clinvestКачественная клиническая практикаKachestvennaya Klinicheskaya Praktika = Good Clinical Practice2588-05192618-8473ООО «Издательство ОКИ10.37489/2588-0519-2022-2-13-20clinvest-613Research ArticleИССЛЕДОВАНИЯ РЕАЛЬНОЙ КЛИНИЧЕСКОЙ ПРАКТИКИREAL-WORLD STUDIESМодель логистической регрессии для прогнозирования летальности в отделении интенсивной терапии: проблемы и решенияA logistic regression-based model to predict ICU mortality: problems and solutionshttps://orcid.org/0000-0002-5016-210XЛучининА. С.LuchininA. S.

Лучинин Александр Сергеевич - кандидат медицинских наук, старший научный сотрудник отдела организации и сопровождения научных исследований КНИИГиПК ФМБА России.

Киров.

Alexander S. Luchinin - PhD, Cand. Sci. Med., Senior Researcher of the Department of Organization and Support of Scientific Research, KRIHBT.

Kirov.

luchinin@niigpk.ruhttps://orcid.org/0000-0001-5509-5308ЛянгузовА. В.LyanguzovA. V.

Лянгузов Алексей В. - кандидат медицинских наук, КНИИГиПК ФМБА России.

Киров.

Alexey V. Lyanguzov - PhD, Cand. Sci. Med., KRIHBT.

Kirov.

lyanguzov@niigpk.ruКировский научно-исследовательский институт гематологии и переливания крови Федерального медико-биологического агентстваРоссияKirov Research Institute of Hematology and Blood Transfusion under the Federal Medical Biological AgencyRussian Federation202202082022021320Copyright © Лучинин А.С., Лянгузов А.В., 20222022Лучинин А.С., Лянгузов А.В.Luchinin A.S., Lyanguzov A.V.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://www.clinvest.ru/jour/article/view/613

Летальность в отделении интенсивной терапии является одной из важнейших метрик качества медицинской помощи. Цель исследования — создание модели прогноза летальности для пациентов с заболеваниями крови с использованием метода логистической регрессии и описание условий его применения при анализе данных. В исследование включили 202 пациента в возрасте от 19 до 82 лет (медиана — 57 лет), из них — 112 (55 %) мужчин и 90 (45 %) женщин. Статистический анализ проводился с использованием языка программирования R (версия 3.4.2). Вероятность смерти в отделении интенсивной терапии равнялась 33 % (67 из 202 пациентов), шансы — 0,496 (67/135). Согласно полученной модели снижение уровня тромбоцитов у пациента в 2 раза повышает шансы смерти в отделении интенсивной терапии на 31 %, или в 1,3 раза, снижение уровня общего белка в 2 раза — увеличивает шансы смерти в 11 раз, а любое нарушение сознания по шкале Глазго — почти в 20 раз при условии, что остальные переменные не изменяются. Чувствительность модели равнялась 82,3 %, специфичность — 80 %, точность — 81,6 % (95 % ДИ 67,9–91,2 %). Общая точность модели оказалась выше существующих шкал прогноза летальности qSOFA и MEWS, несмотря на простой метод анализа данных. Исследования в данной области необходимо продолжать.

The ICU department’s mortality rate is one of the most important indicators of quality of care. Based on real clinical data, we attempted to build a prognostic model for patients with blood diseases in the ICU with using of the logistic regression method. The study included 202 patients in total. The median age was 57 (19–82) years. There were 112 (55 %) males and 90 (45 %) females. The statistical analysis was performed by using R software, version 3.4.2. The absolute risk of death (mortality rate) was 67 from 202 (33 %), odds — 0.496. The odds of death in ICU grow up to ~20 times if the patient has a Glasgow score of less than 15. Also, the odds of death increase by 1.3 and 11 times of PLT, or serum total protein level decreases by 2 times accordingly. Our model for “high-risk of death” detection classified patients in the test dataset with 0.816 accuracy (95 % CI 0.679–0.912), with sensitivity 0.823, and specificity 0.80. Despite the simple method for data analysis, we got a pretty accurate model of mortality prognosis with efficacy more than qSOFA and MEWS scales. Research in this area should continue.

летальностьпрогнозлогистическая регрессияmortalityprognosislogistic regressionReferencesLee J, Dubin JA, Maslove DM. Mortality Prediction in the ICU // Secondary Analysis of Electronic Health Records / ed. MIT Critical Data. Cham: Springer International Publishing, 2016. P. 315–324.Lee J, Dubin JA, Maslove DM. Mortality Prediction in the ICU // Secondary Analysis of Electronic Health Records / ed. MIT Critical Data. Cham: Springer International Publishing, 2016. P. 315–324.Pirracchio R, Petersen ML, Carone M, et al. Mortality prediction in the ICU: can we do better? Results from the Super ICU Learner Algorithm (SICULA) project, a population-based study. Lancet Respir Med. 2015;3(1):42– 52. doi: 10.1016/S2213-2600(14)70239-5Pirracchio R, Petersen ML, Carone M, et al. Mortality prediction in the ICU: can we do better? Results from the Super ICU Learner Algorithm (SICULA) project, a population-based study. Lancet Respir Med. 2015;3(1):42– 52. doi: 10.1016/S2213-2600(14)70239-5Awad A, Bader-El-Den M, McNicholas J, et al. Predicting hospital mortality for intensive care unit patients: Time-series analysis. Health Informatics J. SAGE Publications Ltd, 2020;26(2):1043–59. doi: 10.1177/1460458219850323Awad A, Bader-El-Den M, McNicholas J, et al. Predicting hospital mortality for intensive care unit patients: Time-series analysis. Health Informatics J. SAGE Publications Ltd, 2020;26(2):1043–59. doi: 10.1177/1460458219850323Schober P, Vetter TR. Logistic Regression in Medical Research. Anesth Analg. 2021;132(2):365–6. doi: 10.1213/ANE.0000000000005247Schober P, Vetter TR. Logistic Regression in Medical Research. Anesth Analg. 2021;132(2):365–6. doi: 10.1213/ANE.0000000000005247Ahmed SN, Jhaj R, Sadasivam B, et al. Reversal of hypertensive heart disease: a multiple linear regression model. Discoveries (Craiova). 2021;9(4):e138.Ahmed SN, Jhaj R, Sadasivam B, et al. Reversal of hypertensive heart disease: a multiple linear regression model. Discoveries (Craiova). 2021;9(4):e138.Lunt M. Introduction to statistical modelling 2: categorical variables and interactions in linear regression. Rheumatology (Oxford). 2015;54(7):1141–4. doi: 10.1093/rheumatology/ket172Lunt M. Introduction to statistical modelling 2: categorical variables and interactions in linear regression. Rheumatology (Oxford). 2015;54(7):1141–4. doi: 10.1093/rheumatology/ket172Serdar CC, Cihan M, Yücel D, et al. Sample size, power and effect size revisited: simplified and practical approaches in pre-clinical, clinical and laboratory studies. Biochem Med (Zagreb). 2021;31(1):010502. doi: 10.11613/BM.2021.010502Serdar CC, Cihan M, Yücel D, et al. Sample size, power and effect size revisited: simplified and practical approaches in pre-clinical, clinical and laboratory studies. Biochem Med (Zagreb). 2021;31(1):010502. doi: 10.11613/BM.2021.010502Jenkins DG, Quintana-Ascencio PF. A solution to minimum sample size for regressions. PLoS One. 2020;15(2):e0229345. doi: 10.1371/journal.pone.0229345Jenkins DG, Quintana-Ascencio PF. A solution to minimum sample size for regressions. PLoS One. 2020;15(2):e0229345. doi: 10.1371/journal.pone.0229345Wilson Van Voorhis CR, Morgan BL. Understanding Power and Rules of Thumb for Determining Sample Size. TQMP. 2007;3(2):43–50. doi: 10.20982/tqmp.03.2.p043Wilson Van Voorhis CR, Morgan BL. Understanding Power and Rules of Thumb for Determining Sample Size. TQMP. 2007;3(2):43–50. doi: 10.20982/tqmp.03.2.p043Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373–9. doi: 10.1016/s0895-4356(96)00236-3Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373–9. doi: 10.1016/s0895-4356(96)00236-3Bujang MA, Sa’at N, Sidik TMITAB, Joo LC. Sample Size Guidelines for Logistic Regression from Observational Studies with Large Population: Emphasis on the Accuracy Between Statistics and Parameters Based on Real Life Clinical Data. Malays J Med Sci. 2018;25(4):122–30. doi: 10.21315/mjms2018.25.4.12Bujang MA, Sa’at N, Sidik TMITAB, Joo LC. Sample Size Guidelines for Logistic Regression from Observational Studies with Large Population: Emphasis on the Accuracy Between Statistics and Parameters Based on Real Life Clinical Data. Malays J Med Sci. 2018;25(4):122–30. doi: 10.21315/mjms2018.25.4.12Ajana S, Acar N, Bretillon L, et al. Benefits of dimension reduction in penalized regression methods for high-dimensional grouped data: a case study in low sample size. Bioinformatics. 2019;35(19):3628–34. doi: 10.1093/bioinformatics/btz135Ajana S, Acar N, Bretillon L, et al. Benefits of dimension reduction in penalized regression methods for high-dimensional grouped data: a case study in low sample size. Bioinformatics. 2019;35(19):3628–34. doi: 10.1093/bioinformatics/btz135Santana AC, Barbosa AV, Yehia HC, et al. A dimension reduction technique applied to regression on high dimension, low sample size neurophysiological data sets. BMC Neurosci. 2021;22(1):1. doi: 10.1186/s12868-020-00605-0Santana AC, Barbosa AV, Yehia HC, et al. A dimension reduction technique applied to regression on high dimension, low sample size neurophysiological data sets. BMC Neurosci. 2021;22(1):1. doi: 10.1186/s12868-020-00605-0Ishwaran H, O’Brien R. Commentary: The problem of class imbalance in biomedical data. J Thorac Cardiovasc Surg. 2021;161(6):1940–1. doi: 10.1016/j.jtcvs.2020.06.052Ishwaran H, O’Brien R. Commentary: The problem of class imbalance in biomedical data. J Thorac Cardiovasc Surg. 2021;161(6):1940–1. doi: 10.1016/j.jtcvs.2020.06.052Ameringer S, Serlin RC, Ward S. Simpson’s Paradox and Experimental Research. Nurs Res. 2009;58(2):123–7. doi: 10.1097/NNR.0b013e318199b517Ameringer S, Serlin RC, Ward S. Simpson’s Paradox and Experimental Research. Nurs Res. 2009;58(2):123–7. doi: 10.1097/NNR.0b013e318199b517Senaviratna NaMR, Cooray TMJA. Diagnosing Multicollinearity of Logistic Regression Model. Asian Journal of Probability and Statistics. 2019;1– 9. doi: 10.9734/ajpas/2019/v5i230132Senaviratna NaMR, Cooray TMJA. Diagnosing Multicollinearity of Logistic Regression Model. Asian Journal of Probability and Statistics. 2019;1– 9. doi: 10.9734/ajpas/2019/v5i230132Cutanda Henríquez F. [Outliers and robust logistic regression in Health Sciences]. Rev Esp Salud Publica. 2008;82(6):617–25. doi: 10.1590/s1135-57272008000600003Cutanda Henríquez F. [Outliers and robust logistic regression in Health Sciences]. Rev Esp Salud Publica. 2008;82(6):617–25. doi: 10.1590/s1135-57272008000600003Zhang Y, Zhou X, Wang Q, et al. Quality of reporting of multivariable logistic regression models in Chinese clinical medical journals. Medicine (Baltimore). 2017;96(21):e6972. doi: 10.1097/MD.0000000000006972Zhang Y, Zhou X, Wang Q, et al. Quality of reporting of multivariable logistic regression models in Chinese clinical medical journals. Medicine (Baltimore). 2017;96(21):e6972. doi: 10.1097/MD.0000000000006972Altman DG, Vergouwe Y, Royston P, et al. Prognosis and prognostic research: validating a prognostic model. BMJ. 2009;338:b605. doi: 10.1136/bmj.b605Altman DG, Vergouwe Y, Royston P, et al. Prognosis and prognostic research: validating a prognostic model. BMJ. 2009;338:b605. doi: 10.1136/bmj.b605Sun GW, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J Clin Epidemiol. 1996;49(8):907–16. doi: 10.1016/0895-4356(96)00025-xSun GW, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J Clin Epidemiol. 1996;49(8):907–16. doi: 10.1016/0895-4356(96)00025-xZhang Z. Variable selection with stepwise and best subset approaches. Ann Transl Med. 2016;4(7):136. doi: 10.21037/atm.2016.03.35Zhang Z. Variable selection with stepwise and best subset approaches. Ann Transl Med. 2016;4(7):136. doi: 10.21037/atm.2016.03.35Tibshirani R. The lasso method for variable selection in the Cox model. Stat Med. 1997;16(4):385–95. doi: 10.1002/(sici)1097-0258(19970228)16:4>385::aid-sim380<3.0.co;2-3Tibshirani R. The lasso method for variable selection in the Cox model. Stat Med. 1997;16(4):385–95. doi: 10.1002/(sici)1097-0258(19970228)16:4>385::aid-sim380<3.0.co;2-3Musoro JZ, Zwinderman AH, Puhan MA, et al. Validation of prediction models based on lasso regression with multiply imputed data. BMC Med Res Methodol. 2014;14:116. doi: 10.1186/1471-2288-14-116Musoro JZ, Zwinderman AH, Puhan MA, et al. Validation of prediction models based on lasso regression with multiply imputed data. BMC Med Res Methodol. 2014;14:116. doi: 10.1186/1471-2288-14-116Dziak JJ, Coffman DL, Lanza ST, et al. Sensitivity and specificity of information criteria. Brief Bioinform. 2020;21(2):553–65. doi: 10.1093/bib/bbz016Dziak JJ, Coffman DL, Lanza ST, et al. Sensitivity and specificity of information criteria. Brief Bioinform. 2020;21(2):553–65. doi: 10.1093/bib/bbz016Kim JH. Multicollinearity and misleading statistical results. Korean J Anesthesiol. 2019;72(6):558–69. doi: 10.4097/kja.19087Kim JH. Multicollinearity and misleading statistical results. Korean J Anesthesiol. 2019;72(6):558–69. doi: 10.4097/kja.19087Li B, Martin EB, Morris AJ. Box–Tidwell transformation based partial least squares regression. Computers & Chemical Engineering. 2001;25(9):1219–33.Li B, Martin EB, Morris AJ. Box–Tidwell transformation based partial least squares regression. Computers & Chemical Engineering. 2001;25(9):1219–33.Nattino G, Pennell ML, Lemeshow S. Assessing the goodness of fit of logistic regression models in large samples: A modification of the Hosmer-Lemeshow test. Biometrics. 2020;76(2):549–60. doi: 10.1111/biom.13249Nattino G, Pennell ML, Lemeshow S. Assessing the goodness of fit of logistic regression models in large samples: A modification of the Hosmer-Lemeshow test. Biometrics. 2020;76(2):549–60. doi: 10.1111/biom.13249Fluss R, Faraggi D, Reiser B. Estimation of the Youden Index and its associated cutoff point. Biom J. 2005;47(4):458–72. doi: 10.1002/bimj.200410135Fluss R, Faraggi D, Reiser B. Estimation of the Youden Index and its associated cutoff point. Biom J. 2005;47(4):458–72. doi: 10.1002/bimj.200410135Abdullah SMOB, Grand J, Sijapati A, et al. qSOFA is a useful prognostic factor for 30-day mortality in infected patients fulfilling the SIRS criteria for sepsis. Am J Emerg Med. 2020;38(3):512–6. doi: 10.1016/j.ajem.2019.05.037Abdullah SMOB, Grand J, Sijapati A, et al. qSOFA is a useful prognostic factor for 30-day mortality in infected patients fulfilling the SIRS criteria for sepsis. Am J Emerg Med. 2020;38(3):512–6. doi: 10.1016/j.ajem.2019.05.037Roney JK, Whitley BE, Maples JC, et al. Modified early warning scoring (MEWS): evaluating the evidence for tool inclusion of sepsis screening criteria and impact on mortality and failure to rescue. J Clin Nurs. 2015;24(23–24):3343– 54. doi: 10.1111/jocn.12952Roney JK, Whitley BE, Maples JC, et al. Modified early warning scoring (MEWS): evaluating the evidence for tool inclusion of sepsis screening criteria and impact on mortality and failure to rescue. J Clin Nurs. 2015;24(23–24):3343– 54. doi: 10.1111/jocn.12952Лучинин А. С. Искусственный интеллект в гематологии. Клиническая онкогематология. 2022;15(1):16–27. doi: 10.21320/2500-2139-2022-15-1-16-27Luchinin AS. Artificial Intelligence in Hematology. Clinical oncohematology. 2022;15(1):16–2. (In Russ). doi: 10.21320/2500-2139-2022-15-1-16-27

The authors declare that there are no conflicts of interest present.