DOI QR코드

DOI QR Code

Convergence Study in Development of Severity Adjustment Method for Death with Acute Myocardial Infarction Patients using Machine Learning

머신러닝을 이용한 급성심근경색증 환자의 퇴원 시 사망 중증도 보정 방법 개발에 대한 융복합 연구

  • 백설경 (아주대학교병원) ;
  • 박혜진 (대구가톨릭대학교 국제의료경영학과) ;
  • 강성홍 (인제대학교 보건행정학과) ;
  • 최준영 (청암대학교 병원의료정보과) ;
  • 박종호 (계명대학교 동산의료원)
  • Received : 2018.11.15
  • Accepted : 2019.02.20
  • Published : 2019.02.28

Abstract

This study was conducted to develop a customized severity-adjustment method and to evaluate their validity for acute myocardial infarction(AMI) patients to complement the limitations of the existing severity-adjustment method for comorbidities. For this purpose, the subjects of KCD-7 code I20.0 ~ I20.9, which is the main diagnosis of acute myocardial infarction were extracted using the Korean National Hospital Discharge In-depth Injury survey data from 2006 to 2015. Three tools were used for severity-adjustment method of comorbidities : CCI (charlson comorbidity index), ECI (Elixhauser comorbidity index) and the newly proposed CCS (Clinical Classification Software). The results showed that CCS was the best tool for the severity correction, and that support vector machine model was the most predictable. Therefore, we propose the use of the customized method of severity correction and machine learning techniques from this study for the future research on severity adjustment such as assessment of results of medical service.

본 연구는 기존 동반질환을 이용한 중증도 보정 방법의 제한점을 보완하기 위해 급성심근경색증 환자의 맞춤형 중증도 보정방법을 개발하고, 이의 타당성을 평가하기 위해 수행되었다. 이를 위하여 질병관리본부에서 2006년부터 2015년까지 10년간 수집한 퇴원손상심층조사 자료 중 주진단이 급성심근경색증인 한국표준질병사인분류(KCD-7) 코드 I20.0~I20.9의 대상자를 추출하였고, 동반질환 중증도 보정 도구로는 기존 활용되고 있는 CCI(Charlson comorbidity index), ECI(Elixhauser comorbidity index)와 새로이 제안하는 CCS(Clinical Classification Software)를 사용하였다. 이에 대한 중증도 보정 사망예측모형 개발을 위하여 머신러닝 기법인 로지스틱 회귀분석, 의사결정나무, 신경망, 서포트 벡터 머신기법을 활용하여 비교하였고 각각의 AUC(Area Under Curve)를 이용하여 개발된 모형을 평가하였다. 이를 평가한 결과 중증도 보정도구로는 CCS 가 가장 우수한 것으로 나타났으며, 머신러닝 기법 중에서는 서포트 벡터 머신을 이용한 모형의 예측력이 가장 우수한 것으로 확인되었다. 이에 향후 의료서비스 결과평가 등 중증도 보정을 위한 연구에서는 본 연구에서 제시한 맞춤형 중증도 보정방법과 머신러닝 기법을 활용하도록 하는 것을 제안한다.

Keywords

DJTJBT_2019_v17n2_217_f0001.png 이미지

Fig. 1. KNIME workflow for logistic regression model development

DJTJBT_2019_v17n2_217_f0002.png 이미지

Fig. 2. KNIME Workflow for Decision Tree model development

DJTJBT_2019_v17n2_217_f0003.png 이미지

Fig. 3. Severity-adjusted mortality rate model for AMI patients using desicion tree

DJTJBT_2019_v17n2_217_f0004.png 이미지

Fig. 4. KNIME workflow for neural network model development

DJTJBT_2019_v17n2_217_f0005.png 이미지

Fig. 5. KNIME workflow for support vector machine model development

Table 1. General characteristics of acute stroke inpatients

DJTJBT_2019_v17n2_217_t0001.png 이미지

Table 2. Distribution of CCI

DJTJBT_2019_v17n2_217_t0002.png 이미지

Table 3. Distribution of CCI

DJTJBT_2019_v17n2_217_t0003.png 이미지

Table 4. Distribution of comorbidity disease by ECI

DJTJBT_2019_v17n2_217_t0004.png 이미지

Table 5. Distribution of comorbidity disease by CCS category

DJTJBT_2019_v17n2_217_t0005.png 이미지

Table 6. Logistic regression model assessment using AUC

DJTJBT_2019_v17n2_217_t0006.png 이미지

Table 7. Severity-adjusted mortality rate model for acute stroke patients using logistic regression

DJTJBT_2019_v17n2_217_t0007.png 이미지

Table 8. Decision tree model assessment using AUC

DJTJBT_2019_v17n2_217_t0008.png 이미지

Table 9. Neural network model assessment using AUC

DJTJBT_2019_v17n2_217_t0009.png 이미지

Table 10. Support vector machine model assessment using AUC

DJTJBT_2019_v17n2_217_t0010.png 이미지

References

  1. J. H. Park, Y. M. Kim, S. S. Kim, W. J. Kim & S. H. Kang. (2012). Comparison of Hospital Standardized Mortality Ratio Using National Hospital Discharge Injury Data. Journal of the Korea Academia-Industrial cooperation Society, 13(4), 1739-1750. DOI : 10.5762/KAIS.2012.13.4.1739
  2. J. H. Lim & J. Y. Park. (2011). The impact of comorbidity (the Charlson Comorbidity Index) on the health outcomes of patients with the acute myocardial infarction(AMI). Korean J. of Health Policy & Administration, 21(4), 541-564. DOI : 10.4332/KJHPA.2011.21.4.541
  3. S. J. Kim, S. H. Kang, W. J. Kim & Y. M. Kim. (2011). The Variation Factors of Severity- Adjusted Length of Stay in CABG. Journal of the Korean Society for Quality Management, 39(3), 391-399. DOI : 10.7469/JKSQM.2011.39.3.391
  4. T. K. Chung & S. H. Kang. (2013). The Comparison of Risk-adjusted Mortality Rate between Korea and United States. The Journal of Digital Policy & Management, 11(5), 371-384. DOI : 10.14400/JDPM.2013.11.5.371
  5. S. O. Hong, Y. T. Kim, J. H. Park & S. H. Kang. (2015). The Variation of Factors of Severity- Adjusted Length of Stay (LOS) in Injury of Neck. Health and Social Welfare Review, 35(2), 561-583. DOI : 10.15709/hswr.2015.35.2.561
  6. W. J. Kim, S. S. Kim, E. J. Kim & S. H. Kang. (2013). Severity-Adjusted LOS Model of AMI patients based on the Korean National Hospital Discharge in-depth Injury Survey Data. Journal of the Korea Academia-Industrial cooperation Society, 14(10), 4910-4918. DOI : 10.5762/KAIS.2013.14.10.4910
  7. J. H. Lim & M. H. Nam. (2012). Development of Mortality Model of Severity-Adjustment Method of AMI Patients. Journal of the Korea Academia-Industrial cooperation Society, 13(6), 2672-2679. DOI : 10.5762/KAIS.2012.13.6.2672
  8. S. K. Jin & M. S. Bang. (2018). The Challenges of Public Policy Management for the 4th Industrial Revolution. Journal of Digital Convergence, 16(4), 39-47. DOI : 10.14400/JDC.2018.16.4.039
  9. K. Y. Lee & J. H. Kim. (2016). Artificial Intelligence Technology Trends and IBM Watson References in the Medical Field. Korean Medical Education Review, 18(2), 51-57. DOI : 10.17496/kmer.2016.18.2.51
  10. B. K. Choi, S. W. Ham, C. H. Kim, J. S. Seo, M. H. Park & S. H. Kang. (2018). Development of Predictive Model for Length of Stay(LOS) in Acute Stroke Patients using Artificial Intelligence. Journal of Digital Convergence, 16(1), 231-242. DOI : 10.14400/kmer.2018.16.1.231
  11. X. Cai, O. Perez-Concha, E. Coiera, F. Martin-Sanchez, R. Day, D. Roffe & B. Gallego (2016). Real-time prediction of mortality, readmission, and length of stay using electronic health record data. J Am Med Inform Assoc, 23, 553-561. DOI : 10.1093/jamia/ocv110
  12. A. Awada, M. Bader-El-Dena, J. McNicholasa & J. Briggsa. (2017). Early hospital mortality prediction of intensive care unit patients using an ensemble learning approach. International Journal of Medical Informatics, 108, 185-195. DOI : 10.1016/j.ijmedinf.2017.10.002
  13. H. Nilsaz-Dezfouli, M. R. Abu-Bakar, J. Arasan, M. B. Adam & M. A. Pourhoseingholi. (2017). Improving Gastric Cancer Outcome Prediction Using Single Time-Point Artificial Neural Network Models. Cancer Informatics, 16, 1-11. DOI: 10.1177/1176935116686062
  14. Statistics Korea. (2017). cause of death statistics 2016. Statistics Korea Homepage. http://kostat. go.kr/portal/korea/kor_nw/2/6/1/index.board?bmode=read&aSeq=363268
  15. E. J. Han & J. S. Kim. (2015). Effects of Symptom Recognition and Health Behavior Compliance on Hospital Arrival Time in Patients with Acute Myocardial Infarction. Korean Journal of Adult Nursing, 27(1), 83-93. DOI : 10.7475/kjan.2015.27.1.83
  16. S. O. Hong, Y. T. Kim, Y. H. Choi, J. H. Park & S. H. Kang. (2015). Development of severity- adjusted length of stay in knee replacement surgery. Journal of Digital Convergence, 13(2), 215-225. DOI : 10.14400/JDC.2015.13.2.215
  17. H. H. Lee, S. H. Chung & E. J. Choi. (2016). A Case Study on Machine Learning Applications and Performance Improvement in Learning Algorithm. Journal of Digital Convergence, 14(2), 245-258. DOI : 10.14400/JDC.2016.14.2.245
  18. K. Y. Lee & J. H. Kim. (2016). Artificial Intelligence Technology Trends and IBM Watson References in the Medical Field. Korean Medical Education Review, 18(2), 51-57. DOI : 10.17496/kmer.2016.18.2.51
  19. M. A. Oh et. al. (2017). A Study on Social security Big Data Analysis and Prediction Model based on Machine Learning. Korea Institute for Health and Social Affairs Publishing. https://www.kihasa.re.kr/common/fi ledown.do?seq=39848
  20. H. H. Kim, S. B. Yang, Y. S. Kang, Y. B. Park & J. H. Kim. (2016). Machine Learning Approach to Blood Stasis Pattern Identification Based on Self-reported Symptoms. Korean Journal of Acupuncture, 33(3), 102-113. DOI : 10.14406/acu.2016.011
  21. I. S. Park, W. S. Yong, Y. M. Kim, S. H. Kang & J. T. Han. (2008). A Development of a Tailored Follow up Management Model Using the Data Mining Technique on Hypertension. The Korean journal of applied statistics, 21(4), 639-647. DOI : 10.5351/kjas.2008.21.4.639
  22. D. Y. Choi, K. M. Jeong & D. H. Lim. (2018). Breast Cancer Classification using Deep Learning-based Ensemble. Journal of Health Informatics and Statistics, 43(2), 140-147. DOI : 10.21032/jhis.2018.43.2.140