1
|
Lee H, Moon SJ, Kim SW, Min JW, Park HS, Yoon HE, Kim YS, Kim HW, Yang CW, Chung S, Koh ES, Chung BH. Prediction of intradialytic hypotension using pre-dialysis features-a deep learning-based artificial intelligence model. Nephrol Dial Transplant 2023; 38:2310-2320. [PMID: 37019834 DOI: 10.1093/ndt/gfad064] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/11/2023] [Indexed: 04/07/2023] Open
Abstract
BACKGROUND Intradialytic hypotension (IDH) is a serious complication of hemodialysis (HD) that is associated with increased risks of cardiovascular morbidity and mortality. However, its accurate prediction remains a clinical challenge. The aim of this study was to develop a deep learning-based artificial intelligence (AI) model to predict IDH using pre-dialysis features. METHODS Data from 2007 patients with 943 220 HD sessions at seven university hospitals were used. The performance of the deep learning model was compared with three machine learning models (logistic regression, random forest and XGBoost). RESULTS IDH occurred in 5.39% of all studied HD sessions. A lower pre-dialysis blood pressure (BP), and a higher ultrafiltration (UF) target rate and interdialytic weight gain in IDH sessions compared with non-IDH sessions, and the occurrence of IDH in previous sessions was more frequent among IDH sessions compared with non-IDH sessions. Matthews correlation coefficient and macro-averaged F1 score were used to evaluate both positive and negative prediction performances. Both values were similar in logistic regression, random forest, XGBoost and deep learning models, developed with data from a single session. When combining data from the previous three sessions, the prediction performance of the deep learning model improved and became superior to that of other models. The common top-ranked features for IDH prediction were mean systolic BP (SBP) during the previous session, UF target rate, pre-dialysis SBP, and IDH experience during the previous session. CONCLUSIONS Our AI model predicts IDH accurately, suggesting it as a reliable tool for HD treatment.
Collapse
Affiliation(s)
- Hanbi Lee
- Transplantation Research Center, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
- Division of Nephrology, Department of Internal Medicine, Seoul St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| | | | | | - Ji Won Min
- Department of Internal Medicine, Bucheon St Mary's Hospital, College of Medicine, The Catholic University of Korea, Bucheon, Republic of Korea
| | - Hoon Suk Park
- Department of Internal Medicine, Eunpyeong St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| | - Hye Eun Yoon
- Department of Internal Medicine, Incheon St Mary's Hospital, College of Medicine, The Catholic University of Korea, Incheon, Republic of Korea
| | - Young Soo Kim
- Department of Internal Medicine, Uijeongbu St Mary's Hospital, College of Medicine, The Catholic University of Korea, Uijeongbu, Republic of Korea
| | - Hyung Wook Kim
- Department of Internal Medicine, St Vincent's Hospital, College of Medicine, The Catholic University of Korea, Suwon, Republic of Korea
| | - Chul Woo Yang
- Transplantation Research Center, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
- Division of Nephrology, Department of Internal Medicine, Seoul St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| | - Sungjin Chung
- Division of Nephrology, Department of Internal Medicine, Yeouido St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| | - Eun Sil Koh
- Division of Nephrology, Department of Internal Medicine, Yeouido St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| | - Byung Ha Chung
- Transplantation Research Center, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
- Division of Nephrology, Department of Internal Medicine, Seoul St Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
| |
Collapse
|
2
|
Ju SY, Ma SJ. High C-reactive protein to albumin ratio and the short-term survival prognosis within 30 days in terminal cancer patients receiving palliative care in a hospital setting: A retrospective analysis. Medicine (Baltimore) 2020; 99:e19350. [PMID: 32118773 PMCID: PMC7478418 DOI: 10.1097/md.0000000000019350] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/19/2019] [Revised: 01/27/2020] [Accepted: 01/29/2020] [Indexed: 11/26/2022] Open
Abstract
Survival estimates are very important to patients with terminal cancer. The C-reactive protein (CRP)/albumin ratio is associated with cancer outcomes. However, few studies have investigated the dose-response association in terminal cancer patients. Therefore, we aimed to evaluate the association between the CRP/albumin ratio and mortality in terminal cancer patients using a longitudinal analysis. We retrospectively investigated the electronic medical records of 435 inpatients with terminal cancer admitted to the palliative care unit of Yeouido St. Mary's Hospital between October 8, 2015, and January 17, 2018. In total, 382 patients with terminal cancer were enrolled in the study. The serum CRP/albumin ratio measured at admission had a linear dose-response relationship with the risk of death among the terminal cancer patients (P for linearity = .011). The multivariate analyses showed that the CRP/albumin ratio was an independent prognostic factor (Model 1, CRP/albumin ratio >48.53 × 10: HR = 2.68, 95% CI = 1.82-3.93; Model 2, tertile 2: HR = 1.91, 95% CI = 1.31-2.82 and tertile 3: HR = 3.66, 95% CI = 2.24-5.97). The relationship between a high CRP/albumin ratio and poor survival was a flat L-shape for survival time with an inflection point at approximately 15 days, while the relationship was not significant in terminal cancer patients who survived beyond 30 days. This study demonstrated that high CRP/albumin ratios are significantly and independently associated with the short-term survival prognosis of terminal cancer patients within 30 days.
Collapse
Affiliation(s)
- Sang-Yhun Ju
- Department of Family Medicine, Uijeongbu St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul
| | - Soo-Jin Ma
- Department of Family Medicine, College of Medicine, Chosun University, Gwangju, Republic of Korea
| |
Collapse
|
3
|
Kim Y, Park YH, Lee JY, Choi IY, Yu H. Discovery of prostate specific antigen pattern to predict castration resistant prostate cancer of androgen deprivation therapy. BMC Med Inform Decis Mak 2016; 16 Suppl 1:63. [PMID: 27453983 PMCID: PMC4959354 DOI: 10.1186/s12911-016-0297-0] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/10/2022] Open
Abstract
BACKGROUND Prostate specific antigen (PSA) is an important biomarker to monitor the response to the treatment, but has not been fully utilized as a whole sequence. We used a longitudinal biomarker PSA to discover a new prognostic pattern that predicts castration-resistant prostate cancer (CRPC) after androgen deprivation therapy. METHODS We transformed the longitudinal PSA into a discrete sequence, used frequent sequential pattern mining to find candidate patterns from the sequences, and selected the most predictive and informative pattern among the candidates. RESULTS Patients were less likely to be CRPC if, after PSA values reach nadir, the PSA decreases more than 0.048 ng/ml during a month, and the decrease occurs again. This pattern significantly increased the accuracy of predicting CRPC by supplementing information provided by existing PSA patterns such as pretreatment PSA. CONCLUSIONS This result can help clinicians to stratify men by the risk of CRPC and to determine the patient that needs intensive follow-up.
Collapse
Affiliation(s)
- Yejin Kim
- Department of Creative IT Engineering, POSTECH, Pohang, South Korea
| | - Yong Hyun Park
- Department of Urology, Seoul St. Mary's Hospital, Seoul, South Korea
| | - Ji Youl Lee
- Department of Urology, Seoul St. Mary's Hospital, Seoul, South Korea
| | - In Young Choi
- Department of Medical Informatics, The Catholic University of Korea College of Medicine, Seoul, South Korea
| | - Hwanjo Yu
- Department of Creative IT Engineering, POSTECH, Pohang, South Korea.
| |
Collapse
|
4
|
Ryu HJ, Kim WS, Lee JH, Min SW, Kim SJ, Lee YS, Lee YH, Nam SW, Eo GS, Seo SG, Nam MH. Asan medical information system for healthcare quality improvement. Healthc Inform Res 2010; 16:191-7. [PMID: 21818439 PMCID: PMC3089858 DOI: 10.4258/hir.2010.16.3.191] [Citation(s) in RCA: 21] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/03/2010] [Accepted: 09/20/2010] [Indexed: 11/23/2022] Open
Abstract
Objectives This purpose of this paper is to introduce the status of the Asan Medical Center (AMC) medical information system with respect to healthcare quality improvement. Methods Asan Medical Information System (AMIS) is projected to become a completely electronic and digital information hospital. AMIS has played a role in improving the health care quality based on the following measures: safety, effectiveness, patient-centeredness, timeliness, efficiency, privacy, and security. Results AMIS consisted of several distinctive systems: order communication system, electronic medical record, picture archiving communication system, clinical research information system, data warehouse, enterprise resource planning, IT service management system, and disaster recovery system. The most distinctive features of AMIS were the high alert-medication recognition & management system, the integrated and severity stratified alert system, the integrated patient monitoring system, the perioperative diabetic care monitoring and support system, and the clinical indicator management system. Conclusions AMIS provides IT services for AMC, 7 affiliated hospitals and over 5,000 partners clinics, and was developed to improve healthcare services. The current challenge of AMIS is standard and interoperability. A global health IT strategy is needed to get through the current challenges and to provide new services as needed.
Collapse
Affiliation(s)
- Hyeon Jeong Ryu
- Medical Information-Administration Team, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | | | | | | | | | | | | | | | | | | | | |
Collapse
|