This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Ho, A. M.-H. Articles by Warkentin, T. E. Search for Related Content PubMed PubMed Citation Articles by Ho, A. M.-H. Articles by Warkentin, T. E. Related Collections Cardiovascular Blood

---

CARDIOVASCULAR ANESTHESIA

Agreements Between the Prothrombin Times of Blood Treated In Vitro with Heparinase During Cardiopulmonary Bypass (CPB) and Blood Sampled After CPB and Systemic Protamine

Anthony M.-H. Ho, MSc, MD, FRCPC, FCCP^*, Anna Lee, MPH, PhD^*, Elizabeth Ling, MSc, MD, FRCPC, Alan Daly, CCP, Kevin Teoh, MD, FRCSC, and Theodore E. Warkentin, MD, FRCPC^||

*Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Hong Kong, People’s Republic of China; and Departments of Anaesthesia, Clinical Perfusion, Surgery, and ||Medicine, Hamilton Health Sciences Corporation, McMaster University, Hamilton, Ontario, Canada

Address correspondence and reprint requests to Dr. Anthony Ho, Department of Anaesthesia and Intensive Care, Prince of Wales Hospital, Shatin, NT, Hong Kong SAR, PRC. Address e-mail to hoamh{at}hotmail.com

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The prothrombin time (PT) is useful for identifying coagulation factor deficits after cardiopulmonary bypass (CPB). However, long processing times and the need for fresh frozen plasma (FFP) to be thawed cause delays in factor replacement. We hypothesized that, by treating with heparinase, blood sampled toward the end of CPB can provide PT results that help to determine the requirement for FFP after CPB. Laboratory delays can be eliminated with point-of-care monitors. We studied 158 adults undergoing nonemergent cardiac surgery. Blood taken before separation from CPB was mixed with heparinase, and PT was measured in the laboratory with a HemoTec timer. Agreements between these results and laboratory measurements of blood taken after systemic protamine were compared by using Bland and Altman plots with the threshold of ±1.0 s. We found that the laboratory PT measurements during CPB versus after CPB were compara-ble, but the limits of agreement exceeded these thresholds. Similarly, there was unsatisfactory agreement between the HemoTec and laboratory PT results measured before, during, and after CPB. For each PT measured during CPB, the corresponding confidence interval for the postprotamine PT was calculated. During CPB, a laboratory PT of 16 s or 18 s suggests a 83% or 93% probability of not requiring or potentially requiring, respectively, FFP after CPB. We conclude that the majority of PT measurements obtained from blood taken before weaning from CPB and treated in vitro with heparinase was associated with a high probability of whether or not FFP would be needed after CPB.

IMPLICATIONS: Coagulation dysfunction after cardiopulmonary bypass may contribute to bleeding. Obtaining coagulation tests and fresh frozen plasma requires time and delays treatment in patients who need fresh frozen plasma. We have devised a technique to provide early estimation of postbypass coagulation status.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Coagulopathy is common immediately after cardiopulmonary bypass (CPB). Whole blood prothrombin time (PT) and activated partial thromboplastin time (APTT) are useful in identifying post-CPB coagulation factor deficiencies (1) and in managing adult patients with excessive post-CPB bleeding (1,2). Fresh frozen plasma (FFP) transfusion should ideally be guided in part by coagulation indices measured after IV administration of protamine (3,4). Unfortunately, long turnaround times at laboratories for these indices (5) and the need for thawing cause considerable delays for those who do need FFP. On-site coagulation monitors (1,3,4,6–12) with fast turnaround times have moderate levels of accuracy (4,10,11), and their use remains sporadic. Perhaps partly because of these difficulties, and despite guidelines (5), there is large variability between institutions in transfusion behavior for adult cardiac patients (12). We therefore set out to determine in a group of adult patients whether blood taken before weaning from CPB, after in vitro neutralization of its heparin, could provide a PT result that agrees with that obtained after IV protamine. If coagulopathy after protamine will likely be mild, no FFP needs to be thawed. If it will likely be serious, considerations could be given to the timely request for FFP as soon as it is determined that microvascular bleeding is a problem.

Lyophilized heparinase I (IBEX Technologies Inc., Montreal, Canada) from the bacterium Flavobacterium heparinum catalyzes an eliminase reaction at the antithrombin III site (13). This was chosen to neutralize the heparin in vitro. Unlike protamine, excessive heparinase does not render coagulation tests inaccurate. Previous authors (14–16) have used heparinase to neutralize heparin in blood samples drawn during CPB and have found significant correlation (14) or agreement (15) with blood samples taken after protamine administration. However, these studies were done with thrombelastography, which measures whole blood clotting. This study of agreements of PTs has never been reported.

An accurate point-of-care (on-site) coagulation test would avoid the typical delay caused by processing in the laboratory. As a secondary objective, we investigated the agreement of PTs as measured using the HemoTec automatic coagulation timer (Medtronic-HemoTec Inc., Englewood, CO) and using standard hospital laboratory techniques at various time points.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
This study had institutional ethics approval. Written informed consent was obtained from all patients. They were 158 adult cardiac patients for nonemergent cardiac or aortic surgery requiring CPB in 1998 at the Hamilton Health Sciences Corporation, Ontario, Canada. The patients were chosen according to the order of their presentation into the surgical ward or the preoperative clinic. There were no exclusion criteria.

All patients received an opioid-based anesthetic, supplemented with midazolam, isoflurane, and muscle relaxants. Antifibrinolytics were not used in any patient. CPB was accomplished with a Stöckert SIII Roller Pump (Stöckert Instrumente GmbH, München, Germany) and an Affinity (Medtronic Cardiovascular, Minneapolis, MN) or Monolyth (Sorin BioMedica, Mirandola, Italy) membrane oxygenator. The pump was primed with 2 L of Plasmalyte-148 (Baxter International, Deerfield, IL) containing 5000 IU of porcine intestinal heparin (bioMérieux, Marcy-l’Étoile, France), 100 mEq of sodium bicarbonate, and 24 g of mannitol. CPB was established after IV administration of 400 IU/kg of heparin and an activated clotting time (ACT) >480 s, as measured on a Hemochron 401 (International Technidyne Corp., Edison, NJ), was reached. Mild-to-moderate hypothermia was maintained during cardioplegia. During CPB, additional 5000-IU aliquots of heparin were given as required to maintain an ACT >480 s. Shortly before the aortic cross-clamp was released, 5 mL of blood was taken from the CPB machine. The heparin in this blood was neutralized with heparinase, at a ratio of 1 mL of blood to 4 IU of heparinase [as recommended by IBEX, and used by other authors (15)]. The mixture was then measured for PT, APTT, and ACT by using an on-site HemoTec automatic coagulation timer, and also in the central hospital hematology laboratory (for PT, APTT, and thrombin time [TT]) using standard techniques (henceforth referred to as the "heparinase" results). Upon successful total separation from CPB, the protamine dose based on the heparin dose-ACT response curve (17) was administered. Five minutes later, blood samples were obtained via the arterial catheter (after removal of six dead space volumes) for the same measurements (as at the "heparinase" stage) using the same techniques (henceforth referred to as the "postprotamine" results). This postprotamine blood sample was not treated with heparinase. Transfusion of clotting factors between the sampling times was avoided. Only PT data were analyzed. ACT, PTT, and TT were examined to ensure that no gross omission or inadequate heparin neutralization had occurred.

The HemoTec timer contains a high-range heparinase cartridge with two reaction channels maintained at 37°C, both containing kaolin activator. One of the channels also contains sufficient heparinase to neutralize 6 U/mL heparin. Clot formation is detected by repeatedly raising and releasing a plunger out of the blood and timing its rate of fall. Blood is added to both channels, and the clotting time is determined independently for both channels. The machine then provides the ACT and PT with and without the effects of heparin. The HemoTec device was calibrated according to the manufacturer’s instructions before each case.

In the laboratory, the PT (Amax Mechanical CS 190 Coagulation Analyzer; Sigma Diagnostics, St. Louis, MO) measures the time (normal: 12–15 s) for clot formation of citrated test plasma after incubation with calcified human placenta-derived tissue factor thromboplastin (Thromborel S; Dade Behring, Deerfield, IL). For each batch of thromboplastin, a normal and abnormal control is assayed and graded in potency. The PT is often used to guide FFP therapy in surgical patients (1–3,6,18–21). All postprotamine PT values were measured in the hospital laboratory and were the reference standards for comparison with all heparinase results.

The Bland and Altman plot (22) was used to assess agreement between the PT measurements done during (heparinase) CPB and after (postprotamine) CPB (reference standard). It is a graphical representation (22,23) of the data with between-method difference (y axis) plotted against the average of the data (x axis).

Bias is the mean difference between the two methods of measurement and represents the systematic error. The limits of agreement were defined as mean ± 2 standard deviations, representing the range within which most differences between measurements by different methods will lie (23). A Spearman’s correlation (r) was used to assess whether the differences in measurement values varied systematically over a range of measurements. A significant result indicates that the assumption of uniform variance is not met.

When the mean difference was proportional to the magnitude of the measurement, a regression approach was used (22). The limits of agreement for this regression were estimated by modeling the absolute values of the residuals from the regression as a function of the size of the measurement (22). The expected value of the difference between methods is given by: equation

and the limits of agreement is given by: equation

where D is the mean difference in measurements between the methods, b₀ is an intercept for Equation (1), b₁ is slope of the regression line in Equation (1), c₀ is an intercept for Equation (2), c₁ is slope of the regression line of residuals in Equation (2), and A is any true value of the measurement (average of methods). In practice, when only one method is being used, the observed value by that method provides the value of A (22).

The two methods (PT measurements toward the end of CPB and after IV protamine administration) were judged to be interchangeable if the limits of agreement did not exceed the threshold, set a priori ±1.0 s, deliberately stringent to ensure clinical relevance.

The agreement between the HemoTec and laboratory methods of measuring PT was established by analyzing the data across all time intervals (before CPB, before termination of CPB, after CPB) according to the above graphic method (22,23). The limits of agreement were adjusted for repeated measurements (22). Only data from patients with a complete set of all three pairs of data obtained before and during CPB, and after IV protamine, were analyzed. The 95% confidence intervals (CIs) were calculated around the estimated bias and upper and lower limits of agreement.

Repeatability assessment was not performed because it was not practical to take repeated blood samples at each time interval for each method for each patient. Values were reported as mean and standard deviation or 95% CI, or median and range. The level of significance was set at P < 0.05.

Finally, the percentage level of confidence (probability) of the postprotamine PT either 15 s or 16 s that can be inferred from each heparinase PT, measured in the laboratory or with the HemoTec device, was calculated by using the method outlined by Shakespeare et al. (24).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Of the 158 patients in the study, data from 157 (76% men, mean age 64 ± 11 yr, mean weight 82 ± 15 kg) were analyzed (in 1 patient, there was an inadvertent omission of the heparinase neutralization step). Eight patients had taken warfarin up to 3 days preoperatively. The cases were: coronary artery bypass grafting (CABG): 123, of which 8 were reoperations; repair of ventricular septal defect: 2; repair of atrial septal defect: 1; aortic valve replacement: 13, of which 1 was a reoperation; mitral valve replacement: 9; myomectomy: 1; Bentall procedure: 1; aortic arch repair: 1; combined CABG and mitral valve replacement: 2; and combined CABG and aortic valve replacement: 5. The mean duration of CPB was 95 ± 33 min. Median time between heparinase and postprotamine blood samplings was 53 min (23–200 min). The ACT, APTT, and TT measured after IV protamine indicated that in all cases, systemic heparin was adequately reversed and severe hypofibrinogenemia was not present. No patient required blood product transfusion before postprotamine sampling. All patients were normothermic by the time they came off CPB.

The PT range measured by HemoTec (9–26 s) and laboratory (13–26 s) was wide.

Agreement Between Laboratory PT Measurements During (Heparinase) CPB Versus After (Postprotamine) CPB, and Corresponding Confidence Levels of Postprotamine Indices for Each Heparinase PT
Two patients did not have postprotamine PT measurements. There was a significant relationship between the average and difference in PT measurements (r = -0.27, P = 0.001). Therefore, a regression approach to Bland and Altman plot was used to assess method comparison PT (Fig. 1).

View larger version (18K): [in this window] [in a new window]   Figure 1. Bland and Altman plot (using the regression approach) to show agreement between laboratory prothrombin time (PT) measured before the termination of, and after, cardiopulmonary bypass (CPB). Note the divergence in limits of agreement as PT increases.

Table 1 shows the expected PT after CPB for the range of PT taken during CPB and the probability of a postprotamine PT of 15 s and PT 16 s.

View this table: [in this window] [in a new window]   Table 1. Laboratory PT Measurements During CPB and Expected PT Measurements (95% CI) After CPB Estimated from the Bland and Altman Regression Plot and Percentage Level of Confidence if PT 15 s or if PT 16 s

View larger version (21K): [in this window] [in a new window]   Figure 2. Bland and Altman plot showing lack of agreement between HemoTec and laboratory prothrombin time (PT). Data points are mean of the three time measurements by each method on each subject. The three time points were baseline (before IV heparin), before the end of cardiopulmonary bypass, and after IV protamine.

View larger version (20K): [in this window] [in a new window]   Figure 3. Bland and Altman plot (using the regression approach) to show agreement between prothrombin time (PT) measured by using the HemoTec device before the termination of, and in the laboratory after, cardiopulmonary bypass (CPB). Note the divergence in limits of agreement as PT increases.

View this table: [in this window] [in a new window]   Table 2. PT Measurements During CPB as Measured by HemoTec and Expected Laboratory PT Measurements (95% CI) After CPB Estimated from the Bland and Altman Regression Plot and Percentage Level of Confidence if PT 15 s or if PT 16 s

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The HemoTec device consistently underestimated the PT by 1.2 seconds compared with laboratory PT. Because the limits of agreement exceeded the predetermined level of 1.0 second, the 2 methods were judged not to be interchangeable even if 71% of measurements were within the predetermined limits of agreement. It is conceivable that the use of a threshold of ±1.5 seconds would have found good agreement.

The nonideal agreement between the coagulation tests taken during and after CPB may be attributable to changes in the coagulation status of the patients occurring between the two sampling times. Another cause of disagreement could be the variations in fluids infused (not controlled) and blood loss that occurred between the two sampling times. We did not attempt to derive a formula for the postprotamine PT based on the PT measured during CPB, patient weight, fluid change between sampling times, etc., because such an equation has no practical clinical value. Other causes of disagreement might include the presence in blood of small quantities of very low-molecular-weight heparinase-resistant heparin fragments (14,26).

By using ACT, APTT, and TT, we were able to confirm adequacy of heparin neutralization by heparinase in vitro and by protamine in vivo in all patients. Any residual heparin, if present, was not sufficient to increase the TT, which is much more sensitive than APTT to even small amounts of heparin (3). Likewise, the postprotamine plasma fibrinogen level (not measured) was unlikely to have been grossly subnormal.

It was our opinion that the postprotamine results would not have been influenced by excessive protamine. Compared with a weight-based protamine dosing for heparin neutralization, a heparin dose-ACT response curve calculation (17) resulted in reduced protamine dosages and better-matched neutralization as reflected in reduced transfusion requirements (17,20). Although the use of protamine titration methods might have further reduced protamine use, the Hamilton team, using the heparin dose-ACT response curve technique (17), has an excellent record of minimal use of blood products and small mediastinal re-exploring rate (unpublished data).

During the study period, antifibrinolytics were not standard treatment during cardiac surgery at the Hamilton General Hospital.

The Bland and Altman plot was designed to determine whether two techniques tested under the same conditions are interchangeable. Our aim was to determine whether during CPB the PT of blood treated with heparinase in vitro could replace the PT of blood after IV protamine. Conditions during the two sampling times were similar. To this end, the choice of the Bland and Altman plot was appropriate (as confirmed with Prof. J. M. Bland, St. George’s Hospital Medical School, London, UK, personal communication, 2002). Correlation analysis was not used because, in determining whether two techniques are interchangeable, it does not ensure their comparability (22). The correlation coefficient measures the strength of a relation, not the agreement, between variables (22,23).

Although our PT measurements obtained from blood taken before weaning from CPB and treated in vitro with heparinase did not agree with laboratory-determined PT measurements after CPB, determination of the likelihood for FFP after CPB does not require excellent agreement throughout the range of laboratory values. If the only issue of interest is whether the postprotamine PT will be 15 seconds or 16 seconds, then the technique is useful. We have found that a laboratory-measured PT 16 seconds in blood taken shortly before CPB separation and treated in vitro with heparinase is likely to be associated with a postprotamine laboratory PT 15 seconds. If the PT of the heparinased blood is 18 seconds, there is a high probability that the postprotamine PT will be 16 seconds. The decision on whether FFP is needed after CPB can be further facilitated with the use of a fast and accurate point-of-care device. Our results are consistent with those of a previous study (16), in which heparinase-modified thrombelastography during CPB was used to guide treatment with FFP and resulted in a more than threefold reduction in FFP use with no significant effect on post-CPB chest tube blood losses. By the same principle, using our technique to guide post-CPB FFP transfusion could potentially result in improved hemostasis management. The heparinase used costs US$5 per case, a small price to pay for minimizing delayed and/or unnecessary FFP transfusion. Although the HemoTec timer did not produce PT results that were in adequate agreement with those measured using a standard laboratory technique, when used to measure PT of a heparinase-treated blood sample during CPB, it produced useful data indicative of whether or not the PT would likely be unduly prolonged after separation from CPB.

	Acknowledgments

 
Supported by a Hamilton Health Sciences Corporation grant to AMHH (Project 98–031, Program 94084), and by departmental and institutional resources. The Heparinase enzyme was provided free of charge by IBEX Technologies Inc., Montreal, Quebec, Canada.

We thank Professor J. Martin Bland of St. George’s Hospital Medical School, London, UK, for his statistical advice on the appropriateness of the use of Bland and Altman plots in our measurement of agreement analyses. We also thank the anonymous reviewers for their constructive criticisms.

	Footnotes

 
None of the authors has received remuneration from, or has a financial interest in, any of the makers of the products or their competitors mentioned in this article.

This work was presented in part at the 58th Annual Canadian Anesthesiologists’ Society Meeting in Victoria, BC, Canada, in June 2002, and was published in abstract form: Ho AMH, Lee A, Ling E, Daly A, Teoh K, Warkentin TE. Comparing INR’s measured during (on a heparinase-treated sample) and after CPB. Can J Anesth 2002;44:A44.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Despotis GJ, Santoro SA, Spitznagel E, et al. On-site prothrombin time, activated partial thromboplastin time and platelet count: a comparison of methods with coagulation factor analysis in cardiac surgical patients. Anesthesiology 1994; 80: 338–51.[Web of Science][Medline]
2. Nuttall GA, Oliver WC, Santrach PJ, et al. Efficacy of a simple intraoperative transfusion algorithm for nonerythrocyte component utilization after cardiopulmonary bypass. Anesthesiology 2001; 94: 773–81.[Web of Science][Medline]
3. Horrow JC, Fitch JCK. Management of coagulopathy associated with cardiopulmonary bypass. In: Gravelee GP, Davis RF, Utley JR, eds. Cardiopulmonary bypass: principles and practice. 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2000: 436–66.
4. Weightman WM, Clapin AO, Michalopoulis N, Gibbs NM. Sensitivity and specificity of the Hemochron, Hemotec, and APTT for the detection of low concentrations of heparin. Anaesth Intensive Care 1996; 24: 472–6.[Web of Science][Medline]
5. Goodnough LT, Johnston MF, Ramsey G, et al. Guidelines for transfusion support in patients undergoing coronary artery bypass grafting. Ann Thorac Surg 1990; 50: 675–83.[Abstract]
6. Skubas N, Despotis GJ. Intraoperative diagnosis and therapy of hemostasis abnormalities with cardiac surgery. In: Attar S, ed. Hemostasis in cardiac surgery. Armonk, NY: Futura Publishing Company, 1999: 91–188.
7. Despotis GJ, Gravlee G, Filos K, Levy J. Anticoagulation monitoring during cardiac surgery: a review of current and emerging techniques. Anesthesiology 1999; 91: 1122–51.[Web of Science][Medline]
8. Despotis GJ, Santoro SA, Spitznagel E, et al. Prospective evaluation and clinical utility of on-site coagulation monitoring in cardiac surgical patients. J Thorac Cardiovasc Surg 1994; 107: 271–9.[Abstract/Free Full Text]
9. Despotis GJ, Joist JH, Goodnough LT. Monitoring of hemostasis in cardiac surgical patients: impact of point-of-care testing on blood loss and transfusion outcomes. Clin Chem 1997; 43: 1684–96.[Abstract/Free Full Text]
10. Hirsch J, Wendt T, Kuhly P, Schaffartzik W. Point-of-care testing apparatus. Measurement of coagulation. Anaesthesia 2001; 56: 760–3.[Web of Science][Medline]
11. Samama C, Quezada R, Riou B, et al. Intraoperative measurement of activated partial thromboplastin time and prothrombin time with a new compact monitor. Acta Anaesthesiol Scand 1994; 38: 232–7.[Web of Science][Medline]
12. Stover EP, Siegel LC, Parks R, et al. Variability in transfusion practice for coronary artery bypass surgery persists despite national consensus guidelines: a 24-institution study. Institutions of the Multicenter Study of Perioperative Ischemia Research Group. Anesthesiology 1998; 88: 327–33.[Web of Science][Medline]
13. Dietrich CP, Silva ME, Michelacci YM. Sequential degradation of heparin in Flavobacterium heparinum: purification and properties of five enzymes involved in heparin degradation. J Biol Chem 1973; 248: 6408–15.[Abstract/Free Full Text]
14. Tuman KJ, McCarthy RJ, Djuric M, et al. Evaluation of coagulation during cardiopulmonary bypass with a heparinase-modified thromboelastographic assay. J Cardiothorac Vasc Anesth 1994; 8: 144–9.[Medline]
15. Miller BE, Guzzetta NA, Tosone SR, Levy JH. Rapid evaluation of coagulopathies after cardiopulmonary bypass in children using modified thromboelastography. Anesth Analg 2000; 90: 1324–30.[Abstract/Free Full Text]
16. Royston D, von Kier S. Reduced haemostatic factor transfusion using heparinase-modified thrombelastography during cardiopulmonary bypass. Br J Anaesth 2001; 86: 575–8.[Abstract/Free Full Text]
17. Patla V, Comunale ME, Lowenstein E. Heparin neutralization. In: Gravlee GP, Davis RF, Kurusz M, Utley JR, eds. Cardiopulmonary bypass: principles and practice. Philadelphia: Lippincott Williams & Wilkins, 2000: 471–91.
18. Reiner AP. Massive transfusion. In: Spiess BD, Counts RB, Gould SA, eds. Perioperative transfusion medicine. Baltimore: Williams & Wilkins, 1998: 351–64.
19. Greaves M, Preston FE. Approach to bleeding patient. In: Colman R, Hirsh J, Marder V, Clowes AW, George JN, eds. Thrombosis and hemostasis: basic principles and clinical practice. Philadelphia: JB Lippincott, 2001: 783–93.
20. Jobes DR, Ellison N. Hemotherapy: control of hemostasis and blood replacement. In: Estafanous FG, Barash PG, Reves JG, eds. Cardiac anesthesia: principles and clinical practice. Philadelphia: JB Lippincott, 1994: 597–620.
21. Milas BL, Jobes DR, Gorman RC. Management of bleeding and coagulopathy after heart surgery. Semin Thorac Cardiovasc Surg 2000; 12: 326–36.[Medline]
22. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999; 8: 135–160.[Abstract/Free Full Text]
23. Mantha S, Roizen MF, Fleisher LA, et al. Comparing methods of clinical measurement: reporting standards for Bland and Altman analysis. Anesth Analg 2000; 90: 593–602.[Abstract/Free Full Text]
24. Shakespeare TP, Gebski VJ, Veness MJ, Simes J. Improving interpretation of clinical studies by use of confidence levels, clinical significance curves, and risk-benefit contours. Lancet 2001; 357: 1349–53.[Web of Science][Medline]
25. Katz MH. Multivarable analysis: a practical guide for clinicians. New York: Cambridge University Press, 1999: 145.
26. Despotis GJ, Summerfield AL, Joist JH, et al. In vitro reversal of heparin effect with heparinase: evaluation with whole blood prothrombin time and activated partial thromboplastin time in cardiac surgical patients. Anesth Analg 1994; 79: 670–4.[Abstract/Free Full Text]

This article has been cited by other articles:

				B. S. Donahue Factor V Leiden and Perioperative Risk Anesth. Analg., June 1, 2004; 98(6): 1623 - 1634. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Ho, A. M.-H. Articles by Warkentin, T. E. Search for Related Content PubMed PubMed Citation Articles by Ho, A. M.-H. Articles by Warkentin, T. E. Related Collections Cardiovascular Blood