Prediction Model for Hospital-Acquired Pressure Ulcer Development: Retrospective Cohort Study

Risk assessment instruments for pressure ulcer in adults in critical situation: a scoping review

OBJECTIVE

to map the instruments for risk assessment of pressure ulcers in adults in critical situation in intensive care units; identify performance indicators of the instrument, and the appreciation of users regarding the instruments' use/limitations.

METHOD

a scoping review. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews in the writing of the study. We carried out the searches in the EBSCOhost search tool for 8 databases, resulting in 1846 studies, of which 22 studies compose the sample.

RESULTS

we identified two big instrument groups: generalist [Braden, Braden (ALB), Emina, Norton-MI, RAPS, and Waterlow]; and specific (CALCULATE, Cubbin & Jackson, EVARUCI, RAPS-ICU, Song & Choi, Suriaidi and Sanada, and COMHON index). Regarding the predictive value, EVARUCI and CALCULATE presented better results for performance indicators. Concerning appreciation/limitations indicated by users, we highlight the CALCULATE scale, followed by EVARUCI and RAPS-ICU, although they still need future adjustments.

CONCLUSION

the mapping of the literature showed that the evidence is sufficient to indicate one or more instruments for the risk assessment of pressure ulcers for adults in critical situation in intensive care units. (1) The risk assessment instrument must be applied to the patient's specificities. (2) The instruments are divided into two groups: generalist and specific. (3) The EVARUCI and CALCULATE instruments presented better results. (4) The EVARACI presented better results in terms of performance indicators. (5) The CALCULATE highlights itself for being recent scale, appropriate, simple, and easy to use.

Reducing lower extremity hospital-acquired pressure injuries: a multidisciplinary clinical team approach

OBJECTIVE

Optimal methods of reducing incidence of hospital-acquired pressure injuries (HAPIs) remain to be determined. We assessed changes in yearly incidence of lower extremity HAPIs before and after an intervention aimed at reducing these wounds.

METHOD

In 2012, we implemented a three-pronged intervention to reduce the incidence of HAPIs. The intervention included: a multidisciplinary surgical team; enhanced nursing education; and improved quality data reporting. Yearly incidence of lower extremity HAPIs was tracked.

RESULTS

Pre-intervention, incidence of HAPIs was 0.746%, 0.751% and 0.742% in 2009, 2010 and 2011, respectively. Post-intervention, incidence of HAPIs was 0.002%, 0.051%, 0.038%, 0.000% and 0.006% in 2013, 2014, 2015, 2016 and 2017, respectively. Mean incidence of HAPIs was reduced from 0.746% before the intervention to 0.022% after the intervention (p<0.001).

CONCLUSION

An intervention by a multidisciplinary surgical team enhanced nursing education, and improved quality data reporting reduced the incidence of lower extremity HAPIs.

Y. Risk factors for medical device-related pressure injury in ICU patients: A systematic review and meta-analysis

BACKGROUND

Medical device-related pressure injury (MDRPI) in intensive care unit (ICU) patients is a serious issue. We aimed to evaluate the risk factors for MDRPI associated with ICU patients through systematic review and meta-analysis, and provide insights into the clinical prevention of MDRPI.

METHODS

We searched PubMed, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), WanFang Database, and China BioMedical Literature Database (CBM) (from inception to January 2023) for studies that identified risk factors of MDRPI in ICU patients. In order to avoid the omission of relevant literature, we performed a secondary search of the above database on February 15, 2023. Meta-analysis was performed using Revman 5.3.

RESULTS

Fifteen studies involving 4850 participants were selected to analyze risk factors for MDRPI in ICU patients. While conducting a meta-analysis, we used sensitivity analysis to ensure the reliability of the results for cases with significant heterogeneity among studies. When the source of heterogeneity cannot be determined, we only described the risk factor. The risk factors for MDRPI in ICU patients were elder age (OR = 1.06, 95% CI: 1.03-1.10), diabetes mellitus (OR = 3.20, 95% CI: 1.96-5.21), edema (OR = 3.62, 95% CI: 2.31-5.67), lower Braden scale score (OR = 1.22, 95%CI: 1.11-1.33), higher SOFA score (OR = 4.21, 95%CI: 2.38-7.47), higher APACHE II score (OR = 1.38, 95%CI: 1.15-1.64), longer usage time of medical devices (OR = 1.11, 95%CI: 1.05-1.19), use of vasoconstrictors (OR = 6.07, 95%CI: 3.15-11.69), surgery (OR = 4.36, 95% CI: 2.07-9.15), prone position (OR = 24.71, 95% CI: 7.34-83.15), and prone position ventilation (OR = 17.51, 95% CI: 5.86-52.36). Furthermore, we found that ICU patients who used subglottic suction catheters had a higher risk of MDRPI, whereas ICU patients with higher hemoglobin and serum albumin levels had a lower risk of MDRPI.

CONCLUSION

This study reported the risk factors for MDRPI in ICU patients. A comprehensive analysis of these risk factors will help to prevent and optimize interventions, thereby minimizing the occurrence of MDRPI.

Machine learning-based predictive models for the occurrence of behavioral and psychological symptoms of dementia: model development and validation

The behavioral and psychological symptoms of dementia (BPSD) are challenging aspects of dementia care. This study used machine learning models to predict the occurrence of BPSD among community-dwelling older adults with dementia. We included 187 older adults with dementia for model training and 35 older adults with dementia for external validation. Demographic and health data and premorbid personality traits were examined at the baseline, and actigraphy was utilized to monitor sleep and activity levels. A symptom diary tracked caregiver-perceived symptom triggers and the daily occurrence of 12 BPSD classified into seven subsyndromes. Several prediction models were also employed, including logistic regression, random forest, gradient boosting machine, and support vector machine. The random forest models revealed the highest area under the receiver operating characteristic curve (AUC) values for hyperactivity, euphoria/elation, and appetite and eating disorders; the gradient boosting machine models for psychotic and affective symptoms; and the support vector machine model showed the highest AUC. The gradient boosting machine model achieved the best performance in terms of average AUC scores across the seven subsyndromes. Caregiver-perceived triggers demonstrated higher feature importance values across the seven subsyndromes than other features. Our findings demonstrate the possibility of predicting BPSD using a machine learning approach.

An Integrated System of Braden Scale and Random Forest Using Real-Time Diagnoses to Predict When Hospital-Acquired Pressure Injuries (Bedsores) Occur

BACKGROUND AND OBJECTIVES

Bedsores/Pressure Injuries (PIs) are the second most common diagnosis in healthcare system billing records in the United States and account for 60,000 deaths annually. Hospital-Acquired Pressure Injuries (HAPIs) are one classification of PIs and indicate injuries that occurred while the patient was cared for within the hospital. Until now, all studies have predicted who will develop HAPI using classic machine algorithms, which provides incomplete information for the clinical team. Knowing who will develop HAPI does not help differentiate at which point those predicted patients will develop HAPIs; no studies have investigated when HAPI develops for predicted at-risk patients. This research aims to develop a hybrid system of Random Forest (RF) and Braden Scale to predict HAPI time by considering the changes in patients' diagnoses from admission until HAPI occurrence.

METHODS

Real-time diagnoses and risk factors were collected daily for 485 patients from admission until HAPI occurrence, which resulted in 4619 records. Then for each record, HAPI time was calculated from the day of diagnosis until HAPI occurrence. Recursive Feature Elimination (RFE) selected the best factors among the 60 factors. The dataset was separated into 80% training (10-fold cross-validation) and 20% testing. Grid Search (GS) with RF (GS-RF) was adopted to predict HAPI time using collected risk factors, including Braden Scale. Then, the proposed model was compared with the seven most common algorithms used to predict HAPI; each was replicated for 50 different experiments.

RESULTS

GS-RF achieved the best Area Under the Curve (AUC) (91.20 ± 0.26) and Geometric Mean (G-mean) (91.17 ± 0.26) compared to the seven algorithms. RFE selected 43 factors. The most dominant interactable risk factors in predicting HAPI time were visiting ICU during hospitalization, Braden subscales, BMI, Stimuli Anesthesia, patient refusal to change position, and another lab diagnosis.

CONCLUSION

Identifying when the patient is likely to develop HAPI can target early intervention when it is needed most and reduces unnecessary burden on patients and care teams when patients are at lower risk, which further individualizes the plan of care.

Advances in Zebrafish for Diabetes Mellitus with Wound Model

Diabetic foot ulcers cause great suffering and are costly for the healthcare system. Normal wound healing involves hemostasis, inflammation, proliferation, and remodeling. However, the negative factors associated with diabetes, such as bacterial biofilms, persistent inflammation, impaired angiogenesis, inhibited cell proliferation, and pathological scarring, greatly interfere with the smooth progress of the entire healing process. It is this impaired wound healing that leads to diabetic foot ulcers and even amputations. Therefore, drug screening is challenging due to the complexity of damaged healing mechanisms. The establishment of a scientific and reasonable animal experimental model contributes significantly to the in-depth research of diabetic wound pathology, prevention, diagnosis, and treatment. In addition to the low cost and transparency of the embryo (for imaging transgene applications), zebrafish have a discrete wound healing process for the separate study of each stage, resulting in their potential as the ideal model animal for diabetic wound healing in the future. In this review, we examine the reasons behind the delayed healing of diabetic wounds, systematically review various studies using zebrafish as a diabetic wound model by different induction methods, as well as summarize the challenges and improvement strategies which provide references for establishing a more reasonable diabetic wound zebrafish model.

A Comprehensive and Improved Definition for Hospital-Acquired Pressure Injury Classification Based on Electronic Health Records: Comparative Study

BACKGROUND

Patients develop pressure injuries (PIs) in the hospital owing to low mobility, exposure to localized pressure, circulatory conditions, and other predisposing factors. Over 2.5 million Americans develop PIs annually. The Center for Medicare and Medicaid considers hospital-acquired PIs (HAPIs) as the most frequent preventable event, and they are the second most common claim in lawsuits. With the growing use of electronic health records (EHRs) in hospitals, an opportunity exists to build machine learning models to identify and predict HAPI rather than relying on occasional manual assessments by human experts. However, accurate computational models rely on high-quality HAPI data labels. Unfortunately, the different data sources within EHRs can provide conflicting information on HAPI occurrence in the same patient. Furthermore, the existing definitions of HAPI disagree with each other, even within the same patient population. The inconsistent criteria make it impossible to benchmark machine learning methods to predict HAPI.

OBJECTIVE

The objective of this project was threefold. We aimed to identify discrepancies in HAPI sources within EHRs, to develop a comprehensive definition for HAPI classification using data from all EHR sources, and to illustrate the importance of an improved HAPI definition.

METHODS

We assessed the congruence among HAPI occurrences documented in clinical notes, diagnosis codes, procedure codes, and chart events from the Medical Information Mart for Intensive Care III database. We analyzed the criteria used for the 3 existing HAPI definitions and their adherence to the regulatory guidelines. We proposed the Emory HAPI (EHAPI), which is an improved and more comprehensive HAPI definition. We then evaluated the importance of the labels in training a HAPI classification model using tree-based and sequential neural network classifiers.

RESULTS

We illustrate the complexity of defining HAPI, with <13% of hospital stays having at least 3 PI indications documented across 4 data sources. Although chart events were the most common indicator, it was the only PI documentation for >49% of the stays. We demonstrate a lack of congruence across existing HAPI definitions and EHAPI, with only 219 stays having a consensus positive label. Our analysis highlights the importance of our improved HAPI definition, with classifiers trained using our labels outperforming others on a small manually labeled set from nurse annotators and a consensus set in which all definitions agreed on the label.

CONCLUSIONS

Standardized HAPI definitions are important for accurately assessing HAPI nursing quality metric and determining HAPI incidence for preventive measures. We demonstrate the complexity of defining an occurrence of HAPI, given the conflicting and incomplete EHR data. Our EHAPI definition has favorable properties, making it a suitable candidate for HAPI classification tasks.

Machine Learning Techniques, Applications, and Potential Future Opportunities in Pressure Injuries (Bedsores) Management: A Systematic Review

Pressure Injuries (PI) are one of the most common health conditions in the United States. Most acute or long-term care patients are at risk of developing PI. Machine Learning (ML) has been utilized to manage patients with PI, in which one systematic review describes how ML is used in PI management in 32 studies. This research, different from the previous systematic review, summarizes the previous contributions of ML in PI from January 2007 to July 2022, categorizes the studies according to medical specialties, analyzes gaps, and identifies opportunities for future research directions. PRISMA guidelines were adopted using the four most common databases (PubMed, Web of Science, Scopus, and Science Direct) and other resources, which result in 90 eligible studies. The reviewed articles are divided into three categories based on PI time of occurrence: before occurrence (48%); at time of occurrence (16%); and after occurrence (36%). Each category is further broken down into sub-fields based on medical specialties, which result in sixteen specialties. Each specialty is analyzed in terms of methods, inputs, and outputs. The most relevant and potentially useful applications and methods in PI management are outlined and discussed. This includes deep learning techniques and hybrid models, integration of existing risk assessment tools with ML that leads to a partnership between provider assessment and patients' Electronic Health Records (EHR).

An Integrated System of Multifaceted Machine Learning Models to Predict If and When Hospital-Acquired Pressure Injuries (Bedsores) Occur

Hospital-Acquired Pressure Injury (HAPI), known as bedsore or decubitus ulcer, is one of the most common health conditions in the United States. Machine learning has been used to predict HAPI. This is insufficient information for the clinical team because knowing who would develop HAPI in the future does not help differentiate the severity of those predicted cases. This research develops an integrated system of multifaceted machine learning models to predict if and when HAPI occurs. Phase 1 integrates Genetic Algorithm with Cost-Sensitive Support Vector Machine (GA-CS-SVM) to handle the high imbalance HAPI dataset to predict if patients will develop HAPI. Phase 2 adopts Grid Search with SVM (GS-SVM) to predict when HAPI will occur for at-risk patients. This helps to prioritize who is at the highest risk and when that risk will be highest. The performance of the developed models is compared with state-of-the-art models in the literature. GA-CS-SVM achieved the best Area Under the Curve (AUC) (75.79 ± 0.58) and G-mean (75.73 ± 0.59), while GS-SVM achieved the best AUC (75.06) and G-mean (75.06). The research outcomes will help prioritize at-risk patients, allocate targeted resources and aid with better medical staff planning to provide intervention to those patients.

A Hybrid System of Braden Scale and Machine Learning to Predict Hospital-Acquired Pressure Injuries (Bedsores): A Retrospective Observational Cohort Study

Background: The Braden Scale is commonly used to determine Hospital-Acquired Pressure Injuries (HAPI). However, the volume of patients who are identified as being at risk stretches already limited resources, and caregivers are limited by the number of factors that can reasonably assess during patient care. In the last decade, machine learning techniques have been used to predict HAPI by utilizing related risk factors. Nevertheless, none of these studies consider the change in patient status from admission until discharge. Objectives: To develop an integrated system of Braden and machine learning to predict HAPI and assist with resource allocation for early interventions. The proposed approach captures the change in patients' risk by assessing factors three times across hospitalization. Design: Retrospective observational cohort study. Setting(s): This research was conducted at ChristianaCare hospital in Delaware, United States. Participants: Patients discharged between May 2020 and February 2022. Patients with HAPI were identified from Nursing documents (N = 15,889). Methods: Support Vector Machine (SVM) was adopted to predict patients' risk for developing HAPI using multiple risk factors in addition to Braden. Multiple performance metrics were used to compare the results of the integrated system versus Braden alone. Results: The HAPI rate is 3%. The integrated system achieved better sensitivity (74.29 ± 1.23) and detection prevalence (24.27 ± 0.16) than the Braden scale alone (sensitivity (66.90 ± 4.66) and detection prevalence (41.96 ± 1.35)). The most important risk factors to predict HAPI were Braden sub-factors, overall Braden, visiting ICU during hospitalization, and Glasgow coma score. Conclusions: The integrated system which combines SVM with Braden offers better performance than Braden and reduces the number of patients identified as at-risk. Furthermore, it allows for better allocation of resources to high-risk patients. It will result in cost savings and better utilization of resources. Relevance to clinical practice: The developed model provides an automated system to predict HAPI patients in real time and allows for ongoing intervention for patients identified as at-risk. Moreover, the integrated system is used to determine the number of nurses needed for early interventions. Reporting Method: EQUATOR guidelines (TRIPOD) were adopted in this research to develop the prediction model. Patient or Public Contribution: This research was based on a secondary analysis of patients' Electronic Health Records. The dataset was de-identified and patient identifiers were removed before processing and modeling.

Documentation of pressure ulcers in medical records at an internal medicine ward in university hospital in western Sweden

OBJECTIVES

Pressure ulcers cause suffering, prolong care periods, and increase mortality. The aim was to describe and analyze the documentation of pressure ulcers and focused on the medical records from an internal medicine ward in a university hospital in western Sweden.

METHODS

A quantitative, retrospective review of medical records was conducted for all care events (n = 1,458) with descriptive statistics.

RESULTS

Documentation of the pressure ulcers in care plans was 2.1% (n = 31) compared to 6.7 % (n = 46) within final notes written by registered nurses (RN), a lower result compared to PPM (n = 3/14, 21.4%). Risk assessments were carried out in 68 (4.7%) care events, and 31 care plans included pressure ulcers. Moreover, 198 cases of tissue damage were documented, 43 (21.7%) defined as pressure ulcers, the other 147 (74.2%) lacked definition.

CONCLUSIONS

Differences (2.1%-21.4%) highlight improvements; knowledge and communication of pressure ulcers ensure reliable documentation in medical records.

Application of an infrared thermography-based model to detect pressure injuries: a prospective cohort study

BACKGROUND

It is challenging to detect pressure injuries at an early stage of their development.

OBJECTIVES

To assess the ability of an infrared thermography (IRT)-based model, constructed using a convolution neural network, to reliably detect pressure injuries.

METHODS

A prospective cohort study compared validity in patients with pressure injury (n = 58) and without pressure injury (n = 205) using different methods. Each patient was followed up for 10 days.

RESULTS

The optimal cut-off values of the IRT-based model were 0·53 for identifying tissue damage 1 day before visual detection of pressure injury and 0·88 for pressure injury detection on the day visual detection is possible. Kaplan-Meier curves and Cox proportional hazard regression model analysis showed that the risk of pressure injury increased 13-fold 1 day before visual detection with a cut-off value higher than 0·53 [hazard ratio (HR) 13·04, 95% confidence interval (CI) 6·32-26·91; P < 0·001]. The ability of the IRT-based model to detect pressure injuries [area under the receiver operating characteristic curve (AUC)_{lag 0 days} , 0·98, 95% CI 0·95-1·00] was better than that of other methods.

CONCLUSIONS

The IRT-based model is a useful and reliable method for clinical dermatologists and nurses to detect pressure injuries. It can objectively and accurately detect pressure injuries 1 day before visual detection and is therefore able to guide prevention earlier than would otherwise be possible. What is already known about this topic? Detection of pressure injuries at an early stage is challenging. Infrared thermography can be used for the physiological and anatomical evaluation of subcutaneous tissue abnormalities. A convolutional neural network is increasingly used in medical imaging analysis. What does this study add? The optimal cut-off values of the IRT-based model were 0·53 for identifying tissue damage 1 day before visual detection of pressure injury and 0·88 for pressure injury detection on the day visual detection is possible. Infrared thermography-based models can be used by clinical dermatologists and nurses to detect pressure injuries at an early stage objectively and accurately.

Risk Management In Intensive Care Units With Artificial Intelligence Technologies: Systematic Review of Prediction Models Using Electronic Health Records

Background and aim: Clinical risk assessments should be made to protect patients from negative outcomes, and the definition, frequency and severity of the risk should be determined. The information contained in the electronic health records (EHRs) can use in different areas such as risk prediction, estimation of treatment effect ect. Many prediction models using artificial intelligence (AI) technologies that can be used in risk assessment have been developed. The aim of this study is to bring together the researches on prediction models developed with AI technologies using the EHRs of patients hospitalized in the intensive care unit (ICU) and to evaluate them in terms of risk management in healthcare. Methods: The study restricted the search to the Web of Science, Pubmed, Science Direct, and Medline databases to retrieve research articles published in English in 2010 and after. Studies with a prediction model using data obtained from EHRs in the ICU are included. The study focused solely on research conducted in ICU to predict a health condition that poses a significant risk to patient safety using artificial intellegence (AI) technologies. Results: Recognized prediction subcategories were mortality (n=6), sepsis (n=4), pressure ulcer (n=4), acute kidney injury (n=3), and other areas (n=10). It has been found that EHR-based prediction models are good risk management and decision support tools and adoption of such models in ICUs may reduce the prevalence of adverse conditions. Conclusions: The article results remarks that developed models was found to have higher performance and better selectivity than previously developed risk models, so they are better at predicting risks and serious adverse events in ICU. It is recommended to use AI based prediction models developed using EHRs in risk management studies. Future work is still needed to researches to predict different health conditions risks.

Machine Learning Approaches for Hospital Acquired Pressure Injuries: A Retrospective Study of Electronic Medical Records

Background

Many machine learning heuristics integrate well with Electronic Medical Record (EMR) systems yet often fail to surpass traditional statistical models for biomedical applications.

Objective

We sought to compare predictive performances of 12 machine learning and traditional statistical techniques to predict the occurrence of Hospital Acquired Pressure Injuries (HAPI).

Methods

EMR information was collected from 57,227 hospitalizations acquired from Dartmouth Hitchcock Medical Center (April 2011 to December 2016). Twelve classification algorithms, chosen based upon classic regression and recent machine learning techniques, were trained to predict HAPI incidence and performance was assessed using the Area Under the Receiver Operating Characteristic Curve (AUC).

Results

Logistic regression achieved a performance (AUC = 0.91 ± 0.034) comparable to the other machine learning approaches. We report discordance between machine learning derived predictors compared to the traditional statistical model. We visually assessed important patient-specific factors through Shapley Additive Explanations.

Conclusions

Machine learning models will continue to inform clinical decision-making processes but should be compared to traditional modeling approaches to ensure proper utilization. Disagreements between important predictors found by traditional and machine learning modeling approaches can potentially confuse clinicians and need to be reconciled. These developments represent important steps forward in developing real-time predictive models that can be integrated into EMR systems to reduce unnecessary harm.

A computational framework to support the treatment of bedsores during COVID-19 diffusion

The treatment of pressure ulcers, also known as bedsores, is a complex process that requires to employ specialized field workforce assisting patients in their houses. In the period of COVID-19 or during any other non-trivial emergency, reaching the patients in their own house is impossible. Therefore, as well as in the other sectors, the adoption of digital technologies is invoked to solve, or at least mitigate, the problem. In particular, during the COVID-19, the social distances should be maintained in order to decrease the risk of contagion. The Project Health Management Systems proposes a complete framework, based on Deep Learning, Augmented Reality. Pattern Matching, Image Segmentation and Edge Detection approaches, to support the treatment of bedsores without increasing the risk of contagion, i.e., improving the remote aiding of specialized operators and physicians and involving inexperienced familiars in the process.

Development and Effectiveness of a Clinical Decision Support System for Pressure Ulcer Prevention Care Using Machine Learning: A Quasi-experimental Study

This study was conducted to develop and evaluate the effectiveness of a clinical decision support system for pressure ulcer prevention on clinical (performance, visual discrimination ability, and decision-making ability) and cognitive (knowledge and attitude) workflow. After developing a clinical decision support system using machine learning, a quasi-experimental study was used. Data were collected between January and April 2020. Forty-nine RNs who met the inclusion criteria and worked at seven tertiary and five secondary hospitals participated. A clinical decision support system was provided to the intervention group during the same period. Differences in outcome variables between the two groups were analyzed using t tests. The level of pressure ulcer prevention nursing performance and visual differentiation ability of skin pressure and oral mucosa pressure ulcer showed significantly greater improvement in the experimental group compared with the control group, whereas clinical decision making did not differ significantly. A clinical decision support system using machine learning was partially successful in performance of skin pressure ulcer prevention, attitude, and visual differentiation ability for skin and oral mucosa pressure ulcer prevention. These findings indicated that a clinical decision support system using machine learning needs to be implemented for pressure ulcer prevention.

Dynamic Risk Prediction for Hospital-Acquired Pressure Injury in Adult Critical Care Patients

Supplemental Digital Content is available in the text.

Accurately measuring the risk of pressure injury remains the most important step for effective prevention and intervention. Time-dependent risk factors for pressure injury development in the adult intensive care unit setting are not well understood.

Risk Factors for Pressure Injury Development Among Critical Care Patients

Identification of the appropriate pressure injury (PI) risk factors is the first step in successful PI prevention. Measuring PI risk through formalized PI risk assessment is an essential component of any PI prevention program. Major PI risk factors identified in the empirical literature in the critical care population include age, diabetes, hypotension, mobility, prolonged intensive care unit admission, mechanical ventilation and vasopressor administration. Future risk assessment using sophisticated data analytics available in the electronic medical record may result in earlier, targeted PI prevention and will improve our understanding of risk factors that may contribute to unavoidable PIs.

Pressure Injuries in Critical Care Patients: A Conceptual Schema

GENERAL PURPOSE

To outline a conceptual schema describing the relationships among the empirically supported risk factors, the etiologic factors, and the mitigating measures that influence pressure injury (PI) development in the critical care population.

TARGET AUDIENCE

This continuing education activity is intended for physicians, physician assistants, nurse practitioners, and nurses with an interest in skin and wound care.

LEARNING OBJECTIVES/OUTCOMES

After participating in this educational activity, the participant will: 1. Choose a static intrinsic factor that increases the risk for the development of PI. 2. List several dynamic intrinsic risk factors for developing a PI. 3. Identify dynamic extrinsic risk factors that may predispose a patient to developing a PI. 4. Explain the pathophysiology of PI development.

INCIDENCE OF SKIN INJURIES, RISK AND CLINICAL CHARACTERISTICS OF CRITICAL PATIENTS

ABSTRACT Objective: to analyze the incidence of skin injuries, risk and clinical characteristics of critically ill patients. Method: a retrospective cohort study performed in the intensive care center with a sample of 125 patients whose outcome was skin injury. Results: the overall injury incidence was 28% (n=35), with 36.3% (n=8) being dermatitis associated with urinary and fecal incontinence, 19.2% (n=24) pressure injury, 7.2% (n=9) skin tears, and 0.8% (n=1) medical-adhesive-related skin injury. The appearance time of the injuries varied from 1 to 44 days. The average number of injuries per patient was 1.7. Factors such as enteral nutrition (p<0.001), mechanical ventilation (p=0.001), fecal incontinence (p=0.049), diaper use with a delayed urinary catheter or urinary diversion (p=0.004) were associated with injury onset. Conclusions: incontinence-associated dermatitis and pressure injury had a higher incidence in critically ill patients. Patients who developed pressure injuries were at higher risk.

Exploration of critical care data by using unsupervised machine learning

BACKGROUND AND OBJECTIVE

Identification of subgroups may be useful to understand the clinical characteristics of ICU patients. The purposes of this study were to apply an unsupervised machine learning method to ICU patient data to discover subgroups among them; and to examine their clinical characteristics, therapeutic procedures conducted during the ICU stay, and discharge dispositions.

METHODS

K-means clustering method was used with 1503 observations and 9 types of laboratory test results as features.

RESULTS

Three clusters were identified from this specific population. Blood urea nitrogen, creatinine, potassium, hemoglobin, and red blood cell were distinctive between the clusters. Cluster Three presented the highest blood products transfusion rate (19.8%), followed by Cluster One (15.5%) and cluster Two (9.3%), which was significantly different. Hemodialysis was more frequently provided to Cluster Three while bronchoscopy was done to Cluster One and Two. Cluster Three showed the highest mortality (30.4%), which was more than two-fold compared to Cluster One (14.1%) and Two (12.2%).

CONCLUSION

Three subgroups were identified and their clinical characteristics were compared. These findings may be useful to anticipate treatment strategies and probable outcomes of ICU patients. Unsupervised machine learning may enable ICU multi-dimensional data to be organized and to make sense of the data.

Risk prediction models for the development of oral-mucosal pressure injuries in intubated patients in intensive care units: A prospective observational study

PURPOSE

Oral-mucosal pressure injury (PI) is the most commonly encountered medical device-related PIs. This study was performed to identify risk factors and construct a risk prediction model for oral-mucosal PI development in intubated patients in the intensive care unit.

METHODS

The study design was prospective, observational with medical record review. The inclusion criteria stipulated that 1) participants should be > 18 years of age, 2) there should be ETT use with holding methods including adhesive tape, gauze tying, and commercial devices. Data of 194 patient-days were analysed. The identification and validation of risk model development was performed using SPSS and the SciKit learn platform.

RESULTS

The risk prediction logistic models were composed of three factors (bite-block/airway, commercial ETT holder, and corticosteroid use) for lower oral-mucosal PI development and four factors (commercial ETT holder, vasopressor use, haematocrit, and serum albumin level) for upper oral-mucosal PI development among 10 significant input variables. The sensitivity and specificity for lower oral-mucosal PI development were 85.2% and 76.0%, respectively, and those for upper oral-mucosal PI development were 60.0% and 89.1%, respectively. Based on the results of the machine learning, the upper oral-mucosal PI development model had an accuracy of 79%, F1 score of 88%, precision of 86%, and recall of 91%.

CONCLUSIONS

The development of lower oral-mucosal PIs is affected by immobility-related factors and corticosteroid use, and that of upper oral-mucosal PIs by undernutrition-related factors and ETT holder use. The high sensitivities of the two logit models comprise important minimum data for positively predicting oral-mucosal PIs.

Care complexity individual factors associated with adverse events and in-hospital mortality

Introduction

Measuring the impact of care complexity on health outcomes, based on psychosocial, biological and environmental circumstances, is important in order to detect predictors of early deterioration of inpatients. We aimed to identify care complexity individual factors associated with selected adverse events and in-hospital mortality.

Methods

A multicenter, case-control study was carried out at eight public hospitals in Catalonia, Spain, from January 1, 2016 to December 31, 2017. All adult patients admitted to a ward or a step-down unit were evaluated. Patients were divided into the following groups based on the presence or absence of three adverse events (pressure ulcers, falls or aspiration pneumonia) and in-hospital mortality. The 28 care complexity individual factors were classified in five domains (developmental, mental-cognitive, psycho-emotional, sociocultural and comorbidity/complications). Adverse events and complexity factors were retrospectively reviewed by consulting patients’ electronic health records. Multivariate logistic analysis was performed to identify factors associated with an adverse event and in-hospital mortality.

Results

A total of 183,677 adult admissions were studied. Of these, 3,973 (2.2%) patients experienced an adverse event during hospitalization (1,673 [0.9%] pressure ulcers; 1,217 [0.7%] falls and 1,236 [0.7%] aspiration pneumonia). In-hospital mortality was recorded in 3,996 patients (2.2%). After adjustment for potential confounders, the risk factors independently associated with both adverse events and in-hospital mortality were: mental status impairments, impaired adaptation, lack of caregiver support, old age, major chronic disease, hemodynamic instability, communication disorders, urinary or fecal incontinence, vascular fragility, extreme weight, uncontrolled pain, male sex, length of stay and admission to a medical ward. High-tech hospital admission was associated with an increased risk of adverse events and a reduced risk of in-hospital mortality. The area under the ROC curve for both outcomes was > 0.75 (95% IC: 0.78–0.83).

Conclusions

Several care complexity individual factors were associated with adverse events and in-hospital mortality. Prior identification of complexity factors may have an important effect on the early detection of acute deterioration and on the prevention of poor outcomes.

Predictive Modeling for Frailty Conditions in Elderly People: Machine Learning Approaches

Background

Frailty is one of the most critical age-related conditions in older adults. It is often recognized as a syndrome of physiological decline in late life, characterized by a marked vulnerability to adverse health outcomes. A clear operational definition of frailty, however, has not been agreed so far. There is a wide range of studies on the detection of frailty and their association with mortality. Several of these studies have focused on the possible risk factors associated with frailty in the elderly population while predicting who will be at increased risk of frailty is still overlooked in clinical settings.

Objective

The objective of our study was to develop predictive models for frailty conditions in older people using different machine learning methods based on a database of clinical characteristics and socioeconomic factors.

Methods

An administrative health database containing 1,095,612 elderly people aged 65 or older with 58 input variables and 6 output variables was used. We first identify and define six problems/outputs as surrogates of frailty. We then resolve the imbalanced nature of the data through resampling process and a comparative study between the different machine learning (ML) algorithms – Artificial neural network (ANN), Genetic programming (GP), Support vector machines (SVM), Random Forest (RF), Logistic regression (LR) and Decision tree (DT) – was carried out. The performance of each model was evaluated using a separate unseen dataset.

Results

Predicting mortality outcome has shown higher performance with ANN (TPR 0.81, TNR 0.76, accuracy 0.78, F1-score 0.79) and SVM (TPR 0.77, TNR 0.80, accuracy 0.79, F1-score 0.78) than predicting the other outcomes. On average, over the six problems, the DT classifier has shown the lowest accuracy, while other models (GP, LR, RF, ANN, and SVM) performed better. All models have shown lower accuracy in predicting an event of an emergency admission with red code than predicting fracture and disability. In predicting urgent hospitalization, only SVM achieved better performance (TPR 0.75, TNR 0.77, accuracy 0.73, F1-score 0.76) with the 10-fold cross validation compared with other models in all evaluation metrics.

Conclusions

We developed machine learning models for predicting frailty conditions (mortality, urgent hospitalization, disability, fracture, and emergency admission). The results show that the prediction performance of machine learning models significantly varies from problem to problem in terms of different evaluation metrics. Through further improvement, the model that performs better can be used as a base for developing decision-support tools to improve early identification and prediction of frail older adults.