Predicting the Risk for Hospital-Acquired Pressure Ulcers in Critical Care Patients

BACKGROUND Assessments of risk for pressure ulcers in critical care patients may not include important predictors. OBJECTIVE To construct risk-prediction models of hospital-acquired pressure ulcers in intensive care patients and compare the models’ predictive validities with validity of the Braden Scale. METHODS Data were collected retrospectively on patients admitted to intensive care from October 2011 through October 2013. Logistic regression and decision trees were used to construct the risk-prediction models. Predictive validity was measured by using sensitivity, specificity, positive and negative predictive values, and area under the curve. RESULTS With logistic regression analysis, 6 factors were significant independent predictors. With the decision tree, 4 types of high-risk populations were identified. Predictive validity of Braden Scale scores was lower than the validities of the logistic regression and the decision tree models. CONCLUSION Risk for hospital-acquired pressure ulcers is overpredicted with the Braden Scale, with low specificity and low positive predictive value

Download Full-text

Identifying Risk Factors for Pressure Injury in Adult Critical Care Patients

American Journal of Critical Care ◽

10.4037/ajcc2020243 ◽

2020 ◽

Vol 29 (3) ◽

pp. 204-213 ◽

Cited By ~ 2

Author(s):

Jill Cox ◽

Marilyn Schallom ◽

Christy Jung

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Critical Care ◽

Intensive Care ◽

Decision Tree ◽

Scale Score ◽

Critically Ill Patients ◽

Braden Scale ◽

Pressure Injury ◽

Critical Care Patients

Background Critically ill patients have a variety of unique risk factors for pressure injury. Identification of these risk factors is essential to prevent pressure injury in this population. Objective To identify factors predicting the development of pressure injury in critical care patients using a large data set from the PhysioNet MIMIC-III (Medical Information Mart for Intensive Care) clinical database. Methods Data for 1460 patients were extracted from the database. Variables that were significant in bivariate analyses were used in a final logistic regression model. A final set of significant variables from the logistic regression was used to develop a decision tree model. Results In regression analysis, cardiovascular disease, peripheral vascular disease, pneumonia or influenza, cardiovascular surgery, hemodialysis, norepinephrine administration, hypotension, septic shock, moderate to severe malnutrition, sex, age, and Braden Scale score on admission to the intensive care unit were all predictive of pressure injury. Decision tree analysis revealed that patients who received norepinephrine, were older than 65 years, had a length of stay of 10 days or less, and had a Braden Scale score of 15 or less had a 63.6% risk of pressure injury. Conclusion Determining pressure injury risk in critically ill patients is complex and challenging. One common pathophysiological factor is impaired tissue oxygenation and perfusion, which may be nonmodifiable. Improved risk quantification is needed and may be realized in the near future by leveraging the clinical information available in the electronic medical record through the power of predictive analytics.

Download Full-text

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

American Journal of Critical Care ◽

10.4037/ajcc2022657 ◽

2022 ◽

Vol 31 (1) ◽

pp. 42-50

Author(s):

Phillip Kim ◽

Vamsi K. Aribindi ◽

Amy M. Shui ◽

Sharvari S. Deshpande ◽

Sachin Rangarajan ◽

...

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Critical Care ◽

Scale Score ◽

Arterial Oxygen ◽

Effective Prevention ◽

Braden Scale ◽

Pressure Injury ◽

Hospital Acquired ◽

Critical Care Patients

Background Accurately measuring the risk of pressure injury remains the most important step for effective prevention and intervention. Relative contributions of risk factors for the incidence of pressure injury in adult critical care patients are not well understood. Objective To develop and validate a model to identify risk factors associated with hospital-acquired pressure injuries among adult critical care patients. Methods This retrospective cohort study included 23 806 adult patients (28 480 encounters) with an intensive care unit stay at an academic quaternary care center. Patient encounters were randomly split (7:3) into training and validation sets. The training set was used to develop a multivariable logistic regression model using the least absolute shrinkage and selection operator method. The model’s performance was evaluated with the validation set. Results Independent risk factors identified by logistic regression were length of hospital stay, preexisting diabetes, preexisting renal failure, maximum arterial carbon dioxide pressure, minimum arterial oxygen pressure, hypotension, gastrointestinal bleeding, cellulitis, and minimum Braden Scale score of 14 or less. On validation, the model differentiated between patients with and without pressure injury, with area under the receiver operating characteristic curve of 0.85, and performed better than a model with Braden Scale score alone (P < .001). Conclusions A model that identified risk factors for hospital-acquired pressure injury among adult critical care patients was developed and validated using a large data set of clinical variables. This model may aid in selecting high-risk patients for focused interventions to prevent formation of hospital-acquired pressure injuries.

Download Full-text

Surge Effects and Survival to Hospital Discharge in Critical Care Patients with COVID-19 During the Early Pandemic: A Cohort Study

10.21203/rs.3.rs-111259/v1 ◽

2020 ◽

Author(s):

Christopher Dale ◽

Rachael Starcher ◽

Shu Ching Chang ◽

Ari Robicsek ◽

Guilford Parsons ◽

...

Keyword(s):

Logistic Regression ◽

Critical Care ◽

Cohort Study ◽

Intensive Care ◽

Intensive Care Units ◽

Hospital Discharge ◽

Hospital Capacity ◽

Multivariable Logistic Regression ◽

Critical Care Patients ◽

Over Time

Abstract BackgroundThe early months of the COVID-19 pandemic were fraught with much uncertainty and some resource constraint. We assessed the change in survival to hospital discharge over time for intensive care unit patients with COVID-19 during the first three months of the pandemic and the presence of any surge effects on patient outcomes.MethodsRetrospective cohort study with electronic medical record data of all patients with laboratory-confirmed COVID-19 admitted to intensive care units from February 25, 2020 to May 15, 2020 admitted to intensive care units of 26 hospitals of an integrated delivery system in the Western United States. Patient demographic, comorbidity and severity of illness were measured along with exposure to pharmacologic and medical therapies and hospital outcomes over time. Multivariable logistic regression models were constructed to assess the change in survival to hospital discharge over time during the study period.ResultsOf 620 patients with COVID-19 admitted to the study ICUs (mean age 63.5 years (SD 15.7) and 69% male), 403 (65%) survived to hospital discharge and 217 (35%) died in hospital. Survival to hospital discharge increased over the study period from 60.0% in the first two weeks of patient admission to 67.6% in the last two weeks. In a multivariable logistic regression analysis, the risk-adjusted odds of survival to hospital discharge increased over time (bi-weekly change, adjusted odds ratio [aOR] 1.22, 95%CI 1.04-1.40, P = 0.02). Additionally, an a priori-defined explanatory model showed that after adjusting for both hospital occupancy and COVID positive/PUI percent hospital capacity, and the same set of covariates, the temporal trend in risk-adjusted patient survival to hospital discharge remained the same (bi-weekly change, aOR 1.18, 95% CI 1.00 to 1.38, P = 0.04) and a greater COVID positive/PUI percentage of hospital capacity remained significantly and inversely associated with survival to hospital discharge (aOR 0.95, 95% CI 0.92 to 0.98, P < 0.01).ConclusionsDuring the the early COVID-19 pandemic, risk-adjusted survival to hospital discharge increased over time for critical care patients. This may have been partially explained by surge affects, as measured by a greater COVID positive/PUI percentage of hospital capacity.

Download Full-text

Surge Effects and Survival to Hospital Discharge in Critical Care Patients with COVID-19 During the Early Pandemic: A Cohort Study

10.21203/rs.3.rs-111259/v2 ◽

2021 ◽

Author(s):

Christopher Dale ◽

Rachael Starcher ◽

Shu Ching Chang ◽

Ari Robicsek ◽

Guilford Parsons ◽

...

Keyword(s):

Logistic Regression ◽

Critical Care ◽

Cohort Study ◽

Intensive Care ◽

Hospital Discharge ◽

Multivariable Logistic Regression Analysis ◽

Hospital Capacity ◽

Multivariable Logistic Regression ◽

Critical Care Patients ◽

Over Time

Abstract BackgroundThe early months of the COVID-19 pandemic were fraught with much uncertainty and some resource constraint. We assessed the change in survival to hospital discharge over time for intensive care unit patients with COVID-19 during the first three months of the pandemic and the presence of any surge effects on patient outcomes.MethodsRetrospective cohort study using electronic medical record data for all patients with laboratory-confirmed COVID-19 admitted to intensive care units from February 25, 2020 to May 15, 2020 at one of 26 hospitals within an integrated delivery system in the Western United States. Patient demographics, comorbidities and severity of illness were measured along with medical therapies and hospital outcomes over time. Multivariable logistic regression models were constructed to assess temporal changes in survival to hospital discharge during the study period.ResultsOf 620 patients with COVID-19 admitted to the ICU (mean age 63.5 years (SD 15.7) and 69% male), 403 (65%) survived to hospital discharge and 217 (35%) died in the hospital. Survival to hospital discharge increased over time, from 60.0% in the first two weeks of the study period to 67.6% in the last two weeks. In a multivariable logistic regression analysis, the risk-adjusted odds of survival to hospital discharge increased over time (bi-weekly change, adjusted odds ratio [aOR] 1.22, 95%CI 1.04-1.40, P = 0.02). Additionally, an a priori-defined explanatory model showed that after adjusting for both hospital occupancy and percent hospital capacity by COVID-19 positive individuals and persons under investigation (PUI), the temporal trend in risk-adjusted patient survival to hospital discharge remained the same (bi-weekly change, aOR 1.18, 95% CI 1.00 to 1.38, P = 0.04). The presence of greater rates of COVID-19 positive/PUI as a percentage of hospital capacity was, however, significantly and inversely associated with survival to hospital discharge (aOR 0.95, 95% CI 0.92 to 0.98, P < 0.01). ConclusionsDuring the early COVID-19 pandemic, risk-adjusted survival to hospital discharge increased over time for critical care patients. An association was also seen between a greater COVID-19 positive/PUI percentage of hospital capacity and a lower survival rate to hospital discharge.

Download Full-text

Risk Factors for Hospital-Acquired Pressure Injury in Surgical Critical Care Patients

American Journal of Critical Care ◽

10.4037/ajcc2020810 ◽

2020 ◽

Vol 29 (6) ◽

pp. e128-e134

Author(s):

Jenny Alderden ◽

Linda J. Cowan ◽

Jonathan B. Dimas ◽

Danli Chen ◽

Yue Zhang ◽

...

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Critical Care ◽

Odds Ratio ◽

Academic Medical Center ◽

Skin Irritation ◽

Surgical Critical Care ◽

Pressure Injury ◽

Hospital Acquired ◽

Critical Care Patients

Background Hospital-acquired pressure injuries disproportionately affect critical care patients. Although risk factors such as moisture, illness severity, and inadequate perfusion have been recognized, nursing skin assessment data remain unexamined in relation to the risk for hospital-acquired pressure injuries. Objective To identify factors associated with hospital-acquired pressure injuries among surgical critical care patients. The specific aim was to analyze data obtained from routine nursing skin assessments alongside other potential risk factors identified in the literature. Methods This retrospective cohort study included 5101 surgical critical care patients at a level I trauma center and academic medical center. Multivariate logistic regression using the least absolute shrinkage and selection operator method identified important predictors with parsimonious representation. Use of specialty pressure redistribution beds was included in the model as a known predictive factor because specialty beds are a common preventive intervention. Results Independent risk factors identified by logistic regression were skin irritation (rash or diffuse, nonlocalized redness) (odds ratio, 1.788; 95% CI, 1.404-2.274; P < .001), minimum Braden Scale score (odds ratio, 0.858; 95% CI, 0.818-0.899; P < .001), and duration of intensive care unit stay before the hospital-acquired pressure injury developed (odds ratio, 1.003; 95% CI, 1.003-1.004; P < .001). Conclusions The strongest predictor was irritated skin, a potentially modifiable risk factor. Irritated skin should be treated and closely monitored, and the cause should be eliminated to allow the skin to heal.

Download Full-text

Predictors of Pressure Ulcers in Adult Critical Care Patients

American Journal of Critical Care ◽

10.4037/ajcc2011934 ◽

2011 ◽

Vol 20 (5) ◽

pp. 364-375 ◽

Cited By ~ 149

Author(s):

Jill Cox

Keyword(s):

Risk Factors ◽

Risk Assessment ◽

Intensive Care Unit ◽

Critical Care ◽

Intensive Care ◽

Length Of Stay ◽

Critically Ill ◽

Critically Ill Patients ◽

Pressure Ulcers ◽

Critical Care Patients

BackgroundPressure ulcers are one of the most underrated conditions in critically ill patients. Despite the introduction of clinical practice guidelines and advances in medical technology, the prevalence of pressure ulcers in hospitalized patients continues to escalate. Currently, consensus is lacking on the most important risk factors for pressure ulcers in critically ill patients, and no risk assessment scale exclusively for pressure ulcers in these patients is available.ObjectiveTo determine which risk factors are most predictive of pressure ulcers in adult critical care patients. Risk factors investigated included total score on the Braden Scale, mobility, activity, sensory perception, moisture, friction/shear, nutrition, age, blood pressure, length of stay in the intensive care unit, score on the Acute Physiology and Chronic Health Evaluation II, vasopressor administration, and comorbid conditions.MethodsA retrospective, correlational design was used to examine 347 patients admitted to a medical-surgical intensive care unit from October 2008 through May 2009.ResultsAccording to direct logistic regression analyses, age, length of stay, mobility, friction/shear, norepinephrine infusion, and cardiovascular disease explained a major part of the variance in pressure ulcers.ConclusionCurrent risk assessment scales for development of pressure ulcers may not include risk factors common in critically ill adults. Development of a risk assessment model for pressure ulcers in these patients is warranted and could be the foundation for development of a risk assessment tool.

Download Full-text

Surge effects and survival to hospital discharge in critical care patients with COVID-19 during the early pandemic: a cohort study

Critical Care ◽

10.1186/s13054-021-03504-w ◽

2021 ◽

Vol 25 (1) ◽

Author(s):

Christopher R. Dale ◽

Rachael W. Starcher ◽

Shu Ching Chang ◽

Ari Robicsek ◽

Guilford Parsons ◽

...

Keyword(s):

Logistic Regression ◽

Critical Care ◽

Cohort Study ◽

Intensive Care ◽

Hospital Discharge ◽

Multivariable Logistic Regression Analysis ◽

Hospital Capacity ◽

Multivariable Logistic Regression ◽

Critical Care Patients ◽

Over Time

Abstract Background The early months of the COVID-19 pandemic were fraught with much uncertainty and some resource constraint. We assessed the change in survival to hospital discharge over time for intensive care unit patients with COVID-19 during the first 3 months of the pandemic and the presence of any surge effects on patient outcomes. Methods Retrospective cohort study using electronic medical record data for all patients with laboratory-confirmed COVID-19 admitted to intensive care units from February 25, 2020, to May 15, 2020, at one of 26 hospitals within an integrated delivery system in the Western USA. Patient demographics, comorbidities, and severity of illness were measured along with medical therapies and hospital outcomes over time. Multivariable logistic regression models were constructed to assess temporal changes in survival to hospital discharge during the study period. Results Of 620 patients with COVID-19 admitted to the ICU [mean age 63.5 years (SD 15.7) and 69% male], 403 (65%) survived to hospital discharge and 217 (35%) died in the hospital. Survival to hospital discharge increased over time, from 60.0% in the first 2 weeks of the study period to 67.6% in the last 2 weeks. In a multivariable logistic regression analysis, the risk-adjusted odds of survival to hospital discharge increased over time (biweekly change, adjusted odds ratio [aOR] 1.22, 95% CI 1.04–1.40, P = 0.02). Additionally, an a priori-defined explanatory model showed that after adjusting for both hospital occupancy and percent hospital capacity by COVID-19-positive individuals and persons under investigation (PUI), the temporal trend in risk-adjusted patient survival to hospital discharge remained the same (biweekly change, aOR 1.18, 95% CI 1.00–1.38, P = 0.04). The presence of greater rates of COVID-19 positive/PUI as a percentage of hospital capacity was, however, significantly and inversely associated with survival to hospital discharge (aOR 0.95, 95% CI 0.92–0.98, P < 0.01). Conclusions During the early COVID-19 pandemic, risk-adjusted survival to hospital discharge increased over time for critical care patients. An association was also seen between a greater COVID-19-positive/PUI percentage of hospital capacity and a lower survival rate to hospital discharge.

Download Full-text

Predicting 30-day Hospital Readmission with Publicly Available Administrative Database

Methods of Information in Medicine ◽

10.3414/me14-02-0017 ◽

2015 ◽

Vol 54 (06) ◽

pp. 560-567 ◽

Cited By ~ 11

Author(s):

K. Zhu ◽

Z. Lou ◽

J. Zhou ◽

N. Ballester ◽

P. Parikh ◽

...

Keyword(s):

Heart Failure ◽

Logistic Regression ◽

Decision Tree ◽

Ad Hoc ◽

Prediction Models ◽

Conditional Logistic Regression ◽

Hospital Readmissions ◽

Decision Rules ◽

Classification Models ◽

Standard Classification

SummaryIntroduction: This article is part of the Focus Theme of Methods of Information in Medicine on “Big Data and Analytics in Healthcare”.Background: Hospital readmissions raise healthcare costs and cause significant distress to providers and patients. It is, therefore, of great interest to healthcare organizations to predict what patients are at risk to be readmitted to their hospitals. However, current logistic regression based risk prediction models have limited prediction power when applied to hospital administrative data. Meanwhile, although decision trees and random forests have been applied, they tend to be too complex to understand among the hospital practitioners.Objectives: Explore the use of conditional logistic regression to increase the prediction accuracy.Methods: We analyzed an HCUP statewide in-patient discharge record dataset, which includes patient demographics, clinical and care utilization data from California. We extracted records of heart failure Medicare beneficiaries who had inpatient experience during an 11-month period. We corrected the data imbalance issue with under-sampling. In our study, we first applied standard logistic regression and decision tree to obtain influential variables and derive practically meaning decision rules. We then stratified the original data set accordingly and applied logistic regression on each data stratum. We further explored the effect of interacting variables in the logistic regression modeling. We conducted cross validation to assess the overall prediction performance of conditional logistic regression (CLR) and compared it with standard classification models.Results: The developed CLR models outperformed several standard classification models (e.g., straightforward logistic regression, stepwise logistic regression, random forest, support vector machine). For example, the best CLR model improved the classification accuracy by nearly 20% over the straightforward logistic regression model. Furthermore, the developed CLR models tend to achieve better sensitivity of more than 10% over the standard classification models, which can be translated to correct labeling of additional 400 – 500 readmissions for heart failure patients in the state of California over a year. Lastly, several key predictor identified from the HCUP data include the disposition location from discharge, the number of chronic conditions, and the number of acute procedures.Conclusions: It would be beneficial to apply simple decision rules obtained from the decision tree in an ad-hoc manner to guide the cohort stratification. It could be potentially beneficial to explore the effect of pairwise interactions between influential predictors when building the logistic regression models for different data strata. Judicious use of the ad-hoc CLR models developed offers insights into future development of prediction models for hospital readmissions, which can lead to better intuition in identifying high-risk patients and developing effective post-discharge care strategies. Lastly, this paper is expected to raise the awareness of collecting data on additional markers and developing necessary database infrastructure for larger-scale exploratory studies on readmission risk prediction.

Download Full-text