Early Prediction of Mortality for Septic Patients Visiting Emergency Room Based on Explainable Machine Learning: A Real-World Multicenter Study.

BACKGROUND: Worldwide, sepsis is the leading cause of death in hospitals. If mortality rates in patients with sepsis can be predicted early, medical resources can be allocated efficiently. We constructed machine learning (ML) models to predict the mortality of patients with sepsis in a hospital emergency department. METHODS: This study prospectively collected nationwide data from an ongoing multicenter cohort of patients with sepsis identified in the emergency department. Patients were enrolled from 19 hospitals between September 2019 and December 2020. For acquired data from 3,657 survivors and 1,455 deaths, six ML models (logistic regression, support vector machine, random forest, extreme gradient boosting [XGBoost], light gradient boosting machine, and categorical boosting [CatBoost]) were constructed using fivefold cross-validation to predict mortality. Through these models, 44 clinical variables measured on the day of admission were compared with six sequential organ failure assessment (SOFA) components (PaO2/FIO2 [PF], platelets (PLT), bilirubin, cardiovascular, Glasgow Coma Scale score, and creatinine). The confidence interval (CI) was obtained by performing 10,000 repeated measurements via random sampling of the test dataset. All results were explained and interpreted using Shapley's additive explanations (SHAP). RESULTS: Of the 5,112 participants, CatBoost exhibited the highest area under the curve (AUC) of 0.800 (95% CI, 0.756-0.840) using clinical variables. Using the SOFA components for the same patient, XGBoost exhibited the highest AUC of 0.678 (95% CI, 0.626-0.730). As interpreted by SHAP, albumin, lactate, blood urea nitrogen, and international normalization ratio were determined to significantly affect the results. Additionally, PF and PLTs in the SOFA component significantly influenced the prediction results. CONCLUSION: Newly established ML-based models achieved good prediction of mortality in patients with sepsis. Using several clinical variables acquired at the baseline can provide more accurate results for early predictions than using SOFA components. Additionally, the impact of each variable was identified.

Fully Automated MRI Segmentation and Volumetric Measurement of Intracranial Meningioma Using Deep Learning.

BACKGROUND: Accurate and rapid measurement of the MRI volume of meningiomas is essential in clinical practice to determine the growth rate of the tumor. Imperfect automation and disappointing performance for small meningiomas of previous automated volumetric tools limit their use in routine clinical practice. PURPOSE: To develop and validate a computational model for fully automated meningioma segmentation and volume measurement on contrast-enhanced MRI scans using deep learning. STUDY TYPE: Retrospective. POPULATION: A total of 659 intracranial meningioma patients (median age, 59.0 years; interquartile range: 53.0-66.0 years) including 554 women and 105 men. FIELD STRENGTH/SEQUENCE: The 1.0 T, 1.5 T, and 3.0 T; three-dimensional, T1 -weighted gradient-echo imaging with contrast enhancement. ASSESSMENT: The tumors were manually segmented by two neurosurgeons, H.K. and C.-K.P., with 10 and 26 years of clinical experience, respectively, for use as the ground truth. Deep learning models based on U-Net and nnU-Net were trained using 459 subjects and tested for 100 patients from a single institution (internal validation set [IVS]) and 100 patients from other 24 institutions (external validation set [EVS]), respectively. The performance of each model was evaluated with the Sørensen-Dice similarity coefficient (DSC) compared with the ground truth. STATISTICAL TESTS: According to the normality of the data distribution verified by the Shapiro-Wilk test, variables with three or more categories were compared by the Kruskal-Wallis test with Dunn's post hoc analysis. RESULTS: A two-dimensional (2D) nnU-Net showed the highest median DSCs of 0.922 and 0.893 for the IVS and EVS, respectively. The nnU-Nets achieved superior performance in meningioma segmentation than the U-Nets. The DSCs of the 2D nnU-Net for small meningiomas less than 1 cm3 were 0.769 and 0.780 with the IVS and EVS, respectively. DATA CONCLUSION: A fully automated and accurate volumetric measurement tool for meningioma with clinically applicable performance for small meningioma using nnU-Net was developed. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

Simultaneous determination of L-tryptophan impurities in meat products.

L-tryptophan has been used as a feed additive for swine and poultry and as a nutrient supplement for humans. However, some impurities in L-tryptophan have been reported as causative components of eosinophilia-myalgia syndrome. Therefore, from a safety perspective, it is important to analyze meat samples for these impurities. This study aims to develop an analytical method for the simultaneous detection of L-tryptophan impurities in meat products using LC-MS/MS. Among the various impurities, detection methods for (S)-2-amino-3-(5-hydroxy-1H-indol-3-yl)propanoic acid (5-hydroxytryptophan) (HTP), 1-methyl-1,2,3,4-tetrahydro-ß-carboline-3-carboxylic acid (MTCA), 3a-hydroxy-1,2,3,3a,8,8a-hexahydropyrrolo-[2,3-b]-indole-2-carboxylic acid (PIC), and 1,1'-ethylidenebistryptophan (EBT) and 2-(3-indoylmethyl)-L-tryptophan (IMT) were developed. The developed method allowed simultaneous determination of these four impurities in 5 min. No interferences from the matrix were observed, and the method showed good sensitivity to each analyte. The method detection limit and limit of quantification in meat matrices were below 11.2 and 35.7 µg/kg, respectively.

Development of a deep learning-based auto-segmentation algorithm for hepatocellular carcinoma (HCC) and application to predict microvascular invasion of HCC using CT texture analysis: preliminary results.

BACKGROUND: Automatic segmentation has recently been developed to yield objective data. Prediction of microvascular invasion (MVI) of hepatocellular carcinoma (HCC) using radiomics has been reported. PURPOSE: To develop a deep learning-based auto-segmentation algorithm (DL-AS) for the detection of HCC and to predict MVI using computed tomography (CT) texture analysis. MATERIAL AND METHODS: We retrospectively collected training data from 249 patients with HCC and validation set from 35 patients. Lesions of the training set were manually drawn by radiologist, in the delayed phase. 2D U-Net was selected as the DL architecture. Using the validation set, one radiologist manually drew 2D and 3D regions of interest twice, and the developed DL-AS was performed twice with a one-month time interval. The reproducibility was calculated using intraclass correlation coefficients (ICC). Logistic regression was performed to predict MVI. RESULTS: ICC was in the range of 0.190-0.998/0.341-0.997 in the manual 3D/2D segmentation. In contrast, it was perfect in 3D/2D using DL-AS, with a success rate of 88.6% for the detection of HCC. For predicting MVI, sphericity was a significant parameter (odds ratio <0.001; 95% confidence interval <0.001-0.206; P = 0.020) for predicting MVI using 2D DL-AS. However, 3D DL-AS segmentation did not yield a predictive parameter. CONCLUSION: The auto-segmentation of HCC using DL-AS provides perfect reproducibility, although it failed to detect 11.4% (4/35). However, the extracted parameters yielded different important predictors of MVI in HCC. Sphericity was a significant predictor in 2D DL-AS and 3D manual segmentation, while discrete compactness was a significant predictor in 2D manual segmentation.

The Impact of Union Membership on Nursing Turnover and Job Satisfaction.

OBJECTIVE: The present study examines the scope of nursing unions and the relationship with turnover and job satisfaction among RNs. BACKGROUND: There is no recent empirical literature documenting evidence on workplace performance measures (turnover and job satisfaction) among unionized nurses at the national level. METHODS: This cross-sectional study analyzed secondary data (n = 43 960), the 2018 National Sample Survey of Registered Nurses. RESULTS: Approximately 16% of the sample reported being represented by labor unions. The overall nursing turnover rate for the sample was 12.8%. Unionized nurses were less likely than their counterparts to report turnover (mean, 10.9% vs 13.16%; P = 0.02) and less job satisfaction (mean, 3.20 vs 3.28). Union nurses were more likely than nonunion nurses to be male (12.72% vs 9.46%; P = 0.0004), minorities (37.65% vs 25.67%, P < 0.001), be employed in hospital settings (70.1% vs 57.9%, P = 0.001), but report fewer hours of work per week (mean, 36.73 vs 37.66; P = 0.003). Regression results revealed a positive association between union status and nursing turnover (odds ratio, 0.83; P < 0.05), whereas union status was inversely associated with job satisfaction (B = -0.13, P < 0.001) after adjusting for age, gender, race/ethnicity, care coordination time per week, weekly hours of work, and employment settings. CONCLUSION: Overall, job satisfaction was high among all nurses, regardless of union membership. However, when comparing specifically union and nonunion members, unionized nurses were less likely to report turnover, but were more likely to experience job dissatisfaction.

Structural characteristics of sediment humins from South Korean lakes and their phenanthrene binding compared to other carbon sources.

Sediment humins are extremely important for binding hydrophobic organic contaminants in rivers and lakes. Nonetheless, little is known about their structure and binding. We, therefore, examined the structure and phenanthrene sorption affinity of sediment humin samples upstream, midstream, and downstream from two artificial lakes in South Korea by using the elemental 13C-NMR analysis, Freundlich model, and Langmuir model. The characteristics and phenanthrene sorption affinity of sediment humins were also compared with those of sediment humic acids from similar origins as well as soil humins/humic acids in South Korea from previous studies by using principal component analysis. In both lakes, downstream sediment humins exhibited lower N/C, O/C, and (N + O)/C ratios, lower internal oxidation, and higher aliphaticity due to the presence of long-chain aliphatic compounds generated during anaerobic decomposition. The principal component analysis results also showed that C,H-alkyl, O-alkyl, and polar organic carbon contents were significantly different when comparing the up-mid stream and downstream sediment samples in Daecheong Lake. In addition, midstream sediment humin in Andong Lake presented higher C,H-alkyl and lower polar organic carbon contents compared to those of up-downstream samples. In both lakes, the sorption coefficient and adsorption isotherm linearity were positively correlated with the C,H-alkyl content and negatively correlated with the O-aryl content. Similar to C,H-alkyl and POC, C,H-alkyl, and (N + O)/C had an extremely high correlation coefficient when predicting the sorption coefficient (Freundlich model) and the maximum adsorption capacity (Langmuir model) of sediment humins. Sediment humins had higher C,H-alkyl contents and lower sorption coefficients than those of sediment humic acids and soil humins/humic acids. These findings provide key information for monitoring water quality and polycyclic aromatic hydrocarbon contamination in South Korean lake sediments.

Chemically Triggered Synthesis, Remodeling, and Degradation of Soft Materials.

Polymer topology dictates dynamic and mechanical properties of materials. For most polymers, topology is a static characteristic. In this article, we present a strategy to chemically trigger dynamic topology changes in polymers in response to a specific chemical stimulus. Starting with a dimerized PEG and hydrophobic linear materials, a lightly cross-linked polymer, and a cross-linked hydrogel, transformations into an amphiphilic linear polymer, lightly cross-linked and linear random copolymers, a cross-linked polymer, and three different hydrogel matrices were achieved via two controllable cross-linking reactions: reversible conjugate additions and thiol-disulfide exchange. Significantly, all the polymers, before or after topological changes, can be triggered to degrade into thiol- or amine-terminated small molecules. The controllable transformations of polymeric morphologies and their degradation herald a new generation of smart materials.

Artificial neural network for myelin water imaging.

PURPOSE: To demonstrate the application of artificial neural network (ANN) for real-time processing of myelin water imaging (MWI). METHODS: Three neural networks, ANN-IMWF , ANN-IGMT2 , and ANN-II, were developed to generate MWI. ANN-IMWF and ANN-IGMT2 were designed to output myelin water fraction (MWF) and geometric mean T2 of intra- and extra-cellular water signal (GMT2,IEW ), respectively, whereas ANN-II generates a T2 distribution. For the networks, gradient and spin echo data from 18 healthy controls (HC) and 26 multiple sclerosis patients (MS) were utilized. Among them, 10 HC and 12 MS had the same scan parameters and were used for training (6 HC and 6 MS), validation (1 HC and 1 MS), and test sets (3 HC and 5 MS). The remaining data had different scan parameters and were applied to exam effects of the scan parameters. The network results were compared with those of conventional MWI in the white matter mask and regions of interest. RESULTS: The networks produced highly accurate results, showing averaged normalized root-mean-squared error under 3% for MWF and 0.4% for GMT2,IEW in the white matter mask of the test set. In the region of interest analysis, the differences between ANNs and conventional MWI were less than 0.1% in MWF and 0.1 ms in GMT2,IEW (no statistical difference and R2 > 0.97). Datasets with different scan parameters showed increased errors. The average processing time was 0.68 s in ANNs, gaining 11,702 times acceleration in the computational speed (conventional MWI: 7,958 s). CONCLUSION: The proposed neural networks demonstrate the feasibility of real-time processing for MWI with high accuracy.

Advances in gradient echo myelin water imaging at 3T and 7T.

Gradient echo myelin water imaging (GRE-MWI) is an MRI technique to measure myelin concentration and involves the analysis of signal decay characteristics of the multi-echo gradient echo data. The method provides a myelin water fraction as a quantitative biomarker for myelin. In this work, a new sequence and post-processing methods were proposed to generate high quality GRE-MWI images at 3T and 7T. In order to capture the rapidly decaying myelin water signals, a bipolar readout GRE sequence was designed with "gradient pairing," compensating for the eddy current effects. The flip angle dependency from the multi-compartmental T1 effects was explored and avoided using a 2D multi-slice acquisition with a long TR. Additionally, the sequence was tested for the effects of inflow and magnetization transfer and demonstrated robustness to these error sources. Lastly, the temporal and spatial B0 inhomogeneity effects were mitigated by using the B0 navigator and field inhomogeneity corrections. Using the method, high-quality myelin water images were successfully generated for the in-vivo human brain at both field strengths. When the myelin water fraction at 3T and 7T were compared, they showed a good correlation (R2≥ 0.88; p < 0.001) with a larger myelin water fraction at 7T. The proposed method also opens the possibility of high resolution (isotropic 1.5 mm resolution) myelin water mapping at 7T.

Inhalation toxicity of benzalkonium chloride and triethylene glycol mixture in rats.

Benzalkonium chloride (BAC), a disinfectant, and triethylene glycol (TEG), an organic solvent/sanitizer, are frequently combined in commercially available household sprays. To assess the respiratory effect of this combination, Sprague-Dawley rats were exposed to an aerosol containing BAC (0.5%, w/v) and TEG (10%, w/v) for up to 2 weeks in a whole-body inhalation chamber. BAC (4.1-4.5 mg/m3, sprayed from 0.5% solution) promoted pulmonary cell damage and inflammation as depicted by the increase in total protein, lactate dehydrogenase, polymorphonuclear leukocytes, and macrophage inflammatory protein-2 in the bronchoalveolar lavage fluid, whereas TEG (85.3-94.5 mg/m3, sprayed from 10% solution) did not affect the lung. Rats exposed to the BAC/TEG mixture for 2 weeks showed severe respiratory symptoms (sneezing, wheezing, breath shortness, and chest tightness), but no lung damage or inflammation was observed. However, significant ulceration and degenerative necrosis were observed in the nasal cavities of rats repeatedly exposed to the BAC/TEG mixture. The mass median aerodynamic diameters of the aqueous, BAC, TEG and BAC/TEG aerosols were 1.24, 1.27, 3.11 and 3.24 µm, respectively, indicating that TEG-containing aerosols have larger particles than those of the aqueous and BAC alone aerosols. These results suggest that the toxic effects of BAC and BAC/TEG aerosols on the different respiratory organs may be associated with the difference in particle diameter, since particle size is important in determining the deposition site of inhaled materials.

Evaluation of pulmonary toxicity of benzalkonium chloride and triethylene glycol mixtures using in vitro and in vivo systems.

Benzalkonium chloride (BAC) is a widely used disinfectant/preservative, and respiratory exposure to this compound has been reported to be highly toxic. Spray-form household products have been known to contain BAC together with triethylene glycol (TEG) in their solutions. The purpose of this study was to estimate the toxicity of BAC and TEG mixtures to pulmonary organs using in vitro and in vivo experiments. Human alveolar epithelial (A549) cells incubated with BAC (1-10 µg/mL) for 24 hours showed significant cytotoxicity, while TEG (up to 1000 µg/mL) did not affect cell viability. However, TEG in combination with BAC aggravated cell damage and inhibited colony formation as compared to BAC alone. TEG also exacerbated BAC-promoted production of reactive oxygen species (ROS) and reduction of glutathione (GSH) level in A549 cells. However, pretreatment of the cells with N-acetylcysteine (NAC) alleviated the cytotoxicity, indicating oxidative stress could be a mechanism of the toxicity. Quantification of intracellular BAC by LC/MS/MS showed that cellular distribution/absorption of BAC was enhanced in A549 cells when it was exposed together with TEG. Intratracheal instillation of BAC (400 µg/kg) in rats was toxic to the pulmonary tissues while that of TEG (up to 1000 µg/kg) did not show any harmful effect. A combination of nontoxic doses of BAC (200 µg/kg) and TEG (1000 µg/kg) promoted significant lung injury in rats, as shown by increased protein content and lactate dehydrogenase (LDH) activity in bronchoalveolar lavage fluids (BALF). Moreover, BAC/TEG mixture recruited inflammatory cells, polymorphonuclear leukocytes (PMNs), in terminal bronchioles and elevated cytokine levels, tumor necrosis factor α (TNF-α), and interleukin 6 (IL-6) in BALF. These results suggest that TEG can potentiate BAC-induced pulmonary toxicity and inflammation, and thus respiratory exposure to the air mist from spray-form products containing this chemical combination is potentially harmful to humans.

Enhanced Gas-Sensing Performance of GO/TiO2 Composite by Photocatalysis.

Few studies have investigated the gas-sensing properties of graphene oxide/titanium dioxide (GO/TiO2) composite combined with photocatalytic effect. Room temperature gas-sensing properties of the GO/TiO2 composite were investigated towards various reducing gases. The composite sensor showed an enhanced gas response and a faster recovery time than a pure GO sensor due to the synergistic effect of the hybridization, such as creation of a hetero-junction at the interface and modulation of charge carrier density. However, the issue of long-term stability at room temperature still remains unsolved even after construction of a composite structure. To address this issue, the surface and hetero-junction of the GO/TiO2 composite were engineered via a UV process. A photocatalytic effect of TiO2 induced the reduction of the GO phase in the composite solution. The comparison of gas-sensing properties before and after the UV process clearly showed the transition from n-type to p-type gas-sensing behavior toward reducing gases. This transition revealed that the dominant sensing material is GO, and TiO2 enhanced the gas reaction by providing more reactive sites. With a UV-treated composite sensor, the function of identifying target gas was maintained over a one-month period, showing strong resistance to humidity.

Delays in Seeking Health Care: Comparison of Veterans and the General Population.

OBJECTIVES: Recent reports of delays in receiving care among U.S. Veterans have received national attention. Such delays may have an effect on veterans' propensity to seek healthcare as well, which could be detrimental to their health. There exists no evidence at the national level of the magnitude of perceived care delay in the Veterans Affairs (VA) system and how it compares to populations with other types of insurance coverage in the U.S. STUDY DESIGN AND METHODS: This cross-sectional study analyzed a recent nationally representative survey database (n = 10,907). Descriptive and multivariate regression analyses were performed to determine how care delay for veterans compares with the general population. RESULTS: We found that 17.2% of Americans delayed seeking needed healthcare in 2010-2011, but the figure was 29% for veterans. Also, there was a significant association of care delay with VA health care coverage after adjusting for other personal factors and region of the country. CONCLUSIONS: Study results suggest a possible link between VA access problems and veterans' behavior in seeking needed healthcare, which may be creating disparities in the effectiveness of care for this vulnerable and deserving population. Our study provides evidence of self-reported care delay among veterans. More studies are necessary to further understand factors in relation to delaying seeking care among veterans.

Physician report of industry gifts and quality of care.

BACKGROUND: Physician financial conflict of interest is a concern in the delivery of medicine because of its possible influence on the cost and the quality of patient care. There has been an extensive discussion of the ethical, economic, and legal aspects of this issue but little direct empirical evidence of its magnitude or effects. METHODOLOGY: A nationally representative survey (n = 4,720) was used to empirically examine physician self-report of receipt of financial gifts from the pharmaceutical and medical devices industry and its association with their ability to provide quality care. FINDINGS: Results indicate that the vast majority of physicians receive industry gifts in various forms, and the receipt of gifts is associated with lower perceived quality of patient care. There is also an inverse relationship between the frequency of received gifts and the perceived quality of care. PRACTICE IMPLICATIONS: Physicians need to be aware of the widespread receipt of industry gifts in medical practice and the potential adverse impact of such receipts on the delivery of care.

PERCEIVED RACIAL DISCRIMINATION AMONG HOME HEALTH AIDES: EVIDENCE FROM A NATIONAL SURVEY.

Home health aides are one of our essential human resources in the U.S. long-term care industry but understanding whether home health aides experience racial discrimination in the workplace and, if so, which personal/organizational factors are associated at the national level has been unnoticed. Using a nationally representative sample (n=3377), we attempt to investigate the association between racial discrimination and personal and organizational factors. The study found the 13.5% prevalence rate of racial discrimination. The study findings from multiple regression analysis reveal that black home care aides are more likely than white aides to experience racial discrimination in the workplace, suggesting that racial disparity may be an additional barrier to our home health care industry. National chain affiliation and low income were also found to be associated with perceived racial discrimination.

Effect of Physician IT Use on Practice Performance.

The role of information technology (IT) remains important within the medical community. However, little is known about the extent to which practicing physicians improve practice performance by having and utilizing IT at the national level. The present study, analyzing a national physician survey (n = 4,720), seeks to explore associations of IT availability and utilization with practice performance at the national level. The multivariate regression analysis results suggest that patient information IT functionality upholds physician advantages in annual income but prescription drug IT functionality was reversely linked to annual income. We also found a negative association of patient information IT functionality with patient visit volume. Our study results revealed mixed findings. Not all IT functionalities would offer benefits to practicing physicians. Our data suggest that patient information IT functionality can moderate cost concerns related to IT investment among providers.

Accuracy of blood loss estimations among anesthesia providers.

Research regarding the accuracy of estimating blood loss by anesthesia providers has been limited, generally consisting of small sample sizes with conflicting findings. We conducted a prospective study using a convenience sample of 91 anesthesia providers: 36 Certified Registered Nurse Anesthetists (CRNAs), 48 student registered nurse anesthetists (SRNAs), and 7 anesthesiologists. We applied a known volume of moulage (theater) blood to items at 4 stations and asked each participant to estimate the amount of blood present at each station. Descriptive and inferential statistical analyses were performed. Multivariate regression analysis was conducted to evaluate the association of estimation errors with covariates (provider training, gender, ethnicity, years of education, and years of anesthesia experience). We found no difference in the mean error of blood loss estimation based on provider training, gender, ethnicity, education, or experience. Although no statistically significant differences in blood loss estimations existed among provider groups, CRNAs had smaller mean errors than anesthesiologists in all 4 stations and smaller mean errors than SRNAs in 3 of 4 stations. Accuracy of estimating blood loss, measured by mean error, was not related to type of provider training, years of education or experience, gender, or ethnicity. Anesthesia providers may not estimate blood loss accurately.

A profile of solo/two-physician practices.

Understanding practice behaviors of solo/dual physician ownership and associated factors at the national level is important information for policymakers and clinicians in response to the Affordable Care Act (ACA) of 2010, but poorly understood in the literature. We analyzed nationally representative data (n = 4,720). The study results reveal nearly 33% of the sample reported solo/two-physician practices. Male/minority/older physicians, psychiatrists, favor small practices. Greater market competition was perceived and less charity care was given among solo/two-physician practitioners. The South region was favored by small physician practitioners. Physicians in solo or two-person practices provided fewer services to chronic patients and were dissatisfied with their overall career in medicine. Small practices were favored by international medical graduates (IMGs) and primary care physicians (PCPs). Overall our data suggest that the role of solo/dual physician practices is fading away in the delivery of medicine. Our findings shed light on varied characteristics and practice behaviors of solo/two-physician practitioners, but more research may be needed to reevaluate the potential role of small physician practitioners and find a way to foster a private physician practice model in the context of the newly passed ACA of 2010.

The relationship between arterial carbon dioxide tension and end-tidal carbon dioxide tension in intubated adults with traumatic brain injuries who required emergency craniotomies.

Anesthetic management of patients who have suffered traumatic brain injuries can be challenging. We investigated the relationship between arterial to end-tidal carbon dioxide partial pressure gradients (Pa-etCO2) and 3 predictor variables: (1) injury severity score, (2) use of positive end-expiratory pressure, and (3) presence of rib fractures. Using a convenient sampling method, we sampled 56 patients who arrived to the operating room intubated after traumatic brain injuries between 2005 and 2011. Two groups were compared retrospectively: those with Pa-etCO2 greater than 10 mm Hg (case group) (n = 37) and those with Pa-etCO2 gradients of 10 mm Hg or less (control group) (n = 19). Descriptive and inferential statistics were employed to identify any differences between the groups. Stepwise regression was also performed. Cross tabulation analysis revealed that injury severity score of 30 or more was a predictor of Pa-etCO2 gradient. Stepwise regression analysis revealed the presence of rib fracture and body mass index to be significant predictors of Pa-etCO2 gradient (P < .011). This study identified coexisting conditions in which the patients' Pa-etCO2 gradients were large. Results showed that injury severity score of 30 or more, the presence of rib fractures, and higher body mass index were statistically significant predictors of Pa-etCO2 gradients greater than 10 mm Hg. These observations should be considered when evaluating PetCO2 in conjunction with arterial blood gas analysis to determine optimal ventilation status of these patients.