a b s t r a c t

Objective: The purpose of this study was to analyze the prevalence and significance of incidental findings on com- puted tomography pulmonary angiography (CTPA) studies and to assess the diagnostic yield of CTPA in identify- ing an alternate diagnosis to pulmonary embolism (PE) on PE negative exams.

Methods: All patients who had a CTPA exam for PE evaluation between Jan 2016 and Dec 2018 with a negative PE result were included in the study. A total of 2083 patients were identified. We retrospectively queried the electronic medical record and the Radiology report and recorded the following: Age, Sex, BMI, Patient location and Incidental findings. The incidental findings were classified into type 1 (Alternate diagnosis other than PE which could explain the patient’s symptoms), type 2 (non-emergent findings which needed further work up) and type 3 findings (non-emergent findings which did not need further work up). Logistic regression analysis was performed to determine what factors affected the probability of finding a type 1 incidental (alternate diag- nosis) or a type 2 incidental.

Results: 74.5% of the patients in our study had at least one incidental finding. Type 1 incidental findings (alternate diagnosis to PE) were found in 864 patients (41.5%). The most common type 1 finding was pneumonia followed by fluid overload. Male sex, increased age and lower BMI were significantly associated with increased odds of a type 1 incidental(p < 0.05). Similarly, all the patient locations had significantly different odds of finding a type-1 incidental, with ICU having the highest odds, followed by inpatient, ED and outpatient locations (p < 0.05). 563 patients (27%) had at least one type 2 incidental findings and the most common type 2 findings were progressive lung malignancy/ metastatic disease and new pulmonary nodule. Increased age was significantly associated with the probability of a type 2 finding (p < 0.05).

Conclusions: CTPA may suggest an alternative diagnosis to pulmonary embolism in approximately 40% of the patients with a negative study. The probability of finding an alternate diagnosis (type 1 incidental) is higher in elderly patients and in patients referred from ICU and Inpatient units.

(C) 2022

1. Introduction

Pulmonary Embolism (PE) is the third leading cause of death after myocardial infarction and stroke in the United States [1]. CT pulmonary angiography (CTPA) is the Diagnostic modality of choice for the evalua- tion of PE due to its high specificity and sensitivity [2].

* Corresponding author at: Department of Radiology, University of Iowa Carver College of Medicine, 200 Hawkins Dr, Iowa city, IA 52242, USA.

E-mail address: [email protected] (S. Narayanasamy).

¹ Ali Eskandari and Sabarish Narayanasamy share equal first authorship.

Although the utilization of CTPA has significantly increased over the past two decades in the evaluation of PE, it has not been associ- ated with improvement in clinical outcomes [3-6]. While the overuti- lization of CT has led to more PE’s being diagnosed, studies have suggested that they may be less severe and may not always represent clinically significant Disease burden [3,7]. CTPA is associated with ex- posure to ionizing radiation and risks of acute kidney injury and con- trast induced nephropathy from intravenous contrast administration [8]. In addition, overutilization of CTPA results in detection of large number of incidental findings like lung nodules, which necessitates further follow up studies [9]. Overall, the detection of such incidental

https://doi.org/10.1016/j.ajem.2022.01.064

0735-6757/(C) 2022

findings, some of which are benign yet warrants further evaluation or follow-up exams, contributes to increase in the Healthcare costs, pa- tient anxiety, and risks of complications with procedures performed for work-up of these incidental findings. Previous studies have shown that CTPA ordered from the emergency department are twice likely to find lung nodules or mediastinal adenopathy compared to de- tection of PE [10]. Some authors also argued that one of the drivers for overutilization for CTPA study is that it is often useful to explain pa- tient’s symptoms even in the absence of a PE, because the CT might provide diagnostic clues suggesting an alternative etiology for chest pain. However, many of these diagnoses like pulmonary edema may be suspected based on clinical evaluation, specific laboratory testing, or may also be evident on chest radiographs.

Traditionally, the positive yield of CTPA studies for PE has been shown to vary between 6% and 30% in various studies [11-13]. However, some recent single-center and multicenter studies have shown positiv- ity of testing for PE ranging from 1.7% to 3.1% [6,14]. These recent studies highlight the overuse of imaging leading to overall decrease in positivity of diagnostic testing for pulmonary embolism, potentially leading to in- creased burden of incidental findings. While the difference in yield of CTPA for PE between different patient locations (Emergency depart- ment, Inpatient, ICU and outpatient) has been previously reported [11,15], there is limited evidence comparing the Diagnostic efficacy of CTPA in providing an alternate diagnosis to PE across different patient locations.

The purpose of this study was to measure the current prevalence of incidental findings detected in CTPA studies and to evaluate whether CT can suggest an alternative diagnosis in the absence of a pulmonary embolism; an argument frequently made in favor of CTPA overutiliza- tion. In addition, we also evaluated whether the diagnostic yield of iden- tifying an alternate diagnosis varied with patient characteristics and patient location.

1. Methods
   1. Data collection

This was a retrospective study approved by the local Institutional Review Board. All patients that underwent CTPA for PE evaluation between Jan 2016 and Dec 2018 and had a negative PE result were included in the study. Data was collected by accessing the health information system. Patients who were positive for PE on CTPA were excluded from the study. CTA studies done for other in- dications (such as aneurysm follow up, pulmonary AVMs etc.) and routine CT chest with contrast studies (ordered to evaluate lung mass etc.) were excluded. Any non-diagnostic study due to technical reasons was also excluded from the analysis. There were a total of 2083 CTPA studies which were negative for PE during this time period.

We retrospectively queried the patients electronic medical record and the radiology report for the incidental findings and to determine the significance of those findings. Demographic data including age, gen- der, Patient location at the time of scan (Emergency department, Inpa- tient, ICU and Outpatient) and Body Mass Index (BMI) was recorded.

• 1. Characterization of incidental findings

Incidental findings from the impression of the radiology report were further classified as Type-1, Type-2 and Type-3 based on the clinical significance of the findings, classification similar to a prior study on this topic [16]. Type-1 incidental findings were critical or emergent diagnoses (other than pulmonary embolism) which could explain the patient’s symptoms of chest pain or dyspnea. These in- cluded but were not limited to pneumonia, pulmonary edema, lung collapse (of at least one entire lobe or more), aortic dissection, or large pleural effusion (causing collapse of majority of the lung).

These findings usually demand prompt clinical action. Type-2 inciden- tals were non-emergent findings which would need work-up. These findings included but were not limited to Pulmonary nodules (new or enlarging), worsening or unsuspected cancer or metastatic disease, enlarged pulmonary artery (defined as main pulmonary artery diame- ter > 3 cm without known diagnosis of pulmonary hypertension), in- determinate or unsuspected thyroid nodules, or liver lesions. Type-3 incidental findings were non-emergent findings that were unlikely to be of any clinical significance pertaining to patients’ acute presenta- tion and not needing follow-up imaging (Stable/ small pleural effu- sion, degenerative spinal disease, Known / stable emphysema, stable primary Lung cancer or metastatic disease). If a patient had more than one incidental finding (for example having two Type-1 findings or Type- 1 and Type-2 findings in the same patient), the findings were counted individually.

• 1. Statistical analysis

Categorical data were collected as number and percentages and con- tinuous variables were described as mean and standard deviation (mean +- SD). McNemar’s tests for paired proportions was used to test for differences between the percentages of patients with each Type of incidental finding. To investigate the relationships between covariates of interest and the detection of incidental findings, two logistic regres- sion models were fit. In the first model, the outcome variable was the presence of a Type-1 incidental and in the second model, the outcome variable was the presence of a Type-2 incidental. Both models used the same covariates: gender, age, BMI, and patient location and model selection using AIC was performed. BMI was entered as a continuous covariate. Model results are presented in terms of the odds ratios, 95% confidence intervals (CI) for the odds ratios, and p-values. All analyses were completed with R software (R Foundation for Statistical Comput- ing; Vienna, Austria, 2006).

1. Results
   1. Patient and study characteristics

A total of 2083 patients with negative PE results were included in the study. There were 1119 females (54%) and 964 males (46%) in the study population. The mean age was 58 +- 17.2 years (Mean +- SD). The Mean BMI was 31 +- 9 Kg/m2 and majority of the patients had BMI between 25 and 30 (522/2083, 25.0%). Mean contrast ad- ministered was 88.5 +- 17.4 ml, and the mean radiation effective dose per study was 5.5 +- 4.4 mSv. Majority of the patients were re- ferred from ED (1219/2083, 58.52%) followed by Inpatient (466/ 2083, 22.37%), ICU (226/2083, 10.85%) and Outpatient (172/2083,

8.26%) (Table 1).

Table 1

Demographic data of the study population.

	Mean	Standard deviation
Age	57.9	17.2
BMI	31.3	9.9
Contrast (ml)	88.6	17.4
Dose length product (mGy*cm)	393.8	315.3
Effective dose (mSv)	5.51	4.41
BMI category	BMI	Frequency (%)
Underweight	<18.5	86 (4.1)
Normal	18.5-24.99	499 (24.0)
Pre obese	25-24.99	522 (25.0)
Obese class I	30-34.99	400 (19.2)
Obese class II	35-39.99	233 (11.2)
Obese class III	>=40	343 (16.5)
Total		2083 (100)

• 1. Incidental findings

Among the total population, about 74.5% of the patients (1552/ 2083) had at least one incidental finding and 25.5% did not have any in- cidental findings (531/2083). A total of 864 (41.5%) patients had at least one Type-1 incidental finding i.e. the CT study provided an alternative diagnosis for the patient’s symptoms. 563 (27%) patients had at least one Type-2 incidental finding, and 949 (45.5%) patients had at least one Type 3 incidental finding. A total of 376 (18.1%) patients had only Type-3 incidental findings which did not require any further evaluation or follow-up.

The most common Type-1 finding was pneumonia (57%) followed by fluid overload (21%) and large or moderate pleural effusion (6%) (Table 2). Among the Type-2 findings, the most common findings were progressive primary lung malignancy or metastatic disease (18.5%) and new or enlarging lung nodule (18.5%) followed by pulmo- nary arterial enlargement (16.7%). Among the Type-3 findings, the most common diagnosis was emphysema (14.6%) followed by stable mediastinal lymphadenopathy (11.2%). Table 2 shows the five most prevalent findings in each incidental subtype.

Many patients had more than one incidental finding. For example, 251(12%) patients had both Type-1 and Type-2 findings, and 271 (13%) patients had both Type-2 and Type-3 findings. A small percentage of patients (4%) had two or more Type-1 incidental findings.

• 1. Probability of Type-1 incidental findings

Males were more likely to have Type-1 incidental findings compared to females (OR = 1.36, 95% CI = [1.13, 1.18]). This was statistically sig- nificant (p value = 0.001). Increased age is significantly associated with increased odds of a Type-1 incidental (OR = 1.01, 95% CI = [1.00, 1.01], p = 0.001). Lower BMI is also significantly associated with increased odds of a Type-1 incidental finding (OR = 0.97, 95% CI = [0.96, 0.98], p value <0.001). As per the multivariate logistical regression analysis, significant differences were noted in the probability of Type-1 inciden- tal finding between ICU and ED patients (OR = 1.33, p < 0.001), Inpatients and ED (OR = 1.08, p < 0.001) and Outpatients and ED (OR

= 0.81, p < 0.001). Similarly, there was significant differences in prob- ability of a Type-1 incidental between Inpatients and ICU (OR = 0.85, p < 0.001), Outpatients and ICU (OR = 0.65, p < 0.001) and between Outpatients and Inpatients (OR = 0.76, p < 0.001). To summarize, all the patient locations had significantly different odds of finding a Type- 1 incidental, with ICU having the highest odds, followed by inpatient, ED and outpatients (Table 3, Fig. 1). This indicates that the odds of

Table 2

Top 5 findings in each incidental type.

Type 1	Frequency (%)
Pneumonia	567(27.2)
Fluid Overload	220(10.5)
Pleural Effusion-Large/Moderate	77(3.6)
Lung Collapse	13(0.6)
Pericardial Effusion-Large/Moderate	8(0.3)
Type 2
Progressive cancer or Metastatic Disease	126(6.0)
Lung nodule/mass-New or Indeterminate	125(6.0)
Pulmonary artery enlargement-New or unknown	113(5.4)
Lymphadenopathy-Indeterminate	53(2.5)
Thyroid nodule-Indeterminate	47(2.2)
Type 3
Emphysema	195(9.3)
LAD-Stable/known	150(7.2)
Post-surgical findings	141(6.7)
Pleural Effusion-stable or Small	77(3.6)
Metastatic or primary disease-Stable	69(3.3)

Table 3

Logistic regression results with presence of Type 1 incidental as the outcome.

	type 1 findings
	Variable	Odds ratio	95% CI for odds ratio	p-value
	Gender - Male	1.07	(1.03, 1.12)	<0.001
	Age	1.00	(1.00, 1.00)	0.001
	BMI	0.99	(0.99, 1.00)	<0.001
	Location ICU - ED	1.33	(1.25, 1.42)	<0.001
	IN - ED	1.13	(1.08, 1.19)	<0.001
	OUT - ED	0.87	(0.81, 0.94)	<0.001
	IN - ICU	0.85	(0.79, 0.92)	<0.001
	OUT - ICU	0.65	(0.59, 0.72)	<0.001
	OUT - IN	0.77	(0.71, 0.83)	<0.001

finding a Type-1 incidental finding in ICU patients is 57% greater than ED patients and 59% greater than outpatients.

• 1. Probability of Type-2 incidental findings

As with Type-1 incidentals, we found that increased age is signifi- cantly associated with an increase in the probability of a Type-2 inciden- tal (OR = 1.02, 95% CI = [1.01, 1.03], p < 0.001). But there was no significant difference in the presence of a Type-2 incidental between Males and females (OR = 0.97, 95% CI - [0.94,1.01], p = 0.14). Similarly, there was no significant differences in the presence of Type 2 incidental based on BMI (OR = 1.00, CI - [0.99,1.00], p = 0.30). Multivariate logis- tic regression models also do not show any significant difference in find- ing a Type-2 incidental based on patient location (Table 4, Fig. 2).

1. Discussion

Most of the patients in our study had at least one incidental finding and many of them had more than one incidental finding. Although pre- vious studies have addressed the issue of incidental findings on CTPA, our study is one of the largest cohorts to date including more than 2000 patients [10,16,17]. In addition, while most of the previous studies have focused only on patients referred from the ED, our data includes referrals from ED, ICU, inpatient and outpatient locations. This is impor- tant because there is significant difference in patient characteristics among various patient locations and the yield of CTPA in identifying an alternative diagnosis may not be uniform across the clinical settings. Multidetector CT pulmonary angiography is currently considered as the diagnostic modality of choice for the evaluation of pulmonary em- bolism [12]. In this study, CTPA revealed an alternative diagnosis that may explain Patient symptoms in 40% of our patients. The most com- mon alternative diagnosis was pneumonia followed by fluid overload and large pleural effusion. This is consistent with previous studies where the proportion of Alternative diagnoses varied between 25% and 50% [10,18]. In a large retrospective analysis, Hall et al. identified ra- diologic findings supporting alternative diagnosis in 31% of their patient population [10] and Van es et al. were able to identify an alternative di- agnosis in approximately 43% of their patient population [18]. In both these studies, pneumonia was the most identified alternative diagnosis followed by Pleural effusions and pulmonary atelectasis. However, in a large retrospective study from Britain, authors found that CTPA sug- gested an alternative diagnosis in less than 10% of non-PE positive pa- tients and they concluded that finding an alternative diagnosis does not justify the increased use of CTPA [17]. But their definition of “posi- tive alternative diagnosis” excluded the diagnoses that were previously suspected based on clinical presentation and based on chest radio- graphs and this may explain the low yield of CTPA in the study. A yield of nearly 40% for alternative diagnosis, which may explain patient symptoms, may very well be one of the unconscious drivers for overuti- lization of CTPA exams. Some clinicians have argued that CTPA study is useful even in the absence of a positive result as it often provides an

Fig. 1. Estimated probabilities of a type 1 finding for all locations and both sexes across the range of BMIs and ages most prevalent in the study.

alternative diagnosis for the patient’s symptoms which may guide clin- ical management [12]. But many of these alternate findings may also be identified on the preliminary chest radiograph thus negating the utility of a CT.

Although utilization of CT in the evaluation of pulmonary embolism has increased exponentially over the recent years, it has not been shown to be associated with improved outcomes [5,19,20]. After introduction of CTPA in 1998, the incidence of PE cases in united states almost dou- bled, however, the PE mortality has remained relatively stable and the case fatality rate has decreased [21]. One of the reasons for this finding is that with improvement in CT technology, many clinically insignificant PEs are detected which might not have any led to adverse events even if it were undiagnosed. This phenomenon of over testing and overdiagno- sis is important because apart from radiation risks associated with CT, it can lead to iatrogenic harms from unnecessary anticoagulation [21]. It has been shown that most of the isolated subsegmental pulmonary em- boli are anticoagulated at the same frequency as more proximal embo- lisms [22]. Few small retrospective studies and reviews have suggested that isolated subsegmental PE may not need anticoagulation [23,24]. However, recent Cochrane review concluded that there is not enough evidence from existing trials to assess safety and effectiveness of antico- agulation with isolated subsegmental PE at this time [25]. Hence, the practice decision on this topic may evolve in future. In addition, in- creased utilization of CTPA has disadvantages related to contrast risks, overutilization of resources, increased healthcare costs and problems with unnecessary follow up of incidental findings.

Table 4

Logistic regression results with presence of Type 2 incidental as the outcome.

Type 2 findings

Variable Odds ratio 95% CI for odds ratio p-value

Gender - Male	0.97	(0.94, 1.01)	0.147
Age	1.00	(1.00, 1.01)	<0.001
BMI	1.00	(1.00, 1.00)	0.301302
Location ICU - ED	0.99	(0.93, 1.06)	0.823
IN - ED	1.03	(0.99, 1.09)	0.129
OUT - ED	1.03	(0.96, 1.10)	0.421
IN - ICU	1.05	(0.97, 1.12)	0.222
OUT - ICU	1.04	(0.95, 1.13)	0.420
OUT - IN	0.99	(0.92, 1.07)	0.845

We found that ICU and inpatient location at the time of the exam is associated with increased odds of identifying an alternative diagnosis on the CTPA scan compared to ED or outpatient location. The CTPA yield in ICU patients is also known to be higher, but it is interesting to see higher yield of CT for an alternative diagnosis to explain patient symptoms as well [11]. This finding is likely related to the fact that ICU patients are generally sicker, and exams are only performed when the clinical suspicion for pulmonary pathology is very high. To our knowledge, this association has not been previously studied. Also, in- creased age is associated with increased odds of identifying an alterna- tive diagnosis. Given that pneumonia, fluid overload and pleural effusions were the most common alternative diagnoses in our study, these findings are expected to be more common in the elderly age group and in critically ill patients in the intensive care and inpatient units. It has also been previously shown that ICU and inpatients have a higher yield of positive CTPA exams for pulmonary embolism compared to ED patients and outpatients [11,26]. While our results suggest that CTPA might provide valuable alternative diagnoses in this select sub- group of patients, it is important to remember that patients from the ICU are often critically ill with numerous metabolic derangements and are at a higher risk of contrast induced nephropathy from intravenous contrast administration. On the other hand, CTPA studies referred from the ED and Outpatient had the lowest odds of identifying an alter- native diagnosis and therefore in this clinical setting, CTPA should be ju- diciously used to prevent over testing and overdiagnosis in this group. This is consistent with the best practice advisory laid out by the American College of Emergency Physicians (ACEP) [27] which recom- mends that ED clinicians should use clinical prediction rules to estimate the pre-test probability of PE and CT should be reserved for patients with a high pre-test probability of PE or a positive D-dimer test [27].

One-fourth of our patients had incidental findings that required ra- diographic follow up or further clinical work-up (Type-2) but did not di- rectly contribute to the patient’s acute symptoms. The most common diagnoses in this group were progression of primary malignancy or me- tastatic disease and a new pulmonary nodule. The prevalence of lung nodules in our study population is similar to previously published liter- ature [28]. Previously unknown enlargement of main pulmonary artery, defined as size >3 cm can be seen with pulmonary hypertension. This finding was also seen in a high proportion of patients (16.7%). The pa- tients with suspected pulmonary hypertension are frequently referred for echocardiogram as a screening test. Enlarged lymph nodes which re- quired follow up by size criteria was also a common finding in our

Fig. 2. Estimated probabilities of a type 2 finding for all locations and both sexes across the range of BMIs and ages most prevalent in the study.

population and was identified in 53 patients (2.5%). The risks associated with identification and follow up imaging of these incidental nodules and lymph nodes are not insignificant. Studies have shown that even low levels of radiation exposure from CT are not completely safe and in- crease the lifetime risk of developing a radiation induced cancer. A prior study has shown that for each CTPA exam, patients receive a mean radi- ation dose of 5.5 mSv and each additional follow up CT adds an equiva- lent amount of radiation [11]. Multiple exposures to intravenous contrast media also poses a small risk of kidney injury [29]. Finally, fol- low up of these incidentally pulmonary nodules is also associated with increased healthcare and patient costs, increased psychological stress to patients associated with identification of a “positive finding” and add to an ever-expanding clinical workload and may indirectly contrib- ute to physician burnout [30,31].

About half of our patients had incidental findings which were of no clinical importance (Type-3) and did not require any follow up. Emphy- sema and stable calcified lymph nodes were the most common findings in this category. High incidence of calcified lymphadenopathy is expected in our study population given high prevalence of histoplasmo- sis in our region. This trend of high incidence of findings of no clinical importance has also been observed in multiple previous studies [32,33] and in a previous retrospective study by Perelas et al., authors found Type-3 incidental findings in 48% of their study population [16] which is closely similar to our results.

1. Limitations

First, this is a retrospective study design done at a single tertiary care center and the results may not be generalizable in all clinical settings. Second, we excluded patients who had positive CTPA from our analysis of incidental findings as the aim of our study was to assess the yield of CTPA in providing an alternative diagnosis in patients who did not have a pulmonary embolism. The ability of CT to find an alternative di- agnosis in a patient presenting with chest pain or shortness of breath is frequently cited as a reason for overutilization of CT imaging. To accu- rately measure what (and how much) are the alternative diagnosis which are made on CTPA exams, all CTs positive for PE were excluded. Third, we did not review the chest radiographs of the patients to identify which of the incidental findings were identifiable on the radiograph. One of the reasons was that in our practice, frequently, the patients

are sent for CTPA directly without chest x-ray if clinical suspicion is high. Also, for ICU and in-patients chest radiograph findings are fre- quently non-specific with atelectasis obscuring pneumonia due to re- cumbency. Finally, while we explored the patients charts and prior imaging regarding the incidental findings, we did not consider the over- all clinical picture, chronic symptoms and lab results to understand the significance of these findings in a bigger picture. While all the scans in our institution are read by sub-specialty trained thoracic radiologists, we relied on the reports for these incidental findings and did not per- form a dedicated retrospective review of the imaging for purpose of this study.

1. Conclusions

In conclusion, incidental findings are common in CTPA studies with about 75% of the patients in our study with at least one incidental find- ing with greater than 25% patients having a finding which requires a follow-up testing or a procedure for characterization. CTPA may also suggest an alternative diagnosis in approximately 40% of the patients with a negative study, although some of these diagnosis like pneumonia may be suspected based on clinical exam, or radiographs. The probabil- ity of finding an alternative diagnosis on CTPA is higher in elderly pa- tients and in patients referred from intensive care and inpatient units. In this select sub-group of patients, CTPA may provide additional valu- able information to guide clinical management even in the absence of a pulmonary embolism. On the other hand, CTPA studies in younger pa- tients and studies referred from the emergency and outpatient depart- ment had a much lower probability of identifying an alternative diagnosis and therefore CTPA should be judiciously used in accordance with clinical decision rules to prevent over testing and overdiagnosis in this setting. Emergency physicians and primary care physicians should also be conscious of detection of large number of incidental find- ings that may require downstream testing from CTPA overuse and con- scious efforts should be made to adhere to the clinical guidelines to prevent downstream testing or invasive procedures.

Funding support

None.

CRediT authorship contribution statement

Ali Eskandari: Methodology, Investigation, Data curation, Conceptu- alization. Sabarish Narayanasamy: Writing - original draft, Visualiza- tion, Methodology, Investigation, Formal analysis. Caitlin Ward: Software, Methodology, Investigation. Sarv Priya: Writing - review & editing, Supervision, Data curation, Conceptualization. Tanya Aggarwal: Writing - review & editing, Resources, Project administration, Data curation. Jacob Elam: Investigation, Formal analysis, Data curation. Prashant Nagpal: Writing - review & editing, Validation, Supervision, Investigation, Formal analysis, Conceptualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper.

References

1. Giri J, Sista AK, Weinberg I, Kearon C, Kumbhani DJ, Desai ND, et al. Interventional therapies for acute pulmonary embolism: current status and principles for the de- velopment of novel evidence: a scientific statement from the American Heart Asso- ciation. Circulation. 2019;140(20):e774-801.
2. Moore AJE, Wachsmann J, Chamarthy MR, Panjikaran L, Tanabe Y, Rajiah P. Imaging of acute pulmonary embolism: an update. Cardiovasc Diagnos Ther. 2018;8(3): 225-43.
3. Schissler AJ, Rozenshtein A, Kulon ME, Pearson GD, Green RA, Stetson PD, et al. CT pulmonary angiography: increasingly diagnosing less severe pulmonary emboli. PLoS One. 2013.;8(6):e65669.
4. Bellolio MF, Heien HC, Sangaralingham LR, Jeffery MM, Campbell RL, Cabrera D, et al. Increased computed tomography utilization in the emergency department and its association with hospital admission. West J Emerg Med. 2017;18(5):835-45.
5. Swan D, Hitchen S, Klok FA, Thachil J. The problem of under-diagnosis and over- diagnosis of pulmonary embolism. Thromb Res. 2019;177:122-9.
6. Kline JA, Garrett JS, Sarmiento EJ, Strachan CC, Courtney DM. Over-testing for suspected pulmonary embolism in American emergency departments: the continu- ing epidemic. Circ Cardiovasc Qual Outcomes. 2020.;13(1):e005753.
7. Gerstein L, Koolian M, Schwartz BC. Sub-segmental pulmonary emboli as a marker of CT-pulmonary angiogram overuse in a single-center retrospective cohort study: a dimer a dozen? Thromb Res. 2020;196:463-5.
8. Tao SM, Wichmann JL, Schoepf UJ, Fuller SR, Lu GM, Zhang LJ. Contrast-induced ne- phropathy in CT: incidence, risk factors and strategies for prevention. Eur Radiol. 2016;26(9):3310-8.
9. Anjum O, Bleeker H, Ohle R. Computed tomography for suspected pulmonary embo- lism results in a large number of non-significant incidental findings and follow-up investigations. Emerg Radiol. 2019;26(1):29-35.
10. Hall WB, Truitt SG, Scheunemann LP, Shah SA, Rivera MP, Parker LA, et al. The prev- alence of clinically relevant incidental findings on chest computed tomographic an- giograms ordered to diagnose pulmonary embolism. Arch Intern Med. 2009;169 (21):1961-5.
11. Aggarwal T, Eskandari A, Priya S, Mullan A, Garg I, Siembida J, et al. Pulmonary em- bolism rule out: positivity and factors affecting the yield of CT angiography. Postgrad Med J. 2020;96(1140):594-9.
12. Mountain D. Multislice computed tomographic pulmonary angiography for diagnos- ing pulmonary embolism in the emergency department: has the ‘one-stop shop’arrived? Emerg Med Australas. 2006;18(5-6):444-50.
13. Costantino MM, Randall G, Gosselin M, Brandt M, Spinning K, Vegas CD. CT angiog- raphy in the evaluation of acute pulmonary embolus. AJR Am J Roentgenol. 2008; 191(2):471-4.
14. Dhakal P, Iftikhar MH, Wang L, Atti V, Panthi S, Ling X, et al. Overutilisation of imag- ing studies for diagnosis of pulmonary embolism: are we following the guidelines? Postgrad Med J. 2019;95(1126):420-4.
15. Costa AF, Basseri H, Sheikh A, Stiell I, Dennie C. The yield of CT pulmonary angio- grams to exclude acute pulmonary embolism. Emerg Radiol. 2014;21(2):133-41.
16. Perelas A, Dimou A, Saenz A, Rhee JH, Teerapuncharoen K, Rowden A, et al. Inciden- tal findings on computed tomography angiography in patients evaluated for pulmo- nary embolism. Ann Am Thorac Soc. 2015;12(5):689-95.
17. Chandra S, Sarkar PK, Chandra D, Ginsberg NE, Cohen RI. Finding an alternative diag- nosis does not justify increased use of CT-pulmonary angiography. BMC Pulm Med. 2013;13(1):9.
18. van Es J, Douma RA, Schreuder SM, Middeldorp S, Kamphuisen PW, Gerdes VE, et al. Clinical impact of findings supporting an alternative diagnosis on CT pulmonary an- giography in patients with suspected pulmonary embolism. Chest. 2013;144(6): 1893-9.
19. Boone JM, Brunberg JA. Computed tomography use in a tertiary care university hos- pital. J Am Coll Radiol. 2008;5(2):132-8.
20. Broder J, Warshauer DM. Increasing utilization of computed tomography in the adult

emergency department, 2000-2005. Emerg Radiol. 2006;13(1):25-30.

1. Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med. 2011;171(9):831-7.
2. Raslan IA, Chong J, Gallix B, Lee TC, McDonald EG. Rates of overtreatment and treatment-related adverse effects among patients with subsegmental pulmonary embolism. JAMA Intern Med. 2018;178(9):1272-4.
3. Baumgartner C, Klok FA, Carrier M, Limacher A, Moor J, Righini M, et al. Clinical sur- veillance vs. anticoagulation for low-risk patiEnts with isolated SubSegmental pul- monary embolism: protocol for a multicentre randomised placebo-controlled non- inferiority trial (SAFE-SSPE). BMJ Open. 2020;10(11):e040151.
4. Stein PD, Goodman LR, Hull RD, Dalen JE, Matta F. Diagnosis and management of iso- lated subsegmental pulmonary embolism: review and assessment of the options. Clin Appl Thromb Hemost. 2012;18(1):20-6.
5. Yoo HH, Nunes-Nogueira VS, Boas Fortes Villas, PJ.. anticoagulant treatment for subsegmental pulmonary embolism. Cochrane Database Syst Rev. 2020.;2(2): Cd010222.
6. Woo JK, Chiu RY, Thakur Y, Mayo JR. Risk-benefit analysis of pulmonary CT angiog- raphy in patients with suspected pulmonary embolus. Am J Roentgenol. 2012;198 (6):1332-9.
7. Raja AS, Greenberg JO, Qaseem A, Denberg TD, Fitterman N, Schuur JD. Evaluation of patients with suspected acute pulmonary embolism: best practice advice from the clinical guidelines Committee of the American College of physicians. Ann Intern Med. 2015;163(9):701-11.
8. Sohns C, Amarteifio E, Sossalla S, Heuser M, Obenauer S. 64-multidetector-row spiral CT in pulmonary embolism with emphasis on incidental findings. Clin Imaging. 2008;32(5):335-41.
9. Mitchell A, Kline J. Contrast nephropathy following computed tomography angiogra- phy of the chest for pulmonary embolism in the emergency department. J Thromb Haemost. 2007;5(1):50-4.
10. Goehler A, McMahon PM, Lumish HS, Wu CC, Munshi V, Gilmore M, et al. Cost- effectiveness of follow-up of pulmonary nodules incidentally detected on cardiac computed tomographic angiography in patients with suspected coronary artery dis- ease. Circulation. 2014;130(8):668-75.
11. Chetlen AL, Chan TL, Ballard DH, Frigini LA, Hildebrand A, Kim S, et al. Addressing burnout in radiologists. Acad Radiol. 2019;26(4):526-33.
12. Richman PB, Courtney DM, Friese J, Matthews J, Field A, Petri R, et al. Prevalence and significance of nonthromboembolic findings on chest computed tomography angi- ography performed to rule out pulmonary embolism: a multicenter study of 1,025 emergency department patients. Acad Emerg Med. 2004;11(6):642-7.
13. Van Strijen M, Bloem J, de Monye W, Kieft G, Pattynama P, Van Den Berg-Huijsmans A, et al. helical computed tomography and alternative diagnosis in patients with ex- cluded pulmonary embolism. J Thromb Haemost. 2005;3(11):2449-56.