The Predictive Value of Thromboelastography for Postoperative Blood Loss in Off-Pump Coronary Artery Bypass Grafting

W. S. Ng(1), K. Buchan(2), C. Y. Eng, A. Ronald(3)

1 30 South Mount Street, Aberdeen AB25 2TB

2 Cardiothoracic Unit, Aberdeen Royal Infirmary, Foresterhill Road, Aberdeen,

AB25 2ZN

3 Anaesthetic Unit, Aberdeen Royal Infirmary, Foresterhill Road, Aberdeen, AB25

2ZN

Summary

Previous studies have assessed whether abnormal thromboelastography parameters predict bleeding after on-pump coronary artery bypass grafting but none have determined its predictive value in off-pump coronary artery bypass grafting. We investigated the value of thromboelastography in predicting postoperative bleeding in 41 patients undergoing off-pump coronary artery bypass grafting surgery over a year.

We classed patients as “bleeders” or “non-bleeders” based on their blood loss at 1 hour postoperatively, 6 hours postoperatively, and at time of chest drain removal. Thromboelastography reaction times (R & K-times), maximum amplitude (MA), α-angle and 30-minute lysis (LY30) data was compared for bleeders and non-bleeders.

We could not identify any significant differences in thromboelastography data between them.

Therefore, we conclude that thromboelastography data was not predictive of bleeding after off-pump coronary artery bypass grafting in our group of patients.

Introduction:

Thromboelastography (TEG) is a near-patient coagulation monitor, which measures blood clot rate of formation, maximum strength and fibrinolysis. It is increasingly used in Cardiac Surgical Units to manage coagulopathy and guide blood component therapy following surgery (Nuttall GA et al, 1997, Dorman BH et al, 1993). Whilst several papers have studied its predictive role in identifying “bleeders” in conventional on-pump CABG (Camerer U et al, 2003; Ereth MH et al, 1997), none have evaluated its role in off-pump (OPCAB) surgery. This study aimed to investigate the value of TEG in predicting postoperative bleeding following OPCAB.

Methods and Aims:

We retrospectively studied all OPCAB patients operated on by one surgeon and one anaesthetist over a 12 month period. All patients had baseline and post-Protamine TEG samples performed as part of their standard anaesthetic management using a Haemoscope Thrombelastograph^® Haemostasis Analyser (Hemoscope, Illinois, IL). Reaction times (R & K-time), maximum amplitude (MA), α-angle, and 30-minute lysis (LY30) data was obtained from baseline kaolin activated (BL K) and post-protamine kaolin-activated (PP K) and heparinase-modified kaolin activated (PP KH) samples. All samples were arterial, performed within 5 minutes of taking the sample and were run to at least 30 minutes beyond MA. We retrieved data regarding postoperative blood loss at 1 hour, 6 hours and at chest drain removal from case notes.

All patients were first-time cardiac surgery patients undergoing isolated OPCAB. Most patients had their Aspirin or Clopidigrel discontinued at least 7 days preoperatively and received either Tranexamic acid or Aprotinin depending on the surgeon’s preference. We excluded patients with incomplete blood loss or TEG data from analysis.

Patients were sub-divided into three groups on the basis of their postoperative blood loss. Within each of these three groups, we separated patients into “bleeders” and “non-bleeders”. The limits for this were taken from Camerer (Camerer U et al, 2003) and Ereth (Ereth MH et al, 1997). Hence, Group A “bleeders” were defined according to Ereth as patients with blood loss of more than 100ml in the 1^st hour postoperatively. Group B “bleeders” were defined according to Camerer as those with blood loss of more than 750ml in the 6^th hour postoperatively. A further group called Group C “bleeders” was also created for completion. The upper quartile of “total blood loss”(reading at time of chest drain removal) was taken from our study group as a cut off point for “bleeders”. Subsequently Group C “bleeders” were defined as those with blood loss of more than 1260 ml at the time of chest drain removal. Appendix 1 shows a summary of the definitions used for bleeders in each of the three groups.

We then tried to identify any abnormal TEG values between “bleeders” and “non-bleeders” within these 3 groups. We defined abnormal TEG values according to Royston (Royston D et al, 2001) (see Appendix 2). These are the standard TEG limits we use in our unit for on-pump surgery.

“R-time” values of PPK and PPKH samples were also checked to identify inadequate reversal of heparin following protamine administration.

. The mean values between the “bleeders” and “non-bleeders” were later compared using the Mann-Whitney U test with p<0.05 regarded as significant.

Results:

In the period under study, 54 patients underwent OPCAB surgery. 13 patients were excluded from analysis due to incomplete data. The patient demographics, medication and blood loss data of the remaining 41 patients is summarised in Table 1.

There were 19 “bleeders” and 22 “non-bleeders” in Group A, 2 “bleeders” and 39 “non-bleeders” in Group B and 8 “bleeders” and 33 “non-bleeders in Group C. No patient had abnormal baseline TEG values as defined by Royston (Royston D et al, 2001). The treatment protocol for our unit is based on the same limits; therefore, none of the patients would have received transfusions as none had abnormal baseline TEG values.

There were no significant differences between “bleeders” and “non-bleeders” in Groups A nor C. However, for PP K LY30 data in Group B, there was a significant difference between “bleeders” and “non-bleeders” (3.18±0.53 v 1.88±1.20 respectively, p<0.05). The results of Group B are shown in Table 2a and 2b. No samples showed a residual heparin effect when “R-time” values were compared in PP KH and PP K samples.

Discussion

There are a number of limitations in this study, not least the small sample size. Also, we did not control for Aprotinin or Tranexamic Acid use, both of which were used by our surgeon and which have been shown to modify TEG values in different ways (Avidan MS et al, 1995; Robbins P et al, 1998). Had we segregated our patients on the basis of antifibrinolytic usage, the groups would have been too small to analyse. Future studies should be larger to allow for different antifibrinolytic regimes to be compared.

Our definitions of group A and group B “bleeders”, although similar to that used in previous on-pump studies (Camerer U et al, 2003; Ereth MH et al, 1997), was somewhat empirical for our study population given the fact that these studies were performed during on-pump surgery where there are many contributors to bleeding absent during OPCAB procedures. Group C was defined as the upper quartile of total chest tube drainage and was used as an indicator of excessive bleeding in our own specific study population. Further cohort studies need to be performed to better define a threshold for “abnormal bleeding” in OPCAB surgery.

There are also inherent limitations of the TEG. We used the study by Royston et al (Royston D et al, 2001) to define “abnormal” TEG values. However, this study was performed in patients undergoing complex on-pump cardiac surgery and there is no evidence to indicate that their abnormal TEG data can be extrapolated to off-pump surgery. Indeed, Royston himself acknowledges that there is little evidence that the values chosen in their study are optimal for any patient group! Therefore it may not be surprising that we found no difference between TEG data and OPCAB bleeders using these values. Ideally, the best predictor for OPCAB postoperative bleeding should be sought using Receiver Operating Characteristic (ROC) curves to visualise and identify the best cut-off point for each TEG parameter.

Although there was a difference in PP K LY30 data in Group B between bleeders and non-bleeders, this group was too small to analyse. However, as there was no difference in LY30’s in the paired PP KH sample, this may merely represent an irrelevant statistical blip.

Conclusions:

In conclusion, this study did not find any significant differences in TEG values between “bleeders” and “non-bleeders” using the thresholds described and so TEG data was not predictive of postoperative bleeding. However, this was a small study with a number of limitations and a larger study is needed both to define abnormal blood loss and to establish abnormal TEG values during OPCAB surgery. Once these have been achieved it may be possible to develop TEG-guided evidence-based transfusion protocols for OPCAB surgery.

Appendices

Appendix 1 Definitions of “bleeders”

Group A “bleeders”” → >100ml in 1^st hr

Group B “bleeders”” → > 750ml in 6^th hr

Group C “bleeders”” → > 1260ml(upper quartile) when chest drain removed

Appendix 2 Abnormal TEG® Values according to Royston

“r”time > 14mm

MA < 48

LY30 > 7.5%

Tables

Table 1

Patient demographics, medication and blood loss data

Variable

Demographics

Sex (male) 33(80.5%)

Age (year) 65(±9.6)*

Intraoperative antifibrinolytic therapy:

Patients on Aprotinin (no. of patients) 12(29.3%)

Patients on Tranexamic Acid (no. of patients) 29(70.7%)

Blood Loss 1^st hr (ml): 158(±133)*

Blood Loss 6^th hr (ml) 363(±206)*

Blood Loss Total during chest drain removal 958(±541)*

* Data are expressed as mean±SD

Table 2a

Post Protamine Kaolin-activated Results for Group B

TEG® Variable PPK	“bleeders” (>750mls in 6 hours)	“non-bleeders” (<750mls in 6 hours)	Statistical Significance
“r” time	7.63±4.75	5.71±2.39	ns
“k” time	2.00±1.63	1.81±0.86	ns
MA	60.9±6.02	63.4±8.04	ns
α-Angle	67.2±17.2	65.2±8.16	ns
LY30	3.18±0.53	1.88±1.20	p<0.05

* Data are expressed as mean±SD

Table 2b

Post Protamine Heparinase-modified Kaolin activated Sample Results for Group B

TEG® Variable PPKH	“bleeders” (>750mls in 6 hours)	“non-bleeders” (<750mls in 6 hours)	Statistical Significance
“r” time	7.30±4.52	5.50±2.30	ns
“k” time	1.65±1.02	1.71±0.77	ns
MA	59.9±4.09	60.7±8.37	ns
α-Angle	70.1±13.8	67.2±7.53	ns
LY30	4.60±1.64	2.81±1.64	ns

* Data are expressed as mean±SD

		Click on these buttons to visit our journals
Psychiatry On-Line	Dentistry On-Line	Vet On-Line	Chest Medicine On-Line	GP On-Line	Pharmacy On-Line	Anaesthesia On-Line	Medicine On-Line	Family Medical Practice On-Line

All pages copyright ©Priory Lodge Education Ltd 1994-2006.

References:

1 Avidan MS, Da Fonseca J, Parmar K, Alcock E, Ponte J, Hunt BJ. The Effects of Aprotinin on Thromboelastography with three different activators. (2001) Anesthesiology. 95, 1169-1174.

2 Camerer U, Dietrich W, Rampf T, Braun SL, Richter JA. The predictive value of modified computerized Thromboelastography and platelet function analysis for postoperative blood loss in routine cardiac surgery.(2003) Anesth Analg. 96, 51-57.

3 Dorman BH, Spinale FG, Bailey MK, Kratz JM, Roy RC. Identification of patients at risk for excessive blood loss during coronary artery bypass surgery: Thromboelastography vs. coagulation screen.(1993) Anesth Analg. 76, 694-700.

4 Ereth MH, Nutall GA, Klindworth JT, Macveigh I, Santrach PJ , Orszulak TA, Harmsen WS, Oliver WC. Does the platelet-activated clotting test (HemoSTATUS) predict blood loss and platelet dysfunction associated with cardiopulmonary bypass? (1997) Anaesth Analg. 85, 259-264.

5 Nuttall GA, Oliver WC, Ereth MH, Santrach PJ. Coagulation tests predict bleeding after cardiopulmonary bypass. (1997) J Cardothoracic Vasc Anesth. 11,815-823.

6 Royston D, von Kier S. (2001) Reduced haemostatic factor transfusion using heparinase-modified thromboelastography during cardiopulmonary bypass. British Journal of Anaesthesia 86, 575-578.

7 Robbins P, von Kier S, Forrest M, Royston D.(1998) Activated clotting times are shortened by tranexamic acid. Anesthesiology. 89, (3A): A962.