Thromboelastography for the Prediction of Bleeding After Transplant Renal Biopsy1

Connie L. Davis2 and Wayne L. Chandler

Key Words: Renal transplant, ,

CL. Davis, Department of Medicine, University of Washington Medical Center, Seattle, WA B leeding after renal biopsy has been linked to W.L. Chandler, Department of Laboratory Medicine. biopsy technique, hypertension, renal pathol- University of Washington Medical Center, Seattle. WA ogy. and coagulation abnormalities (1-9). Hemostatic screening before biopsy is intended to detect those (J, Am. Soc. Nephrol. 1995; 6:1250-1255) who have significant abnormalities before the proce- dure so appropriate therapy can be given to prevent bleeding. Prebiopsy evaluation designed to limit the ABSTRACT risk of bleeding has often included a The ability of prebiopsy coagulation assays to predict (PT). partial thromboplastin time (PTF), count, mild postbiopsy bleeding was evaluated in renal and bleeding time (BT) (6.8,10). Recent studies, how- transplant patients undergoing renal allograft biopsy ever, have questioned the utility of the BT; It is inva- (N = 120). The coagulation assays studied included sive and subject to performance variability, and the the bleeding time, prothrombln time, partial thrombo- predictive value for bleeding in individual patients at plastin time, platelet count, and thromboelastograph sites other than the skin has been poor (11,12). In an (TEG). Coagulation results were defined as abnormal effort to determine if the prebiopsy evaluation could be if they fell outside the established normal reference safely simplified, we compared the thromboelasto- range. Bleeding was defined as a drop In the hemat- graph (TEG) with more common hemostatic tests as ocrit equal to or more than 4 points 6 h after the predictors of bleeding after transplant renal biopsy. The TEG was chosen for examination in this study procedure or ultrasound evidence of a new perirenal because it produces an overall evaluation of the coag- hematoma. Overall, 21% of patients showed evi- ulation, platelet. and fibrinolytic systems In one as- dence of mild bleeding. Of those who bled, 78% had say. The TEG measures the strength and shear elas- normal results on all coagulation tests, indicating that ticity of a clot as It forms, matures, and lyses. The TEG most mild bleeding was not associated with coagu- has proved to be more useful than the BT, PT, and PTT lation abnormalities. Of the assays tested, only abnor- for predicting bleeding after cardiac bypass mal TEG:angle (P < 0.01) and TEG:k (P < 0.04) values and liver transplantation (13-16). In addition, it is were associated with an increased risk of bleeding. easy to interpret. The output from the TEG is a graph Bleeding times were not predictive of an Increased (Figure 1). The width of the graph at any point is the risk of postblopsy bleeding; five patients had abnor- clot strength or elasticity. Five graph parameters are measured when the TEG is complete. The “r” value is mal bleeding times ranging from 10 to 20 mm of a measure of the time It takes for clotting to begin; it is whom only one bled. All prothrombin time, partial similar to a whole- clotting time. If the “r” value is thromboplastin time, and platelet count abnormali- prolonged, It indicates a possible coagulation defi- ties were mild (e.g., no prothrombln times longer than ciency that could be further investigated with PT, PTT, 15 s, no platelet counts below 129,000/.tL); none of and factor assays. The “k” value and angle (a#{176})indicate these assays predicted postbiopsy bleeding. Other how fast clot strength is increasing once clotting clinical characteristics, including patient age, sex, starts. Reduced “k” or angle may be due to any serum creatinine, blood pressure (if less than 160/90 combination of coagulation deficiencies, reduced mm Hg), number of biopsy passes, or renal pathol- platelet function, and reduced platelet count. The ogy, did not appear to influence bleeding after maximum amplitude (MA) or maximum width of the biopsy. It was concluded that most bleeding after graph is an indication of the maximum attaInable clot transplant renal biopsy was not associated with co- strength. Decreases In MA are associated with re- duced platelet function or number and with reduced agulation abnormalities and that the lEG was the best fibrinogen levels. The amplitude after 60 minutes (A60) assay for detecting mild coagulation abnormalities is a measure of fibrinolysis; if the A60 is significantly associated with an increased risk of bleeding. less than the MA, it indicates that friulvo clot lysis may be occurring.

Received October 13, 1994. Accepted March 13, 1995. Simple inspection of the graph provides the majority 2 Correspondence to Dr. CL. Davis, Division of Transplantation, RC-43, University of the information. A normal graph similar to that of Washington Medical Center. Seattle, WA 98195. shown in Figure 1 shows a rapid onset of clotting, with 1046.6673/0604-1 250$03.00/0 a sufficient width and no evidence of lysis, indicating Journal of the American Society of Nephroiogy Copyright © 1995 by the American Society of Nephrology no abnormalities of the coagulation, platelet, or fi-

1250 Volume 6 - Number 4 ‘ 1995 Davis and Chandler

NORMAL ABNORMAL Thromboelastograph Thmmboelastograph

Prolonged r value

Reduced MA

Reduced A60 Fibrinolysis

Figure 1. TEG tracing. The width or amplitude of the tracing at Figure 2. An abnormal TEG tracing. Prolonged “r” or ‘k,” any point is proportional to the clot shear modulus (elastic reduced angle, MA, or A all indicate possible coagulation, strength). Five parameters are measured on the curve: (1) platelet, or fibrinolytic abnormalities. The lEG Is a visual test; the r value is the time to the start of clot onset, similar to a often, a simple glance at the curve is sufficient to determine whole-blood clotting time; (2) the k value is the time from the whether the patient Isnormal or requires further workup. start of clotting to an amplitude of 20 mm; (3) the angle (a”) ismeasured from the center lineto a linedrawn from the start of clotting tangential to the clotting curve. Both the “k” and patient treated with fresh frozen plasma before biopsy was the angle provide an indication of the rate of clot formation; eliminated from the study. None of the patients were being (4) the MA Is a measure of maximum clot strength; and (5) treated with heparin or nonsteroidal anti-inflammatory med- the A is compared with the MA; if significantly reduced, it ications; however, 98 of 120 patIents were receiving aspirin, indicates that fibrinolysis is occurring. This diagram shows a 325 mg once per week. normal TEG with a rapid clot onset (short “r”), rapid clot formation (short ‘k” and large angle), normal amplitude, and no change in A. Biopsy Procedure

Patients were not biopsied ifthey had a blood pressure of brinolytic systems. An abnormal graph (Figure 2) with more than 160/100mm Hg. All biopsies were performed with a prolonged onset time, a narrow width, or a reduction a 16-gauge spring-loaded Temno (Bauer; Clearwater. FL) in width at the end indicates abnormal clotting and needle under real-time ultrasound guidance with local anes- suggests the need of further workup. In thesia and without the administration of intravenous fluids. order to determine which coagulation measurements Patients undergoing combined kidney and pancreas allograft and intrabiopsy factors predict bleeding after renal biopsy received less than 500 mL of Intravenous normal biopsy, a prospective study was undertaken compar- saline during the procedure. Patients were observed for 6 h after the procedure. After ing standard coagulation tests, TEG, biopsy tech- the biopsy, the blood pressure was kept at less than 160/90 nique, and renal pathology with the incidence of mm Hg. The was checked 6 h after the biopsy. bleeding in patients undergoing renal biopsy on trans- planted kidneys. Bleeding Endpoint METHODS A drop in hematocrit equal to or more than 4 points 6 h Patient Population after biopsy or ultrasound evidence of a new perirenal hema-

All patients undergoing transplant (N = 120) renal biopsy toma was considered indicative of significant small postbi- by nephrologists in training under the direction of one at- opsy hemorrhage. A drop in the hematocrit of equal to or tending nephrologist from May 1993 through May 1994 were more than 6 points was considered indicative of moderate prospectively studied. None of the patients had a history of hemorrhage. Blood transfusion requirements were also re- bleeding during or after surgery or dental procedures. One corded.

Journal of the American Society of Nephrology 1251 Renal Biopsy and Thromboelastograph

Predictors of bleeding whereas 2 demonstrated ultrasound evidence of a

The following potential predictors of bleeding were pro- hematoma immediately after biopsy. Five percent (6 of spectively evaluated: age. sex, aspirin use (325 mg once per 120) dropped their hematocrit by 6 or more points, week as part of a transplant protocol), number of days indicating moderate bleeding. No transfusions were between the last aspirin and biopsy, serum creatinine, pre- given to any of the patients. In patients with mild and biopsy systolic blood and diastolic blood pressure, number of moderate bleeding after biopsy, 78 and 83%, respec- passes with the biopsy needle, physician performing the tIvely, had normal coagulation results on all assays. biopsy, ultrasonographer. presence of predominantly medul- Thus, the majority of patients showing mild to mod- lary tissue or vascular sclerosis on histopathologic examina- tion of the biopsy sample. PT. PIT, platelet count. TEG. and erate bleeding after renal biopsy did not have coagu- BT. lation abnormalities, on the basis of the assays used in this study. Coagulation Assays Overall, 21% of the transplant patients had at least Blood for coagulation assays was anticoagulated by the one coagulation test (PT, PTT, BT, PLT, or any TEG addition of 4.5 mL of whole blood to 0.5 mL of 130 mmol/L value) outside the normal reference range. Having any sodium citrate. PT, PIT, and fibrinogen were measured in test outside the reference range was not associated cltrated plasma as previously described (17.18). TEG mea- with an increased risk of bleeding. Table 1 shows the surements were performed as previously described (19). Briefly, whole citrated blood samples were recalcifled by the expected normal reference range. the percentage of addition of 250 L of whole citrated blood to 100 L of patients with abnormal results, the range of abnormal prewarmed 0.645 g/dL calcium chloride. After 1 h, five results, and the statistical association between an measurements are made on the TEG tracing: the r value, the abnormal result and bleeding. The only results that k value, the angle. the MA. and the Aea (Figure 1). Platelet were significantly associated with bleeding were ab- counts were performed in EDTA-anticoagulated blood with normal TEG:angle (a#{176})and TEG:k. Abnormal PT, P’VF, an Abbott Cell Dyn 3000 flow-cytometric cell counter (Abbott BT, PLT, TEG:r, TEG:MA, and TEG:A6o results were Laboratories, Santa Clara. CA). Bleeding times were deter- mined by use of the Simplate II method. (Organon Teknika. not associated with bleeding. TEG:angle and TEG:k Durham, NC). parameters measure the rate of clot formation once clotting has begun. TEG:angle and TEG:k results are a Statistical Analysis combination of coagulation function, platelet func- lest results were defined as abnormal if they fell outside tion, and platelet count. In contrast, the TEG:r, PT, the established reference range (Table 1). The f test was and PT’F measure the time to the onset of clotting, not used to determine whether abnormal test results were asso- the formation of the clot Itself. The other assay in ciated with an increased risk of bleeding. The x2 test was also theory able to detect platelet function defects was the used to determine whether other discrete variables (e.g., sex, aspirin use, vascular sclerosis) were associated with an BT. increased risk of bleeding. BT were not associated with an Increased risk of postbiopsy bleeding; five patients had abnormal RESULTS bleeding times ranging from 10 to 20 mm, of whom After renal bIopsy, 27% of the patients met one of only one bled. In reviewing Table 1, it should be noted the criteria for small but significant bleeding. A he- that the majority of PT, PTT, and platelet count abnor- matocrit drop of at least 4 points was noted in 30, malities were mild. None of the transplant patients

TABLE 1. Hemostosis testing versus bleeding after renal biopsy

Normal No. Bleeding Range of No. Bleeding TestR eference With Normal Abnormal With Abnormal Test#{176} Range Test Result Results Test Result

Tests Associated With Bleeding TEG:angle (degrees) 54-88 24 of 94 46-53 5 of 7 p < 0.01 TEG:k (mm) 2-7 26 of 97 8-9 3 of 4 p < 0.04 Tests Not Associated With Bleeding TEG:r (mm) 3-22 29 of 101 NAb NA NSb TEG:MA (mm) 48-70 28 of 97 42-46 1 of 4 NS TEG:A (mm) 45-70 28 of 95 40-44 1 of 6 NS PT (s) 11-13.6 33 of 110 13.9-14.7 1 of 9 NS PIT (s) 24-36 34 of 118 37 0 of 2 NS Platelet count (1,000/ILL) 150-400 33 of 117 129-148 1 of 3 NS BT (mm) 1-9 28 of 100 10-20 1 of 5 NS

#{176}Thenumber of patients bleeding with a normal result was compared with the number of patients bieeding with an abnormal result by use of the x2 test. b NA. not applicable, all resuits were normal; NS. not significant. P> 0.05.

1252 Volume 6 - Number 4 - 1995 Davis and Chandler

had a PT longer than 15 s, a PTT longer than 37 s, or sion requirements, the need for surgical Intervention. a platelet count less than 129,000/L. and the presence of delayed hematoma formation as The sensitivity, specificity, positive predictive value, detected by ultrasound or computed tomographic and negative predictive value for TEG:angle, BT, and scanning (5.8,9,20,23,24,26,27,29.30). The drop in PT for small postbiopsy hemorrhage are shown in hematocrit, transfusion requirement, and ultrasound Table 2. The TEG:angle was the best in all catagories; findings Immediately after biopsy were chosen as the the worst was the PT. An abnormal TEG:angle had a determinants of bleeding in this study because these positive predictive value of 71%, as compared with criteria have been previously well defined and are the 20% for the BT and only 11% for the PT. least time consuming and most cost effective for use in Patient age. sex, serum creatinmne (bleeding was clinical practice. Previous reports in renal transplant noted in 27% of patients with a creatinine of more recipients have found that, on average, the hematocrit than 2.0 mg/dL and 27% of those with a creatmnmne drops between 0 and 3 points within 24 h of the less than or equal to 2.0 mg/dL; P = 0.95), systolic or procedure and transfusion Is required after 0 to 4% of diastolic blood pressure in the ranges kept during and biopsies (20,23.24,26.27,29,30). In this study, a drop postbiopsy, the number of biopsy passes (median of in hematocrit of 4 points or more occurred in 25% of four passes in those who bled and those who did not), patients, ultrasound evidence of hematoma was seen the use and timing of aspirin administration, the in 2%, and transfusion was given in 0%. Gross hema- identity of the ultrasonographer or nephrologist, or tuna was not analyzed because It occurred in only the presence of vascular sclerosis or predominantly three patients, and both the BT and the TEG were medullary tissue (P = 0.098) were not associated with unavailable in two of the three patients; thus, conclu- the risk of bleeding after renal biopsy. sions on the predictability of the coagulation assay for bleeding in the form of hematurla were not possible. DISCUSSION Changes in hematocrit noted in this study appeared to Bleeding after renal biopsy has been reported to be secondary to the biopsy procedure and not to fluid occur more frequently when: (1) large-bore needles administration because most patients had not re- are used; (2) needle penetration reaches the medulla; ceived exogenous fluids and those who had, received (3) radlologic guidance is not used; (4) renal tissue less than 500 mL. demonstrates fibrosis and nephrosclerosis; (5) the Most patients In this study had normal coagulation patient has hypertension; and (6) a coagulopathy Is results. Most reference ranges. including ours, are present (1-9.20-24). To prevent bleeding. hyperten- based on 95% confidence limits for the healthy popu- sion is treated, needle sizes have decreased, biopsies lation. Approximately 5% of healthy subjects will have are performed under ultrasound guidance. and biopsy results outside these limits. This is consistent with the in the presence of a coagulopathy is avoided by the results that we obtained. For each Individual assay, 2 performance of screening coagulation assays to 8% of the patients had an abnormal result. All (8-10,25-27). However, the predictive value for proce- together, 21% had at least one result outside the dural bleeding of mild abnormalities of the standard reference range. The problem with screening assays in coagulation assays used (PT, PTT, platelet count, and this and previous studies is the low incidence of BT) has been shown in the past and is shown in this abnormalities in the population, the low predictive study to be poor (11,12,28). The BT in particular isa value of borderline abnormal results, and the high poor prebiopsy screening test. Recent reviews have number of false-positives. concluded that it should not be used for preoperative Of the coagulation assays evaluated, the TEG:angle screening (11,12). The BT is prone to procedural and TEG:k values were the best predictors of bleeding. problems, is affected by a number of medications For patients with an abnormal TEG:angle, 5(71%) of 7 associated with minimal bleeding risk, and is uncom- bled after biopsy versus 24 (26%) of 94 bleeding with a fortable for the patient. normal TEG:angle (P < 0.01). An abnormal TEG:angle Bleeding after native and transplant renal biopsy predicted an approximately threefold increase in the has been previously identified by changes in the he- risk of mild bleeding after renal biopsy. In contrast, matocrit, the presence of gross hematuria. transfu- the bleeding time was not predictive of postblopsy

TABLE 2. Predictive value of selected parameters for bleeding after biopsya

Sens Spec PPV NPV Assay N (%) (%) (%) (%) TEG:Angle (degrees) 17% 97% 71% 74% 101 BT (mm) 3% 95% 20% 72% 105 PT(s) 3% 91% 11% 70% 119

#{176}Abllityof tests to predict mud hemorrhage after renal biopsy. Sens, sensitivity; Spec. specificity; PPV, positive predictive value: NPV, negative predictive value, N, total number of tests evaluated. Bleeding was defined as a drop In hematocrit of 4 poInts or more 6 h after the procedure or uitrasound evidence of a new perirenal heratoma.

Journal of the American Society of Nephrology 1253 Renal Biopsy and Thromboelastograph

bleeding. Further contrasts between the BT and TEG (13). The charge at our institution for performing a BT are noted by the fact that the BT may be prolonged is $17.75; a platelet count is $7.00, and a PT and PTF when the TEG is normal. Aspirin injestion causes no together are $24.75. The charge for performing all of change in the lEG but results in prolongation of the the above for 120 biopsIes would be $5,940.00. The BT (3 1-33). Uremia also prolongs the BT while in- charge for a TEG ($17.25) for 120 biopsies would be creasing rather than decreasing lEG clot formation $2,070.00. If we add to this the charge for additional (34). Furthermore, uremia, which has long been asso- workups on patients with abnormal TEG, we arrive at ciated with a prolongation of the BT, has not always a total of $2,405, stifi far less than the prior workup been associated with increased bleeding after renal while evaluating additional factors (I.e., fibrinolysis on biopsy (5,23,24,35). As mentioned previously, pro- the TEG). longed BT on the forearm are not predictive of bleeding In conclusion, the TEG appears to be the most at other sites (11.36-39). cost-effective and predictive coagulation test to per- The lack of correlation between mild hemostatic form before a renal transplant biopsy. These tests abnormalities and postbiopsy bleeding can be ex- should be used in conjunction with the bleeding his- plained by the fact that bleeding is probably more tory, medication history, and proper biopsy technique often determined by the vascular injury caused by the in order to minimize bleeding. An abnormality on the procedure than by the coagulation status of the pa- lEG should be further investigated by coagulation tient. If the patient has no history or signs of prior screening. After the coagulation defect Is located, the bleeding, it is unlikely that they have a major undiag- biopsy could be canceled or coagulation product sup- nosed abnormality. The most common occult coagu- port could be arranged. Following these guidelines lation abnormalities are low platelet counts, reduced should help to reduce costs and minimize bleeding platelet function, and reduced levels of vitamin K-de- after renal allograit biopsy. pendent factors, either due to warfarin treatment, diet, or broad-spectrum antibiotic therapy. A simple REFERENCES screen for these potential problems is a TEG along 1. Bennett AR, Wiener SN: Intrarenal arteriovenous flstula with the history and physical examination. More seri- and aneurysm: A complication of percutaneous renal ous deficiencies in coagulation, platelet, or fibrmnolytic biopsy. Am J Roentgenol 1965:95:372-382. factors, however, may increase the risk for postbiopsy 2. Ekelund L: Arteriovenous fistulae secondary to renal biopsy: An experimental study in the rabbit. Acta Radlol bleeding. The number of patients with moderate to 1970; 10:218-224. severe abnormalities in our study is low, less than 1%. 3. Ekelund L, Lindholm T: Arteriovenous fistula following Much larger patient groups would be needed to deter- percutaneous renal biopsy. Acta Radiol 1971;1 1:38-48. 4. K#{246}hlerR,Edgren J: Angiographic abnormalities follow- mine clinical thresholds for bleeding for each test ing percutaneous needle biopsy of the kidney. Acta Ra- type. For example, on the basis of larger liver biopsy diol 1974:15:515-527. studies, it would appear that excess bleeding will 5. Diaz-Buxo JA, Donadlo JV: Complications of percutane- ous renal biopsy: An analysis of 1,000 consecutive biop- probably not occur after liver biopsy If platelet counts sies. Clin Nephrol 1975;4:223-227. are above 50,000/jL (28). Whether this result is 6. Almkuist 1W, Buckalew VM: Techniques of renal biopsy. applicable to native or transplant renal biopsy is Urol Chin North Am 1979:6:503-5 17. 7. Lawen JG. Van Buren CT, Lewis RM, Kahan BD: Arte- unknown. riovenous fistulas after renal allograft biopsy: A serious The TEG is an empiric assessment of overall hemo- complication in patients beyond one year. Chin Trans- static and flbrmnolytic function (19,31). The r value is plant 1990:4:357-369. similar to a whole-blood clotting time; abnormal re- 8. ParrIsh AE: Complications of percutaneous renal biopsy: A review of 37 years’ experience. Chin Nephrol 1992;38: sults indicate low levels of clotting factors and should 135-14 1. be followed up with PT, PTF, and factor assays. The k 9. Mahoney MC, Racadlo JM, Merhar GL. First MR: Safety value and angle indicate the rate at which the plate- and efficacy of kidney transplant biopsy: Tru-cut needle vs sonographically guided biopty gun. AJR Am J Roent- let/flbrin thrombus forms once initiated; abnormal genol 1993; 160:325-326. results should be followed up with a PT. PTT, fibrino- 10. Wickre CG. Golper TA: Complications of percutaneous gen, and platelet count. The MA indicates maximum needle biopsy of the kidney. Am J Nephrol 1982:2:173- 178. clot strength, which is a function of platelet count and 11. Rodgers RPC, Levin J: A critical reappraisal of the fibrinogen. The A should be compared with the MA; bleeding time. Semin Thromb Haemost 1990:16:1-20. if a reduction of more than 5 to 10 mm Is seen, it 12. LInd SE: The bleeding time does not predict surgical indicates the presence of significant tIbrinolysis. A bleeding. Blood 1991:77:2547-2552. 13. Kang YG, Martin DJ, Marquez J, et at.: Intraoperatlve normal TEG indicates adequate levels of clotting fac- changes in blood coagulation and thrombelastographic tors, fibrinogen, and with normal fibrinolytic monitoring in liver transplantation. Anesth Analg 1985; activity. 64:888-896. 14. Spiess BD, Tuman KJ, McCarthy RJ, DeLaria GA, Clinically, the TEG measurement has been shown to Schillo R, Ivankovich AD: Thromboelastography as an be more useful in the determination of bleeding after indicator of post- coagulopa- cardiac surgery and liver transplantation and is more thies. J Chin Monit 1987:3:25-30. reproducible than the BT (13.14.16). The use of the 15. Rettke SR. Janossy TA, Chantigian RC, et at.: Hemody- nasnic and metabolic changes in hepatic transplanta- TEG during and after cardiac and hepatic procedures tIon. Mayo Clin Proc 1989;64:232-240. has also lessened the use of blood product support 16. Essell JH, Martin TJ. Salinas J, Thompson JM. Smith

1254 Volume 6 - Number 4 1995 Davis and Chandler

VC: Comparison of thromboelastography to bleeding 28. McVay PA, Toy PTCY: Lack of increased bleeding after time and standard coagulation tests in patients after liver biopsy in patients with mild hemostatic abnormal- cardiopulmonary bypass. J Cardiothorac Vasc Anesth ities. Am J Chin Pathol 1990:94:747-753. 1993:7:410-415. 29. Horowitz MD. Russell E, Abitbol C. Kyriakides G, Miller 17. Castillo JB, Farag AM, Mammen EF, Sachs RJ: Pro- J: Massive hematuria following percutaneous biopsy of thrombin times and clottable fibrinogen determinations renal allograft. Arch Surg 1984:119:1430-1433. on an automated coagulation laboratory (ACL-8 10). 30. Rosenbauin R, Hoffsten PE, Stanley RJ, Klahr S: Use of Thromb Res 1989:55:213-219. computerized tomography to diagnose complications of 18. Sibley C, Singer JW. Wood RJ: Comparison of activated percutaneous renal biopsy. Kidney mt 1978:14:87-92. partial thromboplastin reagents. Am J Chin Pathol 1975: 31. Malhett SV, Cox DJA: Thrombehastography. Br J Anaesth 59:581-586. 1992:69:307-313. 19. Chandler WL: The thromboelastograph and the throm- 32. Trentalange MJ, Waits LF: A comparison of throm- boehastograph technique. Semin Thromb Hemostas boehastogram and template bleeding time in the evalua- 1995, in press. tion of platelet function after aspirin ingestion. J Clin 20. Matas AJ, Tellis VA. Sablay L. Quinn T. Soberman R. Anesth 1991:3:377-381. Veith FJ: The value of needle renal allograft biopsy. III. A 33. Orhikowski CEP. Payne AJ, Moodley J, Rocke DA: prospective study. Surgery 1985:98:922-925. Thrombehastography after aspirin ingestion in pregnant 21. Rao KV: Urohogical complications associated with a kid- and non-pregnant subjects. Br J Anaesth 1992;69: 159- ney transplant biopsy: Report of 3 cases and review of 161. the literature. J Urol 1986:135:768-770. 34. Holloway DS. Vagher JP, Caprini JA, Simon NM, Mock- 22. Bogan ML, Kopecky KK, Kraft JL, et at.: Needle biopsy ros LF: Thrombelastography of blood from subjects with of renal alhografts: Comparison of two techniques. Radi- chronic renal failure. Thromb Res 1987:45:817-825. ology 1990:174:273-275. 23. Beckingham IJ, Nicholson ML, Bell PRF: Analysis of 35. Linn#{233}I: Prolonged bleeding time in non-uremic acute glomerulonephritis. Thromb Haemostas 1979:42:1060 factors associated with complications following renal - transplant needle core biopsy. Br J Urol 1994:73:13-15. 1061. 24. Kohb LG, Vehosa JA, Bergstrahh EJ. Offord KP: Peruta- 36. Anirein PC, Ellman L, Harris WH: Aspirin-induced pro- neous renal allograft biopsy: A comparison of two needle longation of bleeding time and perioperative blood loss. types and analysis of risk factors. Transplantation 1994: JAMA 1981:245:1825-1828. 57: 1742-1746. 37. O’Laughhin JC, Hoftiezer JW, Mahoney JP, Ivey KJ: 25. Ferrario F, Porn MT, Maldifassi P, et aL: Ultrasound- Does aspirin prolong bleeding from gastric biopsies in guided renal biopsy. Contr Nephrol 1989:69:75-80. man? Gastrointest Endosc 1981:27:1-5. 26. Nicholson ML, Attard AR, Bell A, Donnelly PK, Veitch 38. Ferraris VA, Swanson E: Aspirin usage and periopera- PS, Bell PRF: Renal transplant biopsy using real time tive blood loss in patients undergoing unexpected oper- ultrasound guidance. Br J Urol 1990:65:564-565. atlons. Surg Gynecol Obstet 1983:156:439-442. 27. McDonald MW, Sosnowski JT, Mahin EJ, Willard DA, 39. Horlocker IT, Wedel DJ, Offord KP: Does preoperative Lainm DL: Automatic spring-loaded biopsy gun with antiplatelet therapy increase the risk of hemorrhagic ultrasonic control for renal transplant biopsy. Urology complications associated with regional anesthesia? 1993:42:580-582 Anesth Analg 1990:70:631-634.

Journal of the American Society of Nephrology 1255