- Email: [email protected]
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS
British Journal of Anaesthesia 1993; 71: 854-857
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS P. MEYER, D. RENIER, E. ARNAUD, M.-M. JARREAU, B. CHARRON, E. BUY, C. BUISSON AND G. BARRIER
SUMMARY
KEY WORDS Blood: loss, transfusion. Surgery: paediatric neurosurgery.
Premature closure of skull sutures is a relatively common disorder, with an estimated incidence of 1 per 1000 births [1, 2]. In the absence of early surgical correction, compensatory deformation of the skull occurs during rapid growth of the brain in early infancy and this may lead to increased intracranial pressure, especially in children aged more than 1 yr, particularly when multiple sutures are involved [3]. With single suture involvement (the most frequent situation), effects on brain development are variable and there are mostly cosmetic problems [3, 4]. The goal of surgery is to restore normal anatomy at an early age in order to achieve the best cosmetic result and avoid possible cerebral consequences [1,5]. Surgical procedures for correction of craniosynostosis are performed, therefore, in young infants with a small blood volume and represent major craniofacial surgery with unavoidable and extensive blood loss. Accurate determination and precise restoration of blood losses are important during and after operation. Kearney, Rosales and Howes [6] described an elegant method of assessing per-
PATIENTS AND METHODS
After obtaining parents' consent and Institutional Board approval, from January 1988 to January 1990 we studied patients undergoing correction of craniosynostosis. A note was made of age and weight at operation, perinatal history, type of craniosynostosis and associated anomalies or dysmorphic syndromes, type of surgical procedures performed, duration of surgery, peroperative monitoring, peroperative complications and duration of assisted ventilation after operation. The surgical procedure depended upon the type of synostosis and the age of the patient. Strip craniectomies, often with associated posterior craniectomy, were the most simple procedures used only in very young children with isolated sagittal suture involvement (scaphocephaly); other procedures included forehead reconstruction, floating forehead and complex vault remodelling. All the procedures were performed under general anaesthesia without induced hypotension after skull infiltration with adrenaline 1:200000 in normal saline. Anaesthesia was maintained using continuous i.v. infusion of alfentanil and controlled ventilation with 0.8-1.5% isoflurane and 50% nitrous oxide in oxygen. Monitoring included continuous ECG, core temperature, non-invasive automated arterial pressure and end-tidal carbon dioxide analysis. Invasive monitoring (arterial or central venous pressure) was used if necessary. Preoperative blood samples were obtained for red blood cell and platelet count, measurement of haemoglobin and PCV and coagulation screen. Serial blood samples were obtained 2-hourly during operation and 12 h after operation and as frequently as clinically indicated, for measurement of blood-gas tensions and PCV. Intraoperative management included strict observance of an isovolaemic haemodilution regimen, PHILIPPE MEYER, M.D., MARIE-MADELEINE JARREAU, M.D., BRIGITTE CHARRON, M.D., ERIC BUY, M.D., CHRISTIANE BUISSON, M.D., GENEVIEVE BARRIER, M.D. (Department of Paediatric Anaesthesiology); DOMINIQUE RENIER, M.D., ERIC ARNAUD, M.D.
(Department of Paediatric Neurosurgery); Hopital des EnfantsMalades, 149 rue de Sevres, 75015 Paris, France. Accepted for Publication: June 3, 1993.
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
Surgical repair of craniosynostosis carries a high risk with large blood losses. Over a 2-yr period, we have managed 115 patients undergoing craniosynostosis repair with peroperative haemodilution to achieve a final PCV of 0.28-0.35. Measurements of PCV allowed calculation of estimated blood losses and transfused volumes in terms of red blood cell mass. Total estimated red cell volume lost was 91 ±66% of patient's estimated red blood cell volume during the peroperative period. The type of skull deformation and surgical procedure determined the extent of peroperative bleeding. Peroperative transfusion was satisfactory in 48% of patients and slight overtransfusion was noted in 32%. During the postoperative period, liberal administration of blood led to overtransfusion and possibly unnecessary transfusion in 74% of patients. Because of the well known risks of transmission of infectious disease, strict volume compensation with development of haemodilution and autotransfusion procedures should be used to limit these risks. (Br. J. Anaesth. 1993; 7 1 : 854-857)
operative blood loss in this type of procedure. The aim of this study was to evaluate blood loss and adequacy of transfusion practices in a large series of patients undergoing primary correction of craniosynostosis.
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS
855
TABLE I. Nature of skull deformations, and surgical procedures
Type of skull deformation Scaphocephaly Plagiocephaly Trigonocephaly Oxycephaly Brachycephaly Isolated Crouzon syndrome Apert syndrome Complex
Surgical procedures 49(43%) 21(18%) 10(9%) 9(8%) 10(9%) 6(5%) 5(4%) 5(4%)
RESULTS
In a 2-yr period, we studied 130 patients treated by two surgeons; 15 were subsequently eliminated from analysis because of inadequate data. All patients had primary craniosynostosis repair, mean age at operation was 11.6 months (range 1-154 months; 77.3% of the population less than 12 months old) and mean body weight 8.63 kg (range 4—44 kg;
57(50%)
Strip craniectomy
31(27%)
Floating forehead
18(15%)
Complex
9(8%)
70.4% less than 10 kg). The type of skull deformation and surgical procedure are shown in table I. As in other series, isolated sagittal synostosis resulting in scaphocephaly was the most common type, representing 42.6% of cases [2]. Invasive monitoring was used in 70 patients (60%), arterial cannula alone in 47, central venous catheter alone in three and both arterial and central venous monitoring in 20 patients. Ninety percent of the patients had arterial cannulae during the last year of the study. Duration of surgery differed significantly between surgical procedures, with a mean of 79.4 (SD 33.9) min for strip craniectomies and 137 (41.4) min for more complex procedures. Total estimated red blood cell volume lost during the peroperative period was 91 (66)% of preoperative estimated red cell volume (range 5-400%). ERCV, RCT, ERCD, ERCL and percent of ERCV lost are presented in table II. Only 13 patients (11.3%) undergoing simple surgical procedures did not require peroperative blood transfusion. All other children (88.7%) had intraoperative transfusion with a mean 243.8 (181.4) ml (range 50-900 ml) of packed red blood cells. Fresh frozen plasma was used in only 56 children (48.7%). According to our criteria, 48.7 % of the patients were adequately transfused, 32% were slightly overtransfused and 19% were undertransfused. One patient died during surgery after massive bleeding from a large venous sinus. Venous air embolism occurred in three patients (2.6 % incidence), without further consequences. Transient hypotension was the most frequent peroperative incident, with a 35 % incidence. At the end of surgery, the trachea was extubated in 68.5 % of children. The determinant factors for postoperative assisted ventilation were peroperative blood loss and duration of surgery. Duration of postoperative ventilation was 40-540 min. In the first 24 h after surgery, a mean 232.13 (148.8) ml of blood was lost through surgical drains (table II). The PCV in drainage fluid was, in most instances, less than 0.15. Overtransfusion, according to our criteria, was more frequent in the postoperative period than during surgery (74% vs 32% (P < 0.001)), leading to possible unnecessary transfusion in some patients. Postoperative complications included transient hypotension in 15% of patients, persistent bleeding in excess of 50 % of estimated red cell volume in 12.6% and thrombocytopenia (platelet count less than 65000 ml"1) in three patients. Postoperative haemostasis was judged satisfactory with an activated
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
with fluid replacement, based upon haemodynamic variables and comprising human serum albumin and packed red blood cells in order to maintain PCV in the range 0.28-0.35—a value considered generally as adequate for oxygen transport and tissue delivery [7]. Fresh frozen plasma was used only in patients requiring more than 70 % of estimated blood volume transfusion. At the end of surgery, anaesthesia was discontinued and the trachea extubated or the patient sedated lightly with assisted ventilation. The criteria used for extubation were: rapid recovery with stable spontaneous ventilation, haemodynamic stability and normothermia, short-duration procedure with relatively small blood loss and absence of continuing bleeding via surgical drains. Patients were observed closely for at least 12 h. Full blood cell count and coagulation screen were obtained in the postoperative period in all patients. The type, volume and timing of transfused blood products were noted. In an attempt to simplify estimation of blood loss, all estimated volumes referred to red cell volumes; fresh frozen plasma volumes were analysed separately. Estimated body blood volume (EBV) = 80 ml kg"1 and red cell volume (ERCV) = EBV x PCV were calculated before operation. The following variables were then calculated for the periods during and after operation: estimated red cell volume transfused (RCT) = 0.75 x volume of packed red cells; estimated red cell volume deficit (ERCD) = ERCV x PCV variation; estimated red cell volume lost (ERCL) = ERCD + RCT; percent of ERCV lost = ERCL/ERCV. Data were statistically analysed using the Statistical Package for the Social Sciences (SPSS/PC+ , MJ Norusis Ed. SPSS Inc. Chicago 1992). Tabular data were analysed using contingency tables (chisquare or Fischer's exact test) and differences of means using Student's t test, ANOVA or KruskalWallis, as appropriate, depending on distribution and number of variables. Statistical significance was assumed with P < 0.05.
Forehead reconstruction
BRITISH JOURNAL OF ANAESTHESIA
856
TABLE II. Mean (SO) blood loss during the study. ERCL = estimated red cell volume loss; RCT = packed red cell volume transfused; ERCL/ERCV = % of patient's estimated red cell volume lost; ERCD = final red cell volume deficit. ^Negative value indicates postoperative overtransfusion
ERCL (ml) RCT (ml) ERCL/ERCV (%) ERCD (ml)
During op.
After op.
Total
191.5(124.63) 182.87 (136.48) 91 (65.9) 6.15(41.83)
52.75 (80.28) 89.85 (82.82) 27.4 (42.3) -37.88(48.14)+
240.72 (146.09) 272.72 (137.6) 118.2(86.1) -32(42.2)
TABLE III. Mean (SD) percentage of patient's estimated red cell volume lost and type of skull deformation. ^Negative value indicates postoperative overtransfusion During op.
After op.
Total
Oxycephaly Plagiocephaly Trigonocephaly Brachycephaly Scaphocephaly Complex
49.1 (22.9) 59 (37.4) 92.4 (49.7) 105.3 (48.45) 92.1 (65.2) 198.5 (165)
-3.7(16.8)t 27.7(41.7) 11.7(16.6) 25.5 (56.5) 35.9 (38.4) 44.5 (97.6)
45.4 (36.9) 86.7 (56.2) 104.1 (49.2) 130.9(69.1) 121.7(78.2) 243.1 (259.4)
prothrombin time of 60% or more of the normal value in all patients. No change in PCV was noted 24 h after surgery; no further transfusion was required and PCV at discharge from the hospital was always 0.30 or more. The age of the patient correlated highly with blood loss (99 (73) % of ERCBV lost in patients younger than 6 months, compared with 72 (38) % lost in patients older than 6 months (P < 0.008)). The type of skull deformation correlated with the percentage of estimated red cell volume lost during surgery (P < 0.0001) and there was a correlation between type of surgical procedure and peroperative (P < 0.012), but not postoperative, blood loss (table III). Separate analysis of the correlation between blood loss and type of surgical procedure was performed in the subgroup of scaphocephaly, for which three types of procedures could be used. Strip craniectomies involved less blood loss (60 (24.5)% of ERCV lost) than forehead reconstruction (106.4 (53.7)%) and complex vault remodelling (170.6 (10.3)%) (P < 0.006). DISCUSSION
Craniofacial surgery for craniosynostosis treatment includes extensive osteotomies with considerable blood loss [2, 5, 6, 8-12]. Because of the well known complications of blood transfusion, with a current incidence of about 1 per 40000 and 5-10% of transfusion recipients being at risk of AIDS and hepatitis [13, 14], methods designed to reduce blood loss are of interest. Induced hypotension has been used in the past without clear benefit [10]. Skin infiltration with vasoconstrictors is used frequently, but is not very effective because periosteum and bone, not the scalp, are the major sources of blood loss [8]. Acute normovolaemic haemodilution is now used routinely in many paediatric surgical procedures [7], but not, to our knowledge, during craniosynostosis surgery. In this study, using strict guidelines for transfusion, we found no complications. However, it must be noted that about 10% of the patients could be
satisfactorily managed without any packed red cell transfusion. Acute haemodilution alone is probably insufficient to avoid blood transfusion in this type of surgery and other techniques designed to reduce the need for homologous transfusion should probably be considered. Because of the small blood volume of young children, autologous transfusion with preoperative blood withdrawal is clearly not adequate to provide blood requirements during surgery. Peroperative blood salvage and retransfusion have been reported anecdotally during craniosynostosis repair [15]. In our recent experience with peroperative use of a cell saver in 15 patients, in spite of satisfactory blood salvage and significant reduction in homologous transfusion requirements, abrupt haemorrhage in patients with a small blood volume required associated use of homologous transfusion in nine patients. The factors influencing peroperative blood loss have been analysed previously. In this type of surgery, trapping of large blood volumes in surgical drapes and dilution with irrigating fluids prevent precise measurement of blood loss. There are few published estimates of blood loss during these procedures; most refer to weighing sponges and measurement of suction volume [2, 8-10]. Kearney, Rosales and Howes reported an interesting method of assessing blood loss by serial determinations of PCV and calculation of red cell volume deficit. This method allows more precise evaluation of changes in blood volume and assessment of adequacy of transfusion practices [6]. Our blood losses were slightly greater than the reported values of 20-40% of estimated blood volume lost [6,9-11] and may be explained partly by differences in surgical procedures, those in our series being more complex. Forehead reconstructions and craniectomies with posterior flaps are associated with significantly more blood loss than strip craniectomies. As in other reports, we found that strip craniectomies had the smallest estimated blood loss. Because of the lack of details on surgical techniques in other series [6, 8-10] and the changes in surgical practice over a 15-yr
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
Type of synostosis
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS
istical correlations between estimated postoperative blood loss and determinants such as type of synostosis and surgical procedure. REFERENCES 1. Marchac D, Renier D. Classification of the different types of craniosynostosis. In: Marchac D, Renier D, eds. Craniofacial Surgery for Craniosynostosis. Boston: Little Brown, 1982; 1-8. 2. Shilito J, Matson DD. Craniosynostosis: a survey of 204 cases. Journal of Neurosurgery 1968; 22: 229—240. 3. Renier D. Intracranial pressure in craniosynostosis: pre- and postoperative recordings, correlation with functional results. In: Edgerton MT, Jane SA, eds. Scientific Foundation and Surgical Treatment of Craniosynostosis. Baltimore: Williams and Wilkins, 1989; 263-269. 4. Van Effentere R, Houtteville JP, Philippon J. Interet de la mesure de la pression intra-cranienne dans les craniostenoses decouvertes apres l'age d'un an. Neurochirurgie 1989; 22: 59-67. 5. Loftness SL, Albin RE, O'Donnell RS. Perioperative management of craniofacial surgery in infants and children. In: Marchac D, ed. Craniofacial Surgery. Berlin: Springer Verlag, 1985; 470-173. 6. Kearney RA, Rosales JK, Howes WJ. Craniosynostosis: an assessment of blood loss and transfusion practices. Canadian Journal of Anaesthesia 1989; 36: 473-477. 7. Salem MR, Bikhazi GB. Blood conservation. In: Motoyama EK, Davis PJ, eds. Smith's Anesthesia for Infants and Children. St Louis: C. V. Mosby Co., 1990; 371-392. 8. Hans JF, Trusso R, Levy WJ. Craniosynosectomy with reduced blood loss. Neurosurgery 1981; 8: 209-211. 9. Scholtes JL. Craniofaciosynostosis: anaesthetic and perioperative management. Acta Anaesthesiologica Belgica 1985; 3: 176-185. 10. Diaz JH, Lockhart CH. Hypotensive anaesthesia for craniectomy in infancy. British Journal of Anaesthesia 1979; 51: 233-235. 11. McKersie AM. Paediatric neurosurgery and craniofacial surgery. In: Jewkes DA, ed. Anaesthesia for Neurosurgery, Clinical Anaesthesiology. International Practice and Research. London: Bailliere Tindall, 1987; 70-79. 12. Davies DW, Munro IR. The anesthestic management and intra-operative care of patients undergoing major facial osteotomies. Plastic and Reconstructive Surgery 1975; 55: 50-56. 13. Bove JR. Transfusion-associated hepatitis and AIDS. New England Journal of Medicine 1987; 317: 242-245. 14. Zuck TF. Transfusion-transmitted AIDS reassessed. New England Journal of Medicine 1988; 318: 511-512. 15. Paccagnella F, Longatti GP, Mazzoon D. Emotransfusione autologa nella chirurgia delle craniostenosi: un obiettivo. In: Proceedings of 5 Anuale Riunione Italo-Francese di Neuroanestesia-Reanimazione, Torino, 1988.
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
period [6], comparisons are difficult. However, it is easier to compare our study with the report from the Children's Hospital at Denver, where extensive craniectomies and complex vault reconstruction represent the third most common neurosurgical procedure performed in infancy [5]. Our results were broadly comparable to those reported in that study [5], especially for brachycephaly and plagiocephaly treated with forehead reconstruction. The duration of surgery varied between different types of skull deformation and different surgical procedures. It did not influence significantly the extent of operative blood loss. In these procedures, 20-50 % of the patient's estimated red cell volume may be lost in less than 30 min. For complex procedures the mean duration was similar to that of another series [6], but less than the duration observed by Loftness, Albin and O'Donnell [5]. The age of patients at the time of surgery was associated with blood loss, with a greater percentage of ERBCV lost in infants younger than 6 months. However, it must be noted that cosmetic and neurological outcome are more important in this age group [5, 6]. In our institution, the optimal age for surgery represents a compromise between surgical conditions with easier and shorter operations in younger children, and physiological tolerance to acute blood loss with greater haemodynamic stability in children older than 6 months. Thus most of our children were aged 6-12 months, with body weight 6-10 kg. It appeared that postoperative management did not achieve the goal of controlled haemodilution and attempts to modify transfusion practice during the recovery period are now in place. This difference in management during and after operation, with possibly unnecessary transfusion during the postoperative period, was noted also by Kearney, Rosales and Howes, who did not use deliberate peroperative haemodilution [6]. This demonstrates clearly the inadequacy of transfusion regimens dealing more with measurement of blood drainage than with physiological observation and measurement of PCV. Another shortcoming of postoperative management in this study was the difficulty in establishing correlations between estimated blood losses and their principal determinants. There were no stat-
857
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS P. MEYER, D. RENIER, E. ARNAUD, M.-M. JARREAU, B. CHARRON, E. BUY, C. BUISSON AND G. BARRIER
SUMMARY
KEY WORDS Blood: loss, transfusion. Surgery: paediatric neurosurgery.
Premature closure of skull sutures is a relatively common disorder, with an estimated incidence of 1 per 1000 births [1, 2]. In the absence of early surgical correction, compensatory deformation of the skull occurs during rapid growth of the brain in early infancy and this may lead to increased intracranial pressure, especially in children aged more than 1 yr, particularly when multiple sutures are involved [3]. With single suture involvement (the most frequent situation), effects on brain development are variable and there are mostly cosmetic problems [3, 4]. The goal of surgery is to restore normal anatomy at an early age in order to achieve the best cosmetic result and avoid possible cerebral consequences [1,5]. Surgical procedures for correction of craniosynostosis are performed, therefore, in young infants with a small blood volume and represent major craniofacial surgery with unavoidable and extensive blood loss. Accurate determination and precise restoration of blood losses are important during and after operation. Kearney, Rosales and Howes [6] described an elegant method of assessing per-
PATIENTS AND METHODS
After obtaining parents' consent and Institutional Board approval, from January 1988 to January 1990 we studied patients undergoing correction of craniosynostosis. A note was made of age and weight at operation, perinatal history, type of craniosynostosis and associated anomalies or dysmorphic syndromes, type of surgical procedures performed, duration of surgery, peroperative monitoring, peroperative complications and duration of assisted ventilation after operation. The surgical procedure depended upon the type of synostosis and the age of the patient. Strip craniectomies, often with associated posterior craniectomy, were the most simple procedures used only in very young children with isolated sagittal suture involvement (scaphocephaly); other procedures included forehead reconstruction, floating forehead and complex vault remodelling. All the procedures were performed under general anaesthesia without induced hypotension after skull infiltration with adrenaline 1:200000 in normal saline. Anaesthesia was maintained using continuous i.v. infusion of alfentanil and controlled ventilation with 0.8-1.5% isoflurane and 50% nitrous oxide in oxygen. Monitoring included continuous ECG, core temperature, non-invasive automated arterial pressure and end-tidal carbon dioxide analysis. Invasive monitoring (arterial or central venous pressure) was used if necessary. Preoperative blood samples were obtained for red blood cell and platelet count, measurement of haemoglobin and PCV and coagulation screen. Serial blood samples were obtained 2-hourly during operation and 12 h after operation and as frequently as clinically indicated, for measurement of blood-gas tensions and PCV. Intraoperative management included strict observance of an isovolaemic haemodilution regimen, PHILIPPE MEYER, M.D., MARIE-MADELEINE JARREAU, M.D., BRIGITTE CHARRON, M.D., ERIC BUY, M.D., CHRISTIANE BUISSON, M.D., GENEVIEVE BARRIER, M.D. (Department of Paediatric Anaesthesiology); DOMINIQUE RENIER, M.D., ERIC ARNAUD, M.D.
(Department of Paediatric Neurosurgery); Hopital des EnfantsMalades, 149 rue de Sevres, 75015 Paris, France. Accepted for Publication: June 3, 1993.
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
Surgical repair of craniosynostosis carries a high risk with large blood losses. Over a 2-yr period, we have managed 115 patients undergoing craniosynostosis repair with peroperative haemodilution to achieve a final PCV of 0.28-0.35. Measurements of PCV allowed calculation of estimated blood losses and transfused volumes in terms of red blood cell mass. Total estimated red cell volume lost was 91 ±66% of patient's estimated red blood cell volume during the peroperative period. The type of skull deformation and surgical procedure determined the extent of peroperative bleeding. Peroperative transfusion was satisfactory in 48% of patients and slight overtransfusion was noted in 32%. During the postoperative period, liberal administration of blood led to overtransfusion and possibly unnecessary transfusion in 74% of patients. Because of the well known risks of transmission of infectious disease, strict volume compensation with development of haemodilution and autotransfusion procedures should be used to limit these risks. (Br. J. Anaesth. 1993; 7 1 : 854-857)
operative blood loss in this type of procedure. The aim of this study was to evaluate blood loss and adequacy of transfusion practices in a large series of patients undergoing primary correction of craniosynostosis.
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS
855
TABLE I. Nature of skull deformations, and surgical procedures
Type of skull deformation Scaphocephaly Plagiocephaly Trigonocephaly Oxycephaly Brachycephaly Isolated Crouzon syndrome Apert syndrome Complex
Surgical procedures 49(43%) 21(18%) 10(9%) 9(8%) 10(9%) 6(5%) 5(4%) 5(4%)
RESULTS
In a 2-yr period, we studied 130 patients treated by two surgeons; 15 were subsequently eliminated from analysis because of inadequate data. All patients had primary craniosynostosis repair, mean age at operation was 11.6 months (range 1-154 months; 77.3% of the population less than 12 months old) and mean body weight 8.63 kg (range 4—44 kg;
57(50%)
Strip craniectomy
31(27%)
Floating forehead
18(15%)
Complex
9(8%)
70.4% less than 10 kg). The type of skull deformation and surgical procedure are shown in table I. As in other series, isolated sagittal synostosis resulting in scaphocephaly was the most common type, representing 42.6% of cases [2]. Invasive monitoring was used in 70 patients (60%), arterial cannula alone in 47, central venous catheter alone in three and both arterial and central venous monitoring in 20 patients. Ninety percent of the patients had arterial cannulae during the last year of the study. Duration of surgery differed significantly between surgical procedures, with a mean of 79.4 (SD 33.9) min for strip craniectomies and 137 (41.4) min for more complex procedures. Total estimated red blood cell volume lost during the peroperative period was 91 (66)% of preoperative estimated red cell volume (range 5-400%). ERCV, RCT, ERCD, ERCL and percent of ERCV lost are presented in table II. Only 13 patients (11.3%) undergoing simple surgical procedures did not require peroperative blood transfusion. All other children (88.7%) had intraoperative transfusion with a mean 243.8 (181.4) ml (range 50-900 ml) of packed red blood cells. Fresh frozen plasma was used in only 56 children (48.7%). According to our criteria, 48.7 % of the patients were adequately transfused, 32% were slightly overtransfused and 19% were undertransfused. One patient died during surgery after massive bleeding from a large venous sinus. Venous air embolism occurred in three patients (2.6 % incidence), without further consequences. Transient hypotension was the most frequent peroperative incident, with a 35 % incidence. At the end of surgery, the trachea was extubated in 68.5 % of children. The determinant factors for postoperative assisted ventilation were peroperative blood loss and duration of surgery. Duration of postoperative ventilation was 40-540 min. In the first 24 h after surgery, a mean 232.13 (148.8) ml of blood was lost through surgical drains (table II). The PCV in drainage fluid was, in most instances, less than 0.15. Overtransfusion, according to our criteria, was more frequent in the postoperative period than during surgery (74% vs 32% (P < 0.001)), leading to possible unnecessary transfusion in some patients. Postoperative complications included transient hypotension in 15% of patients, persistent bleeding in excess of 50 % of estimated red cell volume in 12.6% and thrombocytopenia (platelet count less than 65000 ml"1) in three patients. Postoperative haemostasis was judged satisfactory with an activated
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
with fluid replacement, based upon haemodynamic variables and comprising human serum albumin and packed red blood cells in order to maintain PCV in the range 0.28-0.35—a value considered generally as adequate for oxygen transport and tissue delivery [7]. Fresh frozen plasma was used only in patients requiring more than 70 % of estimated blood volume transfusion. At the end of surgery, anaesthesia was discontinued and the trachea extubated or the patient sedated lightly with assisted ventilation. The criteria used for extubation were: rapid recovery with stable spontaneous ventilation, haemodynamic stability and normothermia, short-duration procedure with relatively small blood loss and absence of continuing bleeding via surgical drains. Patients were observed closely for at least 12 h. Full blood cell count and coagulation screen were obtained in the postoperative period in all patients. The type, volume and timing of transfused blood products were noted. In an attempt to simplify estimation of blood loss, all estimated volumes referred to red cell volumes; fresh frozen plasma volumes were analysed separately. Estimated body blood volume (EBV) = 80 ml kg"1 and red cell volume (ERCV) = EBV x PCV were calculated before operation. The following variables were then calculated for the periods during and after operation: estimated red cell volume transfused (RCT) = 0.75 x volume of packed red cells; estimated red cell volume deficit (ERCD) = ERCV x PCV variation; estimated red cell volume lost (ERCL) = ERCD + RCT; percent of ERCV lost = ERCL/ERCV. Data were statistically analysed using the Statistical Package for the Social Sciences (SPSS/PC+ , MJ Norusis Ed. SPSS Inc. Chicago 1992). Tabular data were analysed using contingency tables (chisquare or Fischer's exact test) and differences of means using Student's t test, ANOVA or KruskalWallis, as appropriate, depending on distribution and number of variables. Statistical significance was assumed with P < 0.05.
Forehead reconstruction
BRITISH JOURNAL OF ANAESTHESIA
856
TABLE II. Mean (SO) blood loss during the study. ERCL = estimated red cell volume loss; RCT = packed red cell volume transfused; ERCL/ERCV = % of patient's estimated red cell volume lost; ERCD = final red cell volume deficit. ^Negative value indicates postoperative overtransfusion
ERCL (ml) RCT (ml) ERCL/ERCV (%) ERCD (ml)
During op.
After op.
Total
191.5(124.63) 182.87 (136.48) 91 (65.9) 6.15(41.83)
52.75 (80.28) 89.85 (82.82) 27.4 (42.3) -37.88(48.14)+
240.72 (146.09) 272.72 (137.6) 118.2(86.1) -32(42.2)
TABLE III. Mean (SD) percentage of patient's estimated red cell volume lost and type of skull deformation. ^Negative value indicates postoperative overtransfusion During op.
After op.
Total
Oxycephaly Plagiocephaly Trigonocephaly Brachycephaly Scaphocephaly Complex
49.1 (22.9) 59 (37.4) 92.4 (49.7) 105.3 (48.45) 92.1 (65.2) 198.5 (165)
-3.7(16.8)t 27.7(41.7) 11.7(16.6) 25.5 (56.5) 35.9 (38.4) 44.5 (97.6)
45.4 (36.9) 86.7 (56.2) 104.1 (49.2) 130.9(69.1) 121.7(78.2) 243.1 (259.4)
prothrombin time of 60% or more of the normal value in all patients. No change in PCV was noted 24 h after surgery; no further transfusion was required and PCV at discharge from the hospital was always 0.30 or more. The age of the patient correlated highly with blood loss (99 (73) % of ERCBV lost in patients younger than 6 months, compared with 72 (38) % lost in patients older than 6 months (P < 0.008)). The type of skull deformation correlated with the percentage of estimated red cell volume lost during surgery (P < 0.0001) and there was a correlation between type of surgical procedure and peroperative (P < 0.012), but not postoperative, blood loss (table III). Separate analysis of the correlation between blood loss and type of surgical procedure was performed in the subgroup of scaphocephaly, for which three types of procedures could be used. Strip craniectomies involved less blood loss (60 (24.5)% of ERCV lost) than forehead reconstruction (106.4 (53.7)%) and complex vault remodelling (170.6 (10.3)%) (P < 0.006). DISCUSSION
Craniofacial surgery for craniosynostosis treatment includes extensive osteotomies with considerable blood loss [2, 5, 6, 8-12]. Because of the well known complications of blood transfusion, with a current incidence of about 1 per 40000 and 5-10% of transfusion recipients being at risk of AIDS and hepatitis [13, 14], methods designed to reduce blood loss are of interest. Induced hypotension has been used in the past without clear benefit [10]. Skin infiltration with vasoconstrictors is used frequently, but is not very effective because periosteum and bone, not the scalp, are the major sources of blood loss [8]. Acute normovolaemic haemodilution is now used routinely in many paediatric surgical procedures [7], but not, to our knowledge, during craniosynostosis surgery. In this study, using strict guidelines for transfusion, we found no complications. However, it must be noted that about 10% of the patients could be
satisfactorily managed without any packed red cell transfusion. Acute haemodilution alone is probably insufficient to avoid blood transfusion in this type of surgery and other techniques designed to reduce the need for homologous transfusion should probably be considered. Because of the small blood volume of young children, autologous transfusion with preoperative blood withdrawal is clearly not adequate to provide blood requirements during surgery. Peroperative blood salvage and retransfusion have been reported anecdotally during craniosynostosis repair [15]. In our recent experience with peroperative use of a cell saver in 15 patients, in spite of satisfactory blood salvage and significant reduction in homologous transfusion requirements, abrupt haemorrhage in patients with a small blood volume required associated use of homologous transfusion in nine patients. The factors influencing peroperative blood loss have been analysed previously. In this type of surgery, trapping of large blood volumes in surgical drapes and dilution with irrigating fluids prevent precise measurement of blood loss. There are few published estimates of blood loss during these procedures; most refer to weighing sponges and measurement of suction volume [2, 8-10]. Kearney, Rosales and Howes reported an interesting method of assessing blood loss by serial determinations of PCV and calculation of red cell volume deficit. This method allows more precise evaluation of changes in blood volume and assessment of adequacy of transfusion practices [6]. Our blood losses were slightly greater than the reported values of 20-40% of estimated blood volume lost [6,9-11] and may be explained partly by differences in surgical procedures, those in our series being more complex. Forehead reconstructions and craniectomies with posterior flaps are associated with significantly more blood loss than strip craniectomies. As in other reports, we found that strip craniectomies had the smallest estimated blood loss. Because of the lack of details on surgical techniques in other series [6, 8-10] and the changes in surgical practice over a 15-yr
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
Type of synostosis
BLOOD LOSS DURING REPAIR OF CRANIOSYNOSTOSIS
istical correlations between estimated postoperative blood loss and determinants such as type of synostosis and surgical procedure. REFERENCES 1. Marchac D, Renier D. Classification of the different types of craniosynostosis. In: Marchac D, Renier D, eds. Craniofacial Surgery for Craniosynostosis. Boston: Little Brown, 1982; 1-8. 2. Shilito J, Matson DD. Craniosynostosis: a survey of 204 cases. Journal of Neurosurgery 1968; 22: 229—240. 3. Renier D. Intracranial pressure in craniosynostosis: pre- and postoperative recordings, correlation with functional results. In: Edgerton MT, Jane SA, eds. Scientific Foundation and Surgical Treatment of Craniosynostosis. Baltimore: Williams and Wilkins, 1989; 263-269. 4. Van Effentere R, Houtteville JP, Philippon J. Interet de la mesure de la pression intra-cranienne dans les craniostenoses decouvertes apres l'age d'un an. Neurochirurgie 1989; 22: 59-67. 5. Loftness SL, Albin RE, O'Donnell RS. Perioperative management of craniofacial surgery in infants and children. In: Marchac D, ed. Craniofacial Surgery. Berlin: Springer Verlag, 1985; 470-173. 6. Kearney RA, Rosales JK, Howes WJ. Craniosynostosis: an assessment of blood loss and transfusion practices. Canadian Journal of Anaesthesia 1989; 36: 473-477. 7. Salem MR, Bikhazi GB. Blood conservation. In: Motoyama EK, Davis PJ, eds. Smith's Anesthesia for Infants and Children. St Louis: C. V. Mosby Co., 1990; 371-392. 8. Hans JF, Trusso R, Levy WJ. Craniosynosectomy with reduced blood loss. Neurosurgery 1981; 8: 209-211. 9. Scholtes JL. Craniofaciosynostosis: anaesthetic and perioperative management. Acta Anaesthesiologica Belgica 1985; 3: 176-185. 10. Diaz JH, Lockhart CH. Hypotensive anaesthesia for craniectomy in infancy. British Journal of Anaesthesia 1979; 51: 233-235. 11. McKersie AM. Paediatric neurosurgery and craniofacial surgery. In: Jewkes DA, ed. Anaesthesia for Neurosurgery, Clinical Anaesthesiology. International Practice and Research. London: Bailliere Tindall, 1987; 70-79. 12. Davies DW, Munro IR. The anesthestic management and intra-operative care of patients undergoing major facial osteotomies. Plastic and Reconstructive Surgery 1975; 55: 50-56. 13. Bove JR. Transfusion-associated hepatitis and AIDS. New England Journal of Medicine 1987; 317: 242-245. 14. Zuck TF. Transfusion-transmitted AIDS reassessed. New England Journal of Medicine 1988; 318: 511-512. 15. Paccagnella F, Longatti GP, Mazzoon D. Emotransfusione autologa nella chirurgia delle craniostenosi: un obiettivo. In: Proceedings of 5 Anuale Riunione Italo-Francese di Neuroanestesia-Reanimazione, Torino, 1988.
Downloaded from http://bja.oxfordjournals.org/ at Dalhousie University on June 20, 2015
period [6], comparisons are difficult. However, it is easier to compare our study with the report from the Children's Hospital at Denver, where extensive craniectomies and complex vault reconstruction represent the third most common neurosurgical procedure performed in infancy [5]. Our results were broadly comparable to those reported in that study [5], especially for brachycephaly and plagiocephaly treated with forehead reconstruction. The duration of surgery varied between different types of skull deformation and different surgical procedures. It did not influence significantly the extent of operative blood loss. In these procedures, 20-50 % of the patient's estimated red cell volume may be lost in less than 30 min. For complex procedures the mean duration was similar to that of another series [6], but less than the duration observed by Loftness, Albin and O'Donnell [5]. The age of patients at the time of surgery was associated with blood loss, with a greater percentage of ERBCV lost in infants younger than 6 months. However, it must be noted that cosmetic and neurological outcome are more important in this age group [5, 6]. In our institution, the optimal age for surgery represents a compromise between surgical conditions with easier and shorter operations in younger children, and physiological tolerance to acute blood loss with greater haemodynamic stability in children older than 6 months. Thus most of our children were aged 6-12 months, with body weight 6-10 kg. It appeared that postoperative management did not achieve the goal of controlled haemodilution and attempts to modify transfusion practice during the recovery period are now in place. This difference in management during and after operation, with possibly unnecessary transfusion during the postoperative period, was noted also by Kearney, Rosales and Howes, who did not use deliberate peroperative haemodilution [6]. This demonstrates clearly the inadequacy of transfusion regimens dealing more with measurement of blood drainage than with physiological observation and measurement of PCV. Another shortcoming of postoperative management in this study was the difficulty in establishing correlations between estimated blood losses and their principal determinants. There were no stat-
857