a b s t r a c t

Introduction: There is limited evidence regarding the effects of a pre-existing heart failure on the diagnostic yield of pulmonary embolism evaluation in the Emergency Department (ED).

Methods: Electronic medical record of consecutive adults who underwent a computed tomography pulmonary angiogram (CTPA) in the ED at Loma Linda University Medical Center between June 1, 2019 and March 25, 2022 were reviewed. Repeat studies for the same patient and patients with unspecified HF diagnoses or isolated right ventricular HF were excluded. Key demographics, lab values and vital signs, relevant medications were collected. Primary outcome was the incidence of PE on CTPA compared between patients with and without pre-existing HF. Results: A total of 2846 patients were included in the study (602 patients with HF and 2244 without). In total co- hort, 11.7% (n = 334) of patients had PE found on CTPA. The incidence of PE on CTPA was lower among patients with a history of HF than patients without a history of HF (12.5% vs 9%). A history of pre-existing HF was associ- ated with a lower odds ratio for a positive PE study (OR 0.13, 95%CI: 0.03-0.57) in multivariable analyses.

Conclusions: In this study, we observed that the incidence of PE among patients who undergo CTPA was lower among patients with pre-existing HF compared to those without. Further studies should determine if HF is an important mitigating factor when risk stratifying patients for PE.

(C) 2022

1. Introduction

Symptoms suggestive of pulmonary embolism (PE) are common among patients presenting to the emergency room. It may therefore be challenging to determine the threshold for PE evaluation espe- cially among patients with potential alternate diagnoses. Heart fail- ure (HF) is the most common Medicare readmission diagnosis, and patients with Acute HF (AHF) commonly present to the emergency department (ED) with symptoms that may mimic PE. Differentiating PE and AHF on presentation can be challenging due to considerable overlap in clinical manifestations [1,2]. Although patients with a his- tory of HF may present with dyspnea or other symptoms from AHF, they are also at risk for other concomitant etiologies of dyspnea due to comorbidities and other Predisposing factors. Particularly, HF is a risk factor for venous thromboembolism, which may increase

* Corresponding author at: 2068 Orange Tree Lane, Suite 215, Redlands, CA 92374, USA.

E-mail address: [email protected] (D. Abramov).

predisposition to PE [3]. PE has also been reported among patients presenting with AHF [4] and a diagnosis of PE among patient with HF is associated with an increased risk of adverse events [5]. Patients with HF may also be prescribed anticoagulation for non-PE condi- tions [6] or heart rate lowering medications as part of HF manage- ment which may complicate evaluation for PE.

There is limited data on the comparative incidence of PE in pa- tients who undergo evaluation for PE stratified by a history of pre- existing HF. Specifically, it is unknown whether patients suspected of having a PE in the ED with a history of HF would have a higher like- lihood of having a PE due to associated high risk conditions or a lower likelihood of having a PE because AHF may be an alternate diagnosis. Given the burden of HF in the community and frequent ED visits by those with pre-existing HF, this question has important clinical im- plications that may alter the Diagnostic pathway for these patients. Hence, we sought to evaluate whether a history of pre-existing HF alters the incidence of a positive computed tomography pulmonary angiogram (CTPA) for PE among patients presenting to the ED suspected to have a PE.

https://doi.org/10.1016/j.ajem.2022.11.006

0735-6757/(C) 2022

1. Methods
   1. Study design

We performed a retrospective cohort study of consecutive adults (age >= 18 years old) who underwent a CTPA in the emergency depart- ment (ED) at Loma Linda University Medical Center between June 1, 2019 and March 25, 2022.

• 1. Selection of participants

Patients who underwent a CTPA in the ED at Loma Linda University Medical Center during the mentioned period were identified by a query of electronic medical records.

• 1. Measurements

We queried the electronic medical record system for key associated demographics, lab values and vital signs, relevant medications, and key comorbidities as coded on presentation based on defined fields within the electronic medical record system, with comorbidities based on ICD 10 codes. Thromboembolic disease was defined as a documented his- tory of prior deep vein thrombosis or pulmonary embolism. Patients with any diagnosis code of I50.20-23 and I50.40-43 were classified as having heart failure with reduced ejection fraction (HFrEF). Patients with a diagnosis code of I50.30-33 were identified as having heart fail- ure with preserved ejection fraction (HFpEF). Patients with preexisting isolated right ventricular failure or unspecified ICD 10 HF diagnoses, in- cluding I50.1, I50.8, and I50.9, were excluded. If multiple CTPA studies during serial ED visits were performed on a single patient over the study period, only the first CTPA encounter was included in the study, with subsequent encounters excluded.

• 1. Outcomes

The study’s primary outcome was the incidence of PE on CTPA com- pared between the HF and non-HF groups. The secondary outcomes were factors associated with the risk of a positive CTPA for PE in the total cohort and specifically in the HF group.

• 1. Analysis

For comparison of baseline demographics, comorbidities, vitals, and lab values, we used the ?2 test to compare categorical variables and the

analysis of covariance to compare continuous variables. Univariate and multivariate binary logistic regression was used to evaluate an associa- tion between variables and positive CTPA for PE. We used the “Enter” method by entering all variables simultaneously [6]. Numbers of vari- ables from the collected baseline characteristics entering the logistic regression will be based on numbers of outcomes (one variable per 10 PE events). Both HFpEF and HFrEF were combined into one HF category in most analyses. P-value <0.05 was considered statistically significant for all analyses. SPSS Statistics 22 (IBM Corp. Armonk, NY) was used for t-test, Chi-square and regression statistics.

This retrospective study was approved by the hospital Institutional Review Board and follows the institution’s ethical standards.

The methodology was reviewed to follow best practice of retrospec- tive chart review study [7].

1. Results

A total of 3214 patients were identified. After removing 368 patients who met exclusion criteria, 2846 patients were included in the study and final analysis (602 patients with HF and 2244 without). Among the whole cohort, 11.7% (n = 334) of patients had a finding of PE on CTPA. Characteristics of patients with and without HF are shown in Table 1. Patients with a history of HF were older (63.0 +- 15.9 vs 58.2 +- 17.3) and were more likely to have history of hypertension (62.6% vs 40.9%), diabetes mellitus (38.9% vs 25.5%), chronic obstructive pulmonary disease (18.4% vs 8.6%), chronic kidney disease (26.9% vs 13.9%), and anti-coagulation use (29.7% vs 12.2%) when compared to patients without a history of HF. Of the 179 patients with HF who were on anticoagulation, 79 patients had an indication of prior throm- boembolic disease. Of the 274 patients on anticoagulation without HF, 163 had an indication of prior thromboembolic disease. The incidence of PE on CTPA was 12.5% among patients with no history of HF and 9.0% among patients with a history of HF. When analyzed according to coded HF type, the incidence of PE on CTPA was 10.1% among patients with a history of HFrEF and 6.4% among patients with HFpEF (Supple- mentary Table 1).

Multivariable analysis for predictors of positive PE study for the en- tire cohort is shown in Table 2. As we had a total of 334 positive PE stud- ies, inclusion of 15 variables from the collected baseline characteristics was deemed reasonable. A history of pre-existing HF was associated with a lower odds ratio for a positive PE study (OR 0.13, 0.03-0.57). Variables that were associated with positive study include elevated D-dimer on presentation (OR 1.33, 1.19-1.48) and a lower pulse rate on presentation (OR 0.73, 0.55-0.97).

Table 1

Baseline characteristics of study population.

Variables	HF (n = 602)	No HF (n = 2244)	MD/OR with 95% CI (HF group as reference)
Baseline characteristics and comorbidities Age (years) (mean +- SD)	63.0 +- 15.9	58.2 +- 17.3	4.8 (3.2, 6.3)
Male	322 (53.5%)	1141 (50.8%)	1.11 (0.93, 1.33)
Hypertension	377 (62.6%)	918 (40.9%)	2.42 (2.01, 2.91)
Diabetes mellitus	234 (38.9%)	573 (25.5%)	1.85 (1.53, 2.24)
Chronic obstructive pulmonary disease	111 (18.4%)	194 (8.6%)	2.39 (1.85, 3.08)
Chronic kidney disease	162 (26.9%)	312 (13.9%)	2.28 (1.84, 2.83)
Anti-coagulation use	179 (29.7%)	274 (12.2%)	3.04 (2.45, 3.77)
Pulmonary embolism on CTPA	54 (9.0%)	280 (12.5%)	0.69 (0.51, 0.94)
Admission vitals and laboratory values SBP (mmHg) (mean +- SD)	130 +- 26	129 +- 25	1.1 (-1.2, 3.4)
DBP (mmHg) (mean +- SD)	79 +- 17	76 +- 15	2.5 (1.0, 3.9)
Pulse rate (bpm) (mean +- SD)	97 +- 22	98 +- 22	-1.1 (-3.1. 0.8)
Oximetry (%) (mean +- SD)	96 +- 4	96 +- 4	0.2 (-0.1. 0.6)
NT-proBNP (pg/ml) (mean +- SD)	6842 +- 10,302	2374 +- 6726	4467 (3642, 5293)
Creatinine (mg/dL) (mean +- SD)	1.6 +- 1.9	1.3 +- 1.6	0.33 (0.18, 0.48)
D-dimer (ug/mL) (mean +- SD)	3.3 +- 3.6	2.9 +- 3.5	0.32 (-0.43, 1.07)
Elevated troponin (patients with value >99th percentile)	205/486 (42.2%)	337/1312 (25.7%)	2.11 (1.69, 2.63)

CI: Confidence interval, CTPA: Computed tomography pulmonary angiogram, DBP: Diastolic blood pressure, HF: Heart failure, MD: Mean difference, NT-proBNP: N-Terminal pro B-type natriuretic peptide, OR: Odds ratio, SBP: Systolic blood pressure, SD: Standard deviation.

Table 2

Multivariate analysis of association between variables and incidence of pulmonary embolism.

Variables	OR	95% CI
Baseline characteristics and comorbidities Age (years)	0.98	0.95-1.01
Male	0.95	0.36-2.53
Heart failure	0.13	0.03-0.57
Hypertension	1.18	0.41-3.36
Diabetes mellitus	0.36	0.11-1.21
Chronic obstructive pulmonary disease	0.16	0.01-4.41
Chronic kidney disease	0.64	0.10-4.23
Anti-coagulation use	2.29	0.62-8.44
Admission vitals and laboratory values
SBP (per 10 mmHg)	1.09	0.78-1.52
DBP (per 10 mmHg)	1.31	0.77-2.23
Pulse rate (per 10 bpm)	0.73	0.55-0.97
Oximetry (per 1%)	1.08	0.97-1.20
NT-proBNP (per 100 pg/ml)	1.01	0.99-1.02
Creatinine (mg/dL)	0.45	0.20-1.03
D-dimer (ug/mL)	1.33	1.19-1.48
Elevated troponin (patients with value >99th percentile)	2.61	0.80-8.49

CI: Confidence interval, DBP: Diastolic blood pressure, NT-proBNP: N-terminal pro b-type natriuretic peptide, OR: Odds ratio, SBP: Systolic blood pressure.

Among only patients with pre-existing HF, the only significant uni- variate predictor of a positive CTPA for PE was a history of DM (OR 0.37, 0.18-0.74). There were no variables that were predictive in multi- variable analyses (Supplementary Table 2).

1. Discussion

The current study demonstrates that the positivity rate among pa- tients who underwent a CTPA in the ED is lower among patients with pre-existing HF compared to those without. There is limited prior data on the incidence of PE among patients with concomitant cardiac diag- noses such as HF who present with symptoms of suspected PE [4]. Among a primarily outpatient cohort, patients with a history of HF who underwent a CTPA had a lower incidence of PE (21%) compared to the total study population (31%) [8]. To our knowledge, ours is the first report on the incidence of PE among patients with pre-existing HF who present to the ED and undergo a clinically indicated CTPA.

Despite HF being a predisposing factor for venous thromboembo- lism, there may be multiple reasons why patients with HF have a lower incidence of PE. Patients with AHF may present with signs and symptoms which mirror those of PE [1,2]. This implies that AHF may be a viable alternative diagnosis among some patients with pre- existing HF who present with symptoms that may raise suspicion for PE. Criteria for alternative diagnosis was not well characterized during the original derivation of the Wells score [9], and whether HF is a com- mon alternative diagnosis among patients with suspected PE is un- known. One study of patients presenting to the ED with suspected PE found that HF was one of the most common discharge diagnoses (5%) among patients without a PE diagnosis [10], although incidence of PE based on the history of pre-existing HF was not reported.

While a history of HF is clearly not sufficient in ruling out PE and pa- tients with pre-existing HF who present with dyspnea may have a broad differential, extra consideration for AHF as an alternative cause of pre- senting symptoms could be of value prior to pursuing CTPA for PE eval- uation among this population. Performing less frequent CTPA would reduce the associated risks among patients with possible AHF, including those of contrast and radiation. Additional evaluation is needed to con- firm these findings in larger cohorts, as well as to potentially identify characteristics that differentiate PE from AHF which may lead to im- proved clinical decision rules in this setting. Additionally, a substantial number (8-17% in prior cohorts, and almost 30% in our cohort) of pa- tients with HF are already prescribed anticoagulation, [11] which may affect PE incidence. However, use of anticoagulation may also be a

marker for prior high risk thromboembolic disease, and active use of an- ticoagulation did not seem to affect the incidence of a positive CTPA study in this cohort.

This study has important limitations. This cohort includes patients from a single tertiary center ED, and requires validation in other popu- lations. Physician decision-making, specifically regarding factors con- tributing to PE suspicion or indication for CTPA, is not available in this analysis. Risk scores, including the Wells score, at the time of presenta- tion are not available. Therefore, it is unknown whether patients with a history of HF were more likely than others to undergo a CTPA because of concern for increased thrombotic risk or they were less likely to undergo a CTPA because alternative etiologies such as AHF were suspected. This also prevents us from knowing the incidence of missed PEs because a CTPA study was not ordered, including among patients who may have been considered to have AHF as the suspected etiology for presentation. We did not include repeat CTPA for patients who may have had multiple scans during their admission which potentially could underestimate true incidence of PE in our study population, al- though repeating CTPA studies within the same encounter is not com- mon. Pre-existing diagnoses, including those for HF, were based on entered ICD 10 codes, and therefore are dependent on the quality of sys- tem coding by clinicians. As such, specific diagnosis of HF subtype, eg. acute HF vs chronic HF or HFrEF vs HFpEF, can be challenging [12], which may affect our results regarding this cohort. Further work will also be needed to determine whether incidence of PE on a CTPA study is affected by the type of pre-existing HF (HFpEF vs. HFrEF), or whether these findings are due to frequency with which these cohorts undergo CTPA studies or other factors. Additionally, we were not able to deter- mine the final chief diagnosis for each presentation, including whether acute HF was ultimately though to be the presenting diagnosis among patients with pre-existing HF who underwent a CTPA study. Further- more, chart review for indication for anticoagulation was limited by data documented in the EMR, and the role that anticoagulation or prior thromboembolic disease may have played in the clinical decision to undergo a CTPA is not available in this dataset. Additionally, whether the diagnosis of PE indicates an acute or chronic PE cannot be deter- mined with certainty. alternative methods of PE diagnosis, including V/Q scan and pulmonary angiogram, were not evaluated as they are in- frequently performed in the ED setting at our institution. Nevertheless, this study represents a real-world cohort of ED patients with suspected PE who undergo a CTPA study and raises important questions about the role of pre-existing HF as a possible factor in decision making when evaluating patients for PE.

Grant support

None.

Financial disclosures

Authors have no relevant disclosures.

CRediT authorship contribution statement

Jakrin Kewcharoen: Writing - review & editing, Writing - original draft, Visualization, Validation, Software, Resources, Project administra- tion, Investigation, Formal analysis, Data curation. Paresh Giri: Writing - review & editing, Writing - original draft, Visualization, Validation, Supervision, Project administration, Methodology. M. Reza Amini: Writing - review & editing, Writing - original draft, Visualization, Super- vision, Project administration, Methodology, Investigation, Data curation. Laren Tan: Writing - review & editing, Writing - original draft, Validation, Supervision, Project administration, Methodology. Dafne Moretta: Writing - review & editing, Writing - original draft, Visualization, Valida- tion, Supervision, Resources, Project administration, Methodology, Inves- tigation, Conceptualization. Emily Barrett: Writing - review & editing,

Writing - original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Data curation. E. Lea Walters: Writing - review & editing, Writing - original draft, Visualiza- tion, Validation, Supervision, Software, Project administration, Methodol- ogy, Investigation. Dmitry Abramov: Writing - review & editing, Writing - original draft, Supervision, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.

Declaration of Competing Interest

Authors have no conflict of interests to disclose.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2022.11.006.

References

1. Arrigo M, Huber LC. Pulmonary embolism and heart failure: a reappraisal. Card Fail Rev. 2020;7:e03.
2. Piazza G, Goldhaber SZ. Pulmonary embolism in heart failure. Circulation. 2008;118:

1598-601.

1. Tang L, Wu YY, Lip GY, Yin P, Hu Y. Heart failure and risk of venous thromboembo- lism: a systematic review and meta-analysis. Lancet Haematol. 2016;3:e30-44.
2. Correia LC, Goes C, Ribeiro H, Cunha M, Paula R, Esteves JP. Prevalence and predictors of pulmonary embolism in patients with Acutely decompensated heart failure. Arq Bras Cardiol. 2012;98:120-5.
3. Elkaryoni A, Merchant KR, Brailovsky Y, Bunte MC, Sperry BW, Darki A. Effect of acute pulmonary embolism on the hospitalization rates in patients with heart failure (from a Nationwide cohort sample). Am J Cardiol. 2020;135:185-7.
4. Ranganathan P, Pramesh CS, Aggarwal R. common pitfalls in statistical analysis: lo- gistic regression. Perspect Clin Res. 2017;8:148-51.
5. Kaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. Ann Emerg Med. 2014;64:292-8.
6. Sohne M, Kruip MJ, Nijkeuter M, et al. Accuracy of clinical decision rule, D-dimer and Spiral computed tomography in patients with malignancy, previous venous throm- boembolism, COPD or heart failure and in older patients with suspected pulmonary embolism. J Thromb Haemost. 2006;4:1042-6.
7. Wells PS, Ginsberg JS, Anderson DR, et al. Use of a clinical model for safe manage- ment of patients with suspected pulmonary embolism. Ann Intern Med. 1998;129: 997-1005.
8. Bernard Bagattini S, Bounameaux H, Perneger T, Perrier A. Suspicion of pulmonary embolism in outpatients: nonspecific chest pain is the most frequent alternative di- agnosis. J Intern Med. 2004;256:153-60.
9. Zeitler EP, Eapen ZJ. Anticoagulation in heart failure: a review. J Atrial Fibrill. 2015;8:

1250.

1. Vaidya GN, Abramov D. echocardiographic evaluation of Diastolic function is of lim- ited value in the diagnosis and management of HFpEF. J Card Fail. 2018;24:392-6.