Prediction of Inpatient Pressure Ulcers Based on Routine Healthcare Data Using Machine Learning Methodology

Abstract Despite the relevance of pressure ulcers (PU) in inpatient care, the predictive power and role of care-related risk factors (e.g. surgical anesthesia) remain unclear. We investigated the predictability of PU incidence and its association with multiple care variables. We included all somatic cases between 2014 and 2018 with length of stay ≥2 days in a German university hospital. For regression analyses and prediction we used Bayesian Additive Regression Trees (BART) as nonparametric modeling approach. To assess predictive accuracy, we compared BART and logistic regression (LR) using area under the curve (AUC) and confusion matrices. The analysis of 149,006 cases revealed high predictive variable importance and associations between incident PU and intensive care with ventilation, age, surgical anesthesia (≥1 hour) and number of care-involved wards. Despite high AUCs (LR: 0.89; BART: 0.9), the confusion matrices showed a higher number of false negative (LR: 816; BART: 826) than true positive (LR: 138; BART: 68) predictions. In summary, particularly intensive care with ventilation, age, anesthesia and number of care-involved wards were associated with incident PU. Using surgical anesthesia as a proxy for immobility, our results suggest hourly repositioning. High rates of false negative predictions indicate a general challenge in the predictability of PU.

Download Full-text

Predictive Modeling of Pressure Injury Risk in Patients Admitted to an Intensive Care Unit

American Journal of Critical Care ◽

10.4037/ajcc2020237 ◽

2020 ◽

Vol 29 (4) ◽

pp. e70-e80

Author(s):

Mireia Ladios-Martin ◽

José Fernández-de-Maya ◽

Francisco-Javier Ballesta-López ◽

Adrián Belso-Garzas ◽

Manuel Mas-Asencio ◽

...

Keyword(s):

Machine Learning ◽

Intensive Care Unit ◽

Intensive Care ◽

Injury Risk ◽

Area Under The Curve ◽

University Hospital ◽

Machine Learning Techniques ◽

Real Environment ◽

Learning Techniques ◽

Pressure Injury

Background Pressure injuries are an important problem in hospital care. Detecting the population at risk for pressure injuries is the first step in any preventive strategy. Available tools such as the Norton and Braden scales do not take into account all of the relevant risk factors. Data mining and machine learning techniques have the potential to overcome this limitation. Objectives To build a model to detect pressure injury risk in intensive care unit patients and to put the model into production in a real environment. Methods The sample comprised adult patients admitted to an intensive care unit (N = 6694) at University Hospital of Torrevieja and University Hospital of Vinalopó. A retrospective design was used to train (n = 2508) and test (n = 1769) the model and then a prospective design was used to test the model in a real environment (n = 2417). Data mining was used to extract variables from electronic medical records and a predictive model was built with machine learning techniques. The sensitivity, specificity, area under the curve, and accuracy of the model were evaluated. Results The final model used logistic regression and incorporated 23 variables. The model had sensitivity of 0.90, specificity of 0.74, and area under the curve of 0.89 during the initial test, and thus it outperformed the Norton scale. The model performed well 1 year later in a real environment. Conclusions The model effectively predicts risk of pressure injury. This allows nurses to focus on patients at high risk for pressure injury without increasing workload.

Download Full-text

Using the braden and glasgow scales to predict pressure ulcer risk in patients hospitalized at intensive care units

Revista Latino-Americana de Enfermagem ◽

10.1590/s0104-11692008000600006 ◽

2008 ◽

Vol 16 (6) ◽

pp. 973-978 ◽

Cited By ~ 29

Author(s):

Luciana Magnani Fernandes ◽

Maria Helena Larcher Caliri

Keyword(s):

Intensive Care ◽

Pressure Ulcer ◽

Pressure Ulcers ◽

University Hospital ◽

Major Health ◽

Braden Scale ◽

Hospital Patients ◽

Patients At Risk ◽

Pressure Ulcer Development ◽

Critical Patients

Pressure ulcers remain a major health issue for critical patients. The purpose of this descriptive and exploratory study was to analyze the risk factors for the development of pressure ulcers in patients hospitalized at an intensive care unit of a university hospital. Patients were assessed through the Braden scale to determine the risk for the development of pressure ulcers and to identify individual risks, and the Glasgow scale was used to assess their consciousness. It was found that the risks associated with pressure ulcer development were: low scores on the Braden Scale on the first hospitalization day and low scores on the Glasgow scale. The results showed that these tools can help nurses to identify patients at risk, with a view to nursing care planning.

Download Full-text

Predictive Validity of the Cubbin-Jackson and Braden Skin Risk Tools in Critical Care Patients: A Multisite Project

American Journal of Critical Care ◽

10.4037/ajcc2021669 ◽

2021 ◽

Vol 30 (2) ◽

pp. 140-144

Author(s):

Jill M. Delawder ◽

Samantha L. Leontie ◽

Ralitsa S. Maduro ◽

Merri K. Morgan ◽

Kathie S. Zimbro

Keyword(s):

Critical Care ◽

Intensive Care ◽

Intensive Care Units ◽

Assessment Tool ◽

Predictive Accuracy ◽

Area Under The Curve ◽

Fisher Exact Test ◽

Exact Test ◽

Braden Scale ◽

Skin Integrity

Background Patients in intensive care units are 5 times more likely to have skin integrity issues develop than patients in other units. Identifying the most appropriate assessment tool may be critical to preventing pressure injuries in intensive care patients. Objectives To validate the Cubbin-Jackson skin risk assessment in the critical care setting and to compare the predictive accuracy of the Cubbin-Jackson and Braden scales for the same patients. Methods In 5 intensive care units, the Cubbin-Jackson and Braden assessments were completed by different clinicians within 61 minutes of each other for 4137 patients between October 2017 and March 2018. Bivariate correlations and the Fisher exact test were used to check for associations between the scores. Results The Cubbin-Jackson and Braden scores were significantly and positively correlated (r = 0.80, P < .001). Both tools were significant predictors of skin changes and identified as “at risk” 100% of the patients who had a change in skin integrity occur. The specificity was 18.4% for the Cubbin-Jackson scale and 27.9% for the Braden scale, and the area under the curve was 0.75 (P < .001) for the Cubbin-Jackson scale and 0.76 (P < .001) for the Braden scale. These findings show acceptable construct validity for both scales. Conclusions The predictive validities of the Cubbin-Jackson and Braden scales are similar, but both are sub-optimal because of poor specificity and positive predictive value. Change in practice may not be warranted, because there are no differences between the 2 scales of practical benefit to bedside nurses.

Download Full-text

How to Predict SARS-COV2 Infection in Patient Admitted in Emergency Department: A Prospective Cohort Study.

10.21203/rs.3.rs-122151/v1 ◽

2020 ◽

Author(s):

Frederic BALEN ◽

Sandrine CHAPENTIER ◽

Paul-Henri AUBOIROUX ◽

Elise NOEL-SAVINA ◽

Nicolas SANS ◽

...

Keyword(s):

Respiratory Symptoms ◽

False Negative ◽

False Negative Rate ◽

Area Under The Curve ◽

Unknown Origin ◽

University Hospital ◽

Predictive Tool ◽

Final Model ◽

Negative Rate ◽

Temporal Validation

Abstract Background. In order to rapidly identifying patients with a low probability of being infected by COVID19 to quickly refer them to specialized departments, the objective of our study was to develop a clinical predictive model of infection by COVID19 in patients attending the ED for respiratory symptoms or unexplained fever.Methods. We included all patients over 15 years old, admitted in one of the 2 emergency departments of Toulouse University Hospital between March 13 and March 31 for respiratory symptoms (dyspnea, cough), or fever (or sensation of fever) of unknown origin, and potentially requiring hospitalization. COVID19 infection was assessed by CT-SCAN and RT-PCR. All the candidate predictors were variables collected during the first clinical examination. Internal validation of the final model was performed using the bootstrap procedure. We performed a temporal validation in the same way on patients included between April 1 and April 13.Results. 772 patients were included. The prevalence of COVID19 was 25.5%. There were 19 predictors in the final model. The corrected-by-optimism area under the curve was 0.86 (95%CI = [0.83;0.89]). For a threshold at 10%, the sensitivity was 92%., the specificity was 56%, and the false negative rate was 5%. In secondary data, including 387 patients, the prevalence of COVID19 was 15%. The area under the curve was 0.73 (95%CI = [0.63;0.83]). For the same threshold, the sensitivity was 78%, the specificity was 48%, and the false negative rate was 7%Conclusion. We have developed a predictive tool of COVID19 infection for patients attending the ED. It could safely reduce admission in COVID19 dedicated unit in ED and prevent its overcrowding.Trial registration number: NCT: RC31/20/0149

Download Full-text

Post-analysis Procedures

Habitat Ecology and Analysis ◽

10.1093/oso/9780198829287.003.0010 ◽

2021 ◽

pp. 175-192

Author(s):

Joseph A. Veech

Keyword(s):

Statistical Model ◽

Model Comparison ◽

Predictive Accuracy ◽

False Negative ◽

Area Under The Curve ◽

Predictive Performance ◽

Error Rates ◽

True Negative ◽

Habitat Analysis ◽

Formal Procedure

There are several additional statistical procedures that can be conducted after a habitat analysis. The statistical model produced by a habitat analysis can be assessed for fit to the data. Model fit describes how well the predictor variables explain the variance in the response variable, typically species presence–absence or abundance. When more than one statistical model has been produced by the habitat analysis, these can be compared by a formal procedure called model comparison. This usually involves identifying the model with the lowest Akaike information criterion (AIC) value. If the statistical model is considered a predictive tool then its predictive accuracy needs to be assessed. There are many metrics for assessing the predictive performance of a model and quantifying rates of correct and incorrect classification; the latter are error rates. Many of these metrics are based on the numbers of true positive, true negative, false positive, and false negative observations in an independent dataset. “True” and “false” refer to whether species presence–absence was correctly predicted or not. Predictive performance can also be assessed by constructing a receiver operating characteristic (ROC) curve and calculating area under the curve (AUC) values. High AUC values approaching 1 indicate good predictive performance, whereas a value near 0.5 indicates a poor model that predicts species presence–absence no better than a random guess.

Download Full-text

Profile of patients with pressure ulcers in the intensive care unit

Revista de Enfermagem UFPE on line ◽

10.5205/reuol.2052-14823-1-le.0602201217 ◽

2012 ◽

Vol 6 (2) ◽

pp. 378

Author(s):

Joana Greicy Nascimento dos Santos ◽

Priscila De Oliveira Carvalho ◽

José Cristovam Martins Vieira

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Medical Records ◽

Pressure Ulcers ◽

University Hospital ◽

Cross Sectional ◽

Ventilatory Assistance ◽

Factores De Riesgo ◽

Terapia Intensiva ◽

Validated Instrument

ABSTRACTObjective: to identify the profile of patients with pressure ulcers (PUs) hospitalized in the intensive care unit (ICU) of a university hospital in Recife, Pernambuco, Brazil. Methodology: this is a retrospective study with a cross-sectional cohort and quantitative approach, carried out with the medical records of patients hospitalized in the ICU of Hospital das Clínicas within the period from January to June 2009. The data collection was held between September and November 2010, in Serviço de Arquivo Médico e Estatística (SAME), using the validated instrument in five medical records containing clinical, demographic and epidemiological aspects related to patients with pressure ulcers. This study was approved by the Research Ethics Committee of Universidade Federal de Pernembuco under the Protocol 229/10. Results: out of the 56 medical records of patients who met the inclusion criteria, 24 developed PUs (42.86%). Out of these ones, 62.5% were males and 58.4% were over 60 years. A prevalence of patients in Neurology and Medical Clinics was observed, both with 20.83%. Infectious disorders were the main reason for hospitalization (45.8%). Regarding risk factors, 87.5% underwent mechanical ventilatory assistance, 37.5% were sedated, and 70.8% were fed through nasogastric tube. Conclusion: the incidence of PU detected was high, showing the influence of multiple factors and conditions that increase the risk for its occurrence. Thus, one should emphasize a comprehensive approach from the multidisciplinary team, aiming to provide an adequate prevention of this injury. Descriptors: nursing; pressure ulcers; intensive care unit.RESUMOObjetivo: identificar o perfil de pacientes com úlceras por pressão (UPS) internados na unidade de terapia intensiva (UTI) de um hospital universitário de Recife-PE. Metodologia: Trata-se de estudo retrospectivo com de corte transversal e abordagem quantitativa, realizado com os prontuários de pacientes internados na UTI do Hospital das Clínicas, no período de janeiro a junho de 2009. A coleta de dados foi realizada entre setembro e novembro de 2010, no Serviço de Arquivo Médico e Estatística (SAME), utilizando o instrumento validado em cinco prontuários contendo aspectos clínicos, demográficos e epidemiológicos referentes a pacientes com úlceras por pressão. Este estudo foi aprovado pelo Comitê de Ética em Pesquisa da Universidade Federal de Pernambuco (UFPE), sob o Protocolo n. 229/10. Resultados: dos 56 prontuários de pacientes que preencheram os critérios de inclusão, 24 desenvolveram UPs (42,86%). Desses, 62,5% eram do sexo masculino e 58,4% tinham mais de 60 anos. Verificou-se uma prevalência dos pacientes da Neurologia e Clínica Médica, ambas com 20,83%. Predominaram as disfunções infecciosas como motivo de internação (45,8%). Em relação aos fatores de risco, 87,5% encontravam-se em assistência ventilatória mecânica, 37,5% sedados e 70,8% alimentavam-se por sonda nasoenteral. Conclusão: a incidência de UP detectada foi elevada, evidenciando-se a influência da multiplicidade de fatores e condições que aumentam o risco de sua ocorrência. Assim, deve-se enfatizar uma abordagem abrangente da equipe multidisciplinar, visando à prevenção adequada dessa lesão. Descritores: enfermagem; úlceras por pressão; unidade de terapia intensiva.RESUMENObjetivo: identificar el perfil de pacientes con úlceras por presión (UPs) en la unidad de terapia intensiva (UTI) de un hospital universitario en Recife, Pernambuco, Brasil. Metodología: esto es un estudio retrospectivo de corte transversal y abordaje cuantitativo, realizado con los prontuarios de pacientes internados en la UTI del Hospital das Clínicas en el periodo de enero a junio de 2009. La recogida de datos fue realizada entre septiembre y noviembre de 2010, en el Serviço de arquivo Médico e Estatística (SAME), utilizando validado en cinco prontuários conteniendo aspectos clínicos, demográficos y epidemiológicos referentes a pacientes com úlceras por presión. Este estúdio fue aprobado por el Comité de Ética en Investigación de la Universidade Federal de Pernambuco (UFPE), bajo el Protocolo 229/10. Resultados: de los 56 prontuarios de pacientes que cumplieron los criterios de inclusión, 24 desarrollaron UPs (42,86%). De estes 24 pacientes, 62,5% eran del sexo masculino y 58,4% tenían más de 60 años. Se verifico uma prevalencia de los pacientes de la Neurología y Clínica Médica, con 20,83% cada. Predominaron las disfunciones infecciosas (45,8%) como motivo de internación. Com relación a los factores de riesgo, 87,5% estaban en asistencia ventilatoria mecánica, 37,5% estaban sedados y 70,8% eran alimentados por sonda nasogástrica. Conclusión: la incidencia de UP detectada fue elevada, se evidenciando la influencia de la multiplicidad de factores y condiciones que aumentan el riesgo de su ocurrencia. Así, se debe enfatizar un abordaje abarcador del equipo multidisciplinario, intentando la prevención adecuada de esa lesión. Descriptores: enfermería; úlceras por presión; unidad de terapia intensiva.

Download Full-text

Prevalence and Related Risk Factors of Low Back Pain among Nurses Working in Intensive Care Units

Medical & Clinical Research ◽

10.33140/mcr.04.02.03 ◽

2019 ◽

Vol 4 (2) ◽

Keyword(s):

Risk Factors ◽

Intensive Care ◽

Low Back Pain ◽

Back Pain ◽

Intensive Care Units ◽

Working Hours ◽

University Hospital ◽

Low Back ◽

Related Factors ◽

Related Risk

Studies have shown that low back pain is a common health problem among hospital nurses especially those working in Intensive Care Units. However, prevalence and the related risk factors in intensive care units needs to be widely investigated. The aims: of this study were to identify prevalence of low back pain and determine its related risk factors among nurses working in Intensive Care Units. Subjects: A purposive sample of all nurses who worked in intensive care units and meet the inclusion criteria. Setting: The study was conducted at four intensive care units of Menoufia University hospital. Tools of the study: Two tools were utilized for data collection as follow; Tool I: Interviewing questionnaire and Tool II: Observational checklist. Results: The prevalence of low back pain among studied nurses was 85%. The most important and preventable risk factors for low back pain among studied nurses were higher body mass index, more average working hours/day, not enough working space, lower compliance of nurses with proper body mechanics and range of motion exercises during work. Conclusion: prevalence of low back pain among nurses working in intensive care units was high. There were multi interrlatrelatede risk factors for low back pain among studied nurses: work, patients and personnel related factors. Recommendations: Periodic and continuous in-services training for nurses working in intensive care units on preventing and coping strategies for low back pain should be implemented.

Download Full-text

Evaluation of PRE-DELIRIC (PREdiction of DELIRium in ICu patients) delirium prediction model for the patients in the intensive care unit

Acta medica Lituanica ◽

10.6001/actamedica.v25i1.3699 ◽

2018 ◽

Vol 25 (1) ◽

pp. 14-22 ◽

Cited By ~ 1

Author(s):

Gabrielė Linkaitė ◽

Mantas Riauka ◽

Ignė Bunevičiūtė ◽

Saulius Vosylius

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Characteristic Curve ◽

Area Under The Curve ◽

Confusion Assessment Method ◽

Assessment Method ◽

Preventive Therapy ◽

Apache Ii ◽

University Hospital ◽

Risk Patients

Introduction. Delirium not only compromises patient care, but is also associated with poorer outcomes: increased duration of mechanical ventilation, higher mortality, and greater long-term cognitive dysfunction. The PRE-DELIRIC model is a tool used to calculate the risk of the development of delirium. The classification of the patients into groups by risk allows efficient initiation of preventive measures. The goal of this study was to validate the PRE-DELIRIC model using the CAM-ICU (The Confusion Assessment Method for the Intensive Care Unit) method for the diagnosis of delirium. Materials and methods. Patients admitted to the University Hospital of Vilnius during February 2015 were enrolled. Every day, data were collected for APACHE-II and PRE-DELIRIC scores. Out of 167 patients, 38 (23%) were included and screened using the CAM-ICU method within 24 hours of admission to the ICU. We defined patients as having delirium when they had at least one positive CAM-ICU screening or haloperidol administration due to sedation. To validate the PRE-DELIRIC model, we calculated the area under receiver operating characteristic curve. Results. The mean age of the patients was 69.2 ± 17.2 years, 19 (50%) were male, APACHE-II mean score 18.0 ± 7.4 points. Delirium was diagnosed in 22 (58%) of 38 patients. Data used for validation of the PRE-DELIRIC model resulted in an area under the curve of 0.713 (p < 0.05, 95% CI 0.539–0.887); sensitivity and specificity for the patients with 20% risk were, accordingly, 77.3% and 50%; 40% risk – 45.5% and 81.3%, 60% – 36.4%, and 87.5%. Conclusions. The PRE-DELIRIC model predicted delirium in the patients within 24 hours of admission to the ICU. Preventive therapy could be efficiently targeted at high-risk patients if both of the methods are to be implemented.

Download Full-text