Resident-performed Compression Ultrasonography for the Detection ...

ACAD EMERG MED

d

March 2004, Vol. 11, No. 3

d

319

www.aemj.org

Resident-performed Compression Ultrasonography for the Detection of Proximal Deep Vein Thrombosis: Fast and Accurate Timothy Jang, MD, Martin Docherty, MD, Chandra Aubin, MD, Greg Polites, MD Abstract Objectives: To assess whether emergency medicine residents (EMRs) could quickly perform accurate compression ultrasonography (CUS) for the detection of proximal lower extremity deep vein thromboses (PLEDVTs) with minimal training. Methods: A prospective, observational study using a convenience sample of patients presenting with signs and/or symptoms for PLEDVT. Vascular laboratory and department of radiology studies were considered the criterion standard. CUS of the femoral vessels was performed. Incompressibility or visualized thrombus was considered ‘‘positive.’’ Results: Eight residents with limited ultrasound (US) experience and no prior experience with

deep vein thrombosis (DVT) US volunteered to participate in this study, enrolling 72 patients. Their average scan time was 11.7 minutes (95% CI ¼ 9.4 to 14). There were 23 true positives, 4 false positives, 45 true negatives, and 0 false negatives. The test characteristics for PLEDVT gave a sensitivity of 100% (95% CI ¼ 82.2 to 100) and a specificity of 91.8% (95% CI ¼ 79.5 to 97.4). Conclusion: Emergency medicine residents with limited US experience were able to quickly perform CUS after minimal training for the detection of PLEDVT in a select group of patients. Keywords: ultrasound; deep vein thrombosis; resident. ACADEMIC EMERGENCY MEDICINE 2004; 11:319–322.

Deep vein thrombosis (DVT) is a common disorder with fatal complications.1 Unfortunately, the symptoms of DVT are nonspecific and clinical examination is neither sensitive nor specific,2,3 thus necessitating accurate diagnostic imaging. Contrast venography is the criterion standard for diagnosing DVT, but it is invasive and has been replaced by venous duplex ultrasonography (VDUS) as the initial diagnostic test.4–6 Recently, some centers have begun using computed tomography (CT) venography, with comparable results.7–9 Although most emergency departments (EDs) are able to obtain VDUS during business hours, this often requires patients’ being gone from the ED for several hours.10 Furthermore, some EDs do not have access to such studies ‘‘around the clock,’’ leading to either admission for subsequent testing or presumptive anticoagulation and outpatient follow-up.11 Unfortunately, these two management strategies are associated with increased cost and/or risk for bleeding complications. The purpose of this study was to prospectively assess whether emergency medicine residents (EMRs)

with minimum training could quickly perform accurate compression ultrasonography (CUS) for the detection of proximal lower extremity DVTs (PLEDVTs).

From the Division of Emergency Medicine, Washington University School of Medicine, St. Louis, MO (TJ, MD, CA, GP). Received May 22, 2003; revision received September 9, 2003; accepted September 11, 2003. Address for correspondence and reprints: Timothy Jang, MD, Division of Emergency Medicine, Barnes-Jewish Hospital, Campus Box 8072, St. Louis, MO 63110; e-mail: [email protected]. doi:10.1197/j.aem.2003.09.020

METHODS Study Design. This was a prospective study involving a sample of patients presenting with signs and/or symptoms of PLEDVT from September 2000 to September 2001, when one of eight study residents was working in the ED. This study was approved by the institutional review board at our institution. Study Setting and Population. This study was done at an urban, academic Level 1 ED with a postgraduateyear 1 to 4 (PGY1–4) residency with 48 residents and an annual volume of 80,000 patient visits. The US education program was started in April 2000, consisting of 1) an annual two-hour lecture and demonstration on US physics and knobology, basic focused abdominal sonography for trauma (FAST) technique, and introduction to the transabdominal pregnancy examination; 2) ongoing quality assurance (QA) review by one of two American Registry of Diagnostic Medical Sonographers (RDMS)-certified emergency physician (EP)-sonographers; and 3) an optional US elective. In July 2001, annual US lectures on aortic, pericardial, and renal examinations and vascular access were added to the educational conference curriculum and, in July 2002, a required two-week US rotation was added to the PGY3 curriculum,

320 which was subsequently moved to the PGY2 curriculum in July 2003. Patients were required to have lower-extremity swelling or pain, or tenderness to palpation of Hunter’s canal to be eligible for enrollment. The only exclusion criteria were a prior history of DVT and known results of recent diagnostic studies for DVT. Eight EMRs (four PGY1, two PGY2, and two PGY3) without prior experience performing CUS for PLEDVT volunteered to participate following a onehour lecture given as part of the residency’s weekly conference that included a PowerPoint (Microsoft Corp., Redmond, WA) presentation and demonstration of CUS for PLEDVT on two normal volunteers. Their prior US experience consisted only of an average of four FAST examinations. Study Protocol. Patients were identified by their treating EPs but were enrolled by the EMR performing the CUS. The EMR-performed CUS was done prior to criterion standard evaluation without the supervision of a senior resident or attending physician. Treating clinicians were blinded to the results of the EMR-performed CUS. Subsequent one-year phone follow-up and chart review using the hospital’s computer database and vascular laboratory records were done on all patients by one of two investigators (TJ, GP) who were not blinded to the results of the initial EMR-performed CUS examination. Telephone follow-up was attempted up to three times on each patient using a standardized questionnaire to ask whether the patient was diagnosed as having a ‘‘blood clot, DVT, or pulmonary embolism (PE)’’ or started on heparin, enoxaparin/Lovenox, or warfarin/Coumadin within the last year. Measurements. The CUS was done to assess for PLEDVT (Table 1) using an Aloka SSD-1400 (Aloka, Tokyo, Japan) with a 7.5-MHz linear array probe. Standardized data sheets were completed by the EMR performing CUS at the time of examination completion and compared with subsequent criterion standard evaluations. Variables collected on the data sheets included 1) EMR name, 2) date and time of EMR-performed CUS, 3) approximate time required for CUS examination, 4) presence of lower-extremity swelling or pain or tenderness to palpation of Hunter’s canal, 5) ED scan result (positive or negative), and 6) date/time/result of criterion standard evaluation. The EMR performing the examination was responsible for recording items 1–5 and one of two research assistants obtained item 6 from files kept by the department of radiology or vascular laboratory. Estimations of EMR-performed CUS examination time were done by resident self-report. The results of contrast venography were considered criterion standard when available; otherwise, VDUS or CT venography was used. Equivocal examinations

Jang et al.

d

RESIDENT US FOR DVT

TABLE 1. Steps in Resident-performed Compression Ultrasonography for Proximal Lower Extremity Deep Vein Thrombosis (DVT) 1. The common femoral artery and vein (CFA, CFV, respectively) were identified on the side concerning for DVT. If a thrombus was visualized, the examination was considered ‘‘positive’’ and no further images were obtained. 2. If a clot was not seen, gentle pressure was applied to the probe until the CFV was fully compressed. If the CFV was not fully compressible, then the examination was also considered ‘‘positive’’ and no further images were obtained. Adequate compressions were determined on the basis of the criteria of Frazee et al.13 where pressure was applied until the adjacent artery began to be distorted. 3. If the CFV was compressible, then the operator proceeded to move down the leg following the vessels and performing gentle compressions under direct visualization approximately every 2 cm. If a thrombus was encountered, this was noted and considered ‘‘positive’’ and no further images were obtained. Likewise, if the vein became incompressible at any point, this was also noted and considered ‘‘positive’’ without further images being obtained. 4. If the vein continued to be compressible without any thrombus visualized, the vessels were followed down to the popliteal fossa where the popliteal artery and popliteal vein (PV) were evaluated as in step 1. Thermographic images were obtained of the CFA/CFV and PV/PA—one uncompressed and one compressed at each site—for all negative examinations. If the examination was positive, then images were taken of the CFA/CFV and site of positive findings only.

done by EMRs were considered ‘‘false negatives’’ in patients diagnosed as having a PLEDVT and ‘‘false positives’’ in patients diagnosed as not having PLEDVT by criterion standard evaluations. Data Analysis. Data were kept on a Microsoft Excel 97 spreadsheet (Microsoft Corp., Redmond, WA) and analyzed with VassarStats (Poughkeepsie, NY).

RESULTS A total of 72 patients were enrolled. It is unknown how many patients were offered enrollment but declined. Data sheets were completed on all 72 patients. The only data missing was the time but not date of vascular laboratory studies in 24 patients. The average patient age was 54 years. Patient characteristics are shown in Table 2. In all, 23 patients (32%) had PLEDVT, all of whom were correctly identified by EMRs (for DVT location; see Table 3). There were three equivocal EMR studies, occurring in patients without DVT, and, therefore, considered ‘‘false positives.’’ There was one other ‘‘false-positive’’ study. Thus, the sensitivity and specificity of EMRperformed CUS for PLEDVT were 100% (95% CI ¼ 82.2 to 100) and 91.8% (95% CI ¼ 79.5 to 97.4), respectively. The average scan time by self-report was 11.7 minutes (95% CI ¼ 9.4 to 14; median 10 minutes). Criterion standard evaluations included three contrast

ACAD EMERG MED

d

March 2004, Vol. 11, No. 3

d

TABLE 2. Patient Characteristics Sign/Symptom

PLEDVT (n = 23)

No PLEDVT (n = 49)

14 19 20

26 35 34

Unilateral swelling Unilateral pain TTP of HC

PLEDVT = proximal lower extremity deep venous thrombosis; TTP = tenderness to palpation; HC = Hunter’s canal. Demographics

PLEDVT (n = 23)

Average age (yr) Female (%) White (%) African American (%) Asthma/COPD (%) Cancer (%) Diabetes (%) Heart disease (%) History of cancer (%) History of stroke (%) Hypertension (%) Psychiatric illness (%) Smoker (%) Surgery within 1 mo (%)

56 15 (65) 15 (65) 8 (35) 2 (9) 7 (30) 7 (30) 3 (13) 1 (4) 1 (4) 9 (39) 0 (0) 7 (30) 3 (13)

No PLEDVT (n = 49) 54 33 18 31 12 4 9 15 0 6 20 5 12 1

(67) (37) (63) (24) (8) (18) (31) (0) (12) (41) (10) (24) (2)

PLEDVT = proximal lower extremity deep venous thrombosis; COPD = chronic obstructive pulmonary disease.

venograms performed following equivocal VDUS by the vascular laboratory, 58 unequivocal VDUS examinations, and three CT venograms. The three patients who underwent CT venograms were originally scheduled for VDUS, but they developed chest pain while in the ED and underwent PE-protocol CTs with CT venograms. Because DVTs were confirmed in all three cases, subsequent VDUS was not obtained. Eight patients did not undergo subsequent imaging due to confirmation of diagnostic alternatives (e.g., septic arthritis). Chart review was done on all 72 patients, but only 63 patients were reached for phone followup. The final diagnoses of patients without DVT as determined by chart review and phone follow-ups are shown in Table 4.

DISCUSSION 12

321

www.aemj.org

In 1997, Jolly et al. did a retrospective review of 15 patients evaluated with full-length VDUS by two well-trained EPs with 30 prior examinations for DVT. They found a sensitivity and specificity of 100% and 75% for lower-extremity DVT with an average scan time of more than 30 minutes. From 1997 to 1998, Frazee et al.13 evaluated 76 patients with CUS of the common femoral vein (CFV) and popliteal (PV) only, but not the intervening vasculature. Six EPs with ‘‘extensive experience’’ and hospital privileges for ED US13 had a sensitivity and specificity of 89% and 76%, respectively, for lower-extremity DVT. From 1998 to 1999, Blaivas et al.14 evaluated 112 patients with color VDUS performed by five highly trained attending EPs, three of whom were RDMS-certified and all of whom had performed more than 350 prior US

TABLE 3. Distribution of Proximal Lower Extremity Deep Vein Thrombosis (PLEDVT) PLEDVT Location

n (of 23 total)

CFV SPV PV SFV and PV CFV, SFV, and PV

8 1 2 6 6

CFV = common femoral vein; SFV = superficial femoral vein; PV = popliteal vein.

examinations. Evaluating only the CFV and PV, they had a kappa of 0.9 compared with vascular laboratory results, but no long-term follow-up was provided. To the best of our knowledge, this is the only study evaluating resident-performed CUS for the detection of PLEDVT. The EMRs in this study had less training and experience than the EPs in previous studies12–14 and did not have color VDUS available to them. Our study involved CUS of the entire proximal leg, which may explain the improved sensitivity and specificity compared with those obtained previously with more limited examinations.13 It is likely that color VDUS would have improved our specificity because it facilitates identification of blood vessels in difficult examinations.14 Even though prior studies13,14 did not involve long-term follow-up, a one-year follow-up period is considered necessary to rule out venous thromboembolic disease and was used in our study, consistent with the PIOPED study.15 Finally, this study demonstrates that EMRs with limited training can quickly perform accurate CUS for PLEDVT in selected patients without removing them from the clinical area. This finding has obvious implications for improving patient throughput and ED overcrowding, while avoiding the cost of calling in ‘‘after-hours’’ technicians. In our sample, three equivocal studies occurred in patients without a PLEDVT. In all three cases, the examination was terminated due to patient pain with compression. However, in every case, ‘‘adequate compression pressure’’ had not been applied (defined by Frazee et al.13). It is likely that the use of adequate pressure would have yielded negative examinations. TABLE 4. Final Diagnoses of Patients without Lower Extremity Deep Vein Thrombosis Diagnosis

n (of 49 total)

Nonspecific leg pain Cellulitis/septic arthritis Baker’s cyst ACS with leg swelling PNA with leg swelling ACP with leg pain Distal peroneal vein thrombosis NF with leg swelling

17 8 7 5 5 4 2 1

ACS = acute coronary syndrome; ACP = atypical chest pain; PNA = pneumonia; NF = neutropenic fever.

322

Jang et al.

Furthermore, our only ‘‘false positive’’ occurred in a patient with venous congestion and decreased compressibility but no visualized thrombus on evaluation by the vascular laboratory. At one-year followup, including a repeat vascular laboratory study, the patient’s venous congestion had resolved, and there was no evidence of a thrombus.

LIMITATIONS This study has several limitations. First, a consistent criterion standard was not used. The sensitivity, although not the specificity, of CT venography has been questioned,16 but this is not likely to have adversely affected our assessment because all three cases demonstrated the presence of DVT. Likewise, although contrast venography is considered the criterion standard for DVT, VDUS has become the ‘‘standard of care.’’4–6,13,14 Second, a study of consecutive patients would have eliminated possible selection bias, although the two previous prospective studies were also limited by convenience samples. It may be that those patients whom an EP-sonographer feels comfortable examining with US can be adequately assessed. This was shown in a previous study relating to ED US.17 Finally, only eight of 48 eligible residents (17%) participated in this study. It is unclear why other residents did not participate. This may represent a ‘‘US interest’’ bias. This study suggests several areas for further research. First, is there a learning curve for EPs performing CUS for DVT? This would have implications for credentialing and training of physicians, but it has never been addressed in the literature. It is unknown how many US examinations are required to gain ‘‘proficiency’’ in this application of ED US. Second, is there any advantage to color VDUS over CUS in assessing for DVT? It may be that the increased complexity of interpreting color VDUS would make it more difficult for novice EPs performing US to evaluate for DVT. Finally, would EP-performed CUS alter department length of stay, time to treatment/disposition, or outcome in patients with suspected DVT? We would expect the use of bedside US to improve these parameters, but this has yet to be established.

CONCLUSIONS Emergency medicine residents with limited US experience were able to quickly perform CUS after minimum training for the detection of PLEDVT in a select group of patients.

d

RESIDENT US FOR DVT

References 1. Hirsh J, Hoak J. Management of deep vein thrombosis and pulmonary embolism. Circulation. 1996; 93:2212–45. 2. Wells PS, Hirsh J, Anderson DR, et al. Accuracy of clinical assessment of deep-vein thrombosis. Lancet. 1995; 345: 1326–30. 3. Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis. Fallibility of clinical symptoms and signs. Arch Surg. 1976; 111:34–6. 4. Fard MN, Mostaan M, Anaraki M. Utility of lower-extremity duplex sonography in patients with venous thromboembolism. J Clin Ultrasound. 2001; 29:92–8. 5. Garcia ND, Morasch MD, Ebaugh JL, et al. Is bilateral ultrasound scanning of the legs necessary for patients with unilateral symptoms of deep vein thrombosis? J Vasc Surg. 2001; 34:792–7. 6. White RH, McGahan JP, Daschbach MM, Hartling RP. Diagnosis of deep vein thrombosis using duplex ultrasound. Ann Intern Med. 1989; 111:297–304. 7. Yoshida S, Akiba H, Tamakawa M, Yama N, Takeda M, Hareyama M. Spiral CT venography of the lower extremities by injection via an arm vein in patients with leg swelling. Br J Radiol. 2001; 74:1013–6. 8. Loud PA, Katz DS, Klippenstein DL, Shah RD, Grossman ZD. Combined CT venography and pulmonary angiography in suspected venous thromboembolic disease: diagnostic accuracy for deep venous evaluation. AJR. 2000; 174:61–5. 9. Loud PA, Katz DS, Bruce DA, Klippenstein DL, Grossman ZD. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology. 2001; 219:498–502. 10. Poppitti R, Papanicolaou G, Perese S, Weaver FA. Limited B-mode venous imaging versus complete color-flow duplex venous scanning for detection of proximal deep venous thrombosis. J Vasc Surg. 1995; 22:553–7. 11. The Columbus Investigators. Low-molecular weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med. 1997; 337:657–62. 12. Jolly BT, Massarin E, Pigman EC. Color Doppler ultrasonography by emergency physicians for the diagnosis of acute deep venous thrombosis. Acad Emerg Med. 1997; 4:129–32. 13. Frazee BW, Snoey ER, Levitt A. Emergency department compression ultrasound to diagnose proximal deep vein thrombosis. J Emerg Med. 2001; 20:107–11. 14. Blaivas M, Lambert MJ, Harwood RA, Wood JP, Konicki J. Lower-extremity Doppler for deep venous thrombosis—can emergency physicians be accurate and fast? Acad Emerg Med. 2000; 7:120–6. 15. PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA. 1990; 20:2753–9. 16. Peterson DA, Kazerooni EA, Wakefield TW, et al. Computed tomographic venography is specific but not sensitive for diagnosis of acute lower-extremity deep venous thrombosis in patients with suspected pulmonary embolus. J Vasc Surg. 2001; 34:798–804. 17. Davis DP, Campbell C, Wang J, Poste J. The influence of operator confidence on the accuracy of emergency department ultrasound [abstract]. Acad Emerg Med. 2003; 10:486–7.