Your privacy, your choice

We use essential cookies to make sure the site can function. We also use optional cookies for advertising, personalisation of content, usage analysis, and social media.

By accepting optional cookies, you consent to the processing of your personal data - including transfers to third parties. Some third parties are outside of the European Economic Area, with varying standards of data protection.

See our privacy policy for more information on the use of your personal data.

for further information and to change your choices.
Skip to main content
Download PDF

Development of a predictive nomogram for early identification of pulmonary embolism in hospitalized patients: a retrospective cohort study

BMC Pulmonary Medicine volume 24, Article number: 594 (2024) Cite this article

Abstract

Background

Hospitalized patients often present with complex clinical conditions, but there is a lack of effective tools to assess their risk of pulmonary embolism (PE). Therefore, our study aimed to develop a nomogram model for better predicting PE in hospitalized populations.

Methods

Data from hospitalized patients (aged ≥ 15 years) who underwent computed tomography pulmonary angiography (CTPA) to confirm PE and non-PE were collected from December 2013 to April 2023. Univariate and multivariate stepwise logistic regression analyses were conducted to identify independent predictors of PE, followed by the construction of a predictive nomogram and internal validation. The efficiency and clinical utility of the nomogram model were assessed using receiver operating characteristic (ROC) curve, decision curve analysis (DCA), and clinical impact curve (CIC).

Results

The study included 313 PE and 339 non-PE hospitalized patients. Male gender, dyspnea or shortness of breath, interstitial lung disease, lower limb deep vein thrombosis, elevated fibrin degradation product (FDP), pulmonary arterial hypertension, and tricuspid regurgitation were identified as independent risk factors. The AUC of the predictive nomogram model was 0.956 (95% CI: 0.939–0.974), demonstrating superior performance compared with the simplified Wells score of 0.698 (95% CI: 0.654–0.741) and the modified Geneva score of 0.758 (95% CI: 0.717–0.799).

Conclusion

Our study demonstrated that challenges remain in the accuracy of the Wells score and revised Geneva score in assessing PE in hospitalized patients. Fortunately, the nomogram we developed has shown a favorable ability to discriminate PE cases, providing high reference value for clinical practice. However, given that this was a single-center study, we plan to expand efforts to collect data from additional centers to further validate our model.

Graphical Abstract

Peer Review reports

Background

Pulmonary embolism (PE) is a clinical syndrome primarily characterized by obstruction of the pulmonary arterial system by diverse emboli, resulting in disruptions to pulmonary circulation and respiratory function [1]. In recent years, despite advancements in treatment strategies, the mortality associated with PE has declined. Nevertheless, globally, the annual incidence of PE remains high, reaching 1 per 1000 individuals, thus persisting as a substantial contributor to patient death, particularly following stroke and coronary heart disease [2, 3]. This burden is further exacerbated by the COVID-19 outbreak [4].

PE frequently evades timely diagnosis due to its nonspecific clinical presentation [5]. Studies have indicated that up to 59% of venous thromboembolism (VTE) cases remain undiagnosed until postmortem examination [5, 6]. Thus, early detection and timely diagnosis of PE are of paramount importance. Computed tomography pulmonary angiography (CTPA) has become the predominant method for confirming PE owing to its high diagnostic efficacy. However, its use presents certain challenges, particularly in bedridden or critically ill patients, where the procedure carries elevated risks. Moreover, inappropriate use of CTPA can result in the overdiagnosis of subsegmental PE, unnecessary radiation exposure, and an increased risk of contrast-induced nephropathy [7, 8].

To reduce unnecessary CTPA examinations and minimize scan-related adverse reactions, many risk assessment tools for PE have been developed. Unfortunately, due to various factors such as regional disease variability and differences in the clinical settings of patients, these assessment tools do not consistently demonstrate high diagnostic efficacy, and their clinical generalizability is limited [9, 10]. Especially among hospitalized patients, complex clinical conditions, severity of illness, and other factors may influence the accuracy of scoring. For instance, widely used tools such as the Wells score, revised Geneva score, and Charlotte criteria exhibit lower predictive accuracy for hospitalized patients compared to outpatients and emergency patients [11,12,13].

Therefore, this study was conducted to develop and validate a predictive nomogram for early identification of PE in hospitalized populations by retrospectively collecting and analyzing clinical data from hospitalized patients, thereby offering a reference for early detection of PE in these patients.

Methods

Study population

A retrospective cohort study was carried out by gathering medical records of hospitalized patients who underwent CTPA between December 2013 and April 2023 at Guizhou Provincial People’s Hospital. This was a long-term retrospective cohort study. Throughout the research process, we maintained strict confidentiality of subjects’ medical records and study data. When publishing results, no key information related to the subjects would be disclosed. The risks to participants in this study are no greater than minimal risk, and the absence of this research would not impact patient outcomes. Under these conditions, we followed the ethical guidelines established by the Ethics Committee of Guizhou Provincial People’s Hospital, obtained a waiver of informed consent (ethics no: gzwkj2021-089), and strictly adhered to the principles outlined in the Declaration of Helsinki.

Inclusion criteria: (1) Patients aged ≥ 15 years, hospitalized; (2) Completed CTPA, cardiac color Doppler ultrasound, B-ultrasound examination of lower extremity vessels, coagulation function tests (including D-dimer, fibrin (fibrinogen) degradation product [FDP], fibrinogen [Fbg], prothrombin time [PT], activated partial thromboplastin time [APTT], thrombin time [TT], and international normalized ratio [INR]), blood troponin, and BNP tests; (3) No anticoagulation or thrombolysis treatment was administered prior to PE diagnosis during this hospitalization. Exclusion criteria: (1) Patients aged ≤ 14 years, non-hospitalized; (2) Relevant examinations were not completed.; (3) Received anticoagulation or thrombolysis treatment prior to PE diagnosis during this hospitalization. All patients in this study were admitted to the hospital following evaluation by specialized clinicians. In this study, CTPA, cardiac color Doppler ultrasound, and B-ultrasound examination of lower extremity vessels were each conducted by trained specialists in the respective fields, and the examination results were issued by experienced specialists with diagnostic qualifications.

Grouping and diagnostic criteria

In our study, patients were divided into two groups, with the PE group as the case group and the non-PE group as the control. The distinction between PE and non-PE was made based on the CTPA examination results, which were conducted by radiology specialists. The CTPA results were interpreted and issued by experienced radiologists with diagnostic qualifications.

The diagnostic criteria for PE via CTPA included direct signs such as partial filling defects in the vessel lumen, complete occlusion, floating thrombi, saddle embolism, irregular thickening of the vessel wall, and thrombus calcification. Indirect signs included uneven pulmonary vascular distribution, uneven perfusion of lung parenchyma forming the “mosaic” pattern, signs of pulmonary infarction, enlargement of the main pulmonary artery, right ventricular enlargement, pulmonary hypertension, and manifestations of right heart dysfunction. It is important to note that the indirect signs of PE are not specific and require careful differential diagnosis during the interpretation of CTPA [14]. The diagnostic criteria for PE were based on the 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism [5]. Non-PE cases were defined as hospitalized patients with suspected PE who underwent CTPA and had PE definitively excluded.

Data collection

Data were collected through the hospital’s electronic medical record system, specifically: (1) General clinical data: sex, age, smoking and alcohol consumption history; (2) Comorbidities or underlying diseases: chronic obstructive pulmonary disease (COPD), bronchiectasis, pneumonia, interstitial lung disease, connective tissue disease, respiratory failure, coronary heart disease, hypertension, diabetes mellitus, hypoproteinemia, dyslipidemia, other infections, anemia, active tumors, etc.; (3) Clinical manifestations: dyspnea or shortness of breath, chest pain, syncope, cough, pyrexia, cyanosis, and hemoptysis, as well as heart rate, lower limb deep vein tenderness, and edema; (4) Auxiliary examinations meeting inclusion criteria.

Validation of existing tools

Based on the 2018 edition of the “Chinese Guidelines for the Diagnosis, Treatment, and Prevention of Pulmonary Thromboembolism” and related literature, the simplified Wells score and revised Geneva score (Table S1) [15] were adopted and validated using the collected data in this study.

Statistical analysis

Data statistical analysis was conducted using R v4.3.3. Missing data in our study population were minimal and were addressed using multiple imputation. Continuous variables were expressed as mean ± standard deviation. For between-group comparisons, data with normal distribution were analyzed using the t-test, while data with non-normal distribution were analyzed using non-parametric tests. Categorical data were represented as proportions, with between-group comparisons conducted using the χ2 test. Statistical significance was defined at a two-sided P < 0.05.

Patients were randomly split into a training set and a validation set at a 7:3 ratio. In the training set, variables with a P < 0.1 in univariate analysis were included in a multivariate stepwise logistic regression to identify independent predictors for PE, from which a predictive nomogram was constructed. Model discrimination was evaluated using the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUROC), with an AUC > 0.7 indicating acceptable discrimination. Model calibration for the training and validation sets was validated using the Bootstrap method (200 repetitions). The Hosmer-Lemeshow test and the coefficient of determination were employed to evaluate the goodness of fit of the model. A P value greater than 0.05 indicated no statistically significant difference, suggesting favorable calibration of the standard model. Decision curve analysis (DCA) was utilized for the assessment of the clinical utility of the model, to determine the net benefit based on threshold probabilities. The clinical impact curve (CIC) was used to analyze the clinical efficacy of the model. Lastly, the Delong test was employed to analyze differences in predictive performance between the model and the simplified Wells and revised Geneva scores. Statistical significance was confirmed when a P < 0.05. The schematic diagram of the study procedure is shown in Figure S1.

Results

Comparison of clinical characteristics between training and validation sets

This study included 652 hospitalized patients, with 313 in the PE group (190 males, average age: 63.75 ± 14.84 years; 123 females, average age: 63.77 ± 14.83 years) and 339 in the non-PE group (164 males, average age: 55.76 ± 15.97 years; 175 females, average age: 55.76 ± 16.09 years). The training set comprised 456 patients (241 males and 215 females, average age: 59.69 ± 15.89 years), and the validation set included 196 patients (113 males and 83 females, average age: 59.40 ± 15.99 years) (Table S2).

Analysis of differences in predictors between PE and non-PE patients in the training set

In the training set, univariate analysis revealed statistically significant differences (P < 0.05) in 32 predictors between PE and non-PE patients, including sex, age, dyspnea or shortness of breath, syncope, cough, cyanosis, heart rate, symptoms and signs of deep vein thrombosis (DVT), history of PE or DVT, immobilization or surgery within 4 weeks, pneumonia, interstitial lung disease, type I and II respiratory failure, coronary artery atherosclerotic heart disease, dyslipidemia, active tumor, arterial atherosclerosis, D-dimer, FDP, Fbg, PT, APTT, TT, INR, pleural effusion, lower limb DVT, right ventricular enlargement, pulmonary arterial hypertension (PAH), tricuspid regurgitation, troponin, and brain natriuretic peptide (BNP) (Table S3).

Selection of independent predictors and construction of the predictive nomogram

Multivariate logistic regression analysis of the 32 identified predictors showed that being male, experiencing dyspnea or shortness of breath, having interstitial lung disease, presence of lower limb DVT, elevated FDP, PAH, and tricuspid regurgitation were independent predictors for PE occurrence in hospitalized patients (Table 1). Utilizing these factors, a nomogram was constructed to predict the risk of PE in hospitalized patients (Fig. 1).

Table 1 Independent predictors selection by multivariate logistic regression analysis
Full size table
Fig. 1
figure 1

Nomogram model to predict the risk of Pulmonary Embolism in hospitalized patients. Note: “Points” represents the individual score for each independent predictor at different values, with a total score of 100. “Total Points” represents the sum of all individual scores corresponding to each variable, with a maximum score of 350. “Disease Risk” indicates the predicted probability of PE occurrence based on the total score of all variables, ranging from 10–90%

Full size image

Analysis of discrimination and calibration of the nomogram model

The ROC curves in the training and validation sets exhibited AUC values of 0.956 (95% CI: 0.939–0.974) (Fig. 2A) and 0.929 (95% CI: 0.895–0.926) (Fig. 2E). The optimal predicted probability from the prediction model was 0.58, corresponding to a sensitivity of 0.878 and a specificity of 0.906 in the training set, and a sensitivity of 0.815 and a specificity of 0.865 in the validation set, respectively, indicating favorable consistency of the model relative to the observed outcomes and its ability to distinguish between PE and non-PE. The calibration analysis demonstrated that the calibrated curves in both sets closely approached the reference line (Fig. 2B and F), indicating fair calibration and fit of our model.

Fig. 2
figure 2

The related curves of the model for predicting the risk of pulmonary embolism in hospitalized patients. A. ROC curve of Training set; B. Calibration curve of Training set; C. DCA curve of Training set; D. CIC curve of Training set; E. ROC curve of Validation set; F. Calibration curve of Validation set; G. DCA curve of Validation set; H. CIC curve of Validation set

Full size image

Analysis of clinical utility and efficacy of the predictive model

The DCA curves for patients in both the training and validation sets indicated that the model’s DCA curves were consistently higher than the extreme curves(Fig. 2C and G). When the threshold probability of PE occurrence ranged from 0.1 to 1.0, using this predictive model for clinical decision-making offered a higher net benefit compared to the strategies of “treat all” or “treat none”. The CIC showed that at a threshold probability greater than 0.5, the predictive score probability from the model highly matched the actual population who developed PE(Fig. 2D and H). This confirmed the high clinical efficacy of the predictive model.

Comparative analysis of model discrimination capability

The discrimination capability of this model was compared with the Simplified Wells score and revised Geneva score in diagnosing PE. The results showed that in the training set, the AUROC for the simplified Wells score and revised Geneva score were 0.698 (95% CI: 0.654–0.741) and 0.758 (95% CI: 0.717–0.799), respectively. In the validation set, the AUROC for the simplified Wells score and revised Geneva score were 0.672 (95% CI: 0.603–0.740) and 0.798 (95% CI: 0.738–0.858) respectively (Fig. 3).

Fig. 3
figure 3

Comparison of the discrimination ability of model with Simplified Wells score and Revised Geneva score in the training and validation sets. A. ROC curve of Training set; B. ROC curve of Validation set

Full size image

Discussion

Nomogram is widely acknowledged as a dependable quantitative tool for assessing the risk of clinical events and is utilized in the establishment of various clinical disease models [16, 17]. In this study, we also constructed a nomogram model to predict the probability of PE in hospitalized patients. In the model, 7 variables including sex, clinical manifestations, comorbidity history, and auxiliary examination were included as independent predictors of PE in hospitalized patients. All data were sourced from clinical medical records and were less susceptible to subjective influences from assessors. Furthermore, our predictive risk ultimately yielded the probability of PE, rather than specific threshold values, suggesting potentially improved model acceptability.

To enhance the diagnostic efficiency of PE in hospitalized patients, diagnostic prediction risk-based research has long been spotlighted. Massimo Miniati et al. [18] constructed a diagnostic algorithm for PE in hospitalized patients (i.e., meeting three symptoms (sudden dyspnea, chest pain, and syncope) along with signs of right ventricular overload under ECG, radiological signs of volume depletion, pulmonary artery truncation, and lung consolidation compatible with infarction). Their results exhibited a clinical classification accuracy of 90%, with sensitivity and specificity being 84% and 95%, respectively. While this model exhibited favorable diagnostic efficacy, subsequent literature reviews revealed few reports of its clinical application. This phenomenon may be attributed to the rarity of simultaneous occurrences of sudden dyspnea, chest pain, and syncope in patients, particularly as syncope is more prevalent in patients with unstable blood flow — a condition uncommon in PE [19]. Gina Barnes et al. [20] used machine learning algorithms (gradient boosted tree, XGBoost model), neural networks, and logistic regression models to analyze patient demographics, clinical features, and auxiliary examinations. The AUROC values for diagnosing PE were 0.85, 0.74, and 0.67, with sensitivities of 81% each and specificities of 70%, 44%, and 35%, respectively. Liang ZA et al. [7] constructed a prediction model using seven indicators including D-dimer, APTT, FDP, platelet count, sodium, albumin, and cholesterol to assess the risk of PE in hospitalized patients. While the predicted probability from the nomogram aligned with the actual probability, its AUC value was less than 0.7. In our study, the predictors involved potential clinical manifestations of PE, histories likely to cause thrombosis, indicators of coagulation disorders, endothelial damage, and right heart impairment. This comprehensive approach likely contributed to the ideal predictive efficacy of our model.

Previous research has indicated a higher likelihood of PE occurrence in females compared to males. However, our study revealed a higher risk of PE occurrence in males. The analysis suggested that the occurrence of PE is related to both sex and age. Females under 45 or over 80 years of age experience a higher incidence of venous thrombosis compared to males, possibly due to factors such as estrogen, hormonal changes associated with pregnancy, and the longer life expectancy of elderly women [21, 22]. The average age of male participants in this study was 63.75 ± 14.84 years, while the average age of female participants was 63.77 ± 14.83 years. Within this age range, hospitalized male patients were more likely to develop PE, consistent with previous studies. Additionally, retrospective cohort studies have shown that increased PE risk in males is also associated with factors such as height and marital status [23]. However, the underlying mechanisms contributing to the increased PE risk in males have not been elucidated, and further research is needed to clarify these associations. Hospitalized patients with PE were found to be more likely to develop symptoms of dyspnea or shortness of breath, which are consistent with conclusions from previous guidelines and research [5, 12, 24, 25]. The finding suggested that considering these factors as predictors for the occurrence of PE in hospitalized patients is reasonable. We also observed that hospitalized patients with interstitial lung disease had an elevated risk of developing PE, consistent with previous research findings [26]. The propensity for PE in interstitial lung disease may be linked to factors such as systemic inflammation, hormone exposure during treatment, progressive respiratory failure, and prolonged bed rest due to late-stage activity limitations [27, 28]. These pathological states are directly or indirectly associated with the risk of PE occurrence, and most such patients necessitate hospitalization when these conditions arise. Roughly one-third of venous thromboembolism (VTE) events manifest as PE, either with or without DVT [29]. While the prevalence of PE combined with DVT has been reported to range from 13–93% [30], studies have confirmed that patients with both PE and concurrent DVT face elevated risks of PE-related death, all-cause death, and recurrent VTE compared to those with PE alone [31]. Our study suggested that the prevalence of PE combined with DVT was 51.58%, which was significantly higher than that in the non-PE group, consistent with previous studies [30, 32].

While only approximately 3% of patients will develop the most severe long-term complication of post-pulmonary embolism syndrome [33], chronic thromboembolic pulmonary hypertension (CTEPH), after 6 months of treatment for acute PE, over 87% of patients actually develop symptoms of PAH following an acute PE event. Yet, only 55% of patients underwent etiological investigation. Delays in diagnosis and the presence of baseline PAH have been identified as significant risk factors for CTEPH [34, 35]. In our study, 27.6% of PE patients also had PAH, which was significantly higher than that in non-PE patients, indicating that PAH can be used as a predictor of the occurrence of PE. Therefore, we recommend that when hospitalized patients present with PAH, the possibility of PE should be considered, and a comprehensive assessment should be conducted to avoid missed or delayed diagnoses. The presence of right ventricular strain is an important criterion for risk stratification of PE [36]. Tricuspid regurgitation (TR) associated with PE is linked to PAH induced by PE. Previous studies have reported that the degree of TR is highly consistent among observers when categorizing PAH as mild, moderate, or severe, with an interobserver agreement of up to 90% and a Kappa value of 0.8536. Additionally, the specificity of TR for diagnosing PE can reach 87% [37]. Our study also found a higher incidence of TR events, whether mild, moderate, or severe, in PE patients compared to non-PE patients, further suggesting an increased rate of TR in PE cases. Unfortunately, a review of the literature revealed that studies investigating the relationship between TR and PE are relatively scarce, underscoring the need for further research to clarify this association.

Changes in FDP and D-dimer levels are typically positively correlated, rising in tandem with the degree of coagulation and fibrinolysis activation. The diagnostic significance of a negative D-dimer value in excluding PE has garnered considerable attention. However, with the popularization of coagulation function testing, it has been observed that some patients do not exhibit elevated D-dimer levels despite an increase in FDP levels [38, 39]. In this study, a similar phenomenon was observed, where an elevated FDP level served as a predictor for the occurrence of PE, consistent with previous research findings [7]. However, a literature review indicated limited studies evaluating the correlation between FDP and PE. Additionally, our study did not adjust D-dimer and FDP levels for age or conduct a quantitative analysis. Therefore, further comprehensive research is warranted to elucidate their precise relationship with PE.

Assessing clinical probability is integral to the diagnostic process for PE. Wells and Geneva scores are preferred for their simplicity and ease of acquiring scoring information. These scoring methodologies are continually refined to better meet clinical requirements [40]. This study compared the diagnostic efficacy of our model with the simplified Wells score and the revised Geneva score for hospitalized patients with PE. It was found that our model exhibited superior AUC values, sensitivity, and specificity. This disparity may arise from the fact that the Wells and Geneva scores predominantly rely on clinical manifestations and PE-related risk medical history, without involving objective examinations. Prior multicenter PIOPED studies revealed that the Wells score demonstrated an efficacy of approximately 0.76 in evaluating PE in outpatients and emergency patients, but dropped to 0.65 for hospitalized patients. Additionally, the diagnostic efficacy of the Geneva score was found to be less than 0.7 [41, 42]. In our established model, we incorporated predictors associated with common clinical manifestations of PE, the commonly affected organ (heart), lower limb venous thrombosis, and coagulation function, among others. Hence, as indicated by the literature, Wells and Geneva scores are more suitable for assessing PE in outpatients and emergency patients [43], whereas our model demonstrates greater suitability for hospitalized patients.

This study possesses several limitations. Firstly, it is a single-center study that lacks external validation. Secondly, although we included all eligible patients during the study period, the sample size was still limited. Thirdly, as this is a retrospective analysis, further prospective studies are necessary to validate and confirm our findings. Overall, the selection bias inherent to single-center studies and the limitations associated with retrospective data are important factors that may affect the predictive performance and generalizability of our model. However, these limitations also lay a foundation for future prospective research. Additionally, our clinical database for PE requires further refinement and expansion, and we are actively seeking collaboration with other hospitals to collect more data, optimize the model, and perform external validation to better evaluate its generalizability. Fourthly, due to time constraints for data collection, adjustments in China’s COVID-19 management policies, and the need to control for bias, we did not include COVID-19 infection in our analysis. This is a notable limitation of our study, as venous thromboembolism events are common in patients with COVID-19. Moving forward, we will continue to monitor the incidence of PE in COVID-19 patients and incorporate these cases into our model to explore its predictive performance for PE in this population.

Conclusion

This study established an interpretable, simplified nomogram for predicting the risk of PE in hospitalized patients. This tool may contribute to enhancing clinical vigilance of PE in hospitalized patients and promoting the standardized management of this disease.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

CTPA:

Computed tomography pulmonary angiography

ROC:

Receiver operating characteristic

DCA:

Decision curve analysis

CIC:

Clinical impact curve

FDP:

Fibrin degradation product

PE:

Pulmonary embolism

VTE:

Venous thromboembolism

Fbg:

Fibrinogen

PT:

Prothrombin time

APTT:

Activated partial thromboplastin

TT:

Thrombin time

INR:

International normalized ratio

COPD:

Chronic obstructive pulmonary disease

AUROC:

Area under the ROC curve

DVT:

Deep vein thrombosis

PAH:

Pulmonary arterial hypertension

BNP:

Brain natriuretic peptide

CTEPH:

Chronic thromboembolic pulmonary hypertension

References

  1. Freund Y, Cohen-Aubart F, Bloom B. Acute Pulmonary Embolism. JAMA 2022, 328(13).

  2. Gupta R, Fortman DD, Morgenstern DR, Cooper CJ. Short- and long-term mortality risk after Acute Pulmonary Embolism. Curr Cardiol Rep 2018, 20(12).

  3. Kulka HC, Zeller A, Fornaro J, Wuillemin WA, Konstantinides S, Christ M. Acute Pulmonary Embolism: its diagnosis and treatment from a multidisciplinary viewpoint. Deutsches Ärzteblatt international; 2021.

  4. Miró Ò, Jiménez S, Mebazaa A, Freund Y, Burillo-Putze G, Martín A, Martín-Sánchez FJ, García-Lamberechts EJ, Alquézar-Arbé A, Jacob J, et al. Pulmonary embolism in patients with COVID-19: incidence, risk factors, clinical characteristics, and outcome. Eur Heart J. 2021;42(33):3127–42.

    Article  PubMed  Google Scholar 

  5. Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing G-J, Harjola V-P, Huisman MV, Humbert M, Jennings CS, Jiménez D, et al. 2019 ESC guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J. 2020;41(4):543–603.

    Article  PubMed  Google Scholar 

  6. Kahn SR, Solomon CG, de Wit K. Pulmonary embolism. N Engl J Med. 2022;387(1):45–57.

    Article  PubMed  Google Scholar 

  7. Zhou Q, Xiong X-Y, Liang Z-A. Developing a Nomogram-based Scoring Tool to Estimate the risk of Pulmonary Embolism. Int J Gen Med. 2022;15:3687–97.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. van der Hulle T, Cheung WY, Kooij S, Beenen LFM, van Bemmel T, van Es J, Faber LM, Hazelaar GM, Heringhaus C, Hofstee H, et al. Simplified diagnostic management of suspected pulmonary embolism (the YEARS study): a prospective, multicentre, cohort study. Lancet. 2017;390(10091):289–97.

    Article  PubMed  Google Scholar 

  9. van Es N, Takada T, Kraaijpoel N, Klok FA, Stals MAM, Büller HR, Courtney DM, Freund Y, Galipienzo J, Le Gal G, et al. Diagnostic management of acute pulmonary embolism: a prediction model based on a patient data meta-analysis. Eur Heart J. 2023;44(32):3073–81.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Basu S, Geersing G-J, Takada T, Klok FA, Büller HR, Courtney DM, Freund Y, Galipienzo J, Le Gal G, Ghanima W et al. Ruling out pulmonary embolism across different healthcare settings: a systematic review and individual patient data meta-analysis. PLoS Med 2022, 19(1).

  11. Akhter M, Kline J, Kannan V, Courtney DM, Kabrhel C. Discrepancy between Clinician Gestalt and subjective component of the Wells score in the evaluation of Pulmonary Embolism. Ann Emerg Med. 2018;71(6):796–8.

    Article  PubMed  Google Scholar 

  12. Liu Q, Xiao J, Liu L, Liu J, Zhu H, Lai Y, Wang L, Li X, Wang Y, Feng J. A new nomogram prediction model for pulmonary embolism in older hospitalized patients. Heliyon 2024, 10(3).

  13. Kline JA, Nelson RD, Jackson RE, Courtney DM. Criteria for the safe use of D -dimer testing in emergency department patients with suspected pulmonary embolism: a multicenter us study. Ann Emerg Med. 2002;39(2):144–52.

    Article  PubMed  Google Scholar 

  14. Stein PD, Fowler SE, Goodman LR, Gottschalk A, Hales CA, Hull RD, Leeper KV Jr., Popovich J Jr., Quinn DA, Sos TA, et al. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med. 2006;354(22):2317–27.

    Article  CAS  PubMed  Google Scholar 

  15. Ceriani E, Combescure C, Le Gal G, Nendaz M, Perneger T, Bounameaux H, Perrier A, Righini M. Clinical prediction rules for pulmonary embolism: a systematic review and meta-analysis. J Thromb Haemost. 2010;8(5):957–70.

    Article  CAS  PubMed  Google Scholar 

  16. Liang W, Zhang L, Jiang G, Wang Q, Liu L, Liu D, Wang Z, Zhu Z, Deng Q, Xiong X, et al. Development and validation of a Nomogram for Predicting Survival in patients with Resected non–small-cell Lung Cancer. J Clin Oncol. 2015;33(8):861–9.

    Article  PubMed  Google Scholar 

  17. Cao S, Li H, Xin J, Jin Z, Zhang Z, Li J, Zhu Y, Su L, Huang P, Jiang L, et al. Identification of genetic profile and biomarkers involved in acute respiratory distress syndrome. Intensive Care Med. 2023;50(1):46–55.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Miniati M, Prediletto R, Formichi B, Marini C, Di Ricco C, Tonelli L, Allescia C, Pistolesi M. Accuracy of clinical assessment in the diagnosis of pulmonary embolism. Am J Resp Crit Care. 1999;159(3):864–71.

    Article  CAS  Google Scholar 

  19. Barco S, Ende-Verhaar YM, Becattini C, Jimenez D, Lankeit M, Huisman MV, Konstantinides SV, Klok FA. Differential impact of syncope on the prognosis of patients with acute pulmonary embolism: a systematic review and meta-analysis. Eur Heart J. 2018;39(47):4186–95.

    Article  PubMed  Google Scholar 

  20. Ryan L, Maharjan J, Mataraso S, Barnes G, Hoffman J, Mao Q, Calvert J, Das R. Predicting pulmonary embolism among hospitalized patients with machine learning algorithms. Pulmonary Circulation 2022, 12(1).

  21. Duffett L, Castellucci LA, Forgie MA. Pulmonary embolism: update on management and controversies. BMJ. 2020;370:m2177.

    Article  PubMed  Google Scholar 

  22. Heit JA. Epidemiology of venous thromboembolism. Nat Rev Cardiol. 2015;12(8):464–74.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Brink A, Elf J, Svensson PJ, Engström G, Melander O, Zöller B. Sex-specific risk factors for deep venous thrombosis and pulmonary embolism in a Population-based historical cohort study of Middle-aged and older individuals. J Am Heart Assoc. 2023;12(5):e027502.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Lili X, Shunlan D, Lixu J. Predictive model for pulmonary embolism in pregnant and Postpartum women: a 10-Year retrospective study. Clin Appl Thromb Hemost 2023, 29.

  25. Jin ZY, Li CM, Zheng K, Qu H, Yang WT, Wen JH, Zhang WD, Ren HL. Nomogram for predicting pulmonary embolism in gynecologic inpatients with isolated distal deep venous thrombosis. Int J Gynecol Obstet. 2023;164(1):324–33.

    Article  Google Scholar 

  26. Sprunger DB, Olson AL, Huie TJ, Fernandez-Perez ER, Fischer A, Solomon JJ, Brown KK, Swigris JJ. Pulmonary fibrosis is associated with an elevated risk of thromboembolic disease. Eur Respir J. 2011;39(1):125–32.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Sun H, Liu M, Yang X, Xi L, Xu W, Deng M, Ren Y, Xie W, Dai H, Wang C. Incidence and risk factors of venous thrombotic events in patients with interstitial lung disease during hospitalization. Thromb J 2023, 21(1).

  28. Barsoum MK, Heit JA, Ashrani AA, Leibson CL, Petterson TM, Bailey KR. Is progestin an independent risk factor for incident venous thromboembolism? A population-based case-control study. Thromb Res. 2010;126(5):373–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Linnemann B, Blank W, Doenst T, Erbel C, Isfort P, Janssens U, Kalka C, Klamroth R, Kotzerke J, Ley S, et al. Diagnostics and Therapy of venous thrombosis and pulmonary embolism. The revised AWMF S2k Guideline. Vasa. 2023;52(S111):1–146.

    Article  CAS  PubMed  Google Scholar 

  30. Becattini C, Cohen AT, Agnelli G, Howard L, Castejon B, Trujillo-Santos J, Monreal M, Perrier A, Yusen RD, Jimenez D. Risk stratification of patients with Acute Symptomatic Pulmonary Embolism Based on Presence or absence of Lower Extremity DVT: systematic review and Meta-analysis. Chest. 2016;149(1):192–200.

    Article  PubMed  Google Scholar 

  31. Jimenez D, Aujesky D, Diaz G, Monreal M, Otero R, Marti D, Marin E, Aracil E, Sueiro A, Yusen RD, et al. Prognostic significance of deep vein thrombosis in patients presenting with acute symptomatic pulmonary embolism. Am J Respir Crit Care Med. 2010;181(9):983–91.

    Article  PubMed  Google Scholar 

  32. Van Rossum AB, Van Houwelingen HC, Kieft GJ, Pattynama PMT. Prevalence of deep vein thrombosis in suspected and proven pulmonary embolism: a meta-analysis. Brit J Radiol. 1998;71(852):1260–5.

    Article  PubMed  Google Scholar 

  33. Klok FA, Couturaud F, Delcroix M, Humbert M. Diagnosis of chronic thromboembolic pulmonary hypertension after acute pulmonary embolism. Eur Respir J 2020, 55(6).

  34. Klok FA, van der Hulle T, den Exter PL, Lankeit M, Huisman MV, Konstantinides S. The post-PE syndrome: a new concept for chronic complications of pulmonary embolism. Blood Rev. 2014;28(6):221–6.

    Article  CAS  PubMed  Google Scholar 

  35. Fernandes T, Auger W, Fedullo P. Epidemiology and risk factors for chronic thromboembolic pulmonary hypertension. Thromb Res. 2018;164:145–9.

    Article  CAS  PubMed  Google Scholar 

  36. Torbicki A, Perrier A, Konstantinides S, Agnelli G, Galiè N, Pruszczyk P, Bengel F, Brady AJB, Ferreira D, Janssens U, et al. Guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J. 2008;29(18):2276–315.

    Article  CAS  PubMed  Google Scholar 

  37. Alerhand S, Sundaram T, Gottlieb M. What are the echocardiographic findings of acute right ventricular strain that suggest pulmonary embolism? Anaesth Crit Care Pain Med 2021, 40(2).

  38. Falster C, Hellfritzsch M, Gaist TA, Brabrand M, Bhatnagar R, Nybo M, Andersen NH, Egholm G. Comparison of international guideline recommendations for the diagnosis of pulmonary embolism. Lancet Haematol. 2023;10(11):e922–35.

    Article  CAS  PubMed  Google Scholar 

  39. Sato N, Takahashi H, Shibata A. Fibrinogen fibrin degradation products and D-Dimer in Clinical-Practice - interpretation of discrepant results. Am J Hematol. 1995;48(3):168–74.

    Article  CAS  PubMed  Google Scholar 

  40. Cronin P, Dwamena BA. A clinically meaningful interpretation of the prospective investigation of Pulmonary Embolism diagnosis (PIOPED) II and III data. Acad Radiol. 2018;25(5):561–72.

    Article  PubMed  Google Scholar 

  41. Ollenberger GP, Worsley DF. Effect of patient location on the performance of clinical models to predict pulmonary embolism. Thromb Res. 2006;118(6):685–90.

    Article  CAS  PubMed  Google Scholar 

  42. Shen J-H, Chen H-L, Chen J-R, Xing J-L, Gu P, Zhu B-F. Comparison of the Wells score with the revised Geneva score for assessing suspected pulmonary embolism: a systematic review and meta-analysis. J Thromb Thrombolysis. 2015;41(3):482–92.

    Article  Google Scholar 

  43. Stein PD, Sostman HD, Bounameaux H, Buller HR, Chenevert TL, Dalen JE, Goodman LR, Gottschalk A, Hull RD, Leeper KV, et al. Challenges in the diagnosis Acute Pulmonary Embolism. Am J Med. 2008;121(7):565–71.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

We are grateful to Guizhou Provincial People’s Hospital for providing the clinical data for this study, and also to the Beijing Tuberculosis and Thoracic Tumor Research Institute and Department of Respiratory and Critical Care Medicine, Beijing Chest Hospital, Capital Medical University for providing guidance on the overall content of this study.

Funding

This work was supported by the Beijing Municipal Natural Science Foundation (grant number L234007), the Science and Technology Fund of Guizhou Provincial Health Commission (gzwkj2021-089) and the Guizhou Provincial Science and Technology Agency Project (Qian Ke He Basic-ZK (2021) General 349).

Author information

Author notes

  1. Zhimin Cao, Luyu Yang contributed equally to this work and should be considered as co-first authors.

  2. Xianwei Ye, Huan Ye contributed equally to this work and should be considered as co-corresponding authors.

Authors and Affiliations

  1. Beijing Tuberculosis and Thoracic Tumor Research Institute, Courtyard 1, No.9 Beiguan Street, Tongzhou District, Beijing, 101199, China

    Zhimin Cao & Huan Ye

  2. Department of Respiratory and Critical Care Medicine, Guizhou ProvincialPeoplès Hospital, No.83 Zhongshan East Road, Nanming District, Guiyang, 550002, China

    Jing Han, Xiuzhi Lv, Xiao Wang, Bangyan Zhang & Xianwei Ye

  3. Department of Respiratory and Critical Care Medicine, Beijing Chest Hospital, Capital Medical University, Courtyard 1, No.9 Beiguan Street, Tongzhou District, Beijing, 101199, China

    Luyu Yang & Huan Ye

Authors
  1. Zhimin Cao
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Luyu Yang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Jing Han
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Xiuzhi Lv
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Xiao Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Bangyan Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Xianwei Ye
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Huan Ye
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

CZM: Analysis and Interpretation, Data Collection, Writing the Manuscript, Critical Revision, Statistical Analysis, Agreement to be Accountable, Approval of the Manuscript; YLY: Analysis and Interpretation, Data Collection, Critical Revision, Software, Statistical Analysis, Agreement to be Accountable, Approval of the Manuscript; HJ: Data Collection, Critical Revision, Agreement to be Accountable, Funding acquisition were performed, Agreement to be Accountable, Approval of the Manuscript; LXZ : Data Collection, Critical Revision, Statistical Analysis, Agreement to be Accountable, Approval of the Manuscript; WX: Data Collection, Critical Revision, Statistical Analysis, Agreement to be Accountable, Approval of the Manuscript; ZBY: Data Collection, Critical Revision, Funding acquisition were performed, Agreement to be Accountable, Approval of the Manuscript; YXW: Conception and Design, Supervision, Critical Revision, Provide clinical data of patients, Agreement to be Accountable, Approval of the Manuscript; YH: Conception and Design, Project administration, Supervision, Critical Revision, Agreement to be Accountable, Funding acquisition were performed, Approval of the Manuscript.

Corresponding authors

Correspondence to Xianwei Ye or Huan Ye.

Ethics declarations

Ethics approval

This study was approved by the Ethics Committee of Guizhou Provincial People’s Hospital (ethics no: gzwkj2021-089).

Consent to participate

Not applicable as it is not a clinical study.

Consent for publication

Not applicable.

Conflict of interest

The authors declare that they have no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Cao, Z., Yang, L., Han, J. et al. Development of a predictive nomogram for early identification of pulmonary embolism in hospitalized patients: a retrospective cohort study. BMC Pulm Med 24, 594 (2024). https://doi.org/10.1186/s12890-024-03377-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12890-024-03377-z

Keywords

BMC Pulmonary Medicine

ISSN: 1471-2466

Contact us