Comparison of the Wells score with the revised Geneva score for assessing pretest probability of pulmonary embolism in hospitalized elderly patients

Early Detection of Pulmonary Embolism in a General Patient Population Immediately Upon Hospital Admission Using Machine Learning to Identify New, Unidentified Risk Factors: Model Development Study

BACKGROUND

Under- or late identification of pulmonary embolism (PE)-a thrombosis of 1 or more pulmonary arteries that seriously threatens patients' lives-is a major challenge confronting modern medicine.

OBJECTIVE

We aimed to establish accurate and informative machine learning (ML) models to identify patients at high risk for PE as they are admitted to the hospital, before their initial clinical checkup, by using only the information in their medical records.

METHODS

We collected demographics, comorbidities, and medications data for 2568 patients with PE and 52,598 control patients. We focused on data available prior to emergency department admission, as these are the most universally accessible data. We trained an ML random forest algorithm to detect PE at the earliest possible time during a patient's hospitalization-at the time of his or her admission. We developed and applied 2 ML-based methods specifically to address the data imbalance between PE and non-PE patients, which causes misdiagnosis of PE.

RESULTS

The resulting models predicted PE based on age, sex, BMI, past clinical PE events, chronic lung disease, past thrombotic events, and usage of anticoagulants, obtaining an 80% geometric mean value for the PE and non-PE classification accuracies. Although on hospital admission only 4% (1942/46,639) of the patients had a diagnosis of PE, we identified 2 clustering schemes comprising subgroups with more than 61% (705/1120 in clustering scheme 1; 427/701 and 340/549 in clustering scheme 2) positive patients for PE. One subgroup in the first clustering scheme included 36% (705/1942) of all patients with PE who were characterized by a definite past PE diagnosis, a 6-fold higher prevalence of deep vein thrombosis, and a 3-fold higher prevalence of pneumonia, compared with patients of the other subgroups in this scheme. In the second clustering scheme, 2 subgroups (1 of only men and 1 of only women) included patients who all had a past PE diagnosis and a relatively high prevalence of pneumonia, and a third subgroup included only those patients with a past diagnosis of pneumonia.

CONCLUSIONS

This study established an ML tool for early diagnosis of PE almost immediately upon hospital admission. Despite the highly imbalanced scenario undermining accurate PE prediction and using information available only from the patient's medical history, our models were both accurate and informative, enabling the identification of patients already at high risk for PE upon hospital admission, even before the initial clinical checkup was performed. The fact that we did not restrict our patients to those at high risk for PE according to previously published scales (eg, Wells or revised Genova scores) enabled us to accurately assess the application of ML on raw medical data and identify new, previously unidentified risk factors for PE, such as previous pulmonary disease, in general populations.

Diagnostic management of inpatients with a positive D-dimer test: developing a new clinical decision-making rule for pulmonary embolism

Background

A positive D-dimer test has high sensitivity but relatively poor specificity for the diagnosis of pulmonary embolism, causing difficulty for clinicians unskilled in pulmonary embolism diagnosis in determining whether a patient with a positive D-dimer test needs to undergo computed tomographic pulmonary angiography.

Objectives

We sought to develop a new clinical decision-making rule based on a positive D-dimer result to predict the probability of pulmonary embolism and to guide clinicians in making decisions regarding the need for computed tomographic pulmonary angiography.

Methods

We conducted a prospective, multicenter study in three hospitals in China. A total of 3014 inpatients with positive D-dimer results were included. In the derivation group, we built a multivariate logistic regression model and deduced a regression equation from which our score was derived. Finally, we validated the score in an independent cohort.

Results

Our score included nine variables (points): chest pain (1.4), chest tightness (2.3), shortness of breath (3.6), hemoptysis (3.4), heart rate ≥100 beats/min (3.6), blood gas analysis (2.9), electrocardiogram presenting a typical S1Q3T3 pattern (4.1), electrocardiogram findings (2.4), and ultrasonic cardiogram findings (3.7). The sensitivities and specificities were 100% and 86.94%, respectively, in the derivation group and 100% and 90.82%, respectively, in the validation group. Additionally, the observed and predicted proportions of patients who underwent computed tomographic pulmonary angiography were 16.82% and 10.76%, respectively, in the derivation group and 18.72% and 11.40%, respectively, in the validation group.

Conclusions

The new score can categorize inpatients with a positive D-dimer test as pulmonary embolism-likely or pulmonary embolism-unlikely, thus reducing unnecessary computed tomographic pulmonary angiography examinations.

Comparison of the Wells score and the revised Geneva score as a tool to predict pulmonary embolism in outpatients over age 65

TITLE

Comparison of the Wells score and the revised Geneva score as a tool to predict pulmonary embolism in outpatients over 65 years of age.

INTRODUCTION

The incidence and mortality of pulmonary embolism (PE) is high in the elderly. The Wells score (SW) and the revised Geneva score (RGS) have been validated in patient populations with a large age range. The aim of this study was to compare the predictive accuracy of these two scores in diagnosis of PE in patients over 65 years of age.

METHOD

A prospective multicentre study (nine French and three Belgian centres) was conducted at the same time as the PERCEPIC study. A total of 1757 patients admitted with suspected PE were included and divided into two groups according to age (≥65 years or <65 years). The pre-test probability of PE was assessed prospectively for the RGS. The SW was calculated retrospectively. The predictive accuracy of the two scores was compared by the area under the curve (AUC) of the ROC curves.

RESULTS

The overall prevalence of PE was 11.3%. The prevalence among patients aged ≥65 in the low, moderate and high pre-test probability groups, evaluated using the WS and was respectively 13.5% (CI 95%: CI 9.9-17.3), 28.2% (CI 22.1-34.3), 50% (CI 26-74) and 8.1% (CI 3.2-12.9), 22.3% (CI 18.2-26.3), 43.7% (CI 25.6-61.9) using the RGS. The AUC for the WS and RGS for patients aged ≥65 was 0.632 (CI 0.574-0.691) and 0.610 (CI 0.555-0.666). The difference between the AUCs was not statistically significant (p = .441).

CONCLUSION

In the population for this study, the WS and RGS have the same PE diagnostic accuracy in patients over age 65. This result should be validated in a prospective study that directly compares these scores.

Short and long-term mortality in elderly patients with suspected not confirmed pulmonary embolism

INTRODUCTION

Most patients evaluated for suspected pulmonary embolism(PE) conclude the Emergency Department(ED) work-up with a diagnosis of PE not confirmed(PE excluded;PE-E). We aimed to investigate the clinical features, short and long-term mortality, and prognostic factors for death in elderly with PE-E, and to compare these figures with those of patients with PE confirmed(PE-C).

METHODS

Consecutive patients ≥65 years old evaluated in the ED for clinically suspected hemodynamically stable acute PE were included in this retrospective cohort study.

RESULTS

Study population: 657 patients with suspected PE, PE-C:162(24.65%). When compared with PE-C, patients with PE-E presented a higher prevalence of chronic cardiopulmonary disease (17.37% vs 8.02%, p = 0.003), a lower prevalence of pulse rate >110 (13.13% vs 25.93%; p<0.001), of arterial oxygen saturation <90% (16.16% vs. 25.93%; p = 0.007) and of hospitalized patients (52.93% vs 98.15%; p < 0.001). Thirty-day, 90-day, 1-year, 2-year and 5-year overall mortality was 8.83%, 15.98%, 23.59%, 29.68%, and 51.09%, respectively, differences between PE-E and PE-C non statistically significant. Among patients with PE-E, multivariate analysis showed that simplified Pulmonary Embolism Severity Index score>0 was associated with higher short and long-term mortality (30-day:HR:5.31,p = 0.029; 5 year:HR:2.18, p < 0.001), meanwhile comorbidity (Charlson Comorbidity Index>0) only with higher long-term mortality (30-day: HR:1.60, p = 0.342; 5 year: HR:1.41, p = 0.038).

CONCLUSION

In real world haemodinamically stable elderly patients evaluated in the ED for suspected PE, short and long-term mortality was markedly high regardless whether PE was confirmed or excluded. At the time to set management and follow up strategies, elderly patients with PE excluded should not be considered a low-risk population.

The diagnosis and management of pulmonary embolism

Pulmonary embolism (PE) is a condition characterised by an obstruction of the pulmonary arterial system by one or more emboli. Advanced clinical practitioners are often faced with ruling out a diagnosis of PE in patients with non-specific symptoms such as dyspnoea and pleuritic chest pain, which can be fairly mild and therefore a diagnosis of PE easily missed. PEs can be a challenge to diagnose, especially in elderly people, since it can be difficult to differentiate their symptoms from other less serious illnesses. Widely used scoring tools are helpful to calculate a patient's probability of having a PE. The Wells score is the most widely used pre-test clinical probability indicator of PE used in the UK, which scores the patient's probability of having a PE based on their risk factors. The D-dimer test is a relatively simple investigation to rule out venous thromboembolism (VTE) but can be raised for various reasons other than PE. Computed tomography pulmonary angiography (CTPA) is regarded as the gold standard imaging modality for investigation of acute PE but ventilation-perfusion (VQ) scans can be used as an alternative imaging technique for diagnosing PE in those where CTPA is contraindicated. Thrombolysis is underused in clinical practice due to the fear of adverse bleeding events. Patients without a massive or sub-massive PE are treated with anticoagulant therapy, usually commencing with subcutaneous low-molecular-weight heparin and switching over to a direct oral anticoagulant (DOAC). There has been a shift away from treatment with warfarin for the prevention and treatment of VTE over the past decade.

Predictive Model for Pulmonary Embolism in Patients with Deep Vein Thrombosis

BACKGROUND

To develop and verify a risk predictive model/scoring system for pulmonary embolism (PE) among hospitalized patients with deep venous thrombosis of the lower extremities (LDVT).

METHODS

776 patients with LDVT were enrolled in a case-control study between January 2016 and June 2017 from the Vascular Surgery Department of Shanxi Dayi Hospital, China. They were randomly divided into development (543 patients, 70%) and validation (233 patients, 30%) databases. Based on the results of pulmonary computed tomography arteriography, patients were divided into 2 categories; those with PE were designated as the case group, whereas those without comprised the controls. A logistic regression model and scoring system for PE in patients with LDVT was established in the development database and verified in the validation database. Scoring system (Shanxi Dayi Hospital score [SDH score]) was tabulated as follows: right lower extremity or bilateral lower extremities, 1; surgery or immobilization, 1; malignant tumor, 1; history of venous thromboembolism (VTE), 2; D-dimer >1,000 ng/mL, 2; and unprovoked, 2. Calibration and discrimination of the model were assessed by the Hosmer-Lemeshow goodness of fit test and the area under the receiver operating characteristic curve (AUC). Wells score, the Revised Geneva score, and the SDH score for predictive value of PE by AUC in the validation database were compared.

RESULTS

776 patients with LDVT were divided into 2 risk categories based on the scores from the risk model as follows: PE unlikely (score <3) and PE likely (score ≥3). Sensitivity, specificity, and crude agreement of the SDH score in the development database were 76.39%, 55.89%, and 61.33%, respectively. In the validation database, the logistic regression model showed good calibration and discriminative power. The Hosmer-Lemeshow goodness of fit test P value was >0.05, and the AUC was 0.705 (95% CI: 0.634-0.776, P < 0.001). The SDH score also showed good discriminative power, and the AUC was 0.702 (95% CI: 0.631-0.774, P < 0.001). Sensitivity, specificity, and crude agreement of the SDH score in the validation database were 67.61%, 61.73%, and 63.52%, respectively. AUC for the Wells score and the Revised Geneva score was 0.611 (95% CI: 0.533-0.688, P = 0.007) and 0.585 (95% CI: 0.503-0.666, P = 0.040), respectively. Difference of the AUC was not statistically significant between the Wells score and the SDH score (0.611 vs. 0.702, P = 0.059) but was so between the Revised Geneva score and the SDH score (0.585 vs. 0.702, P = 0.016). Sensitivity of the Wells score, Revised Geneva score, and the SDH score (64.79%, 67.61% vs. 67.61%) was not statistically significant. However, the specificity of the Wells score and Revised Geneva score was significantly lower than that of the SDH score (48.77%, 39.51% vs. 61.73%).

CONCLUSIONS

Our logistic regression model and the SDH score based on 7 risk factors as right lower extremity, bilateral lower extremities, unprovoked, surgery or immobilization, malignant tumor, history of VTE, and D-dimer>1,000 ng/mL showed good calibration and discriminative power for the assessment of PE risk in patients with LDVT. The SDH score is more specific for PE prediction in the Chinese population, compared with the Wells score and the Revised Geneva score.