scholar
  20. Rabe-Hesketh, S., Skrondal, A., Pickles, A.: GLLAMM Manual. University of California-Berkeley Division of Biostatistics Working Paper Series. Working Paper 160. (2004). Accessed 2 July 2014
  21. Render, M.L., Deddens, J., Freybery, R., et al.: Veterans affairs intensive care unit risk adjustment model: validation, updating, recalibration. Crit. Care Med. 36(4), 1031–1042 (2008). doi: 10.1097/CCM.0b013e318169f290 CrossRefPubMedGoogle Scholar
  22. Semykina, A., Wooldridge, J.A.: Estimation of dynamic panel data models with sample selection. J. Appl. Econ. 28(1), 47–61 (2013). doi: 10.1002/jae.1266 CrossRefGoogle Scholar
  23. Skrondal, A., Rabe-Hesketh, S.: Prediction in multilevel generalized linear models. J R Stat Soc 172(Part 3), 659–687 (2009)CrossRefGoogle Scholar
  24. Tomlinson, G., Detsky, A.: Composite end points in randomized trials: there is no free lunch. J. Am. Stat. Assoc. 303(3), 267–268 (2010). doi: 10.1001/jama.2009.2017 CrossRefGoogle Scholar
  25. Wilde, J.: Identification of multiple equation probit models with endogenous dummy regressors. Econ. Lett. 63(3), 309–312 (2000). doi: 10.1016/S0165-1765(00)00320-7 CrossRefGoogle Scholar
  26. Yuan, Y., Yen, S.T.: Alcohol consumption by individuals in the United States: a sample selection approach. Appl. Econ. Lett. 19(14), 1353–1358 (2012). doi: 10.1080/13504851.2011.628290 CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York (outside the USA) 2015

Authors and Affiliations

  • Kyle M. Kepreos
    • 1
  • Wyndy L. Wiitala
    • 1
  • Anne E. Sales
    • 1
    • 2
  • Hallie C. Prescott
    • 3
  • Theodore J. Iwashyna
    • 1
    • 3
  1. 1.VA Center for Clinical Management ResearchAnn Arbor VA Health Services Research and DevelopmentAnn ArborUSA
  2. 2.University of Michigan School of NursingAnn ArborUSA
  3. 3.Division of Pulmonary and Critical Care MedicineUniversity of MichiganAnn ArborUSA

Personalised recommendations