Abstract
There is strong evidence that diabetes is underdiagnosed in the US: the Centers for Disease Control and Prevention (CDC) estimates that approximately 25% of diabetic patients are unaware of their condition. To encourage timely diagnosis of at-risk patients, we develop screening guidelines stratified by body mass index (BMI), age, and prior test history by using a Partially Observed Markov Decision Process (POMDP) framework to provide more personalized screening frequency recommendations. We identify structural results that prove the existence of threshold solutions in our problem and allow us to determine the relative timing and frequency of screening given different risk profiles. We then use nationally representative empirical data to identify a policy that provides the optimal action (screen or wait) every six months from age 45 to 90. We find that the current screening guidelines are suboptimal, and the recommended diabetes screening policy should be stratified by age and by finer BMI thresholds than in the status quo. We identify age ranges and BMI categories for which relatively less or more screening is needed compared to the existing guidelines to help physicians target patients most at risk. Compared to the status quo, we estimate that an optimal screening policy would generate higher net monetary benefits by $3,200-$3,570 and save $120-$1,290 in health expenditures per individual in the US above age 45.
Similar content being viewed by others
References
American Diabetes Association (2018) Economic costs of diabetes in the US in 2017. Diabetes Care 41(5):917–928
American Diabetes Association (2019) 2. Classification and diagnosis of diabetes: Standards of medical care in diabetes—2019. Diabetes Care 42(Supplement 1):S13–S28
Papatheodorou K, Banach M, Bekiari E, Rizzo M, Edmonds M (2018) Complications of diabetes 2017. Journal of Diabetes Research
Heron M (2018) Deaths: Leading causes for 2016. In: National Vital Statistics Reports: From the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System, vol 67.6, pp 1–77
Chen R, Ovbiagele B, Feng W (2016) Diabetes and stroke: Epidemiology, pathophysiology, pharmaceuticals and outcomes. Amer J Med Sci 351(4):380–386
Leon BM, Maddox TM (2015) Diabetes and cardiovascular disease: Epidemiology, biological mechanisms, treatment recommendations and future research. World J Diabetes 6(13):1246
Saran R, Robinson B, Abbott KC, Agodoa Lawrence YC, Bhave N, Bragg-Gresham J, Balkrishnan R, Dietrich X, Eckard A, Eggers PW et al (2018) US Renal Data System 2017 annual data report: Epidemiology of kidney disease in the United States. Am J Kidney Dis 71(3 Suppl 1):A7
Nair P, Aiello LP, Gardner TW, Jampol LM, Ferris III FL (2016) Report from the NEI/FDA diabetic retinopathy clinical trial design and endpoints workshop. Invest Ophthalmol Vis Sci 57(13):5127–5142
Riddle MC, Herman WH (2018) The cost of diabetes care—an elephant in the room. Diabetes Care 41(5):929–932
Whitley HP, Hanson C, Parton JM (2017) Systematic diabetes screening using point-of-care HbA1c testing facilitates identification of prediabetes. Ann Family Med 15(2):162–164
Ley SH, Ardisson Korat AV, Sun Q, Tobias DK, Zhang C, Qi L, Willett WC, Manson JE, Hu FB (2016) Contribution of the Nurses’ Health Studies to uncovering risk factors for type 2 diabetes: Diet, lifestyle, biomarkers, and genetics. Am J Publ Health 106(9):1624–1630
Flegal KM, Kruszon-Moran D, Carroll MD, Fryar CD, Ogden CL (2016) Trends in obesity among adults in the United States, 2005 to 2014. J Am Med Assoc 315(21):2284–2291
Güneş ED, Örmeci EL (2018) OR applications in disease screening. In: Operations Research Applications in Health Care Management. Springer, pp 297–325
Steimle LN, Denton BT (2017) Markov decision processes for screening and treatment of chronic diseases. In: Markov Decision Processes in Practice. Springer, pp 189–222
Alagoz O (2011) Optimizing cancer screening using partially observable Markov decision processes. In: Transforming Research into Action. INFORMS, pp 75–89
Suen S-, Brandeau ML, Goldhaber-Fiebert JD (2018) Optimal timing of drug sensitivity testing for patients on first-line tuberculosis treatment. Health Care Manag Sci 21(4):632–646
Ahuja K, Zame W, van der Schaar M (2017) DPSCREEN: Dynamic personalized screening. In: Advances in Neural Information Processing Systems, pp 1321–1332
Pace LE, Keating NL (2014) A systematic assessment of benefits and risks to guide breast cancer screening decisions. J Am Med Assoc 311(13):1327–1335
Shechter SM, Bailey MD, Schaefer AJ, Roberts MS (2008) The optimal time to initiate HIV therapy under ordered health states. Oper Res 56(1):20–33
Maillart LM, Ivy JS, Ransom S, Diehl K (2008) Assessing dynamic breast cancer screening policies. Oper Res 56(6):1411–1427
Ayer T, Alagoz O, Stout NK (2012) OR Forum—A POMDP approach to personalize mammography screening decisions. Oper Res 60(5):1019–1034
Zhang J, Denton BT, Balasubramanian H, Shah ND, Inman BA (2012) Optimization of prostate biopsy referral decisions. Manuf Serv Oper Manag 14(4):529–547
Sandıkçı B, Maillart LM, Schaefer AJ, Roberts MS (2013) Alleviating the patient’s price of privacy through a partially observable waiting list. Manag Sci 59(8):1836–1854
Li J, Dong M, Ren Y, Yin K (2015) How patient compliance impacts the recommendations for colorectal cancer screening. J Comb Optim 30(4):920–937
Zhang J, Denton BT, Balasubramanian H, Shah ND, Inman BA (2012) Optimization of PSA screening policies: A comparison of the patient and societal perspectives. Med Dec Making 32(2):337–349
Denton BT, Kurt M, Shah ND, Bryant SC, Smith SA (2009) Optimizing the start time of statin therapy for patients with diabetes. Med Dec Making 29(3):351–367
Kurt M, Denton BT, Schaefer AJ, Shah ND, Smith SA (2011) The structure of optimal statin initiation policies for patients with type 2 diabetes. IIE Trans Healthcare Syst Eng 1(1):49–65
Mason JE, Denton BT, Shah ND, Smith SA (2014) Optimizing the simultaneous management of blood pressure and cholesterol for type 2 diabetes patients. Eur J Oper Res 233(3):727–738
Handelsman Y, Bloomgarden ZT, Grunberger G, Umpierrez G, Zimmerman RS, Bailey TS, Blonde L, Bray GA, Cohen AJ, Dagogo-Jack S et al (2015) American Association of Clinical Endocrinologists and American College of Endocrinology–clinical practice guidelines for developing a diabetes mellitus comprehensive care plan–2015. Endocr Pract 21(s1):1–87
Kirkman MS, Briscoe VJ, Clark N, Florez H, Haas LB, Halter JB, Huang ES, Korytkowski MT, Munshi MN, Odegard PS et al (2012) Diabetes in older adults. Diabetes Care 35(12):2650–2664
Siu AL (2015) Screening for abnormal blood glucose and type 2 diabetes mellitus: US Preventive Services Task Force recommendation statement. Ann Intern Med 163(11):861–868
Waugh N, Scotland G, Gillet M, Brennan A, Goyder E, Williams R, John A (2007) Screening for type 2 diabetes: Literature review and economic modelling. Health Technol Assess
O’Brien MJ, Bullard KM, Zhang Y, Gregg EW, Carnethon MR, Kandula NR, Ackermann RT (2018) Performance of the 2015 US Preventive Services Task Force screening criteria for prediabetes and undiagnosed diabetes. J Gen Intern Med 33(7):1100–1108
Fritzen K, Heinemann L, Schnell O (2018) Modeling of diabetes and its clinical impact. J Diabetes Sci Technol 12(5):976–984
Kent S, Becker F, Feenstra T, Tran-Duy A, Schlackow I, Tew M, Zhang P, Ye W, Lizheng S, Herman W, McEwan P, Schramm W, Gray A, Leal J, Lamotte M, Willis M, Palmer AJ, Clarke P (2019) The challenge of transparency and validation in health economic decision modelling: A view from Mount Hood. PharmacoEconomics
Leal J, Morrow LM, Khurshid W, Pagano E, Feenstra T (2019) Decision models of prediabetes populations: A systematic review. Diabetes, Obesity and Metabolism
Palmer AJ, Si L, Tew M, Hua X, Willis MS, Asseburg C, McEwan P, Leal J, Gray A, Foos V et al (2018) Computer modeling of diabetes and its transparency: A report on the eighth Mount Hood challenge. Value Health 21(6):724–731
Sherwani SI, Khan HA, Ekhzaimy A, Masood A, Sakharkar MK (2016) Significance of HbA1c test in diagnosis and prognosis of diabetic patients. Biomark Insights 11:BMI–S38440
Weir CB, Jan A (2019) Bmi classification percentile and cut off points. In: StatPearls [Internet]. StatPearls Publishing
Shimodaira M, Okaniwa S, Hanyu N, Nakayama T (2015) Optimal hemoglobin A1c levels for screening of diabetes and prediabetes in the Japanese population. Journal of Diabetes Research
Messori A, Trippoli S (2017) The results of a pharmacoeconomic study: Incremental cost-effectiveness ratio versus net monetary benefit. Heart 103(21):1746–1746
Trippoli S (2017) Incremental cost-effectiveness ratio and net monetary benefit: Current use in pharmacoeconomics and future perspectives. Eur J Internal Med 43:e36
Prieto L, Sacristán J A (2003) Problems and solutions in calculating quality-adjusted life years (QALYs). Health Qual Life Outcomes 1(1):80
Cohen DJ, Reynolds MR (2008) Interpreting the results of cost-effectiveness studies. J Am Coll Cardiol 52(25):2119–2126
Neumann PJ, Cohen JT, Weinstein MC (2014) Updating cost-effectiveness—the curious resilience of the $50,000-per-QALY threshold. N Engl J Med 371(9):796–797
Curl PK, Kahn JG, Ordovas KG, Elicker BM, Naeger DM (2015) Understanding cost-effectiveness analyses: An explanation using three different analyses of lung cancer screening. Am J Roentgenol 205 (2):344–347
Puterman ML (1994) Markov Decision Processes: Discrete Stochastic Dynamic Programming. Wiley
Sandikçi B (2010) Reduction of a POMDP to an MDP. Wiley Encyclopedia of Operations Research and Management Science
Shaked M, Shanthikumar JG (2007) Stochastic orders. Springer Science & Business Media
Barlow RE, Proschan F (1996) Mathematical Theory of Reliability. SIAM
Krishnamurthy V (2016) Partially Observed Markov Decision Processes : From Filtering to Controlled Sensing. Cambridge University Press
Sandıkçı B, Maillart LM, Schaefer AJ, Alagoz O, Roberts MS (2008) Estimating the patient’s price of privacy in liver transplantation. Oper Res 56(6):1393–1410
Erenay FS, Alagoz O, Said A (2014) Optimizing colonoscopy screening for colorectal cancer prevention and surveillance. Manuf Serv Oper Manag 16(3):381–400
Neumann PJ, Sanders GD, Russell LB, Siegel JE, Ganiats TG (2016) Cost-effectiveness in health and medicine. Oxford University Press
Hoerger TJ, Hicks KA, Sorensen SW, Herman WH, Ratner RE, Ackermann RT, Zhang P, Engelgau MM (2007) Cost-effectiveness of screening for pre-diabetes among overweight and obese us adults. Diabetes Care 30(11):2874–2879
Herman WH, Hoerger TJ, Brandle M, Hicks K, Sorensen S, Zhang P, Hamman RF, Ackermann RT, Engelgau MM, Ratner RE (2005) The cost-effectiveness of lifestyle modification or metformin in preventing type 2 diabetes in adults with impaired glucose tolerance. Ann Intern Med 142(5):323–332
Zhang P, Engelgau MM, Valdez R, Benjamin SM, Cadwell B, Narayan KMV (2003) Costs of screening for prediabetes among US adults: A comparison of different screening strategies. Diabetes Care 26(9):2536–2542
Dall TM, Yang W, Halder P, Pang B, Massoudi M, Wintfeld N, Semilla AP, Franz J, Hogan PF (2014) The economic burden of elevated blood glucose levels in 2012: diagnosed and undiagnosed diabetes, gestational diabetes mellitus, and prediabetes. Diabetes Care 37(12):3172–3179
Willey VJ, Kong S, Wu B, Raval A, Hobbs T, Windsheimer A, Deshpande G, Tunceli O, Sakurada B, Bouchard JR (2018) Estimating the real-world cost of diabetes mellitus in the United States during an 8-year period using 2 cost methodologies. American Health & Drug Benefits 11(6):310
Röckl S, Brinks R, Baumert J, Paprott R, Du Y, Heidemann C, Scheidt-Nave C (2017) All-cause mortality in adults with and without type 2 diabetes: Findings from the national health monitoring in Germany. BMJ Open Diabetes Res Care 5(1):e000451
Sonnega A, Faul JD, Ofstedal MB, Langa KM, Phillips JWR, Weir DR (2014) Cohort profile: The Health and Retirement Study (HRS). Int J Epidemiol 43(2):576–585
Health and Retirement Study Survey Research Center (2017) Aging in the 21st Century: Challenges and Opportunities for Americans. Institute for Social Research, University of Michigan, Ann Arbor
Fisher GG, Ryan LH (2017) Overview of the Health and Retirement Study and introduction to the special issue. Work Aging Retirement 4(1):1–9
Chhatwal J, Jayasuriya S, Elbasha EH (2016) Changing cycle lengths in state-transition models: Challenges and solutions. Med Dec Making 36(8):952–964
Letter TM (2017) Metformin for prediabetes. J Am Med Assoc 317(11):1171–1171
Hayes AJ, Leal J, Gray AM, Holman RR, Clarke PM (2013) UKPDS outcomes model 2: A new version of a model to simulate lifetime health outcomes of patients with type 2 diabetes mellitus using data from the 30 year United Kingdom Prospective Diabetes Study: UKPDS 82. Diabetologia 56(9):1925–1933
Guo F, Moellering DR, Garvey WT (2014) Use of HbA1c for diagnoses of diabetes and prediabetes: Comparison with diagnoses based on fasting and 2-hr glucose values and effects of gender, race, and age. Metab Syndr Relat Disord 12(5):258–268
Karnchanasorn R, Huang J, Ou H-Y, Feng W, Chuang L-M, Chiu KC, Samoa R (2016) Comparison of the current diagnostic criterion of HbA1c with fasting and 2-hour plasma glucose concentration. Journal of Diabetes Research
Cawley J, Meyerhoefer C, Biener A, Hammer M, Wintfeld N (2015) Savings in medical expenditures associated with reductions in Body Mass Index among US adults with obesity, by diabetes status. Pharmacoeconomics 33(7):707–722
Sturm R, Hattori A (2013) Morbid obesity rates continue to rise rapidly in the united states. Int J Obes 37(6):889–891
Agha M, Agha R (2017) The rising prevalence of obesity: part a impact on public health. Int J Surg Oncol 2(7)
Schnell O, Crocker JB, Weng J (2017) Impact of hba1c testing at point of care on diabetes management. J Diabetes Sci Technol 11(3):611–617
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Ethics Approval
No ethics review board approval was needed for the data used in this work, as all data was taken from published articles or publicly available, de-identified datasets.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix:
A Proofs
A.1 Observations on TP2 matrices
We first present properties to be used in Lemmas 1 and 2. We define the determinant of a two by two submatrix of a matrix A to be:
Property 1 For two nonnegative vectors with the first vector ordered from largest to smallest and the second one ordered from smallest to largest, the determinant of any two by two submatrix is nonnegative.
Proof
For any two column vectors where \(a_{i_{1}j_{1}} \geq a_{i_{2}j_{1}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{1}j_{2}} \geq 0\), then \(a_{i_{1}j_{1}}a_{i_{2}j_{2}} - a_{i_{1}j_{2}}a_{i_{2}j_{1}} \geq 0\). Similarly, for any two row vectors where \(a_{i_{1}j_{1}} \geq a_{i_{1}j_{2}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{2}j_{1}} \geq 0\), we have \(a_{i_{1}j_{1}}a_{i_{2}j_{2}} - a_{i_{1}j_{2}}a_{i_{2}j_{1}}\) ≥ 0. □
Property 2 For any nonnegative two by two matrix, if \(a_{i_{1}j_{2}}\) or \(a_{i_{2}j_{1}}\) = 0, then the determinant is nonnegative.
Proof
We have have \(a_{i_{1}j_{1}}, a_{i_{1}j_{2}}, a_{i_{2}j_{1}}, a_{i_{2}j_{2}} \geq 0\) and \(a_{i_{1}j_{2}}\) or \(a_{i_{2}j_{1}}\) = 0. Thus, \(a_{i_{1}j_{1}}a_{i_{2}j_{2}} - a_{i_{1}j_{2}}a_{i_{2}j_{1}} = a_{i_{1}j_{1}}a_{i_{2}j_{2}} \geq 0\). □
Property 3 For any nonnegative two by two matrix, if \(a_{i_{1}j_{1}} \geq a_{i_{2}j_{1}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{1}j_{2}} \geq 0\) or \(a_{i_{1}j_{1}} \geq a_{i_{1}j_{2}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{2}j_{1}} \geq 0\), then the determinant is nonnegative.
Proof
In the first case, we have have \(a_{i_{1}j_{1}}, a_{i_{1}j_{2}},a_{i_{2}j_{1}},\) \( a_{i_{2}j_{2}} \geq 0\) and \(a_{i_{1}j_{1}} \geq a_{i_{2}j_{1}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{1}j_{2}} \geq 0\). Thus, \(a_{i_{1}j_{1}}a_{i_{2}j_{2}} - a_{i_{1}j_{2}}a_{i_{2}j_{1}} \geq 0\). In the second case, we have \(a_{i_{1}j_{1}}, a_{i_{1}j_{2}}, a_{i_{2}j_{1}}, a_{i_{2}j_{2}} \geq 0\) and \(a_{i_{1}j_{1}} \geq a_{i_{1}j_{2}} \geq 0\) and \(a_{i_{2}j_{2}} \geq a_{i_{2}j_{1}} \geq 0\). Thus, \(a_{i_{1}j_{1}}a_{i_{2}j_{2}} - a_{i_{1}j_{2}}a_{i_{2}j_{1}} \geq 0\). □
Lemma 1
If the following conditions hold, then in a problem with three health states (orderable from the healthiest to sickest) and observable death (the fourth health state), the observation matrix, Oa, is TP2 given action a.
-
1.
Observation results have a greater than 50% chance to coincide with the true health state.
-
2.
Oa has the IFR property.
-
3.
\(q^{{a}}_{12} \geq q^{{a}}_{13}\) and \(q^{{a}}_{32} \geq q^{{a}}_{31}\)
-
4.
\(q^{{a}}_{12}q^{{a}}_{23} \geq q^{{a}}_{13}q^{{a}}_{22}\) and \(q^{{a}}_{21}q^{{a}}_{32} \geq q^{{a}}_{22}q^{{a}}_{31}\).
A.2 Proof of Lemma 1
The death state is observable, so our observation matrix, Oa, is fully captured by a three by three matrix. For a matrix to be TP2, all determinants of second-order submatrices need to be greater than or equal to 0. With three combinations of columns and rows, we have nine submatrices.
-
Conditions 1-3 ensure that \(q^{{a}}_{11}\geq q^{{a}}_{21} \geq q^{{a}}_{31}\) and \(q^{{a}}_{33}\geq q^{{a}}_{23} \geq q^{{a}}_{13}\). Property 1 then tells us submatrices formed from columns 1 and 3 are non-negative.
-
Conditions 1-3 also ensure that \(q^{{a}}_{11}\geq q^{{a}}_{12}\), \(q^{{a}}_{11}\geq q^{{a}}_{21}\), \(q^{{a}}_{22} \geq q^{{a}}_{12}\), and \(q^{{a}}_{22} \geq q^{a}_{21}\). Property 3 then ensures that \(q^{a}_{11}q^{a}_{22} - q^{a}_{12}q^{a}_{21} \geq 0\) and \(q^{a}_{22}q^{a}_{33} - q^{a}_{23}q^{a}_{32} \geq 0\) must be satisfied.
-
Condition 4 ensures all remaining submatrices are non-negative: \(q^{{a}}_{12}q^{{a}}_{23} - q^{{a}}_{13}q^{{a}}_{22} \!\geq \! 0\) and \(q^{{a}}_{21}q^{{a}}_{32} - q^{a}_{22}q^{a}_{31} \!\geq \) 0.
Therefore determinants of all second-order submatrices are nonnegative.
The TP2 property enables us to order the health states as we observe test results over time. These conditions specify what it means to be a “good” test, where test results are more likely to be “close” to the true state than otherwise, in a context with multiple health states. This means that in a case with a good diagnosis test, one only needs to verify condition four to establish that the observations are TP2. We next turn to establishing similar conditions for the transition matrix, Pt,BMI.
Lemma 2
If the following conditions hold, then in a problem with three health states (orderable from the healthiest to sickest) and death (the fourth health state), the transition matrix, Pt,BMI, is TP2 given time t and BMI category BMI.
-
1.
Once you are in the sickest (non-dead) state, one cannot recover to healthier states.
-
2.
Health states are more than 50% likely to stay the same between transitions.
-
3.
Pt,BMI has the IFR property.
-
4.
\(p^{t,BMI}_{23} \geq p^{t,BMI}_{13}\).
-
5.
\(p^{t,BMI}_{12} \geq p^{t,BMI}_{13} \geq p^{t,BMI}_{14}\) and \(p^{t,BMI}_{23} \geq p^{t,BMI}_{24}\)
-
6.
\(p^{t,BMI}_{i_{1}j_{1}}p^{t,BMI}_{i_{2}j_{2}} \geq p^{t,BMI}_{i_{1}j_{2}}p^{t,BMI}_{i_{2}j_{1}}\) where i1,i2 ∈{1, 2, 3} and j1,j2 ∈{2, 3, 4} ∀i2 ≥ i1,j2 ≥ j1
A.3 Proof of Lemma 2
For a matrix to be TP2, all determinants of second-order submatrices need to be greater than or equal to 0.
-
The death state is an absorbing state, so the last row of Pt,BMI is comprised of 0s and 1. Using Property 2, the determinants of all submatrices that include this row are greater or equal to zero.
-
Conditions 1-5 ensure that \(p^{t,BMI}_{11}\geq p^{t,BMI}_{21} \geq p^{t,BMI}_{31}\) and \(p^{t,BMI}_{32}\geq p^{t,BMI}_{22} \geq p^{t,BMI}_{12}\) and \(p^{t,BMI}_{33}\geq p^{t,BMI}_{23} \geq p^{t,BMI}_{13}\) and \(p^{t,BMI}_{34}\geq p^{t,BMI}_{24} \geq p^{t,BMI}_{14}\). Property 1 then tells us submatrices formed from columns 1 and any of the remaining columns are non-negative.
-
Condition 1 also ensures that \(p^{t,BMI}_{31} = p^{t,BMI}_{32} = 0\). Then \(p^{t,BMI}_{12}p^{t,BMI}_{33} - p^{t,BMI}_{13}p^{t,BMI}_{32} \geq 0\) and \(p^{t,BMI}_{22}p^{t,BMI}_{33} - p^{t,BMI}_{23}p^{t,BMI}_{32} \geq 0\) and \(p^{t,BMI}_{12}p^{t,BMI}_{34} - p^{t,BMI}_{14}p^{t,BMI}_{32} \geq 0\) and \(p^{t,BMI}_{22}p^{t,BMI}_{34} - p^{t,BMI}_{24}p^{t,BMI}_{32} \geq 0\) must be satisfied according to Property 2.
-
Condition 6 ensures all remaining submatrices are non-negative: \(p^{t,BMI}_{12}p^{t,BMI}_{23} - p^{t,BMI}_{13}p^{t,BMI}_{22} \geq 0\) and \(p^{t,BMI}_{12}p^{t,BMI}_{24} - p^{t,BMI}_{14}p^{t,BMI}_{22} \geq 0\) and \(p^{t,BMI}_{13}\) \(p^{t,BMI}_{24} - p^{t,BMI}_{14}p^{t,BMI}_{23} \geq 0\) and \(p^{t,BMI}_{13}p^{t,BMI}_{34} - p^{t,BMI}_{14}p^{t,BMI}_{33} \geq 0\) and \(p^{t,BMI}_{23}\) \( p^{t,BMI}_{34} -p^{t,BMI}_{24}\) \(p^{t,BMI}_{33} \geq 0\).
Therefore determinants of all second-order submatrices are nonnegative.
We will use this result to show when the optimal policy has a threshold solution. There are more conditions for the transition matrix to be TP2 than the observation matrix because we cannot exclude the death state (but we can in the observation analysis because it is fully observable). Evaluating TP2 directly for a four by four matrix requires thirty-six submatrices; here, we have reduced the conditions so only five submatrices need to be checked. These lemmas allow us to examine when a monotonic value function exists.
A.4 Proof of proposition 1
This problem is a special case studied by Krishnamurthy (2016), where our linear reward function satisfies the assumption that reward function is first-order stochastically decreasing with respect to belief πt for each action at [51]. To see this, define their Hidden Markov model (HMM) filter T(πt, Oa,a) and the normalization measure σ(πt,Oa,a) as:
Therefore, the proposition is true by Theorem 11.2.1 in Krishnamurthy (2016) [51].
A.5 Proof of Theorem 1
A threshold π∗,t can be defined as the π∗,t such that vt(π∗,t,at = a1) = vt(π∗,t,at = a2). We consider two cases: (1) there is no such threshold, or (2) there are one or more thresholds. For the case that there does not exist a threshold π∗,t, we can define the threshold to be [1 0 0 0] or [0 0 0 1] since one action dominates the other over the entire belief simplex (always screen or wait for any belief at time t). We therefore restrict our analysis to the case where there exists at least one π∗,t.
We use a proof by contradiction to show that there cannot be two or more thresholds given the conditions in Theorem 1. We start with an observation on the single threshold case.
Let us start with the single threshold (\(\pi ^{\ast ,t}_{1} \)) case. Recall that we have vt(πt,a1) − vt(πt,a2) is MLR decreasing in πt, by assumption. If the current belief \(\pi ^{t} \leq _{r} \pi ^{\ast ,t}_{1}\), we then must have:
Where the last equality is by the definition of threshold. Thus, we have vt(πt,a1) ≥ vt(πt,a2). Therefore, if \(\pi ^{t} \leq _{r} \pi ^{\ast ,t}_{1}\), then the optimal action is a1.
Similarly, if the current belief \(\pi ^{t} \geq _{r} \pi ^{\ast ,t}_{1}\), then we have:
Thus, we have vt(πt,a1) ≤ vt(πt,a2), so then the optimal action must be a2.
Therefore, if there is a single threshold, the optimal action must differ on either side of it. To show that two or more thresholds cannot exist simultaneously, we use this rationale in a proof by contradiction.
Suppose there are two thresholds: \(\pi ^{\ast ,t}_{1} \leq _{r} \pi ^{\ast ,t}_{2}\). From the observation above, we can see that if \(\pi ^{t} \leq _{r} \pi ^{\ast ,t}_{1}\), then the optimal action is a1, and if \(\pi ^{\ast ,t}_{2} \leq _{r} \pi ^{t}\), then the optimal action is a2.
Let \(\pi ^{\ast ,t}_{1} \leq _{r} \pi ^{t} \leq _{r} \pi ^{\ast ,t}_{2}\). By repeating the previous procedure we can see that if \(\pi ^{\ast ,t}_{1} \leq _{r} \pi ^{t}\), then the optimal action is a2, and if \(\pi ^{t} \leq _{r} \pi ^{\ast ,t}_{2}\), then the optimal action is a1. This is a contradiction since the MLR decreasing assumption is violated and you can not take two actions at the same time. This means that if the optimal action is a1 for \(\pi ^{\ast ,t}_{1} \leq _{r} \pi ^{t} \leq _{r} \pi ^{\ast ,t}_{2}\), then \(\pi ^{\ast ,t}_{2}\) is the single threshold. Otherwise, if the optimal action is a2, then \(\pi ^{\ast ,t}_{1}\) is the single threshold.
Following the same rationale, we see that there cannot be n thresholds for n > 1: \(\pi ^{\ast ,t}_{1} \leq _{r} \pi ^{\ast ,t}_{2} \leq _{r} ... \leq _{r} \pi ^{\ast ,t}_{n}\).
A.6 Proof of Theorem 2
We have monotonic value functions and unique threshold policies for both individuals, so using Theorem 1, we can rewrite our threshold solution: if \({v^{t}_{i}}({\pi ^{t}_{i}},a^{t} = a_{1}) \geq {v^{t}_{i}}(\pi ^{\ast ,t}_{i},a^{t} = a_{1})\) then a1, else a2. Note that by definition we have \({v^{t}_{i}}(\pi ^{\ast ,t}_{i},a^{t} = a_{1}) = {v^{t}_{i}}(\pi ^{\ast ,t}_{i},a^{t} = a_{2})\). Thus, \({v^{t}_{1}}({\pi ^{t}_{1}}, a^{t} = a_{1}) - {v^{t}_{1}}(\pi ^{\ast ,t}_{1}, a^{t} = a_{2}) \geq {v^{t}_{2}}({\pi ^{t}_{2}}, a^{t} = a_{1}) - {v^{t}_{2}}(\pi ^{\ast ,t}_{2}, a^{t} = a_{2})\) ensures that when the optimal action for individual 1 is a2 (\({v^{t}_{1}}({\pi ^{t}_{1}},a^{t} = a_{1}) \leq {v^{t}_{1}}(\pi ^{\ast ,t}_{1},a^{t} = a_{2})\)), individual 2’s optimal action must also be a2 (\({v^{t}_{2}}({\pi ^{t}_{2}},a^{t} = a_{1}) \leq {v^{t}_{2}}(\pi ^{\ast ,t}_{2},a^{t} = a_{2})\)) but not vice versa. Similarly, when individual 2’s optimal action is a1 (\({v^{t}_{2}}({\pi ^{t}_{2}},a^{t} = a_{1}) \geq {v^{t}_{2}}(\pi ^{\ast ,t}_{2},a^{t} = a_{2})\)), individual 1’s optimal must also be a1 (\({v^{t}_{1}}({\pi ^{t}_{1}},a^{t} = a_{1}) \geq {v^{t}_{1}}(\pi ^{\ast ,t}_{1},a^{t} = a_{2})\)) but not vice versa. We also know \(\pi ^{t=0}_{1} \leq _{r} \pi ^{t=0}_{2}\) and \({\pi ^{t}_{1}} \leq _{r} {\pi ^{t}_{2}}\) then \({\pi ^{t}_{1}} P^{t,BMI_{1}}_{1} \leq _{r} {\pi ^{t}_{2}} P^{t,BMI_{2}}_{2}\), which ensures that the belief ordering over time will be preserved given BMI1 ≤ BMI2. Therefore, in expectation, individual 2 will take action a2 earlier in time and more frequently across decision epochs.
A.7 Proof of Corollary 1
\(\pi ^{\ast ,t}_{2} \leq _{r} \pi ^{\ast ,t}_{1}\), and \({v^{t}_{1}}(\pi ^{\ast ,t}_{1}) \leq {v^{t}_{2}}(\pi ^{\ast ,t}_{2})\) for all t ensure that \({v^{t}_{2}}(\pi ^{\ast ,t}_{2}, a^{t} = a_{2}) \geq {v^{t}_{1}}(\pi ^{\ast ,t}_{1}, a^{t} = a_{2})\) for all t. We also have \({v^{t}_{1}}({\pi ^{t}_{1}}, a^{t} = a_{1}) \geq {v^{t}_{2}}({\pi ^{t}_{2}}, a^{t} = a_{1})\) for all t by Theorem 1. Therefore, \({v^{t}_{1}}({\pi ^{t}_{1}}, a^{t} = a_{1}) - {v^{t}_{1}}(\pi ^{\ast ,t}_{1}, a^{t} = a_{2}) \geq {v^{t}_{2}}({\pi ^{t}_{2}}, a^{t} = a_{1}) - {v^{t}_{2}}(\pi ^{\ast ,t}_{2}, a^{t} = a_{2})\), which is condition 2 of Theorem 2. The results of Theorem 2 therefore apply.
B Estimating transition probabilities using multinomial logistic regressions
Since HRS Biomarker waves are every four years, the merged HRS and HRS Biomarker data from 2006-2014 allows us to estimate the 4-year transition probability matrix by fitting multinomial logistic regressions. We use the current health state, age, race, gender, and BMI to estimate the likelihood an individual is in each health state (healthy, prediabetic, diabetic, or dead) every four years. We define categories for BMI = {1, 2, 3} where 1 is for BMI < 25, 2 for 25 ≤ BMI < 30, and 3 for BMI ≥ 30 following the CDC guidelines and the definition in HRS. Each BMI category therefore uses a separate set of transition matrices. To estimate the transition probabilities (staying diabetic in 6 months or being dead within 6 months) given a patient is on treatment, we use the relative risk (treated to untreated) of diabetic mortality found in the medical literature to estimate mortality for undiagnosed diabetic patients [60], assuming all patients with diagnosed diabetes in the HRS are on treatment. After modifying the death probabilities, we renormalize the outcomes to ensure the result produces a stochastic matrix. We then use the eigendecomposition numerical approximation methods proposed by Chhatwal et al. (2016) to estimate half-year transition matrices from the 4-year data by raising them to the 1/8 power [64]. This allows us to generate different transition probability matrices for each age, each BMI category, and each diabetes treatment status (treated or untreated). See Tables 8 and 9 for examples. The transition probabilities for untreated individuals are plotted in Appendix Figs. 7, 8 and 9.
C Numerical diabetes example: the IFR property
The transition probabilities change with age. To verify that the transition matrices used for our diabetes numerical example have the IFR property, we confirm that, for the transition matrix associated with each age:
-
The diagonal elements from the transition matrices are greater than 0.5 (see Fig. 7)
-
\(p^{t,BMI}_{34} \geq p^{t,BMI}_{24} \geq p^{t,BMI}_{14}\) or P(4∣3) ≥ P(4∣2) ≥ P(4∣1) (see Fig. 8)
-
\(p^{t,BMI}_{23} \geq p^{t,BMI}_{13}\) or P(3∣2) ≥ P(3∣1) (see Fig. 9)
This is sufficient to guarantee that each transition matrix has IFR property. The following graphs show the transition probabilities for white males with BMI ≥ 30.
D Numerical verification of a threshold solution
We now evaluate our problem for the existence of a threshold policy. Our theoretical results show that we will have a threshold policy for any problem with two actions, a monotonic value function, and that the difference in vt is MLR decreasing in πt with two actions. There are only two actions in our problem, so we turn first to establishing that the value function is monotonic.
We use Proposition 1, which states that if rt(πt,at) is linear and decreasing in πt for each action at, and the product of Pt,BMI and Oa is TP2, then the value function is monotonic. In this problem, rt(πt,at) is the NMB associated with the patient’s health state, which is a linear function of QALYs and costs: λ QALY(πt,at) - Cost(πt,at). Note that the linear combination of linear functions is still linear (see notation section for definition of NMB given πt). Our QALY values are decreasing in health states and our costs are increasing in health states, so we know that rt(πt,at) is linear and decreasing.
The second condition in Proposition 1 requires the product of the transition matrix and the observation matrix to be TP2 for a fixed action at. Our observation matrix, Oa, is fully captured by a three by three matrix since the death state is observable. To show the product is TP2, it suffices to show the transition matrix for the partially observable states to be TP2. Then, we directly verify these matrices are TP2 numerically using the conditions in Lemma 1 and Lemma 2. Our problem therefore has a monotonic value function.
To show that the policy has a unique threshold solution, we additionally need the difference in vt to be MLR decreasing in πt with two actions for all t. We verify this result numerically by sampling from the belief simplex and calculating the value function for both actions at each time period given that sampled belief (similar to plotting Fig. 4 for each decision epoch).
We then repeat the above process for all three BMI groups. This establishes that our problem for BMI-stratified diabetes screening in the US has a threshold solution, ensuring that our optimal policy is more easily interpretable and communicable to policy-makers. Additionally, a grid approximation method is appropriate for this problem [23].
Rights and permissions
About this article
Cite this article
Wu, CC., Suen, Sc. Optimizing diabetes screening frequencies for at-risk groups. Health Care Manag Sci 25, 1–23 (2022). https://doi.org/10.1007/s10729-021-09575-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10729-021-09575-z