Anaesthesia 2015, 70, 135–141
doi:10.1111/anae.12911
Original Article The effect of serial in vitro haemodilution with maternal cerebrospinal fluid and crystalloid on thromboelastographic (TEGâ) blood coagulation parameters, and the implications for epidural blood patching* S. Armstrong,1 R. Fernando,2 P. Tamilselvan,3 A. Stewart2 and M. Columb4 1 2 3 4
Consultant, Consultant, Consultant, Consultant,
Department Department Department Department
of of of of
Anaesthesia, Anaesthesia, Anaesthesia, Anaesthesia,
Frimley Health NHS Foundation Trust, Frimley, UK University College London Hospitals NHS Trust, London, UK The Princess Alexandra Hospital, Harlow, UK South Manchester University Hospitals NHS Trust, Wythenshawe, UK
Summary Epidural blood patches may be used to treat post-dural puncture headache following accidental dural puncture in parturients. Their mode of action and the optimum volume of blood for injection remain controversial, with the interaction between injected blood and cerebrospinal fluid unknown. We aimed to establish the effects of serial haemodilution of whole blood with cerebrospinal fluid from 34 pregnant patients compared with serial haemodilution with Hartmann’s solution, using the thromboelastogram. Haemodilution with either cerebrospinal fluid or Hartmann’s solution had significant procoagulant and clot destabilising effects, enhanced with progressive haemodilution up to 30%. The effect of cerebrospinal fluid was greater compared with Hartmann’s solution (p < 0.001). Cerebrospinal fluid led to a mean (95% CI) decrease in r-time by 2.4 (1.6–3.2) min, a decrease in k-time by 0.6 (0.4–0.8) min, an increase in alpha angle by 7.3 (5.5–9.0)°, and a decrease in maximum amplitude by 2.0 (0.6–3.4) mm. This may have implications for epidural blood patch, as success may be reduced near the time of dural puncture when cerebrospinal fluid leak is at its greatest, and large volumes of blood may be required to reduce haemodilution and clot destabilisation by cerebrospinal fluid. In addition, blood patching should be performed at the level of the dural puncture in order to ensure that the maximum volume of blood comes into contact with the cerebrospinal fluid.
.................................................................................................................................................................
Correspondence to: R. Fernando Email: [email protected] *Presented in part at the annual meeting of the Society for Obstetric Anesthesiology and Perinatology, May 2005, California, USA, and at the annual meeting of the Obstetric Anaesthetists’ Association, May 2005, London, UK. This paper is accompanied by an editorial: Anaesthesia 2015; 70: 119–121. Accepted: 18 September 2014
Introduction It has been reported that up to 80% of parturients who have an accidental dural puncture during epidural insertion develop a post-dural puncture headache © 2014 The Association of Anaesthetists of Great Britain and Ireland
(PDPH) [1]. Characteristically, the headache occurs within 24 h of the dural puncture. In most cases, PDPH is moderate and self-limiting, but in some cases it may be severe and debilitating. There is also the 135
Anaesthesia 2015, 70, 135–141
Armstrong et al. | Haemodilution with CSF and coagulation
potential for PDPH to result in morbidity and even death. The injection of autologous blood into the epidural space (epidural blood patch) was initially suggested by Gormley in 1960, and remains the most effective treatment to date. Success rates of epidural blood patch to treat PDPH have been quoted to be as high as 90%, yet other studies have shown success rates of only 60–70% [2]. The effectiveness of the epidural blood patch is thought to be twofold: first, injecting blood into the epidural space causes movement of the dura anteriorly, resulting in an increase in cerebrospinal fluid (CSF) pressure in the subarachnoid space and immediate relief of symptoms [3]. Second, formation of a blood clot at and around the site of the dural puncture prevents any further CSF leak, giving longterm relief of symptoms [4]. Despite the fact that the epidural blood patch has been used as a valuable treatment for over 50 years, there are still many controversies regarding its application and effectiveness. The volume of blood to be injected is controversial; some authors advocate 8–10 ml, and others 15–20 ml, although 20 ml has now become the standard volume [5, 6]. Thromboelastography has been used as a monitor of whole blood coagulation in a variety of clinical settings, including obstetric anaesthesia [7]. It is easy to use, and reflects the adequacy of whole blood coagulation within 30–40 min, by delineating the interaction between platelets, fibrinogen and other clotting factors [8]. We postulated that increasing the degree of haemodilution by CSF in an epidural blood patch may increase both the firmness and speed of clot formation. This might be significant in establishing the effectiveness of the blood patch. The aim of this in-vitro study was therefore to assess the effects of serial haemodilution of whole blood with CSF from pregnant healthy patients on blood coagulation using the thromboelastogram (TEG).
pre-existing prothrombotic disorders; pregnancyrelated hypertensive disease; antiplatelet medication; height < 150 cm or > 180 cm; weight < 50 kg or > 100 kg; and macroscopic evidence of accidental epidural venous cannulation. A 16-G peripheral intravenous cannula was sited under local anaesthesia in a forearm vein. Initially, 2 ml blood was taken and discarded; following this, another 10 ml whole blood was collected, and subsequently the cannula was connected to 1000 ml intravenous Hartmann’s solution (Baxter, Thetford, Norfolk, UK) as part of the routine management for caesarean section. The blood sample was placed in a standard coagulation sample bottle containing buffered citrate (Vacutainer; Becton-Dickinson, Franklin Lakes, NJ, USA). All citrated samples were inverted five times and assessed within 60 min of sampling. Combined spinal-epidural anaesthesia was initiated in the sitting position, at the L3-4 or L4-5 interspace, using a locking kit (Portexâ Combined Spinal/Epidural Minipack with Lock Pencil Point Spinal Needle, Smiths Medical ASD Inc., Keene, NH, USA), consisting of a 16-G Tuohy needle and a 26-G pencil-point spinal needle. After puncturing the dura with the spinal needle, 1 ml CSF was collected by passive flow of drops into a graduated glass tube, before giving the standardised injection of 11 mg hyperbaric bupivacaine 0.5% and 15 ll fentanyl. Serial dilutions of whole blood with CSF were then carried out, calculated to produce final CSF dilutions of 10%, 20% and 30%. A 10% dilution represented 0.9 ml blood and 0.1 ml CSF; a 20% dilution represented 0.8 ml blood and 0.2 ml CSF; and a 30% dilution represented 0.7 ml blood and 0.3 ml CSF. Similar dilutions of whole blood with Hartmann’s solution at body temperature were used as controls. All the reconstituted blood samples were recalcified by mixing 0.33 ml blood with 0.3 ml calcium chloride 0.2 mol within the testing cup, warmed to 37 °C. Thromboelastography was performed on all the samples with the TEG 5000â (Hemoscope Corp., Niles, IL, USA). Quality control was regularly performed on the TEG by the local point-of-care testing department in accordance with the manufacturer’s guidelines. Using the TEG, r- and k-times, the alpha angle and the maximum amplitude (MA) were measured for
Methods After institutional ethical approval, written informed consent was obtained from 34 healthy term parturients. All were awaiting planned elective caesarean section under spinal anaesthesia at the Royal Free Hospital in London, UK. Exclusion criteria included: 136
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
Table 1 Thromboelastograph (TEG) terms of reference. r-Time k-Time
Clotting factor activity Clot formation time
Alpha angle
Platelet and fibrinogen activity Clot strength
MA
Time from initiation of the test to initial fibrin formation and pin movement Time from the beginning of clot formation until the amplitude of the TEG reaches 20 mm Angle between the line in the middle of the TEG tracing and the line tangential to the developing ‘body’ of the TEG Greatest amplitude on the TEG trace
MA, maximum amplitude.
each sample (Table 1). Thromboelastographic variables for serial blood-CSF dilutions were compared with the corresponding serial dilutions with Hartmann’s solution. Repeated measures ANOVA with Geisser-Greenhouse adjustment was used to examine the effects of dilution, diluent and interaction. Post-tests included linear and quadratic trends to examine for a doseresponse effect of dilution and Tukey–Kramer tests for paired comparisons. Significance was defined at p < 0.05 (two-sided). Data were analysed using Number Cruncher Statistical Systems (NCSS) 2004 (NCSS Inc., Kaysville, UT, USA). Sample size calculations were based on the report by Ruttmann et al., and it was estimated that 34 patients would be required to find within-subject differences of 25% (SD 45%) in the r-time, at 90% power between diluents and at 80% power with Bonferroni corrections for three dilutions [9].
The mean (SD) gestational age of the participants was 38 + 5 weeks (1 day). Dilution with CSF had increasing effect on all TEG variables with increasing amounts of CSF (Figs. 1–4). At 30% dilution, there were maximum procoagulant effects as seen by decreased r-time and ktime, and increased alpha angle (Table 2). Dilution with CSF also reduced MA, and therefore had a clot destabilising effect (Table 2). The effect of dilution with Hartmann’s solution was more variable; at 30% dilution, r-time, k-time and MA were reduced (Fig. 1). The effect of CSF was consistently greater than that of Hartmann’s solution (Fig. 1 and Table 2).
Discussion
Analyses were complete from all samples for the 34 patients. The mean (SD) age was 34.7 (4.6) years, height was 163.3 (5.2) cm, and weight 71.8 (10.1) kg.
In this study, we demonstrated that serial haemodilution with Hartmann’s solution reduced r- and k-times by up to 30%, demonstrating an increase in the coagulability of whole blood when measured using TEG. We found that progressive dilution of whole blood with CSF also increased the coagulability of whole blood. Analysis of the data show increased coagulability with CSF compared with Hartmann’s solution for all dilutions. This
Figure 1 Mean r-time for blood diluted with cerebrospinal fluid (CSF; s) or Hartmanns solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p = 0.015; diluent: p < 0.001.
Figure 2 Mean k-time for blood diluted with cerebrospinal fluid (CSF; s) or Hartmanns solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p = 0.02; diluent: p < 0.001.
Results
© 2014 The Association of Anaesthetists of Great Britain and Ireland
137
Anaesthesia 2015, 70, 135–141
Figure 3 Mean maximum amplitude for blood diluted with cerebrospinal fluid (CSF; s) or Hartmann’s solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p < 0.001; diluent: p = 0.005.
Figure 4 Alpha angle for blood diluted with cerebrospinal fluid (CSF; s) or Hartmann’s solution () Error bars are SD. Dilution: CSF p = 0.001, Hartmann’s solution p = 0.30; diluent: p < 0.001.
Armstrong et al. | Haemodilution with CSF and coagulation
suggests that the increase in coagulability seen with the addition of CSF cannot be attributed only to its dilutional effect on whole blood, but that addition of CSF has a procoagulant effect that is independent of dilution. There was also a significant increase in the alpha angle with CSF, but not with Hartmann’s solution, which would indicate that the presence of CSF increases the speed of formation of the clot. Both Hartmann’s solution and CSF significantly reduced the MA with progressive haemodilution, indicating that both Hartmann’s solution and CSF reduce the strength of the clot formed. Our findings suggest that, when performing an epidural blood patch, the presence of a CSF leak could speed up the activation of the coagulation pathway, but clot strength may be reduced. However, we found up to a 50% reduction in r- and k-time with the addition of CSF, whereas the reduction in MA was only 10%. This may explain why blood patches are so effective despite the slight loss of clot strength, and may also explain the recurrence of the PDPH, after initial resolution of symptoms following epidural blood patch. Post-dural puncture headache is an important iatrogenic cause of morbidity in obstetric anaesthesia. Although autologous epidural blood patch is the definitive treatment for PDPH and the failure of this procedure is relatively rare, it may not be effective in a
Table 2 Effects of dilution of whole blood with Hartmann’s solution (HS) or cerebrospinal fluid (CSF). Data are mean difference and percentage change with 95% CI. TEG variable
HS (0% vs 30% dilution)
CSF (0% vs 30% dilution)
HS vs CSF across 10%, 20% and 30% dilutions
r-Time; min
1.8 ( 3.2 to 0.4) 17% ( 30 to 4%) p = 0.004 0.5 ( 1.0 to 0.0) 20% ( 41 to 0%) p = 0.050 2.1 ( 6.5 to 2.3) 4% ( 4 to 11%) p = 0.600 5.3 ( 7.5 to 3.2) 8% ( 11 to 5) p < 0.001
5.1 ( 6.2 to 4.0) 47% ( 57 to 37%) p < 0.001 1.3 ( 1.7 to 0.8) 49% ( 66 to 32%) p < 0.001 10.4 (6.3–14.6) 17% (11–25%) p < 0.001 9.2 ( 12.2 to 6.2) 9% ( 12 to 6%) p < 0.001
2.4 ( 3.2 to 1.6) 27% ( 36 to 18%) p < 0.001 0.6 ( 0.8 to 0.4) 30% ( 39 to 21%) p < 0.001 7.3 (5.5–9.0) 12% (9–15%) p < 0.001 2.0 ( 3.4 to 0.6) 3% ( 5 to 1%) p = 0.005
k-Time; min
Alpha angle; °
MA; mm
Within-diluent p values are Tukey–Kramer tests for 0% vs 30% dilution. Between diluent p values across 10%, 20% and 30% dilutions are Geisser-Greenhouse adjusted. TEG, thromboelastography; MA, maximum amplitude. 138
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
significant proportion of patients [6, 10–12]. Baraz et al. undertook a postal questionnaire of UK obstetric units, and found that only 26% treated PDPH with an epidural blood patch as soon as it was diagnosed, whereas in 71% of units, the blood patch was only performed once conservative measures had failed [13]. There are two widely held theories, which are somewhat competing, to account for the onset of the headache. Firstly, CSF leaks from the subarachnoid space into the epidural space, facilitated by the subatmospheric pressure within this space, allowing the brain and its supporting anatomy to descend and cause traction on pain-sensitive intracranial structures and meninges. The second theory suggests that CSF leak from the subarachnoid space also results in loss of CSF volume and CSF pressure, causing intracranial hypotension, compensatory cerebral vasodilatation and migraine-like pain [14–16]. This theory is supported by the fact that some magnetic resonance imaging (MRI) studies have shown enhanced cerebral blood flow in PDPH [15]. However, MRI has not shown an association between the volume of CSF that escapes and the presence of PDPH [17]. Epidural blood patching is thought to reduce PDPH by two mechanisms. Firstly, compression of the intrathecal space with blood increases the subarachnoid pressure by forcing CSF cephalad. Secondly, it is thought that a fibrin clot forms over the dural puncture site, which prevents further CSF leak. It has been shown by MRI in humans that, at 7–13 h, there is clot resolution, leaving a thick layer of mature clot over the dorsal part of the thecal sac, and animal studies have demonstrated widespread fibroblastic activity and collagen formation after seven days [18–20]. The TEG gives a graphic display of the stages of whole blood clot formation, and in normal pregnancy it demonstrates a reduction in r- and k-times, and an increase in alpha angle and MA, in keeping with a hypercoagulable picture [21]. It should be borne in mind with any TEG study that there may be variability between r- and k-times undertaken even by the same operator, possibly due to variable mixing of blood and kaolin in the sample tube [22]. Previous studies on the use of the epidural blood patch for accidental dural puncture have discovered that CSF has procoagulant activity. Cook et al. per-
formed a controlled animal model study, which simulated the mixing of blood and CSF as in an epidural blood patch and recorded the coagulation of the CSFblood mixture using TEG [23]. They compared 360 ll blood with 180 ll blood mixed with 180 ll CSF. They found that the r-time, the k-time and r+k value all showed significant decreases when compared with unmixed blood. The alpha angle was measured, but not shown to differ significantly between groups. In this study, MA was not measured. Acceleration of the onset of coagulation and a stronger clot were both observed. Ruttmann et al. investigated the effect of a 20% haemodilution of an in-vitro blood sample with either a warmed physiological balanced salt solution (plasmalyte B), or CSF from parturients undergoing spinal anaesthesia for caesarean section using TEG. They demonstrated that haemodilution of whole blood enhanced coagulation as before, but also that CSF had a profound hypercoagulable effect, greater than that seen with plasmalyte B, supporting the presence of a procoagulant in CSF [9]. In our study, we demonstrated that serial haemodilution with Hartmann’s solution resulted in an increase in the coagulability of whole blood when measured using TEG. This would support work again done by Ruttmann et al., who demonstrated that haemodilution with 0.9% sodium chloride and other crystalloid and colloid solutions caused enhanced coagulation [24]. In their study, blood was taken from a peripheral cannula in patients undergoing vascular surgery following crystalloid or colloid administration. Exact percentages of haemodilution could not be established due to individual differences in total blood volume, although both groups showed a decrease in haematocrit of approximately 10%. We believe our study has greater significance in the context of epidural blood patch for accidental dural puncture. No other investigators have established serial blood-CSF dilutions and its effect on coagulation. Many studies have concentrated on the optimal volume of blood to be injected when performing an epidural blood patch, with the consensus suggesting a volume of 15–20 ml. Paech et al. concluded that anaesthetists should attempt to administer 20 ml autologous blood, limited by back pain, based on their
© 2014 The Association of Anaesthetists of Great Britain and Ireland
139
Anaesthesia 2015, 70, 135–141
Armstrong et al. | Haemodilution with CSF and coagulation
trial [6]. Some authors have advocated prophylactic epidural blood patch after accidental dural puncture, and a number of reviews have looked at this area, concluding that previous studies were flawed in their methodology and further large studies were required [25, 26]. One hundred and sixteen parturients were randomly assigned by Stein et al. to receive either prophylactic or therapeutic epidural blood patch with 15–20 ml autologous blood after accidental dural puncture [27]. They found that prophylactic blood patch decreased the incidence of PDPH and also decreased the intensity of the headache and accompanying symptoms. When considering the effect of CSF on whole blood, we only concentrated on a dilution of up to 30%. From previous work by Roche et al., with lactated Ringer’s solution, a procoagulant effect was seen with up to 40% dilution, yet further dilution resulted in a reduction in coagulation [28]. This may suggest that as large a volume of blood as possible (20 ml) should be injected, as even if the CSF leak is small (up to 10 ml), this may represent up to a maximum of 30% haemodilution, allowing for uniform distribution of CSF. Further work is required to look at greater dilutions of whole blood with CSF using TEG to assess whether this has an effect on whole blood coagulation. The timing of epidural blood patching is also controversial. Several studies have shown improved outcomes if epidural blood patch is performed later, and combined results support postponement of epidural blood patch for at least 24–48 h after the onset of symptoms [29]. Paech et al. and more recently Kokki et al. supported the use of epidural blood patch more than 48 h after the onset of symptoms [6, 12]. We propose that the effect of CSF on blood may also affect the optimal timing for performance of an epidural blood patch. It may not be desirable to perform the patch when the leak is maximal, such as at the time of dural puncture, as this too may result in a greater than 30% haemodilution of the injected blood. Magnetic resonance imaging of extradural blood patches have demonstrated that the main bulk of the extradural clot extends on average three to five spinal segments from the site of injection, and the spread of injection is mainly cephalad [30]. Therefore, if epidural blood patch is to be performed, and the procoagulant
effect of the CSF is thought to be desirable, we suggest that epidural blood patch should be performed at the level of the dural puncture to ensure maximal effect of CSF on the injected blood. In summary, we have shown that haemodilution of whole blood with Hartmann’s solution or CSF led to both procoagulant and clot destabilising effects, and these increased with progressive haemodilution up to 30%. These effects were significantly greater with CSF than with Hartmann’s solution. Therefore, the addition of increasing amounts of CSF to whole blood, up to 30% dilution, speeds up both clotting activation and clot formation, but decreases clot strength. We therefore surmise that: there may be a reduced success rate performing epidural blood patches when CSF leak is greatest, such as at the time of dural puncture; the volume of blood to be injected should be as large as possible; and epidural blood patching should be performed at the level of the accidental dural puncture.
140
Acknowledgements PT was supported by an obstetric anaesthesia research fellowship grant from Smiths Medical, UK. RF was supported by the University College London Hospitals/ University College London Comprehensive Biomedical Research Centre, which receives a proportion of its funding from the National Institute of Health Research Biomedical Research Centre funding scheme.
Competing interests No conflicts of interest declared.
References 1. Scavone BM, Wong CA, Sullivan JT, Yaghmour E, Sherwani SS, McCarthy RJ. Efficacy of a prophylactic epidural blood patch in preventing post dural puncture headache in parturients after inadvertent dural puncture. Anesthesiology 2004; 101: 1422– 7. 2. Stocks GM, Wooller DJ, Young JM, Fernando R. Postpartum headache after epidural blood patch: investigation and diagnosis. British Journal of Anaesthesia 2000; 84: 407–10. 3. Vakharia SB, Thomas PS, Rosenbaum AE, Wasenko JJ, Fellows DG. Magnetic resonance imaging of cerebrospinal fluid leak and tamponade effect of blood patch in postdural puncture headache. Anesthesia and Analgesia 1997; 84: 585–90. 4. Siau C, Ng HP, Tan GM, Ho BS, Pua HL. In vitro effects of local anaesthetics on the thromboelastographic profile of parturients. British Journal of Anaesthesia 2005; 94: 117–20. 5. Crawford JS. Experiences with epidural blood patch. Anaesthesia 1980; 35: 513–5. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
6. Paech MJ, Doherty DA, Christmas T, Wong CA; Epidural Blood Patch Trial Group. The volume of blood for epidural blood patch in obstetrics: a randomized, blinded clinical trial. Anesthesia and Analgesia 2011; 113: 126–33. 7. Robinson LS, Gorton H, Columb MO, Lyons G. Thromboelastograhy: validation of TEG 3000 and TEG 5000. British Journal of Anaesthesia 2001; 86: 309P. 8. Sharma SK, Philip J, Wiley J. Thromboelastographic changes in healthy parturients and postpartum women. Anesthesia and Analgesia 1997; 85: 94–8. 9. Ruttmann TG, James MF, Wells KF. Effect of 20% in vitro haemodilution with warmed buffered salt solution and cerebrospinal fluid on coagulation. British Journal of Anaesthesia 1999; 82: 110–1. 10. Van de Velde M, Schepers R, Berends N, Vandermeersch E, De Buck F. Ten years of experience with accidental dural puncture and post-dural puncture headache in a tertiary obstetric anaesthesia department. International Journal of Obstetric Anesthesia 2008; 17: 329–35. 11. Buettner A, Popham P, Morgan D. Incidence of epidural blood patch following obstetric regional analgesia in private Australian anaesthetic practice. International Journal of Obstetric Anesthesia 2005; 14: 5–8. €vall S, Kein€anen M, Kokki H. The influence of 12. Kokki M, Sjo timing on the effectiveness of epidural blood patches in parturients. International Journal of Obstetric Anesthesia 2013; 22: 303–9. 13. Baraz R, Collis RE. The management of accidental dural puncture during labour epidural analgesia: a survey of UK practice. Anaesthesia 2005; 60: 673–9. 14. Longo S. Postdural puncture: implications and complications. Current Opinion in Anesthesiology 1999; 12: 271–5. 15. Bakshi R, Mechtler LL, Kamran S, et al. MRI findings in lumbar puncture headache syndrome: abnormal dural-meningeal and dural venous sinus enhancement. Clinical Imaging 1999; 23: 73–6. 16. Sechzer PH. Post-spinal anesthesia headache treated with caffeine. Part II: intracranial vascular distension. A key factor. Survey of Anesthesiology 1981; 25: 55–6. 17. Iqbal J, Davis LE, Orrison WW. An MRI study of lumbar puncture headaches. Headache 1995; 35: 420–2. 18. Lander CJ, Korbon GA. Histopathologic consequences of epidural blood patch and epidurally administered Dextran 40. Anesthesiology 1988; 69: A410.
19. DiGiovanni AJ, Galbert MW, Wahle WM. Epidural injection of autologous blood for postlumbar-puncture headache. II. Additional clinical experiences and laboratory investigation. Anesthesia and Analgesia 1972; 51: 226–32. 20. Turnbull DK, Shepherd DB. Post-dural puncture headache: pathogenesis, prevention and treatment. British Journal of Anaesthesia 2003; 91: 718–29. 21. Macafee B, Campbell JP, Ashpole K, et al. Reference ranges for thromboelastography (TEGâ) and traditional coagulation tests in term parturients undergoing caesarean section under spinal anaesthesia. Anaesthesia 2012; 67: 741–7. 22. Quarterman C, Shaw M, Johnson I, Agarwal S. Intra- and intercentre standardisation of thromboelastography (TEGâ). Anaesthesia 2014; 69: 883–90. 23. Cook MA, Watkins-Pitchford JM. Epidural blood patch: a rapid coagulation response. Anesthesia and Analgesia 1990; 70: 567–8. 24. Ruttmann TG, James MFM, Finlayson J. Effects on coagulation of intravenous crystalloid or colloid in patients undergoing peripheral vascular surgery. British Journal of Anaesthesia 2002; 89: 226–30. 25. Apfel CC, Saxena A, Cakmakkaya OS, Gaiser R, George E, Radke O. Prevention of postdural puncture headache after accidental dural puncture: a quantitative systematic review. British Journal of Anaesthesia 2010; 105: 255–63. 26. Bradbury CL, Singh SI, Badder SR, Wakely LJ, Jones PM. Prevention of post-dural puncture headache in parturients: a systematic review and meta-analysis. Acta Anaesthesiologica Scandinavica 2013; 57: 417–30. 27. Stein MH, Cohen S, Mohiuddin MA, Dombrovskiy V, Lowenwirt I. Prophylactic vs therapeutic blood patch for obstetric patients with accidental dural puncture-a randomised controlled trial. Anaesthesia 2014; 69: 320–6. 28. Roche AM, James MFM, Grocott MPW, Mythen MG. Coagulation effects of in vitro serial haemodilution with a balanced electrolyte hetastarch solution compared with a saline-based hetastarch solution and lactated Ringer’s solution. Anaesthesia 2002; 57: 950–5. P, et al. Effectiveness of 29. Safa-Tisseront V, Thormann F, Malassine epidural blood patch in the management of post-dural puncture headache. Anesthesiology 2001; 95: 334–9. 30. Beards SC, Jackson A, Griffiths AG, Horsman EL. Magnetic resonance imaging of extradural blood patches: appearances from 30 min to 18 h. British Journal of Anaesthesia 1993; 71: 182–8.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
141
doi:10.1111/anae.12911
Original Article The effect of serial in vitro haemodilution with maternal cerebrospinal fluid and crystalloid on thromboelastographic (TEGâ) blood coagulation parameters, and the implications for epidural blood patching* S. Armstrong,1 R. Fernando,2 P. Tamilselvan,3 A. Stewart2 and M. Columb4 1 2 3 4
Consultant, Consultant, Consultant, Consultant,
Department Department Department Department
of of of of
Anaesthesia, Anaesthesia, Anaesthesia, Anaesthesia,
Frimley Health NHS Foundation Trust, Frimley, UK University College London Hospitals NHS Trust, London, UK The Princess Alexandra Hospital, Harlow, UK South Manchester University Hospitals NHS Trust, Wythenshawe, UK
Summary Epidural blood patches may be used to treat post-dural puncture headache following accidental dural puncture in parturients. Their mode of action and the optimum volume of blood for injection remain controversial, with the interaction between injected blood and cerebrospinal fluid unknown. We aimed to establish the effects of serial haemodilution of whole blood with cerebrospinal fluid from 34 pregnant patients compared with serial haemodilution with Hartmann’s solution, using the thromboelastogram. Haemodilution with either cerebrospinal fluid or Hartmann’s solution had significant procoagulant and clot destabilising effects, enhanced with progressive haemodilution up to 30%. The effect of cerebrospinal fluid was greater compared with Hartmann’s solution (p < 0.001). Cerebrospinal fluid led to a mean (95% CI) decrease in r-time by 2.4 (1.6–3.2) min, a decrease in k-time by 0.6 (0.4–0.8) min, an increase in alpha angle by 7.3 (5.5–9.0)°, and a decrease in maximum amplitude by 2.0 (0.6–3.4) mm. This may have implications for epidural blood patch, as success may be reduced near the time of dural puncture when cerebrospinal fluid leak is at its greatest, and large volumes of blood may be required to reduce haemodilution and clot destabilisation by cerebrospinal fluid. In addition, blood patching should be performed at the level of the dural puncture in order to ensure that the maximum volume of blood comes into contact with the cerebrospinal fluid.
.................................................................................................................................................................
Correspondence to: R. Fernando Email: [email protected] *Presented in part at the annual meeting of the Society for Obstetric Anesthesiology and Perinatology, May 2005, California, USA, and at the annual meeting of the Obstetric Anaesthetists’ Association, May 2005, London, UK. This paper is accompanied by an editorial: Anaesthesia 2015; 70: 119–121. Accepted: 18 September 2014
Introduction It has been reported that up to 80% of parturients who have an accidental dural puncture during epidural insertion develop a post-dural puncture headache © 2014 The Association of Anaesthetists of Great Britain and Ireland
(PDPH) [1]. Characteristically, the headache occurs within 24 h of the dural puncture. In most cases, PDPH is moderate and self-limiting, but in some cases it may be severe and debilitating. There is also the 135
Anaesthesia 2015, 70, 135–141
Armstrong et al. | Haemodilution with CSF and coagulation
potential for PDPH to result in morbidity and even death. The injection of autologous blood into the epidural space (epidural blood patch) was initially suggested by Gormley in 1960, and remains the most effective treatment to date. Success rates of epidural blood patch to treat PDPH have been quoted to be as high as 90%, yet other studies have shown success rates of only 60–70% [2]. The effectiveness of the epidural blood patch is thought to be twofold: first, injecting blood into the epidural space causes movement of the dura anteriorly, resulting in an increase in cerebrospinal fluid (CSF) pressure in the subarachnoid space and immediate relief of symptoms [3]. Second, formation of a blood clot at and around the site of the dural puncture prevents any further CSF leak, giving longterm relief of symptoms [4]. Despite the fact that the epidural blood patch has been used as a valuable treatment for over 50 years, there are still many controversies regarding its application and effectiveness. The volume of blood to be injected is controversial; some authors advocate 8–10 ml, and others 15–20 ml, although 20 ml has now become the standard volume [5, 6]. Thromboelastography has been used as a monitor of whole blood coagulation in a variety of clinical settings, including obstetric anaesthesia [7]. It is easy to use, and reflects the adequacy of whole blood coagulation within 30–40 min, by delineating the interaction between platelets, fibrinogen and other clotting factors [8]. We postulated that increasing the degree of haemodilution by CSF in an epidural blood patch may increase both the firmness and speed of clot formation. This might be significant in establishing the effectiveness of the blood patch. The aim of this in-vitro study was therefore to assess the effects of serial haemodilution of whole blood with CSF from pregnant healthy patients on blood coagulation using the thromboelastogram (TEG).
pre-existing prothrombotic disorders; pregnancyrelated hypertensive disease; antiplatelet medication; height < 150 cm or > 180 cm; weight < 50 kg or > 100 kg; and macroscopic evidence of accidental epidural venous cannulation. A 16-G peripheral intravenous cannula was sited under local anaesthesia in a forearm vein. Initially, 2 ml blood was taken and discarded; following this, another 10 ml whole blood was collected, and subsequently the cannula was connected to 1000 ml intravenous Hartmann’s solution (Baxter, Thetford, Norfolk, UK) as part of the routine management for caesarean section. The blood sample was placed in a standard coagulation sample bottle containing buffered citrate (Vacutainer; Becton-Dickinson, Franklin Lakes, NJ, USA). All citrated samples were inverted five times and assessed within 60 min of sampling. Combined spinal-epidural anaesthesia was initiated in the sitting position, at the L3-4 or L4-5 interspace, using a locking kit (Portexâ Combined Spinal/Epidural Minipack with Lock Pencil Point Spinal Needle, Smiths Medical ASD Inc., Keene, NH, USA), consisting of a 16-G Tuohy needle and a 26-G pencil-point spinal needle. After puncturing the dura with the spinal needle, 1 ml CSF was collected by passive flow of drops into a graduated glass tube, before giving the standardised injection of 11 mg hyperbaric bupivacaine 0.5% and 15 ll fentanyl. Serial dilutions of whole blood with CSF were then carried out, calculated to produce final CSF dilutions of 10%, 20% and 30%. A 10% dilution represented 0.9 ml blood and 0.1 ml CSF; a 20% dilution represented 0.8 ml blood and 0.2 ml CSF; and a 30% dilution represented 0.7 ml blood and 0.3 ml CSF. Similar dilutions of whole blood with Hartmann’s solution at body temperature were used as controls. All the reconstituted blood samples were recalcified by mixing 0.33 ml blood with 0.3 ml calcium chloride 0.2 mol within the testing cup, warmed to 37 °C. Thromboelastography was performed on all the samples with the TEG 5000â (Hemoscope Corp., Niles, IL, USA). Quality control was regularly performed on the TEG by the local point-of-care testing department in accordance with the manufacturer’s guidelines. Using the TEG, r- and k-times, the alpha angle and the maximum amplitude (MA) were measured for
Methods After institutional ethical approval, written informed consent was obtained from 34 healthy term parturients. All were awaiting planned elective caesarean section under spinal anaesthesia at the Royal Free Hospital in London, UK. Exclusion criteria included: 136
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
Table 1 Thromboelastograph (TEG) terms of reference. r-Time k-Time
Clotting factor activity Clot formation time
Alpha angle
Platelet and fibrinogen activity Clot strength
MA
Time from initiation of the test to initial fibrin formation and pin movement Time from the beginning of clot formation until the amplitude of the TEG reaches 20 mm Angle between the line in the middle of the TEG tracing and the line tangential to the developing ‘body’ of the TEG Greatest amplitude on the TEG trace
MA, maximum amplitude.
each sample (Table 1). Thromboelastographic variables for serial blood-CSF dilutions were compared with the corresponding serial dilutions with Hartmann’s solution. Repeated measures ANOVA with Geisser-Greenhouse adjustment was used to examine the effects of dilution, diluent and interaction. Post-tests included linear and quadratic trends to examine for a doseresponse effect of dilution and Tukey–Kramer tests for paired comparisons. Significance was defined at p < 0.05 (two-sided). Data were analysed using Number Cruncher Statistical Systems (NCSS) 2004 (NCSS Inc., Kaysville, UT, USA). Sample size calculations were based on the report by Ruttmann et al., and it was estimated that 34 patients would be required to find within-subject differences of 25% (SD 45%) in the r-time, at 90% power between diluents and at 80% power with Bonferroni corrections for three dilutions [9].
The mean (SD) gestational age of the participants was 38 + 5 weeks (1 day). Dilution with CSF had increasing effect on all TEG variables with increasing amounts of CSF (Figs. 1–4). At 30% dilution, there were maximum procoagulant effects as seen by decreased r-time and ktime, and increased alpha angle (Table 2). Dilution with CSF also reduced MA, and therefore had a clot destabilising effect (Table 2). The effect of dilution with Hartmann’s solution was more variable; at 30% dilution, r-time, k-time and MA were reduced (Fig. 1). The effect of CSF was consistently greater than that of Hartmann’s solution (Fig. 1 and Table 2).
Discussion
Analyses were complete from all samples for the 34 patients. The mean (SD) age was 34.7 (4.6) years, height was 163.3 (5.2) cm, and weight 71.8 (10.1) kg.
In this study, we demonstrated that serial haemodilution with Hartmann’s solution reduced r- and k-times by up to 30%, demonstrating an increase in the coagulability of whole blood when measured using TEG. We found that progressive dilution of whole blood with CSF also increased the coagulability of whole blood. Analysis of the data show increased coagulability with CSF compared with Hartmann’s solution for all dilutions. This
Figure 1 Mean r-time for blood diluted with cerebrospinal fluid (CSF; s) or Hartmanns solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p = 0.015; diluent: p < 0.001.
Figure 2 Mean k-time for blood diluted with cerebrospinal fluid (CSF; s) or Hartmanns solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p = 0.02; diluent: p < 0.001.
Results
© 2014 The Association of Anaesthetists of Great Britain and Ireland
137
Anaesthesia 2015, 70, 135–141
Figure 3 Mean maximum amplitude for blood diluted with cerebrospinal fluid (CSF; s) or Hartmann’s solution (). Error bars are SD. Dilution: CSF p < 0.001, Hartmann’s solution p < 0.001; diluent: p = 0.005.
Figure 4 Alpha angle for blood diluted with cerebrospinal fluid (CSF; s) or Hartmann’s solution () Error bars are SD. Dilution: CSF p = 0.001, Hartmann’s solution p = 0.30; diluent: p < 0.001.
Armstrong et al. | Haemodilution with CSF and coagulation
suggests that the increase in coagulability seen with the addition of CSF cannot be attributed only to its dilutional effect on whole blood, but that addition of CSF has a procoagulant effect that is independent of dilution. There was also a significant increase in the alpha angle with CSF, but not with Hartmann’s solution, which would indicate that the presence of CSF increases the speed of formation of the clot. Both Hartmann’s solution and CSF significantly reduced the MA with progressive haemodilution, indicating that both Hartmann’s solution and CSF reduce the strength of the clot formed. Our findings suggest that, when performing an epidural blood patch, the presence of a CSF leak could speed up the activation of the coagulation pathway, but clot strength may be reduced. However, we found up to a 50% reduction in r- and k-time with the addition of CSF, whereas the reduction in MA was only 10%. This may explain why blood patches are so effective despite the slight loss of clot strength, and may also explain the recurrence of the PDPH, after initial resolution of symptoms following epidural blood patch. Post-dural puncture headache is an important iatrogenic cause of morbidity in obstetric anaesthesia. Although autologous epidural blood patch is the definitive treatment for PDPH and the failure of this procedure is relatively rare, it may not be effective in a
Table 2 Effects of dilution of whole blood with Hartmann’s solution (HS) or cerebrospinal fluid (CSF). Data are mean difference and percentage change with 95% CI. TEG variable
HS (0% vs 30% dilution)
CSF (0% vs 30% dilution)
HS vs CSF across 10%, 20% and 30% dilutions
r-Time; min
1.8 ( 3.2 to 0.4) 17% ( 30 to 4%) p = 0.004 0.5 ( 1.0 to 0.0) 20% ( 41 to 0%) p = 0.050 2.1 ( 6.5 to 2.3) 4% ( 4 to 11%) p = 0.600 5.3 ( 7.5 to 3.2) 8% ( 11 to 5) p < 0.001
5.1 ( 6.2 to 4.0) 47% ( 57 to 37%) p < 0.001 1.3 ( 1.7 to 0.8) 49% ( 66 to 32%) p < 0.001 10.4 (6.3–14.6) 17% (11–25%) p < 0.001 9.2 ( 12.2 to 6.2) 9% ( 12 to 6%) p < 0.001
2.4 ( 3.2 to 1.6) 27% ( 36 to 18%) p < 0.001 0.6 ( 0.8 to 0.4) 30% ( 39 to 21%) p < 0.001 7.3 (5.5–9.0) 12% (9–15%) p < 0.001 2.0 ( 3.4 to 0.6) 3% ( 5 to 1%) p = 0.005
k-Time; min
Alpha angle; °
MA; mm
Within-diluent p values are Tukey–Kramer tests for 0% vs 30% dilution. Between diluent p values across 10%, 20% and 30% dilutions are Geisser-Greenhouse adjusted. TEG, thromboelastography; MA, maximum amplitude. 138
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
significant proportion of patients [6, 10–12]. Baraz et al. undertook a postal questionnaire of UK obstetric units, and found that only 26% treated PDPH with an epidural blood patch as soon as it was diagnosed, whereas in 71% of units, the blood patch was only performed once conservative measures had failed [13]. There are two widely held theories, which are somewhat competing, to account for the onset of the headache. Firstly, CSF leaks from the subarachnoid space into the epidural space, facilitated by the subatmospheric pressure within this space, allowing the brain and its supporting anatomy to descend and cause traction on pain-sensitive intracranial structures and meninges. The second theory suggests that CSF leak from the subarachnoid space also results in loss of CSF volume and CSF pressure, causing intracranial hypotension, compensatory cerebral vasodilatation and migraine-like pain [14–16]. This theory is supported by the fact that some magnetic resonance imaging (MRI) studies have shown enhanced cerebral blood flow in PDPH [15]. However, MRI has not shown an association between the volume of CSF that escapes and the presence of PDPH [17]. Epidural blood patching is thought to reduce PDPH by two mechanisms. Firstly, compression of the intrathecal space with blood increases the subarachnoid pressure by forcing CSF cephalad. Secondly, it is thought that a fibrin clot forms over the dural puncture site, which prevents further CSF leak. It has been shown by MRI in humans that, at 7–13 h, there is clot resolution, leaving a thick layer of mature clot over the dorsal part of the thecal sac, and animal studies have demonstrated widespread fibroblastic activity and collagen formation after seven days [18–20]. The TEG gives a graphic display of the stages of whole blood clot formation, and in normal pregnancy it demonstrates a reduction in r- and k-times, and an increase in alpha angle and MA, in keeping with a hypercoagulable picture [21]. It should be borne in mind with any TEG study that there may be variability between r- and k-times undertaken even by the same operator, possibly due to variable mixing of blood and kaolin in the sample tube [22]. Previous studies on the use of the epidural blood patch for accidental dural puncture have discovered that CSF has procoagulant activity. Cook et al. per-
formed a controlled animal model study, which simulated the mixing of blood and CSF as in an epidural blood patch and recorded the coagulation of the CSFblood mixture using TEG [23]. They compared 360 ll blood with 180 ll blood mixed with 180 ll CSF. They found that the r-time, the k-time and r+k value all showed significant decreases when compared with unmixed blood. The alpha angle was measured, but not shown to differ significantly between groups. In this study, MA was not measured. Acceleration of the onset of coagulation and a stronger clot were both observed. Ruttmann et al. investigated the effect of a 20% haemodilution of an in-vitro blood sample with either a warmed physiological balanced salt solution (plasmalyte B), or CSF from parturients undergoing spinal anaesthesia for caesarean section using TEG. They demonstrated that haemodilution of whole blood enhanced coagulation as before, but also that CSF had a profound hypercoagulable effect, greater than that seen with plasmalyte B, supporting the presence of a procoagulant in CSF [9]. In our study, we demonstrated that serial haemodilution with Hartmann’s solution resulted in an increase in the coagulability of whole blood when measured using TEG. This would support work again done by Ruttmann et al., who demonstrated that haemodilution with 0.9% sodium chloride and other crystalloid and colloid solutions caused enhanced coagulation [24]. In their study, blood was taken from a peripheral cannula in patients undergoing vascular surgery following crystalloid or colloid administration. Exact percentages of haemodilution could not be established due to individual differences in total blood volume, although both groups showed a decrease in haematocrit of approximately 10%. We believe our study has greater significance in the context of epidural blood patch for accidental dural puncture. No other investigators have established serial blood-CSF dilutions and its effect on coagulation. Many studies have concentrated on the optimal volume of blood to be injected when performing an epidural blood patch, with the consensus suggesting a volume of 15–20 ml. Paech et al. concluded that anaesthetists should attempt to administer 20 ml autologous blood, limited by back pain, based on their
© 2014 The Association of Anaesthetists of Great Britain and Ireland
139
Anaesthesia 2015, 70, 135–141
Armstrong et al. | Haemodilution with CSF and coagulation
trial [6]. Some authors have advocated prophylactic epidural blood patch after accidental dural puncture, and a number of reviews have looked at this area, concluding that previous studies were flawed in their methodology and further large studies were required [25, 26]. One hundred and sixteen parturients were randomly assigned by Stein et al. to receive either prophylactic or therapeutic epidural blood patch with 15–20 ml autologous blood after accidental dural puncture [27]. They found that prophylactic blood patch decreased the incidence of PDPH and also decreased the intensity of the headache and accompanying symptoms. When considering the effect of CSF on whole blood, we only concentrated on a dilution of up to 30%. From previous work by Roche et al., with lactated Ringer’s solution, a procoagulant effect was seen with up to 40% dilution, yet further dilution resulted in a reduction in coagulation [28]. This may suggest that as large a volume of blood as possible (20 ml) should be injected, as even if the CSF leak is small (up to 10 ml), this may represent up to a maximum of 30% haemodilution, allowing for uniform distribution of CSF. Further work is required to look at greater dilutions of whole blood with CSF using TEG to assess whether this has an effect on whole blood coagulation. The timing of epidural blood patching is also controversial. Several studies have shown improved outcomes if epidural blood patch is performed later, and combined results support postponement of epidural blood patch for at least 24–48 h after the onset of symptoms [29]. Paech et al. and more recently Kokki et al. supported the use of epidural blood patch more than 48 h after the onset of symptoms [6, 12]. We propose that the effect of CSF on blood may also affect the optimal timing for performance of an epidural blood patch. It may not be desirable to perform the patch when the leak is maximal, such as at the time of dural puncture, as this too may result in a greater than 30% haemodilution of the injected blood. Magnetic resonance imaging of extradural blood patches have demonstrated that the main bulk of the extradural clot extends on average three to five spinal segments from the site of injection, and the spread of injection is mainly cephalad [30]. Therefore, if epidural blood patch is to be performed, and the procoagulant
effect of the CSF is thought to be desirable, we suggest that epidural blood patch should be performed at the level of the dural puncture to ensure maximal effect of CSF on the injected blood. In summary, we have shown that haemodilution of whole blood with Hartmann’s solution or CSF led to both procoagulant and clot destabilising effects, and these increased with progressive haemodilution up to 30%. These effects were significantly greater with CSF than with Hartmann’s solution. Therefore, the addition of increasing amounts of CSF to whole blood, up to 30% dilution, speeds up both clotting activation and clot formation, but decreases clot strength. We therefore surmise that: there may be a reduced success rate performing epidural blood patches when CSF leak is greatest, such as at the time of dural puncture; the volume of blood to be injected should be as large as possible; and epidural blood patching should be performed at the level of the accidental dural puncture.
140
Acknowledgements PT was supported by an obstetric anaesthesia research fellowship grant from Smiths Medical, UK. RF was supported by the University College London Hospitals/ University College London Comprehensive Biomedical Research Centre, which receives a proportion of its funding from the National Institute of Health Research Biomedical Research Centre funding scheme.
Competing interests No conflicts of interest declared.
References 1. Scavone BM, Wong CA, Sullivan JT, Yaghmour E, Sherwani SS, McCarthy RJ. Efficacy of a prophylactic epidural blood patch in preventing post dural puncture headache in parturients after inadvertent dural puncture. Anesthesiology 2004; 101: 1422– 7. 2. Stocks GM, Wooller DJ, Young JM, Fernando R. Postpartum headache after epidural blood patch: investigation and diagnosis. British Journal of Anaesthesia 2000; 84: 407–10. 3. Vakharia SB, Thomas PS, Rosenbaum AE, Wasenko JJ, Fellows DG. Magnetic resonance imaging of cerebrospinal fluid leak and tamponade effect of blood patch in postdural puncture headache. Anesthesia and Analgesia 1997; 84: 585–90. 4. Siau C, Ng HP, Tan GM, Ho BS, Pua HL. In vitro effects of local anaesthetics on the thromboelastographic profile of parturients. British Journal of Anaesthesia 2005; 94: 117–20. 5. Crawford JS. Experiences with epidural blood patch. Anaesthesia 1980; 35: 513–5. © 2014 The Association of Anaesthetists of Great Britain and Ireland
Armstrong et al. | Haemodilution with CSF and coagulation
Anaesthesia 2015, 70, 135–141
6. Paech MJ, Doherty DA, Christmas T, Wong CA; Epidural Blood Patch Trial Group. The volume of blood for epidural blood patch in obstetrics: a randomized, blinded clinical trial. Anesthesia and Analgesia 2011; 113: 126–33. 7. Robinson LS, Gorton H, Columb MO, Lyons G. Thromboelastograhy: validation of TEG 3000 and TEG 5000. British Journal of Anaesthesia 2001; 86: 309P. 8. Sharma SK, Philip J, Wiley J. Thromboelastographic changes in healthy parturients and postpartum women. Anesthesia and Analgesia 1997; 85: 94–8. 9. Ruttmann TG, James MF, Wells KF. Effect of 20% in vitro haemodilution with warmed buffered salt solution and cerebrospinal fluid on coagulation. British Journal of Anaesthesia 1999; 82: 110–1. 10. Van de Velde M, Schepers R, Berends N, Vandermeersch E, De Buck F. Ten years of experience with accidental dural puncture and post-dural puncture headache in a tertiary obstetric anaesthesia department. International Journal of Obstetric Anesthesia 2008; 17: 329–35. 11. Buettner A, Popham P, Morgan D. Incidence of epidural blood patch following obstetric regional analgesia in private Australian anaesthetic practice. International Journal of Obstetric Anesthesia 2005; 14: 5–8. €vall S, Kein€anen M, Kokki H. The influence of 12. Kokki M, Sjo timing on the effectiveness of epidural blood patches in parturients. International Journal of Obstetric Anesthesia 2013; 22: 303–9. 13. Baraz R, Collis RE. The management of accidental dural puncture during labour epidural analgesia: a survey of UK practice. Anaesthesia 2005; 60: 673–9. 14. Longo S. Postdural puncture: implications and complications. Current Opinion in Anesthesiology 1999; 12: 271–5. 15. Bakshi R, Mechtler LL, Kamran S, et al. MRI findings in lumbar puncture headache syndrome: abnormal dural-meningeal and dural venous sinus enhancement. Clinical Imaging 1999; 23: 73–6. 16. Sechzer PH. Post-spinal anesthesia headache treated with caffeine. Part II: intracranial vascular distension. A key factor. Survey of Anesthesiology 1981; 25: 55–6. 17. Iqbal J, Davis LE, Orrison WW. An MRI study of lumbar puncture headaches. Headache 1995; 35: 420–2. 18. Lander CJ, Korbon GA. Histopathologic consequences of epidural blood patch and epidurally administered Dextran 40. Anesthesiology 1988; 69: A410.
19. DiGiovanni AJ, Galbert MW, Wahle WM. Epidural injection of autologous blood for postlumbar-puncture headache. II. Additional clinical experiences and laboratory investigation. Anesthesia and Analgesia 1972; 51: 226–32. 20. Turnbull DK, Shepherd DB. Post-dural puncture headache: pathogenesis, prevention and treatment. British Journal of Anaesthesia 2003; 91: 718–29. 21. Macafee B, Campbell JP, Ashpole K, et al. Reference ranges for thromboelastography (TEGâ) and traditional coagulation tests in term parturients undergoing caesarean section under spinal anaesthesia. Anaesthesia 2012; 67: 741–7. 22. Quarterman C, Shaw M, Johnson I, Agarwal S. Intra- and intercentre standardisation of thromboelastography (TEGâ). Anaesthesia 2014; 69: 883–90. 23. Cook MA, Watkins-Pitchford JM. Epidural blood patch: a rapid coagulation response. Anesthesia and Analgesia 1990; 70: 567–8. 24. Ruttmann TG, James MFM, Finlayson J. Effects on coagulation of intravenous crystalloid or colloid in patients undergoing peripheral vascular surgery. British Journal of Anaesthesia 2002; 89: 226–30. 25. Apfel CC, Saxena A, Cakmakkaya OS, Gaiser R, George E, Radke O. Prevention of postdural puncture headache after accidental dural puncture: a quantitative systematic review. British Journal of Anaesthesia 2010; 105: 255–63. 26. Bradbury CL, Singh SI, Badder SR, Wakely LJ, Jones PM. Prevention of post-dural puncture headache in parturients: a systematic review and meta-analysis. Acta Anaesthesiologica Scandinavica 2013; 57: 417–30. 27. Stein MH, Cohen S, Mohiuddin MA, Dombrovskiy V, Lowenwirt I. Prophylactic vs therapeutic blood patch for obstetric patients with accidental dural puncture-a randomised controlled trial. Anaesthesia 2014; 69: 320–6. 28. Roche AM, James MFM, Grocott MPW, Mythen MG. Coagulation effects of in vitro serial haemodilution with a balanced electrolyte hetastarch solution compared with a saline-based hetastarch solution and lactated Ringer’s solution. Anaesthesia 2002; 57: 950–5. P, et al. Effectiveness of 29. Safa-Tisseront V, Thormann F, Malassine epidural blood patch in the management of post-dural puncture headache. Anesthesiology 2001; 95: 334–9. 30. Beards SC, Jackson A, Griffiths AG, Horsman EL. Magnetic resonance imaging of extradural blood patches: appearances from 30 min to 18 h. British Journal of Anaesthesia 1993; 71: 182–8.
© 2014 The Association of Anaesthetists of Great Britain and Ireland
141
