|Year : 2023 | Volume
| Issue : 1 | Page : 44-51
Poor adherence to the recommended pulmonary embolism diagnostic pathway in the emergency department: A retrospective cohort study
Alia M Hadid1, Ala Jalabi1, Mahmoud Anka1, Arif Alper Cevik2
1 Department of Emergency Medicine, Tawam Hospital, Al Ain, UAE
2 Department of Emergency Medicine, Tawam Hospital; Department of Internal Medicine, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, UAE
|Date of Submission||19-Aug-2022|
|Date of Decision||09-Oct-2022|
|Date of Acceptance||28-Oct-2022|
|Date of Web Publication||02-Jan-2023|
Alia M Hadid
Department of Emergency Medicine, Tawam Hospital, Al Ain
Source of Support: None, Conflict of Interest: None
OBJECTIVES: Pulmonary embolus (PE) is a form of venous thromboembolism associated with increased morbidity and mortality if not diagnosed and treated early. Variations in clinical presentation make the diagnosis challenging. The gold standard for diagnosing PE is a computed tomography pulmonary angiogram (CTPA). Physicians show a low threshold for over-investigating PE. The evaluation of patients with suspected PE should be efficient, including but not limited to the use of risk stratification methods. This study aims to assess the adherence to the recommended diagnostic pathways of ordering CTPAs in patients with suspected PE.
METHODS: This retrospective cohort study included patients above 18 years of age who received a CTPA for a suspected PE in the emergency department (ED) of a hospital between 2015 and 2019. Patient demographic data, chief complaint, variables of the Wells and pulmonary emboli rule-out criteria scores, pregnancy status, investigations, and the patient's final PE diagnosis were extracted from the hospital electronic medical records. Diagnostic pathways that took place were compared to the internationally recommended pathway.
RESULTS: Four hundred and eighty-six patients were included in this study. The mean age was 51.01 (±19.5) years, and 377 (69.3%) patients were female. The recommended PE diagnostic pathway to order CTPA was incorrectly followed in 288 patients (59.3%). Seventy-five (15.4%) patients received an unnecessary CTPA. D-dimer test was ordered unnecessarily in 144 patients (29.6%). The overall prevalence of PE in our population was 9.47% (n: 46). Out of the 75 unnecessarily ordered CTPAs, 2 (2.7%) showed PE, while CTPAs ordered using the correct pathway showed 31 (10.9%) PEs.
CONCLUSION: Our study revealed that approximately two-thirds of all CTPA requests did not adhere to the recommended PE clinical decision pathway. There was a significant improper and unnecessary utilization of CTPA imaging and D-dimer testing. Improvements seem imperative to enhance physicians' clinical approach to PE diagnosis.
Keywords: Clinical decision rules, computed tomography pulmonary angiograms, D-dimer, diagnosis, emergency department, pulmonary emboli rule-out criteria, pulmonary embolism, wells
|How to cite this article:|
Hadid AM, Jalabi A, Anka M, Cevik AA. Poor adherence to the recommended pulmonary embolism diagnostic pathway in the emergency department: A retrospective cohort study. Turk J Emerg Med 2023;23:44-51
|How to cite this URL:|
Hadid AM, Jalabi A, Anka M, Cevik AA. Poor adherence to the recommended pulmonary embolism diagnostic pathway in the emergency department: A retrospective cohort study. Turk J Emerg Med [serial online] 2023 [cited 2023 Mar 28];23:44-51. Available from: https://www.turkjemergmed.org/text.asp?2023/23/1/44/366486
| Introduction|| |
Pulmonary embolus (PE) is a form of venous thromboembolism with about 2 million cases reported annually and estimated mortality of 3% from PE alone. The clinical presentation of PE is variable and often nonspecific, making the diagnosis challenging and may cause high morbidity and mortality., Therefore, physicians tend to have a low threshold for over-investigating PE., This low threshold usually results in unnecessary testing, contributing to prolonged emergency department (ED) lengths of stay, financial burdens, and patients' exposure to the adverse effects of investigation. Computed tomography pulmonary angiography (CTPA) remains the modality of choice in evaluating patients with suspected PE because of its high sensitivity and accuracy. Although CTPA has become the standard investigation modality of PE, it carries risks such as radiation exposure and contrast-induced nephropathy.,, Therefore, efficient clinical decision-making strategies, including risk stratification methods, are recommended.
The European Society of Cardiology and the American College of Physicians' guidelines on diagnosing PE suggest using validated clinical prediction rules such as the Wells and Pulmonary Emboli Rule Out Criteria (PERC) scores to estimate pretest probability in patients for whom acute PE is being considered., However, non-adherence to clinical decision-making pathways and unnecessary testing in the diagnosis of PE is widely reported.,,,, Middle eastern data are still limited. A recent study in the Gulf region found that one-third of CTPA studies did not adhere to recommended pathways.
This study aims to assess the adherence to the recommended diagnostic pathway of ordering CTPA for ED patients with a suspected PE in our setting.,
| Methods|| |
Study design, setting, and population
The study was a retrospective cohort study conducted in the ED of a community-based teaching hospital located in the city of Alain in the United Arab Emirates. The ED operates 24 h a day, 365 days a year, and has a 50-bed capacity, with an average of 115,000 patients annually. Patients coming to the ED would initially be assessed in the triaging area by a licensed emergency medicine physician or general practitioner and allocated to different areas of the ED based on the acuity and clinical condition of the patient. Full-capacity blood testing and specialty consultations are 24-h availability in this tertiary-care hospital. In addition, a CT scanner is located in the ED.
All patients above 18 years of age who presented to the ED with a suspected PE and underwent a CTPA in the ED between January 1st, 2015, and December 31st, 2019, were reviewed. Because our study focused on the adherence of ED physicians, CTPAs ordered by admitting specialty physicians, such as internal medicine and critical care physicians, were excluded. CTPAs ordered by ED physicians were analyzed.
The patient data were extracted from the electronic medical record system in the hospital, Citrix Cerner®. Individual patient charts were opened using the medical record number, and the data were coded into an Excel sheet. A random sample of patients was selected to check data accuracy at regular intervals. Data included patient demographics of age, sex, pregnancy status, D-dimer result, CTPA results, and continuous and categorical data, including the variables of both the Wells and PERC scoring systems [Appendix 1].,
The Siemens Sysmex CS-5100 System with a cutoff level of 0.5 mg/L was used for D-dimer. A licensed radiology technician performed the CTPA scans using the Siemens Somatom Definition AS 128-slice Scanner, injecting 50 ml of Optiray® (Ioversol) 350 contrast at 4.0 ml/s. The CTPA images were interpreted and reported by a licensed radiologist physician.
The Wells and PERC scores were retrospectively calculated., The two-tiered Wells score was used., If the patient was classified as “PE Likely” as per the Wells criteria (>4 points), CTPA should have been commenced directly. If the patient was classified as “PE unlikely” as per the Wells criteria (0–4 points), the PERC score should have been implemented. If PERC was calculated to be negative, PE was unlikely, and no further testing should have been required. If PERC was calculated to be positive, a D-dimer test should have been done. If the D-dimer result was negative, PE should have been ruled out. If the D-dimer result was positive, the patient should have undergone a CTPA. If the result of the CTPA was negative, then PE should have been ruled out. If the result of the CTPA was positive, then the PE diagnosis should have been made.
Clinical decision-making pathways that took place were compared to the recommended international clinical decision-making pathway mentioned above.
This study aimed to find (1) the amount of non-adherence to the recommended clinical decision-making pathway when investigating PE. (2) The amount of necessary and unnecessary CTPAs. Necessary CTPA was defined as the proportion of CTPAs deemed required per the Wells and PERC scores when following the proper pathway. (3) The amount of necessary and unnecessary D-dimer tests. Necessary D-dimer testing was defined as the proportion of D-dimer tests required as per the Wells and PERC scores when following the proper pathway. (4) The PE detection rate in necessary and unnecessary CTPAs.
The data were inserted into an Excel datasheet (Microsoft Excel, 2021, Version 16.57 (Microsoft, Redmond, Washington, USA), [Computer software]). Descriptive and frequency analyses were done using JASP (2022), Version 0.16.2 (University of Amsterdam, Amsterdam, North Holland, Netherlands), (Computer software).
The local ethics committee approved this study on December 20th, 2020 (#MF2058-2020-760).
| Results|| |
Five hundred and seventy-one patients received a CTPA scan in the ED. Eighty-five patients were excluded as the CTPA was ordered by the admitting specialty team. Four hundred 86 patients were enrolled for analysis. Most of the patients were females (n: 377, 69.3%). The mean age (standard deviation) was 51.0 (19.54) years (minimum: 18, maximum: 114). Twenty patients were pregnant (4.1%). Eighty-one (16.67%) patients described being immobile for 3 days or had surgery in the past 4 weeks. Seventy-nine (16.26%) patients had a previous PE or deep vein thrombosis. One hundred and eight (22.22%) patients had malignancies. Only 8 (1.65%) patients were under treatment of hormone therapy. The most common chief complaints were dyspnea (n: 259, 52.3%), chest pain (n: 142, 29.2%), and palpitations (n: 13, 2.7%).
[Figure 1] provides the study flow diagram. One hundred and fifty-nine (32.7%) patients were classified as “PE Likely,” and 327 (67.3%) patients were classified as “PE Unlikely” based on the Wells scoring system. The PERC was applied to “PE unlikely” patients.
|Figure 1: Flow chart depicting the correct (green line) and incorrect (red line) pathways followed through PE evaluation. Red D-dimer boxes represent unnecessary D-dimers. Green D-dimer box represents necessary D-dimers. Green CTPA boxes represent necessary CTPAs. Red CTPA boxes represent unnecessary CTPAs. The orange box represents CTPAs where a necessary or unnecessary decision could not be made on the path because those patients required D-dimer before CTPA. White boxes represent PE diagnosis by CTPA in each path. CTPA: Computerized Tomography Pulmonary Angiography, PE: Pulmonary Embolus, PERC: Pulmonary Embolism Rule Out Criteria|
Click here to view
Outcome 1: Amount of nonadherence to the recommended clinical decision-making pathway
Of 486 patients, the recommended PE clinical decision-making pathway was not followed correctly in 288 patients (59.3%). [Table 1] provides information about the types of nonadherence.
|Table 1: Types of nonadherence to recommended pulmonary embolism clinical decision-making pathways|
Click here to view
According to the correct pathway, all PE-likely patients (n: 159) should have gotten a direct CTPA. However, 86 (54.1%) of PE likely patients received an unnecessary D-dimer test before the CTPA.
Of the 327 “PE Unlikely” patients, PERC was positive in 259 (79.2%) and negative in 68 (20.8%). PERC-negative patients should have been ruled out of PE according to the recommended clinical decision-making pathway. However, 10 (14.7%) patients in PERC negative group underwent the unnecessary CTPA directly, and 58 (85.3%) received an unnecessary D-dimer before the unnecessary CTPA. All PERC-positive patients (n: 259) should have received a D-dimer test. One hundred and twenty-seven (49.0%) patients of the PERC-positive group had incorrectly undergone a direct CTPA with no preceding D-dimer test. One hundred and thirty-two (51%) patients in PERC positive group underwent the correct pathway and got a D-dimer test before the CTPA. D-dimer test was negative in 7 (5.3%) patients, and they should have been ruled out of PE. However, these patients received an unnecessary CTPA.
Outcome 2: Amount of necessary and unnecessary computed tomography pulmonary angiography
Only 284 (58.4%) CTPAs were considered necessary in our setting. 75 (15.4%) CTPAs were unnecessary according to the recommended PE clinical decision-making pathway. 127 (26.1%) patients who received CTPA were in an undetermined group (neither necessary nor unnecessary) because they were in PERC-positive group and should have received a D-dimer test to determine the need to proceed to CTPA instead of a direct CTPA.
Outcome 3: Amount of necessary and unnecessary D-dimer tests
Two hundred and seventy-six patients (56.7%) received a D-dimer test, which was unnecessary in 144 patients (29.6%). D-dimer was ordered unnecessarily in 86 patients (54.1%) when Wells score showed a “PE Likely” and in 58 patients (85.3%) when PERC was negative in PE unlikely patients.
Outcome 4: Pulmonary embolus detection rate in necessary and unnecessary computed tomography pulmonary angiography ordered
In our study, the overall prevalence of PE was 9.5% (n: 46). PE was diagnosed in 31 patients (10.9%) who received a CTPA categorized in the necessary group (n: 284). CTPA categorized as unnecessary (n: 75) diagnosed PE with a rate of 2.7% (n: 2).
| Discussion|| |
In our setting, the PE clinical decision-making pathway was incorrectly followed for most patients. In addition to ordering unnecessary CTPA for patients, there was a considerable amount of incorrect usage of D-dimer testing in the ED. As expected, CTPAs ordered after correctly following the PE clinical decision-making pathway diagnosed more PEs as compared to PEs diagnosed by incorrectly ordered CTPAs.
CTPA is considered the gold standard diagnostic modality for PE. International guidelines suggest using validated clinical prediction rules such as the Wells and PERC scores to estimate pretest probability in patients suspected to have an acute PE.,, Wells score alone is used to rule in PE as it has a relatively high specificity. It was also reported that if paired with D-dimer testing, specificity would increase up to 99.7%.
The overuse of the D-dimer to screen for possible PE has been found to add to the cost of healthcare and increase the frequency of testing for PE among very low-risk patients, leading to large numbers of negative CTPA in young patients, increasing the risk of malignancy secondary to radiation exposure.,, Therefore, Kline et al. derived and tested the PERC rule to justify not ordering a D-dimer. Applying this rule in the low-risk populations yielded sensitivities of 96%100% and specificities of 15%27%. The combination of Wells and PERC to diagnose PE provides additional accuracy and protects against unnecessary investigations.
Our study identifies the overutilization of CTPA in the ED, which is congruent with several other studies globally,,, although no studies provide optimum adherence levels. Our study shows that the majority (59.3%) of CTPA being ordered for patients suspected to have a PE did not adhere to internationally validated guidelines.,, The amount of inappropriately ordered CTPA was on the higher end when compared to studies, which report rates from 16.8% to 43%.,,,, Out of the 75 unnecessarily ordered CPAs, only 2 (2.7%) were positive for PE, subjecting 73 (97.3%) patients out of 75 who underwent the unnecessary CTPA to unnecessary ED waiting times, costs, and an increased risk of CTPA side effects. We could not compare our results to other literature as no literature was found delineating results of unnecessary CTPAs from CTPAs done with nonadherence to the pathway. We believe this finding is significant and should be addressed for institutional improvements.
D-dimer was ordered 29.6% unnecessarily in our study. Al Dandan et al. and Osman et al. reported unnecessary D-Dimer order at 41.2% and 15%, respectively., They also reported the issue of underutilizing the D-dimer test with PERC-positive patients and skipping straight to CT, which was 57% of all PE unlikely cases for both studies. Underutilization of D-dimer testing when required and skipping straight to CTPA happened in 38.8% of all our PE Unlikely cases, which is almost half the amount reported by Al Dandan et al. and Osman et al.,
The overall yielding of positive PE in CTPA in our population is 9.5%, similar to the generally accepted threshold of 10% for the positivity rate for CTPA. In our study, when the pathway was not appropriately used and unnecessary CTPAs were ordered, the PE detection rate was four times lower than when the CTPAs were ordered according to PE clinical decision-making pathway. This implies significantly reduced sensitivity when not following the pathway, which was also shown by Roy et al. and Yan et al. who highlighted that implementing using the tools and following clinical decision-making pathways before ordering CTPAs resulted in a significant reduction in CTPA utilization with a resultant increased positive CTPA rate., All our PERC-negative patients did not have a PE on the CTPA that was subsequently performed, confirming the safety of not performing CTPA scans in these patients. This adds further support to the PERC tool developed by Kline et al., which states that low-risk patients who satisfy all criteria do not need D-dimer testing or further imaging investigation.
The direct cause for our unnecessary CTPA and D-dimer was improperly following the recommended pathway. Indirectly attributing causes could have been individual factors such as the different methods of approach between physicians, distrust of clinical decision rules, fear of litigation, and insufficient understanding of radiation exposure. Environmental factors such as ED patient load and flow and shift timings and duration could also play a role. An important contributing factor to over-utilizing CTPA could be the availability of CTPA with a rapid turnaround time that is available free of charge for all citizens, which was also noted in a similar study in the Gulf. The underutilization of D-dimer assay testing when necessary is due to non-adherence to the recommended pathway, which could be attributed to the reluctance to order the test due to the high frequency of false-positive results.
All the studies reviewed regarding investigating the use of risk stratification tools for PE had obtained data pre-COVID. Knowing that COVID-19 infection increases the risk of pulmonary embolism could manipulate the way physicians estimate this risk, which could alter the course of diagnosing patients presenting with mimicking symptoms.,, There are currently no specifically developed or validated risk stratification tools for estimating the risk of PE in COVID-19 patients. This is when ordering a CTPA directly or using the available risk stratification tools heavily depends on clinical gestalt.
A combination of in-house training sessions, adopting computerized physician order entry systems, and utilizing performance feedback reports have been proven to improve adherence to recommended diagnostic pathways, reduce the overutilization of imaging, and enhance physicians' approach to diagnosing PE.,,, Lacking a written diagnostic algorithm to score the clinical probability of PE in suspected patients was also an obstacle adding to the inappropriate workup.
This was the first study in the country to assess the utility of risk stratification tools in diagnosing pulmonary emboli in the ED. Being a single-center study may limit the generalization of the results to other settings. Another limitation is the study's retrospective nature, as scores were calculated retrospectively, and there was the risk of over-calculating or under-calculating the Wells score. However, researchers did their best to apply scores with available objective data in the files and randomly cross-checked their accuracy. An important limitation was using a generalized D-dimer cut-off level and not an age-adjusted cut-off level. Finally, this study was conducted on data obtained before the emergence of the COVID-19 pandemic. Therefore, the pandemic might affect and change current practice and usage of PE clinical decision-making pathway in many settings.
| Conclusion|| |
Our study revealed that about two-thirds of CTPAs ordered did not adhere to the proper recommended clinical decision-making pathway, with a substantial amount of unnecessary CTPAs and improper utilization of D-dimer testing. The study also highlighted that adherence to PE clinical decision-making pathway increases the chance of diagnosing PE with CTPAs. Further studies are required to elaborate on the influencing factors for the evident nonadherence.
Author contributions statement
AH: Project administration; conceptualization; methodology; investigation; formal analysis; writing- original draft, review, and editing; visualization. AJ: Conceptualization; methodology; investigation; formal analysis; writing- original draft, review, and editing. MA: Conceptualization; investigation; supervision; writing- review and editing. AAC: Conceptualization; formal analysis; supervision; writing- review and editing; visualization.
Conflicts of interest
The Tawam-Office of Research Governance and Research Ethics Committee approved this study on December 20th, 2020 (#MF2058-2020-760)
| References|| |
Al Sayegh F, Almahmeed W, Al Humood S, Marashi M, Bahr A, Al Mahdi H, et al.
Global risk profile verification in patients with venous thromboembolism (GRIP VTE) in 5 gulf countries. Clin Appl Thromb Hemost 2009;15:289-96.
Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing GJ, Harjola VP, et al.
2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS): The Task Force for the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur Respir J 2019;54:1901647.
Le Gal G, Bounameaux H. Diagnosing pulmonary embolism: Running after the decreasing prevalence of cases among suspected patients. J Thromb Haemost 2004;2:1244-6.
Hsu N, Soo Hoo GW. Underuse of clinical decision rules and d-dimer in suspected pulmonary embolism: A nationwide survey of the Veterans Administration Healthcare System. J Am Coll Radiol 2020;17:405-11.
Perrier A, Roy PM, Sanchez O, Le Gal G, Meyer G, Gourdier AL, et al.
Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 2005;352:1760-8.
Reagle Z, Tringali S, Gill N, Peterson MW. Diagnostic yield and renal complications after computed tomography pulmonary angiograms performed in a community-based academic hospital. J Community Hosp Intern Med Perspect 2012;2:17722.
Stojanovska J, Carlos RC, Kocher KE, Nagaraju A, Guy K, Kelly AM, et al.
CT pulmonary angiography: Using decision rules in the emergency department. J Am Coll Radiol 2015;12:1023-9.
Crichlow A, Cuker A, Mills AM. Overuse of computed tomography pulmonary angiography in the evaluation of patients with suspected pulmonary embolism in the emergency department. Acad Emerg Med 2012;19:1219-26.
Fesmire FM, Brown MD, Espinosa JA, Shih RD, Silvers SM, Wolf SJ, et al.
Critical issues in the evaluation and management of adult patients presenting to the emergency department with suspected pulmonary embolism. Ann Emerg Med 2011;57:628-52.e75.
Al Dandan O, Hassan A, Alnasr A, Al Gadeeb M, AbuAlola H, Alshahwan S, et al.
The use of clinical decision rules for pulmonary embolism in the emergency department: A retrospective study. Int J Emerg Med 2020;13:23.
Ferguson C, Low G, Fung C. Retrospective analysis of the computed tomography pulmonary angiogram utilization patterns in the emergency department. Can Assoc Radiol J 2019;70:388-93.
Osman M, Subedi SK, Ahmed A, Khan J, Dawood T, Ríos-Bedoya CF, et al.
Computed tomography pulmonary angiography is overused to diagnose pulmonary embolism in the Emergency Department of Academic Community Hospital. J Community Hosp Intern Med Perspect 2018;8:6-10.
Roy PM, Meyer G, Vielle B, Le Gall C, Verschuren F, Carpentier F, et al.
Appropriateness of diagnostic management and outcomes of suspected pulmonary embolism. Ann Intern Med 2006;144:157-64.
Wells PS, Anderson DR, Rodger M, Stiell I, Dreyer JF, Barnes D, et al.
Excluding pulmonary embolism at the bedside without diagnostic imaging: Management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001;135:98-107.
Kline JA, Courtney DM, Kabrhel C, Moore CL, Smithline HA, Plewa MC, et al.
Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6:772-80.
Bass AR, Fields KG, Goto R, Turissini G, Dey S, Russell LA. Clinical decision rules for pulmonary embolism in hospitalized patients: A systematic literature review and meta-analysis. Thromb Haemost 2017;117:2176-85.
Wells PS, Anderson DR, Rodger M, Ginsberg JS, Kearon C, Gent M, et al.
Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: Increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000;83:416-20.
Yan Z, Ip IK, Raja AS, Gupta A, Kosowsky JM, Khorasani R. Yield of CT pulmonary angiography in the emergency department when providers override evidence-based clinical decision support. Radiology 2017;282:717-25.
Kauppi JM, Airaksinen KE, Saha J, Bondfolk A, Pouru JP, Purola P, et al.
Adherence to risk-assessment protocols to guide computed tomography pulmonary angiography in patients with suspected pulmonary embolism. Eur Heart J Qual Care Clin Outcomes 2022;8:461-8.
Chen Z, Deblois S, Toporowicz K, Boldeanu I, Francoeur MO, Sadouni M, et al.
Yield of CT pulmonary angiography in the diagnosis of acute pulmonary embolism: Short report. BMC Res Notes 2019;12:41.
Miró Ò, Jiménez S, Mebazaa A, Freund Y, Burillo-Putze G, Martín A, et al.
Pulmonary embolism in patients with COVID-19: Incidence, risk factors, clinical characteristics, and outcome. Eur Heart J 2021;42:3127-42.
Riyahi S, Dev H, Behzadi A, Kim J, Attari H, Raza SI, et al.
Pulmonary embolism in hospitalized patients with COVID-19: A multicenter study. Radiology 2021;301:E426-33.
Grillet F, Behr J, Calame P, Aubry S, Delabrousse E. Acute pulmonary embolism associated with COVID-19 pneumonia detected with pulmonary CT angiography. Radiology 2020;296:E186-8.
Raja AS, Ip IK, Dunne RM, Schuur JD, Mills AM, Khorasani R. Effects of performance feedback reports on adherence to evidence-based guidelines in use of CT for evaluation of pulmonary embolism in the emergency department: A randomized trial. AJR Am J Roentgenol 2015;205:936-40.
Deblois S, Chartrand-Lefebvre C, Toporowicz K, Chen Z, Lepanto L. Interventions to reduce the overuse of imaging for pulmonary embolism: A systematic review. J Hosp Med 2018;13:52-61.