Development of non-invasive diabetes risk prediction models as decision support tools designed for application in the dental clinical environment

Harshad Hegde, Neel Shimpi, Aloksagar Panny, Ingrid Glurich, Pamela Christie, Amit Acharya, Harshad Hegde, Neel Shimpi, Aloksagar Panny, Ingrid Glurich, Pamela Christie, Amit Acharya

Abstract

The objective was to develop a predictive model using medical-dental data from an integrated electronic health record (iEHR) to identify individuals with undiagnosed diabetes mellitus (DM) in dental settings. Retrospective data retrieved from Marshfield Clinic Health System's data-warehouse was pre-processed prior to conducting analysis. A subset was extracted from the preprocessed dataset for external evaluation (Nvalidation) of derived predictive models. Further, subsets of 30%-70%, 40%-60% and 50%-50% case-to-control ratios were created for training/testing. Feature selection was performed on all datasets. Four machine learning (ML) classifiers were evaluated: logistic regression (LR), multilayer perceptron (MLP), support vector machines (SVM) and random forests (RF). Model performance was evaluated on Nvalidation. We retrieved a total of 5319 cases and 36,224 controls. From the initial 116 medical and dental features, 107 were used after performing feature selection. RF applied to the 50%-50% case-control ratio outperformed other predictive models over Nvalidation achieving a total accuracy (94.14%), sensitivity (0.941), specificity (0.943), F-measure (0.941), Mathews-correlation-coefficient (0.885) and area under the receiver operating curve (0.972). Future directions include incorporation of this predictive model into iEHR as a clinical decision support tool to screen and detect patients at risk for DM triggering follow-ups and referrals for integrated care delivery between dentists and physicians.

Keywords: Decision-support systems; Dental informatics; Electronic health records; Evidence-based practice; Machine leaning; Modeling healthcare services.

Conflict of interest statement

Declaration of competing interest The authors do not have any conflict of interest.

Figures

Fig. 1.
Fig. 1.
Shows the study flow.
Fig. 2.
Fig. 2.
Catalogs all retained variables identified across the 69 studies that met eligibility for data abstraction and were included for modeling.
Fig. 3.
Fig. 3.
Shows ROC (AUC) of all the four classifiers with varied case-control distribution.

References

    1. American Diabetes Association AD. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes—2018. January 1 [cited 2018 Sep 19] Diabetes Care [Internet] 2018;41(Supplement 1). S13–27. Available from: .
    1. Bergman M Pathophysiology of prediabetes and treatment implications for the prevention of type 2 diabetes mellitus. June 7 [cited 2018 Jul 20] Endocrine [Internet] 2013;43(3). 504–13. Available from: .
    1. American Diabetes Association. 5. Prevention or delay of type 2 diabetes. January 1 [cited 2018 Jul 20] Diabetes Care [Internet] 2015;38(Supplement 1). S31–2. Available from: .
    1. Ogurtsova K, da Rocha Fernandes JD, Huang Y, Linnenkamp U, Guariguata L, Cho NH, et al. IDF Diabetes Atlas: global estimates for the prevalence of diabetes for 2015 and 2040. June 1 [cited 2018 Sep 17] Diabetes Res Clin Pract [Internet] 2017;128:40–50. Available from: .
    1. National diabetes statistics report. 2017 [Internet]. 2017. [cited 2017 Sep 27]. Available from: .
    1. Lamster IB, Cheng B, Burkett S, Lalla E. Periodontal findings in individuals with newly identified pre-diabetes or diabetes mellitus [cited 2017 Oct 31] J Clin Periodontol [Internet] 2014. November;41(11). 1055–60. Available from: .
    1. Wang T-F, Jen I-A, Chou C, Lei Y-P. Effects of periodontal therapy on metabolic control in patients with type 2 diabetes mellitus and periodontal disease: a meta-analysis [cited 2018 Sep 19] Medicine (Baltimore) [Internet] 2014. December;93(28). e292 Available from: .
    1. Corbella S, Francetti L, Taschieri S, De Siena F, Fabbro M Del. Effect of periodontal treatment on glycemic control of patients with diabetes: a systematic review and meta-analysis. September 13 [cited 2018 Sep 19] J Diabetes Investig [Internet] 2013;4(5). 502–9. Available from: .
    1. Panny A, Krueger K, Acharya A. Achieving the ‘True’ triple aim in healthcare [cited 2019 Jul 10]. In: Amit Acharya, Valerie Powell, Torres-Urquidy Miguel H, Posteraro Robert Hugh, Paul Thyvalikakath Thankam, editors. Integration of medical and dental care and patient data [Internet]. Second. Switzerland: Springer International Publishing; 2019. p. 11–32. Available from:. .
    1. Lalla E, Kunzel C, Burkett S, Cheng B, Lamster IB. Identification of unrecognized diabetes and pre-diabetes in a dental setting [cited 2017 Sep 27] J Dent Res [Internet] 2011. July 29;90(7). 855–60. Available from: .
    1. Lamster IB, Kunzel C, Lalla E. Diabetes mellitus and oral health care: time for the next step. Mar [cited 2017 Sep 27] J Am Dent Assoc [Internet] 2012;143(3). 208–10. Available from: .
    1. Glurich I, Nycz G, Acharya A. Status update on translation of integrated primary dental-medical care delivery for management of diabetic patients [cited 2018 Sep 19] Clin Med Res [Internet] 2017. June 1;15(1–2):21–32. Available from: .
    1. Acharya A, Cheng B, Koralkar R, Olson B, Lamster IB, Kunzel C, et al. Screening for diabetes risk using integrated dental and medical electronic health record data [cited 2018 Nov 1] JDR Clin Transl Res [Internet] 2018. April 26;3(2). 188–94. Available from: .
    1. Sohler N, Matti-Orozco B, Young E, Li X, Gregg EW, Ali MK, et al. Opportunistic screening for diabetes and prediabetes using hemoglobin A1C in an urban primary care setting [cited 2018 Jul 20] Endocr Pract [Internet] 2016. February;22(2). 143–50. Available from: .
    1. Glurich I, Bartkowiak B, Berg RL, Acharya A. Screening for dysglycaemia in dental primary care practice settings: systematic review of the evidence. December 1 [cited 2018 Nov 28] Int Dent J [Internet] 2018;68(6):369–77. 10.1111/idj.12405.
    1. Genco RJ, Schifferle RE, Dunford RG, Falkner KL, Hsu WC, Balukjian J. Screening for diabetes mellitus in dental practices [cited 2017 Oct 31] J Am Dent Assoc [Internet] 2014. January;145(1):57–64. Available from: .
    1. Franco et al. National center for health Statistics.Health. United States. 2016. Available from: .
    1. Nasseh K, Vujicic M. Dental care utilization steady among working-age adults and children, up slightly among the elderly [cited 2018 Jul 20]; Available from: ; 2016.
    1. Shimpi N, Schroeder D, Ph C, Glurich I, Acharya. Assessment of dental providers’ knowledge, behavior and attitude towards incorporating chairside screening for medical conditions: a pilot study [cited 2018 Sep 25]; Available from: .
    1. Shimpi N, Bharatkumar A, Jethwani M, Chyou P-H, Glurich I, Blamer J, et al. Knowledgeability, attitude and behavior of primary care providers towards oral cancer: a pilot study [cited 2018 Aug 2] J Cancer Educ [Internet] 2018. April 23;33 (2). 359–64. Available from: .
    1. Glurich I, Schwei KM, Lindberg S, Shimpi N, Acharya A. Integrating medical-dental care for diabetic patients: qualitative assessment of provider perspectives. July 26 [cited 2018 Aug 2] Health Promot Pract [Internet] 2018;19(4). 531–41. Available from: .
    1. Shimpi N, Glurich I, Acharya A. Integrated care case study: Marshfield clinic health system [cited 2019 Jul 10], 315–26. Available from: ; 2019.
    1. Shimpi N, Ye Z, Koralkar R, Glurich I, Acharya A. Need for diagnostic-centric care in dentistry [cited 2018 Sep 25] J Am Dent Assoc [Internet] 2018. February;149(2). 122–31. Available from: .
    1. Acharya A Marshfield clinic health system: integrated care case study. March [cited 2018 Sep 25] J Calif Dent Assoc [Internet] 2016;44(3). 177–81. Available from: .
    1. Bhaskaran K, Douglas I, Forbes H, dos-Santos-Silva I, Leon DA, Smeeth L. Body-mass index and risk of 22 specific cancers: a population-based cohort study of 5·24 million UK adults. August [cited 2018 Nov 14] Lancet [Internet] 2014;384(9945). 755–65. Available from: .
    1. Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Dennison Himmelfarb C, et al. New ACC/AHA high blood pressure guidelines lower definition of hypertension. May [cited 2018 Nov 14] J Am Coll Cardiol [Internet] 2018;71(19). e127–248. Available from: .
    1. Berry J, Murrell D. High HDL levels: recommendations, balance, and tips [Internet]. Medical News Today; 2017. [cited 2018 Nov 14]. Available from: .
    1. Naushad H, Marion S. Leukocyte count (WBC): reference range, interpretation, collection and panels [Internet] MedScape 2015. [cited 2018 Nov 14]. Available from: .
    1. Jepsen S, Caton JG, Albandar JM, Bissada NF, Bouchard P, Cortellini P, et al. Periodontal manifestations of systemic diseases and developmental and acquired conditions: consensus report of workgroup 3 of the 2017 world Workshop on the classification of periodontal and peri-implant diseases and conditions [cited 2019 Oct 10] J Clin Periodontol [Internet] 2018. June;45 10.1111/jcpe.12951.
    1. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: machine learning in Python [cited 2018 Aug 2] J Mach Learn Res [Internet] 2011; 12 (Oct): 2825–30. Available from: .
    1. Aksoy S, Haralick RM. Feature normalization and likelihood-based similarity measures for image retrieval. April 1 [cited 2018 Aug 2] Pattern Recognit Lett [Internet] 2001;22(5). 563–82. Available from: .
    1. R Core Team. R: a language and environment for statistical computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2018. Available from: .
    1. Base SAS ® 9.4. Procedures guide statistical procedures [Internet]. Cary, NC, USA: second ed. 2013. [cited 2019 Jul 11]. Available from: .
    1. Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work?. March [cited 2018 Aug 20] Int J Methods Psychiatr Res [Internet] 2011;20(1). 40–9. Available from: .
    1. Shah AD, Bartlett JW, Carpenter J, Nicholas O, Hemingway H. Comparison of random forest and parametric imputation models for imputing missing data using MICE: a CALIBER study. March 15 [cited 2018 Aug 20] Am J Epidemiol [Internet] 2014;179(6). 764–74. Available from: .
    1. Tipping ME, Christopher MB. Probabilistic principal component analysis [cited 2018 Aug 20] J R Stat [Internet] 1999;61(3):611–22. Available from: .
    1. Witten IH, Ian H, Frank E, Hall MA, Mark A, Pal CJ. Data mining: practical machine learning tools and techniques. Fourth. Morgan Kaufmann; 2016. 621 pp.
    1. Contreras I, Vehi J. Artificial intelligence for diabetes management and decision support: literature review. May 30 [cited 2018 Aug 3] J Med Internet Res [Internet] 2018;20(5). e10775 Available from: .
    1. Breiman L Random forests [cited 2019 Jul 19] Mach Learn [Internet] 2001;45(1): 5–32. Available from: .
    1. Vapnik VN. The nature of statistical learning theory [Internet]. New York, NY: Springer New York; 2000. [cited 2019 Jul 19]. Available from: .
    1. Lee J-H, Kim D-H, Jeong S-N, Choi S-H. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm. October 1 [cited 2018 Nov 28] J Dent [Internet] 2018;77 106–11. Available from: .
    1. Shimpi N, McRoy S, Zhao H, Wu M, Acharya A. Development of a periodontitis risk assessment model for primary care providers in an interdisciplinary setting [cited 2019 Jul 10];Preprint(Preprint) Technol Heal Care [Internet] 2019. July 1:1–12. Available from: .
    1. Naushad SM, Hussain T, Indumathi B, Samreen K, Alrokayan SA, Kutala VK. Machine learning algorithm-based risk prediction model of coronary artery disease. Mol Biol Rep [Internet] 2018. July 11 [cited 2018 Aug 3]; Available from: .
    1. Wong CKH, Siu S-C, Wan EYF, Jiao F-F, Yu EYT, Fung CSC, et al. Simple non-laboratory- and laboratory-based risk assessment algorithms and nomogram for detecting undiagnosed diabetes mellitus. May 1 [cited 2019 Oct 10] J Diabetes [Internet] 2016;8(3):414–21. 10.1111/1753-0407.12310.
    1. Li W, Xie B, Qiu S, Huang X, Chen J, Wang X, et al. Non-lab and semi-lab algorithms for screening undiagnosed diabetes: a cross-sectional study. September [cited 2019 Oct 10] EBioMedicine [Internet] 2018;35 307–16. Available from: .
    1. Li S, Williams PL, Douglass CW. Development of a clinical guideline to predict undiagnosed diabetes in dental patients [cited 2017 Sep 27] J Am Dent Assoc [Internet] 2011. January;142(1):28–37. Available from: .
    1. Borrell LN, Kunzel C, Lamster I, Lalla E. Diabetes in the dental office: using NHANES III to estimate the probability of undiagnosed disease [cited 2017 Sep 27] J Periodontal Res [Internet] 2007. December 1;42(6):559–65. 10.1111/j.1600-0765.2007.00983.x.
    1. Hegde H, Shimpi N, Glurich I, Acharya A. Tobacco use status from clinical notes using Natural Language Processing and rule based algorithm. June 29 [cited 2019 Jul 8] Technol Heal Care [Internet] 2018;26(3). 445–56. Available from: .

Source: PubMed

Sponsorzy i współpracownicy

Warunki medyczne

Interwencje lekowe

3
Subskrybuj