American Journal of Emergency Medicine 32 (2014) 609–613
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Pulmonary Embolism Rule-out Criteria vs D-dimer testing in low-risk patients for pulmonary embolism: a retrospective study☆,☆☆,★ J. Bokobza, MD a, A. Aubry, MD a, N. Nakle, MD b, C. Vincent-Cassy, MD b, D. Pateron, MD, PhD c, f, C. Devilliers, MD e, B. Riou, MD, PhD a, f, P. Ray, MD, PhD d, f, Y. Freund, MD a, f,⁎ a
Emergency Department, Hôpital Pitié-Salpêtrière, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France Emergency Department, Hôpital Bicêtre, Assistance Publique – Hôpitaux de Paris (APHP), Le Kremlin-Bicêtre, France c Emergency Department, Hôpital Saint-Antoine, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France d Emergency Department, Hôpital Tenon, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France e Biochemistry Department, Hôpital Pitié-Salpêtrière, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France f Sorbonne Universités, UPMC Univ Paris 06, F-75005, Paris, France b
a r t i c l e
i n f o
Article history: Received 10 January 2014 Received in revised form 9 March 2014 Accepted 10 March 2014
a b s t r a c t Study objectives: The Pulmonary Embolism Rule-out Criteria (PERC) score has shown excellent negative predictive value; however, its use in the European population with high prevalence of PE is controversial. In Europe, PERC is not part of routine practice. For low-risk patients, guidelines recommend D-dimer testing, followed if positive by imaging study. We aimed to study the rate of diagnosis of PE after D-dimer testing in PERC-negative patients that could have been discharged if PERC was applied. Method: This was a multicenter retrospective study in Paris, France. We included all patients with a suspicion of PE who had D-dimer testing in the emergency department, low pre-test probability, and a negative PERC score (that was retrospectively calculated). Patients with insufficient record to calculate PERC score were excluded. The primary end point was the rate of PE diagnosis before discharge in this population. Secondary end points included rate of invasive imaging studies and subsequent adverse events. Results: We screened 4301 patients who had D-dimer testing, 1070 of whom were PERC negative and could be analyzed. The mean age was 35 years and 46% were men. D-dimer was positive (N500 ng/L) in 167 (16%) of them; CTPA or V/Q scan was performed in 153 (14%) cases. PE was confirmed in 5 cases (total rate 0.5%, 95% confidence interval 0.1%-1.1%). Fifteen patients (1%) experienced non-severe adverse events. Conclusion: D-dimer testing in PERC-negative patients led to a diagnosis of PE in 0.5% of them, with 15% of patients undergoing unnecessary irradiative imaging studies. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Diagnosis of pulmonary embolism (PE) in the emergency department (ED) is a crucial matter. This potentially life-threatening condition has many different presentations, including any complaint of chest pain, dyspnea, or syncope, which represent a
☆ This article has been presented at the 2013 research forum of the ACEP scientific assembly, Seattle, WA. ☆☆ Author Contribution Statement: YF, PR and BR conceived the study and designed the trial. JB, AA, NN, CVC and DP collected the data. CD provided biochemical expertise. YF and BR undertook statistical analysis. YF drafted the manuscript. BR, PR and DP provided extensive reviewing and substantial revisions. ★ The authors would like to thank Dr EC Baker, Barts Health NHS trust, for extensive review and support. ⁎ Corresponding author. Service d’Accueil des Urgences, Hôpital Pitié-Salpêtriere, 47-83 boulevard de l’ Hôpital, 75013 Paris, France. Tel.: 00 33 1 84827129. E-mail address: [email protected] (Y. Freund). http://dx.doi.org/10.1016/j.ajem.2014.03.008 0735-6757/© 2014 Elsevier Inc. All rights reserved.
volume of more than 10 million patients a year in the United States to assess for potential PE [1]. The fear of missing a diagnosis of this life-threatening disease has led to an increase in the use of invasive diagnostic strategies, with a significant rise of computed tomographic pulmonary angiography (CTPA) use over the last decade [2,3]. Subsequently, it has also been reported that PE has become overdiagnosed: more small emboli and mild PEs are detected on CT, with no clear benefit of this diagnosis in terms of outcomes [4-6]. In the workup of PE, several clinical decision rules have been proposed to guide proper use of diagnostic studies, depending on the pretest probability for the condition. The PE rule-out criteria (PERC) is an 8-item decision rule that was initially reported to safely rule out PE in the ED in patients with low pretest probability [7]. Although further studies reported conflicting results [8], 2 recent meta-analyses conclude that PERC has sufficient sensitivity to rule out PE in patients with low pre-test probability [9,10]. Despite these results, PERC
610
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
decision-based policies are not widespread [11], especially in Europe where overall prevalence of PE is higher [8]. In patients who could have been safely discharged if the PERC rule had been applied, consequences of resource overutilization include longer ED stay, increased costs, and excessive radiation. In the lowrisk population, the prevalence of PE for PERC-negative patients varies from 1% to 5% [7,8,12], and emergency physicians may still prefer to rely on D-dimer testing to rule out PE. However, as raised by Self et al [13], due to its imperfect specificity, this strategy could lead to further testing and may be responsible for overdiagnosis of PE and adverse events from both unnecessary imaging studies and treatment. We sought to study the consequences of D-dimer testing in this very low-risk population in a European sample, a setting where previous reports have called into question the validity of the PERC approach and where this rule is not used in routine practice. The goal of this investigation was not to report the rate of PE in PERC-negative patients but to determine the actual rate of diagnosed PE in this population with the use of D-dimer. We also report the rate of potentially avoidable ionizing radiation studies and adverse events in this low-probability sample. 2. Methods 2.1. Study setting and population We conducted a multicenter retrospective chart review in 4 urban academic emergency departments in the Paris metropolitan area in 2012. Institutional review board (Comité de Protection des Personnes – Paris ile de France 6) approved the study and waived informed consent as the study was retrospective. The study was performed from January 1, 2012, for a 12-month period. The four centers have an annual census of 45 000 to 55 000 ED visits each and are affiliated with Assistance Publique-Hopitaux de Paris (Paris, France). We screened all patients who attended one of these centers during the study period and had D-dimer testing. All these EDs use the same electronic medical record software (Urqual; McKesson, San Francisco, CA). We extracted all medical charts pertaining to those screened patients. 2.2. Routine practice and selection of participants Diagnostic strategies for ruling out PE are similar in the 4 participating centers and likely generalisable to most French EDs. As recommended by European guidelines [14], national expert recommendations [15] and local policies, patients suspect of having a PE are first assessed for pre-test probability either with gestalt, Revised Geneva Score (RGS) or Wells Score. For low and intermediate risk, routine practice includes D-dimer testing, with imaging studies if positive. Thereby all patients with a suspicion of PE and a low pretest probability should have D-dimer testing in the ED. We screened and electronically retrieved all patients with D-dimer tests in the 4 EDs in the study period. All extracted charts were reviewed by trained abstractors, and the following patients were excluded from analysis: - Patients with incomplete charts and missing data to calculate PERC score - Patients with D-dimer testing for any other reason than excluding PE (suspicion of dural sinus thrombosis, deep vein thrombosis, septic patients with coagulation screen, etc.) - Patients with a positive PERC score, i.e., any positive answer to the following: age above 49 years, pulse rate above 99 beats per minute, pulse oxymetry less than 95% on room air, history of hemoptysis, exogenous estrogen intake, prior diagnosis of PE or deep vein thrombosis, recent surgery or trauma in the last 4 weeks, and unilateral leg swelling. - Patients with an intermediate or high pre-test probability using RGS [16,17].
2.3. Data collection and objective For each included patient, we collected clinical baseline characteristics on arrival in the ED including blood pressure, heart rate, oxygen saturation, respiratory rate, and Glasgow Coma Scale. History of the presenting complaint, past medical history and prior medication were also retrieved from the electronic medical chart, and subsequently, PERC score and RGS could be retrospectively calculated (Table 1). PERC-negative patients were those with all items of PERC score negative, ie, a score of zero. Chart abstraction was performed according to the recommendation of Gilbert et al. [18] including: - Training of abstractors before the study starts with practice medical charts and examples. - Explicit protocols and criteria for selection and inclusion, or exclusion of screened patients. - Definition of every variable recorded - Regular meeting between abstractors and study coordinators during the study, and after completion of data collection at each site - Blinding initial chart reviews to the tested hypothesis, as PERC score was calculated prior to the collection of diagnosis with the exception of a double blind chart abstractions with interrater reliability calculation. In our study, every chart underwent a second review by the study coordinator, who was not blind to the first review. In case of disagreement regarding PERC score or the adjudication diagnosis, consensus was made upon discussion between the 2 abstractors. The primary objective was to assess the rate of diagnosed PE with D-dimer testing in the PERC-negative patient population. The primary end point was the diagnosis of PE in the ED after D-dimer testing in PERC-negative patients. This rate corresponds to the true-positive rate with D-dimer testing and false-negative rate with PERC score, and will express the added value of D-dimer testing amongst patients with very low pretest probability, with negative PERC score and RGS less than 3. Secondary end points included the rate of invasive imaging studies, the rate of anticoagulation treatment commencement and the rate of unnecessary hospitalization of patients for further testing, and adverse events associated with these. 2.4. Outcome measure D-dimers were analyzed using the ELISA method on MiniVidas automat (Biomerrieux, Marcy l’Etoile, France) with normal range Table 1 Revised Geneva Score for Pulmonary Embolism. Adapted from Ann Intern Med. 2006;144(3):165-171 Risk factors
Points
Age older than 65 y Previous DVT or pulmonary embolism Surgery under general anesthesia or fracture of lower limbs within 1 month Active malignant condition (solid or hematologic, currently active or considered cured b1 y) Symptoms Unilateral lower limb pain Hemoptysis Clinical signs Heart rate 75-94 beats/min Heart rate ≥95 beats/min Pain on lower limb deep venous palpation and unilateral edema Clinical probability Low Intermediate High
1 3 2 2
3 2 3 5 4 Score 0-3 total 4-10 total ≥11 total
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
defined as under 500 ng/mL, coefficient of variation 5.9%, and limits of detection from 45 to 10 000 ng/mL. All imaging studies, namely CTPA or V/Q scan, were interpreted by a senior radiologist and corroborated by a second senior radiologist blinded to PERC score. Diagnostic adjudication of PE was later made by two experts (one chart abstractor and the study principal investigator), based on these imaging study reports and all medical records pertaining to the patient stay. Pre-defined adverse events included allergic reaction to contrast agent infusion, anticoagulation treatment in the absence of PE and undue hospitalization. 2.5. Sample size calculation and statistical analysis According to the cross-sectional studies by Hansson et al [19], at least 0.5% of men aged 50 years and older have a PE. We made the hypothesis that the rate of PE in PERC-negative patients would be similar to the one in the general population and calculated our sample size so that the upper limit of the 95% confidence interval (CI) would be less than 1%: 1110 patients should be included in the analysis. Based on preliminary analysis of our database, a 12-month study period in the 4 centers would be sufficient to achieve this goal. Normally distributed variables are expressed as mean (SD), other continuous variables as median (interquartile range [IQR] 25%-75%). Normality was assessed using Kolmogorov-Smirnov test. Confidence intervals (CI) of proportions were calculated using the “exact” binomial method of Clopper and Pearson [20]. Agreement between chart abstractor and study principal investigator for the diagnosis of PE was calculated with the Cohen κ method. Analyses were performed with SPSS software (version 20.0, IBM, Armonk, NY). 3. Results Within the study period there were 223 287 ED visits in the 4 participating EDs, from which 4301 patients had D-dimer testing.
611
Table 2 Main characteristics Characteristics
N
All patients Center Pitie Saint-Antoine Bicêtre Tenon Age, mean (SD) Heart rate, mean (SD) Systolic BP, mean (SD) Diastolic BP, mean (SD) Pulse oxymetry, median [IQR] Temperature, mean (SD) Tests in ED ECG Arterial blood gas Chest x-ray CT pulmonary angiogram V/Q scan D-dimer N500 ng/L Confirmed PE Hospital admission
1070
100%
510 89 273 198 35.3 80 130 80 99 36.8
48% 8% 26% 19% (9) (12) (24) (17) [98;100] (0.9)
394 456 829 143 11 167 5 94
37% 43% 77% 13% 1% 16% 0.5% 9%
ECG, electrocardiogram; CT: computed tomography.
Amongst these 4301 patients, 442 were excluded as they were not tested to rule out PE. Of the 3859 patients with suspicion of PE who had D-dimer testing, 1070 (27%) had a PERC score of zero (Fig.). Main characteristics of the studied sample are summarized in Table 2. Mean age of the patients analyzed was 35 years (SD, 9), and 496 (46%) were men. One hundred sixty-seven patients (16%) had a D-dimer value above normal range, ie, N500 ng/mL. Amongst these patients, median D-dimer value was 725 ng/L [IQR 659; 887]. Subsequently, 153 (14%) invasive studies were performed including 143 CTPAs and 11 V/Q scans (one patient had both). Six patients had a positive CTPA: one was considered as “inconclusive” and subsequently PE was ruled out after negative V/Q scan. In total, 5 patients were diagnosed with a PE in this population (Cohen κ 1.0), a rate of 0.5% (95% CI 0.1%-1.1%). In our sample of very low-risk patients, D-dimer testing had a positive predictive value of 3%. Amongst the five patients diagnosed with PE who had D-dimer testing, 2 of them were subsegmental; 2, segmental; and the fifth one, lobar. D-dimer testing in PERC-negative patients led to the completion of invasive studies for 153 (14%) of patients; 15 of them (1% in total) experienced adverse events. All adverse events were non severe, including two allergic reactions to iodine contrast agent (generalized urticaria), and 13 patients were started on an anticoagulation regimen and underwent unnecessary admission (all of them while waiting for imaging studies that were not rapidly available). No bleeding was reported, and no severe adverse events resulted from this management. 4. Discussion
Fig. Flowchart. PERC: Pulmonary Embolism Rule out Criteria. RGS: Revised Geneva Score. PE: Pulmonary Embolism.
In this study, we undertook a retrospective analysis of patients with a negative PERC score and low pre-test probability of PE, who have had D-dimer testing in the ED. The objectives were to assess the yield and adverse events subsequent to further testing in this very low-risk population, rather than to determine the exact rate of PE. We found that 0.5% of PERC negative patients were subsequently diagnosed with a PE after D-dimer testing. Although it is not the real prevalence of PE in this population, this very low rate represents the percentage of patients who are diagnosed with PE when PERC rule is not applied. First validated in 2004 by Kline et al [7], PERC score has a very high sensitivity and negative predictive value for the diagnosis of PE [9,10].
612
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
However, to rely solely on PERC score, even associated with a low pretest probability, remains controversial. As raised by Schwartz et al [21], there is still no intervention trial in which PERC score is used for clinical decision making; hence, reliability and clinical application of this tool still need confirmation. Rationale for its use appears weaker in Europe where prevalence of PE has been reported to be higher than in the United States, and where a large trial has questioned the reliability of PERC rule [8]. In a recent analysis, Self and Barett simulated the impact of Ddimer testing in PERC-negative patients, based on previous theoretical diagnostic performances [13]. They assessed the risk/benefit ratio of D-dimer testing, derived from reported values of diagnostic performances of PERC score, D-dimer, and CTPA. They calculated that in a sample of 1,000 PERC negative patients, 30 will have a PE; and that Ddimer testing strategy would lead to the completion of 621 CTPAs, 6 PEs still being missed, 24 being correctly diagnosed, and 24 treated in the absence of a PE. The benefit in terms of outcome for these low risk PEs are unclear, and their conclusion reached the one from Newman et al. [22] who suggested a higher rate of death resulting from a liberal strategy of D-dimer testing and thus anticoagulation-related complications and consequences of invasive imaging. Our results differ slightly from these as we did not test all PERC negative patients with D-dimer. Due to the design of this study, only a fraction of them were tested, hence the different results and possible overestimation of the rate of PE in this population. The strength of our study lies in its retrospective design, which allowed us to describe what changes would occur with a PERC based strategy. Indeed, even in the absence of a PERC based decision rule (as it was the case in our centers), some patients were still not tested for PE with a D-dimer, and our study focuses only on those who could have been weaned from this test if the PERC rule was used. We report a rate of PE of 0.5% (95% CI 0.1-1.1%). This rate is similar to that reported in the general population [19,23]. Previous studies have highlighted the high rate (13%-23%) of invasive studies that could have been avoided with a PERC-based decision making policy [11,24]. In our sample, 14% of patients underwent invasive imaging studies. The five patients with a PE were all diagnosed with low-risk PE according to the recent classification [25], with an estimated mortality rate approaching 1% [26-28]. Furthermore, Kline et al [12] found in their retrospective studies that mortality of PERC-negative patients with a PE was 0% at 30 days in their sample. The benefits of these supplemental investigations are unclear, and several studies report that PE tends to be overdiagnosed, with unclear added value. More frequent use of CTPA led to increased diagnosis of less fatal PE [3,4]. PERC score has shown excellent sensitivity and negative predictive value[7,9,10,29]. However controversy remains on whether it should be applied alone in patients with a low pre-test probability [8]. Its very low positive predictive value, the high rate of invasive studies, and the unknown benefits in terms of outcome in those with very low-risk PE advocate against the use of D-dimer testing in PERCnegative patients. 5. Limitations The main limitation to this study was the absence of follow-up due to its retrospective design. Sensitivity, specificity and negative predictive values could not have been calculated as we do not know how many false-negative patients were discharged. However, the objective of our study was to determine the real clinical added value of D-dimer testing in those PERC-negative patients, rather than to assess the performances of this clinical score. Indeed two recent metaanalyses provide these data, with very high sensitivity and negative predictive value, and acceptable specificity [9,10]. Another limitation is that we electronically retrieved all patients with D-dimer testing in our EDs for screening thereafter excluding those with no clear
suspicion of PE hence creating an inclusion bias. We cannot guarantee that all patients with a suspected PE were included for analysis. However, we believe that this bias is very limited, as we stated above, since guidelines and local policies recommend the use of D-dimer for all patients suspected for PE with low pre-test probability. Lastly, in our sample PERC score was retrospectively calculated after chart abstraction. Even though we discarded any incomplete chart with missing data regarding PERC score, we cannot be sure that some of our PERC-negative patients did not in fact have one positive item that was not clearly reported on the medical chart. In summary, the analysis of our cohort reports that D-dimer testing for PERC-negative patients with low pretest probability led to nearly 15% of irradiative imaging studies, for a rate of newly diagnosed PED of 0.5%. This very low rate if close to that of the general population, suggesting that further testing for PERC negative patients should be abandoned. Intervention studies are warranted to validate this hypothesis.
References [1] McCaig LF, Nawar EW. National hospital ambulatory medical care survey: 2004 emergency department summary. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics 2006. http://198.246.124.29/nchs/data/ad/ad372.pdf (accessed 26 Jul2013). [2] Smith-Bindman R, Miglioretti DL, Johnson E, et al. Use of diagnostic imaging studies and associated radiation exposure for patients enrolled in large integrated health care systems, 1996-2010. JAMA J Am Med Assoc 2012;307:2400–9. http:// dx.doi.org/10.1001/jama.2012.5960. [3] Schissler AJ, Rozenshtein A, Kulon ME, et al. CT pulmonary Angiography: increasingly diagnosing less severe pulmonary emboli. PLoS ONE 2013;8: e65669. http://dx.doi.org/10.1371/journal.pone.0065669. [4] Sheh SH, Bellin E, Freeman KD, et al. Pulmonary embolism diagnosis and mortality with pulmonary CT angiography versus ventilation-perfusion scintigraphy: evidence of overdiagnosis with CT? AJR Am J Roentgenol 2012;198:1340–5. http://dx.doi.org/10.2214/AJR.11.6426. [5] 5 Wiener RS, Schwartz LM, Woloshin S. When a test is too good: how CT pulmonary angiograms find pulmonary emboli that do not need to be found. BMJ 2013;347:f3368–f3368. doi:10.1136/bmj.f3368. [6] Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med 2011;171:831–7. http://dx.doi.org/10.1001/archinternmed.2011.178. [7] Kline JA, Mitchell AM, Kabrhel C, et al. Clinical criteria to prevent unnecessary diagnostic testing in emergency department patients with suspected pulmonary embolism. J Thromb Haemost JTH 2004;2:1247–55. http://dx.doi.org/10.1111/ j.1538-7836.2004.00790.x. [8] Hugli O, Righini M, Le Gal G, et al. The pulmonary embolism rule-out criteria (PERC) rule does not safely exclude pulmonary embolism. J Thromb Haemost JTH 2011;9:300–4. http://dx.doi.org/10.1111/j.1538-7836.2010.04147.x. [9] Singh B, Parsaik AK, Agarwal D, et al. Diagnostic accuracy of pulmonary embolism rule-out criteria: a systematic review and meta-analysis. Ann Emerg Med 2012;59: 517–520.e1–4. http://dx.doi.org/10.1016/j.annemergmed.2011.10.022. [10] Singh B, Mommer SK, Erwin PJ, et al. Pulmonary embolism rule-out criteria (PERC) in pulmonary embolism–revisited: A systematic review and meta-analysis. Emerg Med J EMJ Published Online First: 20 October 2012. http://dx.doi.org/10.1136/ emermed-2012-201730. [11] Dachs RJ, Kulkarni D, Higgins 3rd GL. The Pulmonary Embolism Rule-Out Criteria rule in a community hospital ED: a retrospective study of its potential utility. Am J Emerg Med 2011;29:1023–7. http://dx.doi.org/10.1016/j.ajem.2010.05.018. [12] Kline JA, Slattery D, O’Neil BJ, et al. Clinical features of patients with pulmonary embolism and a negative PERC rule result. Ann Emerg Med 2013;61:122–4. http:// dx.doi.org/10.1016/j.annemergmed.2012.06.494. [13] Self WH, Barrett TW. Is ‘PERC negative’ adequate to rule out pulmonary embolism in the emergency department? Evaluating meta-analysis for studies of clinical prediction models: answers to the July 2012 journal club questions. Ann Emerg Med 2012;60:803–14. http://dx.doi.org/10.1016/j.annemergmed.2012.07.121. [14] Torbicki A, Perrier A, Konstantinides S, et al. Guidelines on the diagnosis and management of acute pulmonary embolism: the Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC). Eur Heart J 2008;29:2276–315. http://dx.doi.org/10.1093/eurheartj/ ehn310. [15] Roy P-M, Bichri A, Bouet R, et al. Embolie pulmonaire: algorithmes diagnostiques. 2007.http://www.sfar.org/acta/dossier/archives/mu07/html/mu07_07/urg07_07. htm (accessed 23 Dec 2013). [16] Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006;144: 165–71. [17] Klok FA, Kruisman E, Spaan J, et al. Comparison of the revised Geneva score with the Wells rule for assessing clinical probability of pulmonary embolism. J Thromb Haemost JTH 2008;6:40–4. http://dx.doi.org/10.1111/j.1538-7836.2007.02820.x.
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613 [18] Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27:305–8. [19] Hansson PO, Welin L, Tibblin G, et al. Deep vein thrombosis and pulmonary embolism in the general population. ‘The Study of Men Born in 1913’. Arch Intern Med 1997;157:1665–70. [20] Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934;26:404–13. http://dx.doi.org/10.1093/biomet/26.4.404. [21] Precautions Schwartz DT, With PERC. Ann Emerg Med 2013;62:197. http://dx.doi. org/10.1016/j.annemergmed.2013.02.030. [22] Newman DH, Schriger DL. Rethinking testing for pulmonary embolism: less is more. Ann Emerg Med 2011;57:622–627.e3. http://dx.doi.org/10.1016/j. annemergmed.2011.04.014. [23] White RH. The Epidemiology of Venous Thromboembolism. Circulation 2003;107: I–4–I–8. http://dx.doi.org/10.1161/01.CIR.0000078468.11849.66. [24] 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 Off J Soc Acad Emerg Med 2012;19:1219–26. http://dx.doi.org/10.1111/acem.12012. [25] Jaff MR, McMurtry MS, Archer SL, et al. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis, and chronic thrombo-
[26]
[27]
[28]
[29]
613
embolic pulmonary hypertension: a scientific statement from the American Heart Association. Circulation 2011;123:1788–830. http://dx.doi.org/10.1161/ CIR.0b013e318214914f. Bova C, Pesavento R, Marchiori A, et al. Risk stratification and outcomes in hemodynamically stable patients with acute pulmonary embolism: a prospective, multicentre, cohort study with three months of follow-up. J Thromb Haemost JTH 2009;7:938–44. http://dx.doi.org/10.1111/j.1538-7836.2009.03345.x. Palmieri V, Gallotta G, Rendina D, et al. Troponin I and right ventricular dysfunction for risk assessment in patients with nonmassive pulmonary embolism in the Emergency Department in combination with clinically based risk score. Intern Emerg Med 2008;3:131–8. http://dx.doi.org/10.1007/s11739008-0134-2. Post F, Mertens D, Sinning C, et al. Decision for aggressive therapy in acute pulmonary embolism: implication of elevated troponin T. Clin Res Cardiol Off J Ger Card Soc 2009;98:401–8. http://dx.doi.org/10.1007/s00392-0090017-1. Hogg K, Dawson D, Kline J. Application of pulmonary embolism rule-out criteria to the UK Manchester Investigation of Pulmonary Embolism Diagnosis (MIOPED) study cohort. J Thromb Haemost JTH 2005;3:592–3. http://dx.doi.org/10.1111/ j.1538-7836.2005.01212.x.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Pulmonary Embolism Rule-out Criteria vs D-dimer testing in low-risk patients for pulmonary embolism: a retrospective study☆,☆☆,★ J. Bokobza, MD a, A. Aubry, MD a, N. Nakle, MD b, C. Vincent-Cassy, MD b, D. Pateron, MD, PhD c, f, C. Devilliers, MD e, B. Riou, MD, PhD a, f, P. Ray, MD, PhD d, f, Y. Freund, MD a, f,⁎ a
Emergency Department, Hôpital Pitié-Salpêtrière, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France Emergency Department, Hôpital Bicêtre, Assistance Publique – Hôpitaux de Paris (APHP), Le Kremlin-Bicêtre, France c Emergency Department, Hôpital Saint-Antoine, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France d Emergency Department, Hôpital Tenon, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France e Biochemistry Department, Hôpital Pitié-Salpêtrière, Assistance Publique – Hôpitaux de Paris (APHP), Paris, France f Sorbonne Universités, UPMC Univ Paris 06, F-75005, Paris, France b
a r t i c l e
i n f o
Article history: Received 10 January 2014 Received in revised form 9 March 2014 Accepted 10 March 2014
a b s t r a c t Study objectives: The Pulmonary Embolism Rule-out Criteria (PERC) score has shown excellent negative predictive value; however, its use in the European population with high prevalence of PE is controversial. In Europe, PERC is not part of routine practice. For low-risk patients, guidelines recommend D-dimer testing, followed if positive by imaging study. We aimed to study the rate of diagnosis of PE after D-dimer testing in PERC-negative patients that could have been discharged if PERC was applied. Method: This was a multicenter retrospective study in Paris, France. We included all patients with a suspicion of PE who had D-dimer testing in the emergency department, low pre-test probability, and a negative PERC score (that was retrospectively calculated). Patients with insufficient record to calculate PERC score were excluded. The primary end point was the rate of PE diagnosis before discharge in this population. Secondary end points included rate of invasive imaging studies and subsequent adverse events. Results: We screened 4301 patients who had D-dimer testing, 1070 of whom were PERC negative and could be analyzed. The mean age was 35 years and 46% were men. D-dimer was positive (N500 ng/L) in 167 (16%) of them; CTPA or V/Q scan was performed in 153 (14%) cases. PE was confirmed in 5 cases (total rate 0.5%, 95% confidence interval 0.1%-1.1%). Fifteen patients (1%) experienced non-severe adverse events. Conclusion: D-dimer testing in PERC-negative patients led to a diagnosis of PE in 0.5% of them, with 15% of patients undergoing unnecessary irradiative imaging studies. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Diagnosis of pulmonary embolism (PE) in the emergency department (ED) is a crucial matter. This potentially life-threatening condition has many different presentations, including any complaint of chest pain, dyspnea, or syncope, which represent a
☆ This article has been presented at the 2013 research forum of the ACEP scientific assembly, Seattle, WA. ☆☆ Author Contribution Statement: YF, PR and BR conceived the study and designed the trial. JB, AA, NN, CVC and DP collected the data. CD provided biochemical expertise. YF and BR undertook statistical analysis. YF drafted the manuscript. BR, PR and DP provided extensive reviewing and substantial revisions. ★ The authors would like to thank Dr EC Baker, Barts Health NHS trust, for extensive review and support. ⁎ Corresponding author. Service d’Accueil des Urgences, Hôpital Pitié-Salpêtriere, 47-83 boulevard de l’ Hôpital, 75013 Paris, France. Tel.: 00 33 1 84827129. E-mail address: [email protected] (Y. Freund). http://dx.doi.org/10.1016/j.ajem.2014.03.008 0735-6757/© 2014 Elsevier Inc. All rights reserved.
volume of more than 10 million patients a year in the United States to assess for potential PE [1]. The fear of missing a diagnosis of this life-threatening disease has led to an increase in the use of invasive diagnostic strategies, with a significant rise of computed tomographic pulmonary angiography (CTPA) use over the last decade [2,3]. Subsequently, it has also been reported that PE has become overdiagnosed: more small emboli and mild PEs are detected on CT, with no clear benefit of this diagnosis in terms of outcomes [4-6]. In the workup of PE, several clinical decision rules have been proposed to guide proper use of diagnostic studies, depending on the pretest probability for the condition. The PE rule-out criteria (PERC) is an 8-item decision rule that was initially reported to safely rule out PE in the ED in patients with low pretest probability [7]. Although further studies reported conflicting results [8], 2 recent meta-analyses conclude that PERC has sufficient sensitivity to rule out PE in patients with low pre-test probability [9,10]. Despite these results, PERC
610
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
decision-based policies are not widespread [11], especially in Europe where overall prevalence of PE is higher [8]. In patients who could have been safely discharged if the PERC rule had been applied, consequences of resource overutilization include longer ED stay, increased costs, and excessive radiation. In the lowrisk population, the prevalence of PE for PERC-negative patients varies from 1% to 5% [7,8,12], and emergency physicians may still prefer to rely on D-dimer testing to rule out PE. However, as raised by Self et al [13], due to its imperfect specificity, this strategy could lead to further testing and may be responsible for overdiagnosis of PE and adverse events from both unnecessary imaging studies and treatment. We sought to study the consequences of D-dimer testing in this very low-risk population in a European sample, a setting where previous reports have called into question the validity of the PERC approach and where this rule is not used in routine practice. The goal of this investigation was not to report the rate of PE in PERC-negative patients but to determine the actual rate of diagnosed PE in this population with the use of D-dimer. We also report the rate of potentially avoidable ionizing radiation studies and adverse events in this low-probability sample. 2. Methods 2.1. Study setting and population We conducted a multicenter retrospective chart review in 4 urban academic emergency departments in the Paris metropolitan area in 2012. Institutional review board (Comité de Protection des Personnes – Paris ile de France 6) approved the study and waived informed consent as the study was retrospective. The study was performed from January 1, 2012, for a 12-month period. The four centers have an annual census of 45 000 to 55 000 ED visits each and are affiliated with Assistance Publique-Hopitaux de Paris (Paris, France). We screened all patients who attended one of these centers during the study period and had D-dimer testing. All these EDs use the same electronic medical record software (Urqual; McKesson, San Francisco, CA). We extracted all medical charts pertaining to those screened patients. 2.2. Routine practice and selection of participants Diagnostic strategies for ruling out PE are similar in the 4 participating centers and likely generalisable to most French EDs. As recommended by European guidelines [14], national expert recommendations [15] and local policies, patients suspect of having a PE are first assessed for pre-test probability either with gestalt, Revised Geneva Score (RGS) or Wells Score. For low and intermediate risk, routine practice includes D-dimer testing, with imaging studies if positive. Thereby all patients with a suspicion of PE and a low pretest probability should have D-dimer testing in the ED. We screened and electronically retrieved all patients with D-dimer tests in the 4 EDs in the study period. All extracted charts were reviewed by trained abstractors, and the following patients were excluded from analysis: - Patients with incomplete charts and missing data to calculate PERC score - Patients with D-dimer testing for any other reason than excluding PE (suspicion of dural sinus thrombosis, deep vein thrombosis, septic patients with coagulation screen, etc.) - Patients with a positive PERC score, i.e., any positive answer to the following: age above 49 years, pulse rate above 99 beats per minute, pulse oxymetry less than 95% on room air, history of hemoptysis, exogenous estrogen intake, prior diagnosis of PE or deep vein thrombosis, recent surgery or trauma in the last 4 weeks, and unilateral leg swelling. - Patients with an intermediate or high pre-test probability using RGS [16,17].
2.3. Data collection and objective For each included patient, we collected clinical baseline characteristics on arrival in the ED including blood pressure, heart rate, oxygen saturation, respiratory rate, and Glasgow Coma Scale. History of the presenting complaint, past medical history and prior medication were also retrieved from the electronic medical chart, and subsequently, PERC score and RGS could be retrospectively calculated (Table 1). PERC-negative patients were those with all items of PERC score negative, ie, a score of zero. Chart abstraction was performed according to the recommendation of Gilbert et al. [18] including: - Training of abstractors before the study starts with practice medical charts and examples. - Explicit protocols and criteria for selection and inclusion, or exclusion of screened patients. - Definition of every variable recorded - Regular meeting between abstractors and study coordinators during the study, and after completion of data collection at each site - Blinding initial chart reviews to the tested hypothesis, as PERC score was calculated prior to the collection of diagnosis with the exception of a double blind chart abstractions with interrater reliability calculation. In our study, every chart underwent a second review by the study coordinator, who was not blind to the first review. In case of disagreement regarding PERC score or the adjudication diagnosis, consensus was made upon discussion between the 2 abstractors. The primary objective was to assess the rate of diagnosed PE with D-dimer testing in the PERC-negative patient population. The primary end point was the diagnosis of PE in the ED after D-dimer testing in PERC-negative patients. This rate corresponds to the true-positive rate with D-dimer testing and false-negative rate with PERC score, and will express the added value of D-dimer testing amongst patients with very low pretest probability, with negative PERC score and RGS less than 3. Secondary end points included the rate of invasive imaging studies, the rate of anticoagulation treatment commencement and the rate of unnecessary hospitalization of patients for further testing, and adverse events associated with these. 2.4. Outcome measure D-dimers were analyzed using the ELISA method on MiniVidas automat (Biomerrieux, Marcy l’Etoile, France) with normal range Table 1 Revised Geneva Score for Pulmonary Embolism. Adapted from Ann Intern Med. 2006;144(3):165-171 Risk factors
Points
Age older than 65 y Previous DVT or pulmonary embolism Surgery under general anesthesia or fracture of lower limbs within 1 month Active malignant condition (solid or hematologic, currently active or considered cured b1 y) Symptoms Unilateral lower limb pain Hemoptysis Clinical signs Heart rate 75-94 beats/min Heart rate ≥95 beats/min Pain on lower limb deep venous palpation and unilateral edema Clinical probability Low Intermediate High
1 3 2 2
3 2 3 5 4 Score 0-3 total 4-10 total ≥11 total
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
defined as under 500 ng/mL, coefficient of variation 5.9%, and limits of detection from 45 to 10 000 ng/mL. All imaging studies, namely CTPA or V/Q scan, were interpreted by a senior radiologist and corroborated by a second senior radiologist blinded to PERC score. Diagnostic adjudication of PE was later made by two experts (one chart abstractor and the study principal investigator), based on these imaging study reports and all medical records pertaining to the patient stay. Pre-defined adverse events included allergic reaction to contrast agent infusion, anticoagulation treatment in the absence of PE and undue hospitalization. 2.5. Sample size calculation and statistical analysis According to the cross-sectional studies by Hansson et al [19], at least 0.5% of men aged 50 years and older have a PE. We made the hypothesis that the rate of PE in PERC-negative patients would be similar to the one in the general population and calculated our sample size so that the upper limit of the 95% confidence interval (CI) would be less than 1%: 1110 patients should be included in the analysis. Based on preliminary analysis of our database, a 12-month study period in the 4 centers would be sufficient to achieve this goal. Normally distributed variables are expressed as mean (SD), other continuous variables as median (interquartile range [IQR] 25%-75%). Normality was assessed using Kolmogorov-Smirnov test. Confidence intervals (CI) of proportions were calculated using the “exact” binomial method of Clopper and Pearson [20]. Agreement between chart abstractor and study principal investigator for the diagnosis of PE was calculated with the Cohen κ method. Analyses were performed with SPSS software (version 20.0, IBM, Armonk, NY). 3. Results Within the study period there were 223 287 ED visits in the 4 participating EDs, from which 4301 patients had D-dimer testing.
611
Table 2 Main characteristics Characteristics
N
All patients Center Pitie Saint-Antoine Bicêtre Tenon Age, mean (SD) Heart rate, mean (SD) Systolic BP, mean (SD) Diastolic BP, mean (SD) Pulse oxymetry, median [IQR] Temperature, mean (SD) Tests in ED ECG Arterial blood gas Chest x-ray CT pulmonary angiogram V/Q scan D-dimer N500 ng/L Confirmed PE Hospital admission
1070
100%
510 89 273 198 35.3 80 130 80 99 36.8
48% 8% 26% 19% (9) (12) (24) (17) [98;100] (0.9)
394 456 829 143 11 167 5 94
37% 43% 77% 13% 1% 16% 0.5% 9%
ECG, electrocardiogram; CT: computed tomography.
Amongst these 4301 patients, 442 were excluded as they were not tested to rule out PE. Of the 3859 patients with suspicion of PE who had D-dimer testing, 1070 (27%) had a PERC score of zero (Fig.). Main characteristics of the studied sample are summarized in Table 2. Mean age of the patients analyzed was 35 years (SD, 9), and 496 (46%) were men. One hundred sixty-seven patients (16%) had a D-dimer value above normal range, ie, N500 ng/mL. Amongst these patients, median D-dimer value was 725 ng/L [IQR 659; 887]. Subsequently, 153 (14%) invasive studies were performed including 143 CTPAs and 11 V/Q scans (one patient had both). Six patients had a positive CTPA: one was considered as “inconclusive” and subsequently PE was ruled out after negative V/Q scan. In total, 5 patients were diagnosed with a PE in this population (Cohen κ 1.0), a rate of 0.5% (95% CI 0.1%-1.1%). In our sample of very low-risk patients, D-dimer testing had a positive predictive value of 3%. Amongst the five patients diagnosed with PE who had D-dimer testing, 2 of them were subsegmental; 2, segmental; and the fifth one, lobar. D-dimer testing in PERC-negative patients led to the completion of invasive studies for 153 (14%) of patients; 15 of them (1% in total) experienced adverse events. All adverse events were non severe, including two allergic reactions to iodine contrast agent (generalized urticaria), and 13 patients were started on an anticoagulation regimen and underwent unnecessary admission (all of them while waiting for imaging studies that were not rapidly available). No bleeding was reported, and no severe adverse events resulted from this management. 4. Discussion
Fig. Flowchart. PERC: Pulmonary Embolism Rule out Criteria. RGS: Revised Geneva Score. PE: Pulmonary Embolism.
In this study, we undertook a retrospective analysis of patients with a negative PERC score and low pre-test probability of PE, who have had D-dimer testing in the ED. The objectives were to assess the yield and adverse events subsequent to further testing in this very low-risk population, rather than to determine the exact rate of PE. We found that 0.5% of PERC negative patients were subsequently diagnosed with a PE after D-dimer testing. Although it is not the real prevalence of PE in this population, this very low rate represents the percentage of patients who are diagnosed with PE when PERC rule is not applied. First validated in 2004 by Kline et al [7], PERC score has a very high sensitivity and negative predictive value for the diagnosis of PE [9,10].
612
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613
However, to rely solely on PERC score, even associated with a low pretest probability, remains controversial. As raised by Schwartz et al [21], there is still no intervention trial in which PERC score is used for clinical decision making; hence, reliability and clinical application of this tool still need confirmation. Rationale for its use appears weaker in Europe where prevalence of PE has been reported to be higher than in the United States, and where a large trial has questioned the reliability of PERC rule [8]. In a recent analysis, Self and Barett simulated the impact of Ddimer testing in PERC-negative patients, based on previous theoretical diagnostic performances [13]. They assessed the risk/benefit ratio of D-dimer testing, derived from reported values of diagnostic performances of PERC score, D-dimer, and CTPA. They calculated that in a sample of 1,000 PERC negative patients, 30 will have a PE; and that Ddimer testing strategy would lead to the completion of 621 CTPAs, 6 PEs still being missed, 24 being correctly diagnosed, and 24 treated in the absence of a PE. The benefit in terms of outcome for these low risk PEs are unclear, and their conclusion reached the one from Newman et al. [22] who suggested a higher rate of death resulting from a liberal strategy of D-dimer testing and thus anticoagulation-related complications and consequences of invasive imaging. Our results differ slightly from these as we did not test all PERC negative patients with D-dimer. Due to the design of this study, only a fraction of them were tested, hence the different results and possible overestimation of the rate of PE in this population. The strength of our study lies in its retrospective design, which allowed us to describe what changes would occur with a PERC based strategy. Indeed, even in the absence of a PERC based decision rule (as it was the case in our centers), some patients were still not tested for PE with a D-dimer, and our study focuses only on those who could have been weaned from this test if the PERC rule was used. We report a rate of PE of 0.5% (95% CI 0.1-1.1%). This rate is similar to that reported in the general population [19,23]. Previous studies have highlighted the high rate (13%-23%) of invasive studies that could have been avoided with a PERC-based decision making policy [11,24]. In our sample, 14% of patients underwent invasive imaging studies. The five patients with a PE were all diagnosed with low-risk PE according to the recent classification [25], with an estimated mortality rate approaching 1% [26-28]. Furthermore, Kline et al [12] found in their retrospective studies that mortality of PERC-negative patients with a PE was 0% at 30 days in their sample. The benefits of these supplemental investigations are unclear, and several studies report that PE tends to be overdiagnosed, with unclear added value. More frequent use of CTPA led to increased diagnosis of less fatal PE [3,4]. PERC score has shown excellent sensitivity and negative predictive value[7,9,10,29]. However controversy remains on whether it should be applied alone in patients with a low pre-test probability [8]. Its very low positive predictive value, the high rate of invasive studies, and the unknown benefits in terms of outcome in those with very low-risk PE advocate against the use of D-dimer testing in PERCnegative patients. 5. Limitations The main limitation to this study was the absence of follow-up due to its retrospective design. Sensitivity, specificity and negative predictive values could not have been calculated as we do not know how many false-negative patients were discharged. However, the objective of our study was to determine the real clinical added value of D-dimer testing in those PERC-negative patients, rather than to assess the performances of this clinical score. Indeed two recent metaanalyses provide these data, with very high sensitivity and negative predictive value, and acceptable specificity [9,10]. Another limitation is that we electronically retrieved all patients with D-dimer testing in our EDs for screening thereafter excluding those with no clear
suspicion of PE hence creating an inclusion bias. We cannot guarantee that all patients with a suspected PE were included for analysis. However, we believe that this bias is very limited, as we stated above, since guidelines and local policies recommend the use of D-dimer for all patients suspected for PE with low pre-test probability. Lastly, in our sample PERC score was retrospectively calculated after chart abstraction. Even though we discarded any incomplete chart with missing data regarding PERC score, we cannot be sure that some of our PERC-negative patients did not in fact have one positive item that was not clearly reported on the medical chart. In summary, the analysis of our cohort reports that D-dimer testing for PERC-negative patients with low pretest probability led to nearly 15% of irradiative imaging studies, for a rate of newly diagnosed PED of 0.5%. This very low rate if close to that of the general population, suggesting that further testing for PERC negative patients should be abandoned. Intervention studies are warranted to validate this hypothesis.
References [1] McCaig LF, Nawar EW. National hospital ambulatory medical care survey: 2004 emergency department summary. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics 2006. http://198.246.124.29/nchs/data/ad/ad372.pdf (accessed 26 Jul2013). [2] Smith-Bindman R, Miglioretti DL, Johnson E, et al. Use of diagnostic imaging studies and associated radiation exposure for patients enrolled in large integrated health care systems, 1996-2010. JAMA J Am Med Assoc 2012;307:2400–9. http:// dx.doi.org/10.1001/jama.2012.5960. [3] Schissler AJ, Rozenshtein A, Kulon ME, et al. CT pulmonary Angiography: increasingly diagnosing less severe pulmonary emboli. PLoS ONE 2013;8: e65669. http://dx.doi.org/10.1371/journal.pone.0065669. [4] Sheh SH, Bellin E, Freeman KD, et al. Pulmonary embolism diagnosis and mortality with pulmonary CT angiography versus ventilation-perfusion scintigraphy: evidence of overdiagnosis with CT? AJR Am J Roentgenol 2012;198:1340–5. http://dx.doi.org/10.2214/AJR.11.6426. [5] 5 Wiener RS, Schwartz LM, Woloshin S. When a test is too good: how CT pulmonary angiograms find pulmonary emboli that do not need to be found. BMJ 2013;347:f3368–f3368. doi:10.1136/bmj.f3368. [6] Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med 2011;171:831–7. http://dx.doi.org/10.1001/archinternmed.2011.178. [7] Kline JA, Mitchell AM, Kabrhel C, et al. Clinical criteria to prevent unnecessary diagnostic testing in emergency department patients with suspected pulmonary embolism. J Thromb Haemost JTH 2004;2:1247–55. http://dx.doi.org/10.1111/ j.1538-7836.2004.00790.x. [8] Hugli O, Righini M, Le Gal G, et al. The pulmonary embolism rule-out criteria (PERC) rule does not safely exclude pulmonary embolism. J Thromb Haemost JTH 2011;9:300–4. http://dx.doi.org/10.1111/j.1538-7836.2010.04147.x. [9] Singh B, Parsaik AK, Agarwal D, et al. Diagnostic accuracy of pulmonary embolism rule-out criteria: a systematic review and meta-analysis. Ann Emerg Med 2012;59: 517–520.e1–4. http://dx.doi.org/10.1016/j.annemergmed.2011.10.022. [10] Singh B, Mommer SK, Erwin PJ, et al. Pulmonary embolism rule-out criteria (PERC) in pulmonary embolism–revisited: A systematic review and meta-analysis. Emerg Med J EMJ Published Online First: 20 October 2012. http://dx.doi.org/10.1136/ emermed-2012-201730. [11] Dachs RJ, Kulkarni D, Higgins 3rd GL. The Pulmonary Embolism Rule-Out Criteria rule in a community hospital ED: a retrospective study of its potential utility. Am J Emerg Med 2011;29:1023–7. http://dx.doi.org/10.1016/j.ajem.2010.05.018. [12] Kline JA, Slattery D, O’Neil BJ, et al. Clinical features of patients with pulmonary embolism and a negative PERC rule result. Ann Emerg Med 2013;61:122–4. http:// dx.doi.org/10.1016/j.annemergmed.2012.06.494. [13] Self WH, Barrett TW. Is ‘PERC negative’ adequate to rule out pulmonary embolism in the emergency department? Evaluating meta-analysis for studies of clinical prediction models: answers to the July 2012 journal club questions. Ann Emerg Med 2012;60:803–14. http://dx.doi.org/10.1016/j.annemergmed.2012.07.121. [14] Torbicki A, Perrier A, Konstantinides S, et al. Guidelines on the diagnosis and management of acute pulmonary embolism: the Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC). Eur Heart J 2008;29:2276–315. http://dx.doi.org/10.1093/eurheartj/ ehn310. [15] Roy P-M, Bichri A, Bouet R, et al. Embolie pulmonaire: algorithmes diagnostiques. 2007.http://www.sfar.org/acta/dossier/archives/mu07/html/mu07_07/urg07_07. htm (accessed 23 Dec 2013). [16] Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006;144: 165–71. [17] Klok FA, Kruisman E, Spaan J, et al. Comparison of the revised Geneva score with the Wells rule for assessing clinical probability of pulmonary embolism. J Thromb Haemost JTH 2008;6:40–4. http://dx.doi.org/10.1111/j.1538-7836.2007.02820.x.
J. Bokobza et al. / American Journal of Emergency Medicine 32 (2014) 609–613 [18] Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27:305–8. [19] Hansson PO, Welin L, Tibblin G, et al. Deep vein thrombosis and pulmonary embolism in the general population. ‘The Study of Men Born in 1913’. Arch Intern Med 1997;157:1665–70. [20] Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934;26:404–13. http://dx.doi.org/10.1093/biomet/26.4.404. [21] Precautions Schwartz DT, With PERC. Ann Emerg Med 2013;62:197. http://dx.doi. org/10.1016/j.annemergmed.2013.02.030. [22] Newman DH, Schriger DL. Rethinking testing for pulmonary embolism: less is more. Ann Emerg Med 2011;57:622–627.e3. http://dx.doi.org/10.1016/j. annemergmed.2011.04.014. [23] White RH. The Epidemiology of Venous Thromboembolism. Circulation 2003;107: I–4–I–8. http://dx.doi.org/10.1161/01.CIR.0000078468.11849.66. [24] 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 Off J Soc Acad Emerg Med 2012;19:1219–26. http://dx.doi.org/10.1111/acem.12012. [25] Jaff MR, McMurtry MS, Archer SL, et al. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis, and chronic thrombo-
[26]
[27]
[28]
[29]
613
embolic pulmonary hypertension: a scientific statement from the American Heart Association. Circulation 2011;123:1788–830. http://dx.doi.org/10.1161/ CIR.0b013e318214914f. Bova C, Pesavento R, Marchiori A, et al. Risk stratification and outcomes in hemodynamically stable patients with acute pulmonary embolism: a prospective, multicentre, cohort study with three months of follow-up. J Thromb Haemost JTH 2009;7:938–44. http://dx.doi.org/10.1111/j.1538-7836.2009.03345.x. Palmieri V, Gallotta G, Rendina D, et al. Troponin I and right ventricular dysfunction for risk assessment in patients with nonmassive pulmonary embolism in the Emergency Department in combination with clinically based risk score. Intern Emerg Med 2008;3:131–8. http://dx.doi.org/10.1007/s11739008-0134-2. Post F, Mertens D, Sinning C, et al. Decision for aggressive therapy in acute pulmonary embolism: implication of elevated troponin T. Clin Res Cardiol Off J Ger Card Soc 2009;98:401–8. http://dx.doi.org/10.1007/s00392-0090017-1. Hogg K, Dawson D, Kline J. Application of pulmonary embolism rule-out criteria to the UK Manchester Investigation of Pulmonary Embolism Diagnosis (MIOPED) study cohort. J Thromb Haemost JTH 2005;3:592–3. http://dx.doi.org/10.1111/ j.1538-7836.2005.01212.x.
