Surgical Techniques to Reduce Blood Loss


Blood loss can occur in any surgical procedure. However, certain types of surgery are associated with blood loss of a magnitude that frequently leads to the need for transfusion. These include hepatic, orthopedic, and vascular operations. In the following sections, we will examine technical approaches that can help reduce transfusion need both in general and in specific areas of surgery. This review does not include a discussion of coagulation disorders, topical hemostatic agents or the use of antifibrinolytics, topics covered elsewhere in this volume. Minimally invasive surgical techniques and issues dealing with vascular surgery also are dealt with in other chapters.


Surgeons have a variety of instruments available to reduce bleeding during surgery. Cautery has been a mainstay in preventing blood loss from surgical incisions for many years. Coagulating vessels in an open wound with direct application of heated irons mercifully has given way in the 20th century to the use of electrocautery and electrosurgery devices. Electrocautery utilizes direct current to heat a wire tip or knife blade. During electrocautery, current does not enter the patient's body. Direct application of the wire tip produces thermal injury, coagulating tissue. The extent of thermal tissue damage is of concern and depends upon the strength of the current and the time of application. Electrosurgical instruments use alternating current, which enters the patient's body as part of the circuit. Both have been shown to coagulate small vessels and reduce blood loss although little information exists about the magnitude of their effect on blood loss. Recently, some authors have attempted to quantify the impact of cautery on surgical blood loss. Kurtz and Frost compared use of the scalpel versus electrocautery in two groups of women who underwent modified radical mastectomy for breast cancer <|[1]|>. The mean blood loss for the cautery group was less than one-half that of the scalpel group (134.1 ml vs. 331.6 ml). The choice of technique was the strongest predictor of blood loss. Newer modifications to electrocautery include the use of an argon beam-enhanced device that produces a stream of argon gas around the cautery tip. This is designed to produce less tissue adhesion and damage. These devices use the argon gas to conduct a high-frequency current to tissue that can coagulate vessels up to 3 mm in diameter while minimizing tissue trauma.

The use of the laser may either reduce or increase bleeding depending on the disease being treated and the operation being performed. In a review of 300 patients who underwent either laser-assisted cholecystectomy or laparoscopic cholecystectomy with electrocautery, Corbitt suggested that bleeding was more excessive in the laser-assisted group <|[2]|>. Operating time was also longer. Impact on blood loss may depend on the type of LASER used. Wyman and Rogers performed a prospective, randomized trial of modified radical mastectomy using a neodymium yttrium-aluminum-garnet laser scalpel compared to a conventional scalpel <|[3]|>. Forty consecutive women were randomized to one or the other group. Operative blood loss was significantly less with the laser (149 vs. 421 ml) but the mean operating time was approximately 25 minutes longer, leading the authors to question the use of the laser scalpel.

Other devices that rely on water, gas, sound and microwaves as dissecting media are available. Those using ultrasonic energy to dissect and coagulate tissues have been used in a variety of both open and laparoscopic procedures, but little information about blood loss has been reported. The Cavitational Ultrasonic Surgical Aspirator, or CUSA, (Valleylabs, Boulder, CO) uses an acoustic vibrating tip that translates electrical energy into mechanical energy that destroys and emulsifies tissue. One advantage to these instruments is that they operate at much lower temperatures than either electrocautery or electrosurgery when coagulating, resulting in less tissue damage and smoke. CUSA has been used as a means of identifying intrahepatic vessels during liver resections <|[4]|>. Vessels can be ligated separately with only minimal bleeding during dissection.

Microwave tissue dissectors use a similar process relying on generation of concentrated microwave energy to produce heat at the instrument tip. Water jet dissectors pump a fine stream of physiologic saline thorough a nozzle consisting of a sapphire tip with a 0.15 mm opening in the tip of a dissection wand at pressures typically between 15-20 kg/cm2 (Handy-jet, Saphir Medical, Schwaig, Germany) <|[Figure 1]|>. These devices are able to dissect through both normal and sclerotic liver tissue while exposing intrahepatic vessels for ligation <|[5]|>. Baer et al studied the value of a water jet dissector in reducing blood loss and subsequent transfusion need in 67 patients who underwent major liver resection <|[6]|>. Fifty-one patients (76%) were treated using the tissue fracture technique and results were compared to 16 (24%) in whom dissection was done with the water jet. Both blood loss (1,386 ml water jet vs. 2,450 ml tissue fracture) and transfusion requirements (mean 2 units water jet vs. 5.2 units tissue fracture) were less in the water jet group but only the amount of blood transfused was significantly different (p = 023). Rau et al found that water jet dissection resulted in one-half the blood loss produced with the ultrasonic aspirator in a prospective trial involving 60 patients undergoing liver resection <|[7]|>. This group at Ludwig Maximilians University in Munich has recently reported their results in a rabbit model of liver resection using a high frequency current along with the water jet cutter <|[8]|>. They predict on the basis of their results that this combination will lead to further reductions in blood loss. Wu et al found the use of intraoperative ultrasonography and intermittent hepatic inflow vascular occlusion led to increased resectability rates and both decreased blood loss and transfusion rates compared to patients treated without these techniques <|[9]|>.

Two separate studies from Asia have reported on the usefulness of hepatic dissection using microwave tissue coagulation in reducing blood loss. Zhou et al report on the effectiveness of microwave dissecting via a monopolar electrode in 50 patients who underwent hepatic surgery for hepatocellular cancer <|[10]|>. Blood loss and transfusion need were both reduced in these patients when compared to a similar group of 46 patients who underwent surgery using conventional techniques. Blood loss averaged 215 189 ml using microwave dissection for hepatectomy compared to 652 1,008 ml in conventional hepatectomy. 39.1% of the former group avoided allogeneic transfusion. Lau et al studied 20 consecutive hepatic resections performed with microwave technique <|[11]|>. Mean blood loss was 1,132 ml. when one patient with an hepatic vein injury was excluded. Five patients (20%) avoided allogeneic blood transfusion. Several authors have reported some success in controlling troublesome presacral hemorrhage encountered during pelvic surgery. Xu and Lin have modified cautery in their approach <|[12]|>. Rather than apply cautery directly to the bleeding venous complex, they interpose a piece of muscle between the cautery tip and the area of hemorrhage. This produces a coagulum that controls bleeding more successfully than direct electrocautery. Berman has used surgical staples and Appleton et al have used bone nails <|[13, 14]|>. Cosman and colleagues successfully controlled bleeding by the use of a tissue expander <|[15]|>.


Laparoscopic approaches to many operations result in less blood loss and transfusion requirements <|[16-20]|>. Successes have been reported for splenectomy, renal and adrenal surgery, colon resections and other major surgical procedures. The benefit gained from a minimally invasive, laparoscopic surgical approach is discussed in the chapter dedicated to this topic.


Hepatic surgery, whether resection for cancer or transplantation, frequently leads to blood transfusion and, consequently, increased morbidity. The negative impact of blood loss during hepatic surgery is demonstrated by two large-scale retrospective studies. Nagasue et al analyzed 229 patients operated on for primary hepatocellular cancer over an 11-year period. Operative death was associated with Child's class, bromosulphalein (BSP) test, and the amount of intraoperative blood loss <|[21]|>. Elias et al reported on their series of 279 patients treated with hepatectomy for cancer over 8 years. 46 patients underwent repeat operations. Intraoperative blood loss greater than 1,500 ml. in this subgroup was significantly related to a higher rate of postoperative complications <|[22]|>.

Technical factors and operative approach have a critical effect on hepatic blood loss. Surgeons throughout the world have demonstrated the ability of isolating vascular inflow to the liver during resections to prevent blood loss and to diminish transfusion need <|[23-28]|>. Experience with these techniques can produce excellent results. Jamieson and colleagues at the Center for Digestive Diseases at the Hopital Pontchaillou in Rennes, France reviewed blood loss and transfusion rates in 75 patients who underwent liver resections for both benign (N = 26) or malignant (N = 49) disease <|[28]|>. Operations ranged from wedge resection (N = 5) through segmentectomy (N = 49) to right or left hepatectomy (N = 21). Blood loss ranged from 0 to 15 units, and 64% of patients did not need transfusion. Of interest is the fact the mean length of hospital stay was 5 days shorter in the nontransfused group. The authors conclude that major liver surgery can be done successfully without blood transfusion in the majority of patients. Their data also suggest that this type of surgery may be more cost effective when done without transfusion. In a similar review, the same authors report their experience with 49 patients who underwent resection of one or more segments of the liver for either benign or malignant disease. Transfusion requirement ranged from 0-15 units, with 63% of patients avoiding allogeneic blood transfusion <|[29]|>.

Although individual techniques vary somewhat, the general approach of producing temporary normothermic ischemia through vascular isolation should be considered a standard at this point in time. Concerns still exist over the possibility of liver damage during extended vascular occlusion. Hannoun et al describe 18 patients in whom normothermic ischemia of greater than one hour was tolerated with a mean transfusion rate of 5.3 units of blood <|[30]|>. They do not show a specific effect on blood loss but document the ability to employ vascular occlusion safely for prolonged periods of time. Kim et al did a similar analysis of 47 patients who underwent hepatic resection, dividing them into two groups based on the length of time of vascular occlusion Group 1 = 50 -75 mins; Group 2 = 30 - 42 mins <|[31]|>. Mean operative blood loss in each group was significantly less when compared to similar patients treated with conventional techniques (Group 3) and no normothermic ischemia (Group 1: 819 ml, Group 2: 1,523 ml, Group 3: 1,652 ml; difference between Group 1 or 2 and Group 3 p <|[32]|>. Yamanaka and colleagues use an ice slush placed on lobes that they map with temporary occlusion of vessels <|[33]|>. Several surgeons have also used stapling devices successfully to control hepatic vascular inflow during major liver resection. Yanaga et al reduced blood loss from a mean of 4,792 ml to 2,071 ml, more than one-half, by using stapler occlusion of Glissons pedicle <|[34]|>. Ramacciato and colleagues had similar results using a vascular stapler to ligate hepatic veins <|[35]|>.

Similar results can be obtained for liver transplantation. Deakin et al analyzed both patient and technical factors in a group of 300 adult liver transplant patients treated over 8 years in an attempt to identify those factors associated with increased blood loss <|[36]|>. Among patient factors, only blood urea and platelet count were found to be independent predictors of increased blood loss. Technical factors were more important as shown by the reduction of intraoperative blood transfusion over eight years from 23.5 median units of blood for the first 50 patients to 8 units for the most recent fifty. These results were attained as the transplant team gained experience with the use of venovenous bypass and the argon beam coagulator. Bilik et al compared the use of standard, orthotopic liver transplantation (OLT) to reduced-size liver transplantation (RLST) in pediatric patients with hepatic failure <|[37]|>. RLST was devised to help improve the chances of these patients who had long waiting periods for a compatibly sized donor organ. RLST can be done successfully but at the expense of significantly higher intraoperative and postoperative blood loss and transfusion rates. RLST patients received a mean of 515.7 490.9 ml of blood products during surgery compared to only 177.3 278.3 ml for the OLT group. Starzls group at Pittsburgh has performed liver transplants without blood products in four Jehovahs Witnesses <|[38]|>.

Although not specifically hepatic surgery, the treatment of varices often involves approaches to the liver. The mortality from variceal bleeding remains high inspite of aggressive medical and interventional therapy. Rapid institution of therapy aimed at stopping the bleeding is essential if we are to reverse this trend. The use of transjugular intrahepatic portosystemic shunt (TIPS) in the treatment of variceal hemorrhage has been reviewed recently by McCormick et al <|[39]|>. The authors note that the shunt can be inserted successfully in 90% of patients and that TIPS can reduce blood transfusion requirements in bleeding patients. Questions remain concerning the risks and long-term patency of TIPS as a primary treatment for variceal bleeding. Orloff's extensive experience in treating patients with variceal bleeding points out that success may be obtained not just by modifying operative technique but by moving quickly to the operating room to prevent further hemorrhage <|[40]|>.

Orloff et al report on their success with emergency portocaval shunting in a group of unselected cirrhotic patients, treated and followed over a period of ten years. Patients were prospectively randomized after initial medical therapy into two treatment groups, one that continued with medical treatment followed by interval shunt and the second that underwent emergency portocaval shunting within eight hours. Bleeding was controlled completely in all the emergency shunt patients, but 55% of the medically treated patients had recurrent or persistent hemorrhage. Total blood transfusion requirements in the emergency shunt group were significantly less than in the comparison group 7.1 2.6 units vs. 21.4 2.6 units. Eighty-one percent of emergency shunt patients were discharged compared to only 45% of those treated with initial medical therapy, demonstrating the effectiveness of this approach.


Information regarding blood loss in thoracic surgery done using the video-assisted thoracic surgery approach (VATS) is limited but encouraging. VATS has been used in a variety of procedures with reductions in blood loss. Walker et al reported on major pulmonary resection in 83 pts. performed with VATS <|[41]|>. Procedures ranged from simple lobectomies to extensive resections. Median blood loss was 80 mL. for lobectomy, with an overall average of 1,050 ml (range 50 3,000 ml) including 21 patients converted to open operations. VATS lung volume reduction for emphysema produced significantly less blood loss (209 versus 82 mL, p = 0.0000017) than a median sternotomy approach in a series of patients treated by Roberts et al <|[42]|>. Vigneswaran and colleagues had similar results, keeping blood loss less than 150 ml in 19 similar patients <|[43]|>. Ravini reviewed the results of four years of open lung biopsy versus VATS for patients treated at the Milan Sarcoid Clinic, finding that the minimally invasive approach produced less blood (61+/-58 versus 156+/-84 mL) in the first postoperative day <|[44]|>. OBrien and colleagues performed thoracoscopic drainage and decortication procedures for empyema, a typically bloody operation, in 8 patients with a median blood loss of 200 mL <|[45]|>.

Esophageal resections have been performed successfully using VATS with reductions in blood loss to one unit or less in most cases. Kawahara and colleagues reported a mean blood loss of 163 +/- 122 ml in 23 patients who underwent VATS esophageal resections <|[46]|>. Swanstrom and Hansen performed VATS and laparoscopic esophageal resections for cancer in 9 patients with an average blood loss of 290 mL <|[47]|>. Blood loss was significantly less in 18 patients treated by Law and associates at the Queen Mary Hospital in Hong Kong with esophageal resections via VATS when compared to 63 who underwent open thoracotomy <|[48]|>. Median blood loss for VATS was 450 ml (range, 200 to 800 ml) compared to a median of 700 ml (range, 300 to 2,500 ml) for thoracotomy (p <|[49]|> Estimated blood loss was 270 +/- 157 ml. for this group.

VATS has also been used as an alternate approach to the thoracic spine for resection and tumor excision. Blood losses of up to 2500 mL have been reported with this operation, depending on the extent of bony metastases and epidural bleeding <|[50, 51]|>. Intrathoracic blood collections can be suctioned from the pleural cavity via a VATS approach in trauma patients <|[52, 53]|>. The potential then exists to process this blood for autotransfusion.


Bleeding sufficient to require transfusion during major urologic procedures is not uncommon. Ekengren and Hahn have studied technical factors related to blood loss during transurethral resection of the prostate in a series of 700 patients <|[54]|>. Both the length of the operation and the amount of resected tissue were independent predictors of blood loss, which ranged from 10 to 3,825 ml with a median of 300 ml. Regional anesthesia was associated with a higher blood loss than general anesthesia, but losses could be reduced with the regional approach if hypotensive techniques (systolic blood pressure maintained at 100 mmHg or less) were used. Strup et al analyzed the impact of intermittent pneumatic compression devices on intraoperative blood loss during radical prostatectomy or cystectomy in 91 patients operated on over a 5 year period <|[55]|>. Fifty-nine patients were treated with intermittent pneumatic compression; 32 were not, with comparable numbers in each group of prostatectomy and cystectomy patients. Blood loss in the compression group ranged from 700 to 8,850 ml with a mean of 2,541 ml. In the no compression group, blood loss was significantly lower with a mean of 1,807 ml (range 450 to 5,100 ml). Compressed patients received 0.6 units of blood more on average than those who were not compressed.

Hrebinko and ODonnell used cautery in radical prostatectomy to reduce the amount of blood lost from the dorsal venous complex <|[56]|>. Tsujihata et al employed a stapling device to minimize bleeding from the deep dorsal vein complex in their patients undergoing prostatectomy <|[57]|>.


Blood loss during liposuction can be considerable, and often requires predonation of blood to meet transfusion needs. Courtiss et al estimated the amount of blood present in tissue aspirated during large volume liposuction to be 30 to 45% depending on the site of fat removal <|[58]|>. They transfused their 108 patients with 227 units of autologous blood and only 2 units of allogeneic blood. Klein reported on the role of tumescent anesthetic technique in reducing blood loss to virtually zero during large volume liposuction (>1,500 ml of fat) (54). In 112 patients, the mean volume of blood suctioned was 18.5 ml, 0.5-9.7 ml for each liter of fat removed <|[59]|>. Mandel found the use of a syringe to perform large-volume liposuction led to less blood loss than vacuum liposuction by machine <|[60]|>. Two prospective trials of tumescent anesthesia for large volume liposuction reported by Klein show that blood loss is essentially eliminated with this technique <|[59]|>.


The choice of anesthetic agent and technique may have an effect on surgical blood loss. Several studies comparing isoflurane to propofol have shown that use of the latter reduces operative blood loss in a variety of settings ranging from endoscopic sinus surgery to therapeutic abortion. Blackwell et al studied the effect on blood loss of general anesthesia with isoflurane inhalation in 13 patients versus monitored sedation with a continuous intravenous infusion of propofol in 12 similar patients undergoing endoscopic sinus surgery <|[61]|>. Blood loss in the isoflurane group was 251 ml compared to a 101 ml loss in the propofol group (p <|[62, 63]|>.

Epidural anesthesia has been shown to reduce blood loss in lumbar laminectomy procedures and in hip arthroplasty. Kakiuchi reduced blood loss in lumbar, but not cervical spine procedures with an initial epidural blockade <|[64]|>. He attributed the success in the lumbar region as well as the failure in the cervical region to the reduction of intraosseus pressures and lumbar venous hypotension by the epidural. In a trial of combinations of epidural anesthesia and aprotinin antifibrinolysis in hip arthroplasty, epidural anesthesia alone was found to reduce blood loss by 21.3% when compared to general anesthesia and placebo <|[65]|>. Shir et al came to a somewhat different conclusion in analyzing their results of a comparative trial of epidural anesthesia alone, general anesthesia alone and a combination of the two in men undergoing radical prostatectomy <|[66]|>. Blood loss was significantly less (mean = 1,490 ml.) when compared to either general anesthesia (mean = 1,940 ml.) or both techniques (mean = 1,810 ml). The authors believe that epidural anesthesia did not reduce blood loss but that general anesthesia increased it. No matter how it is said, it appears that epidural anesthesia does decrease blood loss when compared to general techniques.

Lowering central venous pressure has been shown to reduce blood loss during hepatic surgery in two large series <|[67, 68]|>. Melendez et al reviewed their results with 542 patients undergoing hepatic surgery. Median blood loss in this group was 645 ml.; median transfusion amount was 2 units of packed RBCs. Transfusion of allogeneic blood was avoided during surgery and up to 12 hours after surgery in 67% of patients. Moreover, lowering the CVP allowed for easier control of the hepatic vein. Jones et al reviewed their results in 100 similar patients. When CVP was kept below 5 cm H2O, median blood loss was 200 ml. Only 2 of these 40 patients were transfused. When CVP was kept above 5 cm H2O, median blood loss was 1,000 ml in 52 patients, 25 of whom were transfused. This technique, although proven to be helpful only in hepatic surgery, should be applicable to other types of procedures where the risk of venous bleeding is present, i.e., pancreaticoduodenectomy. Reduction mammoplasty may result in significant blood loss. Kuipers et al reduced blood loss in this procedure by changing their anesthetic technique with the addition of sufentanil to reduce the dose of enflurane employed, thereby decreasing heart rate and diminishing enflurane-enduced changes in vascular tone <|[69]|>. Adding adrenaline to local anesthetics reduces blood loss from skin flaps in reduction mammoplasty <|[70]|>.

In this brief review, we have attempted to focus on recent changes in surgical technique directed at preventing or controlling operative blood loss, thereby reducing the need for transfusion. The best and simplest way to avoid the need for transfusion is to avoid blood loss keep the blood in the patient. Any surgeon can accomplish this goal with attention to detail, careful planning and modification of surgical approach.