OpenLB
Open Library of Bioscience
Advanced Search
BMJ open diabetes research & care
03, 2020;8 (1):

Predictive models of medication non-adherence risks of patients with T2D based on multiple machine learning algorithms.

Xing-Wei Wu, Heng-Bo Yang, Rong Yuan, En-Wu Long, Rong-Sheng Tong
Author Information
  1. Xing-Wei Wu: Personalized Drug Therapy Key Laboratory of Sichuan Province, School of Medicine, University of Electronic Science and Technology of China, Chengdu, China. ORCID
  2. Heng-Bo Yang: School of Pharmacy, Chengdu Medical College, Chengdu, China.
  3. Rong Yuan: Endocrine Department, Sichuan Academy of Medical Sciences and Sichuan Provincial People's Hospital, Chengdu, China.
  4. En-Wu Long: Personalized Drug Therapy Key Laboratory of Sichuan Province, School of Medicine, University of Electronic Science and Technology of China, Chengdu, China tongrs@126.com 1043743338@qq.com.
  5. Rong-Sheng Tong: Personalized Drug Therapy Key Laboratory of Sichuan Province, School of Medicine, University of Electronic Science and Technology of China, Chengdu, China tongrs@126.com 1043743338@qq.com.
PMID: 32156739 DOI: 10.1136/bmjdrc-2019-001055

Abstract

OBJECTIVE: Medication adherence plays a key role in type 2 diabetes (T2D) care. Identifying patients with high risks of non-compliance helps individualized management, especially for China, where medical resources are relatively insufficient. However, models with good predictive capabilities have not been studied. This study aims to assess multiple machine learning algorithms and screen out a model that can be used to predict patients' non-adherence risks.
METHODS: A real-world registration study was conducted at Sichuan Provincial People's Hospital from 1 April 2018 to 30 March 2019. Data of patients with T2D on demographics, disease and treatment, diet and exercise, mental status, and treatment adherence were obtained by face-to-face questionnaires. The medication possession ratio was used to evaluate patients' medication adherence status. Fourteen machine learning algorithms were applied for modeling, including Bayesian network, Neural Net, support vector machine, and so on, and balanced sampling, data imputation, binning, and methods of feature selection were evaluated by the area under the receiver operating characteristic curve (AUC). We use two-way cross-validation to ensure the accuracy of model evaluation, and we performed a posteriori test on the sample size based on the trend of AUC as the sample size increase.
RESULTS: A total of 401 patients out of 630 candidates were investigated, of which 85 were evaluated as poor adherence (21.20%). A total of 16 variables were selected as potential variables for modeling, and 300 models were built based on 30 machine learning algorithms. Among these algorithms, the AUC of the best capable one was 0.866±0.082. Imputing, oversampling and larger sample size will help improve predictive ability.
CONCLUSIONS: An accurate and sensitive adherence prediction model based on real-world registration data was established after evaluating data filling, balanced sampling, and so on, which may provide a technical tool for individualized diabetes care.

Keywords

adherence personality prediction and prevention type 2 diabetes

References

  1. Ann Transl Med. 2016 Feb;4(3):56 [PMID: 26904578]
  2. Patient Prefer Adherence. 2016 Aug 16;10:1573-81 [PMID: 27574406]
  3. Endocr Pract. 2004 Nov-Dec;10(6):502-8 [PMID: 16033724]
  4. Stud Health Technol Inform. 2019;261:174-178 [PMID: 31156111]
  5. Lancet. 2018 Feb 21;391 Suppl 2:S13 [PMID: 29553410]
  6. Value Health Reg Issues. 2017 Sep;13:82-91 [PMID: 29073997]
  7. Diabetes Obes Metab. 2018 Apr;20(4):1040-1043 [PMID: 29135080]
  8. BMJ Open Diabetes Res Care. 2017 Jul 31;5(1):e000429 [PMID: 28878942]
  9. BMJ Open Diabetes Res Care. 2016 May 09;4(1):e000201 [PMID: 27239318]
  10. Fam Pract. 2019 Oct 8;36(5):581-586 [PMID: 30534941]
  11. Diabetes Educ. 2016 Feb;42(1):34-71 [PMID: 26637240]
  12. JAMA Intern Med. 2013 Feb 11;173(3):210-8 [PMID: 23277199]
  13. J Family Med Prim Care. 2019 Mar;8(3):919-922 [PMID: 31041225]
  14. Med Care. 2015 Aug;53(8):720-8 [PMID: 26147866]
  15. Int J Clin Pract. 2008 Jan;62(1):76-87 [PMID: 17983433]
  16. Nucleic Acids Res. 2006 Mar 20;34(5):1608-19 [PMID: 16549873]
  17. J Manag Care Spec Pharm. 2018 Nov;24(11):1146-1155 [PMID: 30362915]
  18. Am J Med. 2010 Mar;123(3 Suppl):S3-11 [PMID: 20206730]
  19. JAMA Intern Med. 2017 Mar 1;177(3):371-379 [PMID: 28114642]
  20. Br J Clin Pharmacol. 2019 Jul 4;: [PMID: 31271668]
  21. Diabetes Res Clin Pract. 2017 Jun;128:40-50 [PMID: 28437734]
  22. Patient Prefer Adherence. 2013 Dec 05;8:1-13 [PMID: 24353407]
  23. Medicine (Baltimore). 2016 Jun;95(26):e3994 [PMID: 27368004]
  24. BMJ Open Diabetes Res Care. 2018 May 5;6(1):e000512 [PMID: 29755756]
  25. Healthc Technol Lett. 2016 Sep 27;3(3):165-170 [PMID: 27733922]
  26. Crit Care. 2019 Apr 8;23(1):112 [PMID: 30961662]
  27. Diabetes Care. 2004 May;27(5):1218-24 [PMID: 15111553]
  28. IEEE Trans Neural Netw Learn Syst. 2015 Oct;26(10):2381-95 [PMID: 25608311]
  29. BMJ Open Diabetes Res Care. 2018 Dec 22;6(1):e000593 [PMID: 30622720]
  30. Eur Heart J. 2013 Oct;34(38):2940-8 [PMID: 23907142]
  31. Ann Transl Med. 2016 Jan;4(1):9 [PMID: 26855945]
  32. BMJ Open Diabetes Res Care. 2019 Jun 28;7(1):e000685 [PMID: 31321061]
  33. Int J Med Inform. 2019 Sep;129:122-132 [PMID: 31445246]
  34. Diabetes Care. 2002 Mar;25(3):471-5 [PMID: 11874932]
  35. Physiol Meas. 2019 Oct 14;40(9):095006 [PMID: 31480025]

MeSH Term

Adult
Aged
Aged, 80 and over
Diabetes Mellitus, Type 2
Female
Humans
Machine Learning
Male
Middle Aged
Patient Compliance
Risk Factors
Sensitivity and Specificity
Journal Article Research Support, Non-U.S. Gov't
Article has an altmetric score of 10

Word Cloud

Created with Highcharts 10.0.0adherencemachinealgorithmspatientslearningbaseddiabetesT2DrisksmodelsmodelmedicationdataAUCsamplesizetype2careindividualizedpredictivestudymultipleusedpatients'non-adherencereal-worldregistration30treatmentstatusmodelingbalancedsamplingevaluatedtotalvariablespredictionOBJECTIVE:MedicationplayskeyroleIdentifyinghighnon-compliancehelpsmanagementespeciallyChinamedicalresourcesrelativelyinsufficientHowevergoodcapabilitiesstudiedaimsassessscreencanpredictMETHODS:conductedSichuanProvincialPeople'sHospital1April2018March2019Datademographicsdiseasedietexercisementalobtainedface-to-facequestionnairespossessionratioevaluateFourteenappliedincludingBayesiannetworkNeuralNetsupportvectorimputationbinningmethodsfeatureselectionareareceiveroperatingcharacteristiccurveusetwo-waycross-validationensureaccuracyevaluationperformedposterioritesttrendincreaseRESULTS:401630candidatesinvestigated85poor2120%16selectedpotential300builtAmongbestcapableone0866±0082ImputingoversamplinglargerwillhelpimproveabilityCONCLUSIONS:accuratesensitiveestablishedevaluatingfillingmayprovidetechnicaltoolPredictivepersonalityprevention

Similar Articles

Cited By