Incremental diagnostic quality gain of CTA over V/Q scan in the assessment of pulmonary embolism by means of a Wells score Bayesian model: results from the ACDC collaboration

Assessment of a Comparative Bayesian-Enhanced Population-Based Decision Model for COVID-19 Critical Care Prediction in the Dominican Republic Social Security Affiliates

Introduction: The novel coronavirus disease 2019 (COVID-19) has been a major health concern worldwide. This study aims to develop a Bayesian model to predict critical outcomes in patients with COVID-19.

Methods: Sensitivity and specificity were obtained from previous meta-analysis studies. The complex vulnerability index (IVC-COV2 index for its abbreviation in Spanish) was used to set the pretest probability. Likelihood ratios were integrated into a Fagan nomogram for posttest probabilities, and IVC-COV2 + National Early Warning Score (NEWS) values and CURB-65 scores were generated. Absolute and relative diagnostic gains (RDGs) were calculated based on pretest and posttest differences.

Results: The IVC-COV2 index was derived from a population of 1,055,746 individuals and was based on mortality in high-risk (71.97%), intermediate-risk (26.11%), and low-risk (1.91%) groups. The integration of models in which IVC-COV2 intermediate + NEWS ≥ 5 and CURB-65 > 2 led to a "number needed to (NNT) diagnose" that was slightly improved in the CURB-65 model (2 vs. 3). A comparison of diagnostic gains revealed that neither the positive likelihood ratio (P = 0.62) nor the negative likelihood ratio (P = 0.95) differed significantly between the IVC-COV2 NEWS model and the CURB-65 model.

Conclusion: According to the proposed mathematical model, the combination of the IVC-COV2 intermediate score and NEWS or CURB-65 score yields superior results and a greater predictive value for the severity of illness. To the best of our knowledge, this is the first population-based/mathematical model developed for use in COVID-19 critical care decision-making.

A Bayesian decision support sequential model for severity of illness predictors and intensive care admissions in pneumonia

Background

Community-acquired pneumonia (CAP) is one of the leading causes of morbidity and mortality in the USA. Our objective was to assess the predictive value on critical illness and disposition of a sequential Bayesian Model that integrates Lactate and procalcitonin (PCT) for pneumonia.

Methods

Sensitivity and specificity of lactate and PCT attained from pooled meta-analysis data. Likelihood ratios calculated and inserted in Bayesian/ Fagan nomogram to calculate posttest probabilities. Bayesian Diagnostic Gains (BDG) were analyzed comparing pre and post-test probability. To assess the value of integrating both PCT and Lactate in Severity of Illness Prediction we built a model that combined CURB65 with PCT as the Pre-Test markers and later integrated the Lactate Likelihood Ratio Values to generate a combined CURB 65 + Procalcitonin + Lactate Sequential value.

Results

The BDG model integrated a CUBR65 Scores combined with Procalcitonin (LR+ and LR-) for Pre-Test Probability Intermediate and High with Lactate Positive Likelihood Ratios. This generated for the PCT LR+ Post-test Probability (POSITIVE TEST) Posterior probability: 93% (95% CI [91,96%]) and Post Test Probability (NEGATIVE TEST) of: 17% (95% CI [15–20%]) for the Intermediate subgroup and 97% for the high risk sub-group POSITIVE TEST: Post-Test probability:97% (95% CI [95,98%]) NEGATIVE TEST: Post-test probability: 33% (95% CI [31,36%]) . ANOVA analysis for CURB 65 (alone) vs CURB 65 and PCT (LR+) vs CURB 65 and PCT (LR+) and Lactate showed a statistically significant difference (P value = 0.013).

Conclusions

The sequential combination of CURB 65 plus PCT with Lactate yielded statistically significant results, demonstrating a greater predictive value for severity of illness thus ICU level care.

Wells and Geneva Scores Are Not Reliable Predictors of Pulmonary Embolism in Critically Ill Patients: A Retrospective Study

BACKGROUND

Critically ill patients are at high risk for pulmonary embolism (PE). Specific PE prediction rules have not been validated in this population. The present study assessed the Wells and revised Geneva scoring systems as predictors of PE in critically ill patients.

METHODS

Pulmonary computed tomographic angiograms (CTAs) performed for suspected PE in critically ill adult patients were retrospectively identified. Wells and revised Geneva scores were calculated based on information from medical records. The reliability of both scores as predictors of PE was determined using receiver operating characteristic (ROC) curve analysis.

RESULTS

Of 138 patients, 42 (30.4%) were positive for PE based on pulmonary CTA. Mean Wells score was 4.3 (3.5) in patients with PE versus 2.7 (1.9) in patients without PE (P < .001). Revised Geneva score was 5.8 (3.3) versus 5.1 (2.5) in patients with versus without PE (P = .194). According to the Wells and revised Geneva scores, 56 (40.6%) patients and 49 (35.5%) patients, respectively, were considered as low probability for PE. Of those considered as low risk by the Wells score, 15 (26.8%) had filling defects on CTA, including 2 patients with main pulmonary artery embolism. The area under the ROC curve was 0.634 for the Wells score and 0.546 for the revised Geneva score. Wells score >4 had a sensitivity of 40%, specificity of 87%, positive predictive value of 59%, and negative predictive value of 77% to predict risk of PE.

CONCLUSIONS

In this population of critically ill patients, Wells and revised Geneva scores were not reliable predictors of PE.

Computed tomography for suspected pulmonary embolism results in a large number of non-significant incidental findings and follow-up investigations

INTRODUCTION

Computed tomographic pulmonary angiograms (CTPAs) are often ordered to evaluate pulmonary embolism (PE) in the emergency department (ED). The increase use of CTPA has led to an increase in incidental findings, often of low clinical significance. Our objectives were to (1) assess the prevalence and clinical significance of incidental findings identified in patients evaluated with CTPAs for PE in the ED, (2) evaluate follow-up investigations for these incidental findings, and (3) assess the utility of routine chest X-rays done prior to CTPA.

METHODS

This is a historical cohort study of adult patients, presenting to two tertiary care EDs from January-December 2015, evaluated with CTPA for possible PE. Two reviewers' extracted data from electronic CT records in a standardized fashion with inter-rater reliability reported using the kappa statistic. We measured the prevalence of PE and stratified non-PE findings according to alternative diagnoses and incidental findings. Data were reported as mean and standard deviation (SD). Univariate analyses were performed with t test for continuous variables.

RESULTS

A total of 1708 studies were included (mean 62 years (SD 16.7), 56.9% female). PE was found in 233 (13.6%) patients. A total of 223 (13.1%) patients had an incidental finding, the majority of which included pulmonary nodules (n = 83, 37.2%) and adenopathy (n = 26, 11.6%). Of the incidental findings, 197 (88.3%) were non-significant and led to no definitive diagnosis of cancer. In patients who underwent both CTPA and chest X-ray, X-ray reports revealed the same diagnosis in 77% of PE-negative patients without missing a clinically significant incidental finding.

CONCLUSIONS

Incidental findings are as common as a diagnosis of PE in patients undergoing CTPA. They are rarely clinically significant. Chest radiograph remains a reasonable initial investigation as it can aid in identifying alternative diagnoses especially in the setting of a low pre-test probability for PE.