Abstract
A multi-state Markov model is calibrated to Austrian data on recipients of long-term care payments the amount of which depends on defined frailty state levels. In contrast to a predecessor paper by one of the authors (see Fleischmann in Eur Actuar J 5(2):327–354, 2015), we are able to allow for different mortality intensities for different frailty states. A correction term is introduced in the mortality intensities’ functional representation to deal with observed mortality humps around the retirement age for certain frailty levels. Parameter calibration is done using MCMC methods (adaptive Metropolis–Hastings-within-Gibbs). The results reveal a considerably better fit of refined to raw prevalence units than the original model of Fleischmann (Eur Actuar J 5(2):327–354, 2015). Finally, the results are used to estimate the remaining healthy lifetime for certain ages, indicating slight but significant increases over the last 4 years.
Similar content being viewed by others
Notes
We requested in the specification of the data report, that a person deceased during a year was reported with frailty level i at the beginning of the year.
These units are obtained from comparing the values of ages x and \(x+1\) for a given year.
“Acceptance rate” refers to the number of steps in the algorithm in which the candidate value from the proposal distribution is accepted vs. the number of total steps.
Note that 2014 is the earliest year for which such a comparison is possible due to the restrictions on the data for years before 2014, cf. the discussion in the introduction.
References
Akaike H (1974) A new look at the statistical model identification. IEEE Trans Autom Control 19(6):716–723. https://doi.org/10.1109/TAC.1974.1100705
Biessy G (2017) Continuous time semi-Markov inference of biometric laws associated with a Long-Term Care Insurance portfolio. ASTIN Bull 47:527–561. https://doi.org/10.1017/asb.2016.41
Brooks SP, Roberts GO (1998) Convergence assessment techniques for Markov chain Monte Carlo. Stat Comput 8(4):319–335. https://doi.org/10.1023/A:1008820505350
Esquível ML, Guerreiro GR, Oliveira MC, Corte Real P (2021) Calibration of transition intensities for a multistate model: application to long-term care. Risks. https://doi.org/10.3390/risks9020037
Fleischmann A (2015) Calibrating intensities for long-term care multiple-state Markov insurance model. Eur Actuar J 5(2):327–354. https://doi.org/10.1007/s13385-015-0117-4
Fong JH, Sherris M, Yap J (2017) Forecasting disability: application of a frailty model. Scand Actuar J 2:125–147. https://doi.org/10.1080/03461238.2015.1092168
Fries JF, Bruce B, Chakravarty E (2011) Compression of morbidity 1980–2011: a focused review of paradigms and progress. J Aging Res 2011:1–10. https://doi.org/10.4061/2011/261702
Fuino M, Wagner J (2018) Long-term care models and dependence probability tables by acuity level: new empirical evidence from Switzerland. Insur Math Econ 81:51–70. https://doi.org/10.1016/j.insmatheco.2018.05.002
Geman S, Geman D (1984) Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE Trans Pattern Anal Mach Intell 6(6):721–741. https://doi.org/10.1109/TPAMI.1984.4767596
Greville TNE (1956) Laws of mortality which satisfy a uniform seniority principle. J Inst Actuar 82(1):114–122. https://doi.org/10.1017/S0020268100046205
Hastings WK (1970) Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57(1):97–109. https://doi.org/10.1093/biomet/57.1.97
Heligman L, Pollard JH (1980) The age pattern of mortality. J Inst Actuar 107(1):49–80. https://doi.org/10.1017/S0020268100040257
Hellweger V, Fischer JT, Kofler A, Huber A, Fellin W, Oberguggenberger M (2016) Stochastic methods in operational avalanche simulation - from back calculation to prediction. In: Proceedings, international snow science workshop, Breckenridge, Colorado, Montana State University Library, pp 1375–1381
Hirz J (2015) Advanced conditional risk measurement and risk aggregation with applications to credit and life insurance. PhD thesis, Vienna University of Technology
Iman R, Helton J, Campbell J (1981a) An approach to sensitivity analysis of computer models: part I-introduction, input variable selection and preliminary variable assessment. J Qual Technol 13(3):174–183. https://doi.org/10.1080/00224065.1981.11978748
Iman R, Helton J, Campbell J (1981b) An approach to sensitivity analysis of computer models: part II-ranking of input variables, response surface validation, distribution effect and technique synopsis. J Qual Technol 13(4):232–240. https://doi.org/10.1080/00224065.1981.11978763
Koller M (2010) Stochastische Modelle in der Lebensversicherung. Springer, Berlin. https://doi.org/10.1007/978-3-642-11252-2
Manton KG, Gu X (2001) Changes in the prevalence of chronic disability in the United States black and nonblack population above age 65 from 1982 to 1999. Proc Natl Acad Sci 98(11):6354–6359. https://doi.org/10.1073/pnas.111152298
McKay MD, Beckman RJ, Conover WJ (1979) A comparison of three methods for selecting values of input variables in the analysis of output from a computer code. Technometrics 21(2):239–245. https://doi.org/10.2307/1268522
Metropolis N, Rosenbluth AW, Rosenbluth MN, Teller AH, Teller E (1953) Equation of state calculations by fast computing machines. J Chem Phys 21(6):1087–1092. https://doi.org/10.1063/1.1699114
Nummelin E (1984) General irreducible Markov chains and non-negative operators. Cambridge Tracts in Mathematics, Cambridge University Press. https://doi.org/10.1017/CBO9780511526237
Oberguggenberger M (2014) Sensitivity and reliability analysis of engineering structures: sampling based methods. In: Hofstetter G (ed) Computational engineering. Springer, Cham. https://doi.org/10.1007/978-3-319-05933-4_4
Roberts GO, Rosenthal JS (2007) Coupling and ergodicity of adaptive Markov chain Monte Carlo algorithms. J Appl Probab 44(2):458–475. https://doi.org/10.1239/jap/1183667414
Roberts GO, Smith AFM (1994) Simple conditions for the convergence of the Gibbs sampler and Metropolis-Hastings algorithms. Stoch Process Appl 49(2):207–216. https://doi.org/10.1016/0304-4149(94)90134-1
Gelman A, Gilks WR, Roberts GO (1997) Weak convergence and optimal scaling of random walk Metropolis algorithms. Ann Appl Probab 7(1):110–120. https://doi.org/10.1214/aoap/1034625254
Rosenthal JS (2011) Optimal proposal distributions and adaptive MCMC. In: Handbook of Markov chain Monte Carlo, chap 4. CRC Press. http://probability.ca/jeff/ftpdir/galinart.pdf
Schwarz G (1978) Estimating the dimension of a model. Ann Stat 6(2):461–464. https://doi.org/10.1214/aos/1176344136
Serfling RE (1963) Methods for current statistical analysis of excess pneumonia-influenza deaths. Public Health Rep 78(6):494–506. https://doi.org/10.2307/4591848
Stallard E (2016) Compression of morbidity and mortality: new perspectives. N Am Actuar J 20(4):341–354. https://doi.org/10.1080/10920277.2016.1227269
Thiele TN, Sprague TB (1871) On a mathematical formula to express the rate of mortality throughout the whole of life. J Inst Actuar Assur Mag 16(5):313–329. https://doi.org/10.1017/S2046167400043688
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Table 8 contains the numbers of recipients of long-term care benefits as of January 2018 by frailty level v and age x. Table 9 contains numbers of fatalities by frailty level v and age x for 2018. The data in these tables is provided by Dachverband der Sozialversicherungsträger, an umbrella association of Austrian social security institutions. Table 10 is the Austrian unisex mortality table for 2018. Table 11 contains average Austrian population numbers for 2018. The data in these tables is provided by Statistik Austria.
Rights and permissions
About this article
Cite this article
Fleischmann, A., Hirz, J. & Sirianni, D. A long-term care multi-state Markov model revisited: a Markov chain Monte Carlo approach. Eur. Actuar. J. 12, 215–247 (2022). https://doi.org/10.1007/s13385-021-00285-y
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13385-021-00285-y