Ultrasonic and Nonultrasonic Instrumentation: A Systematic Review and Meta-analysis

Objective  To compare the efficacy and safety of ultrasonic surgical instrumentation with nonultrasonic traditional surgical techniques in various types of surgery.

Data Sources  Electronic searches of MEDLINE, Current Contents, and the Cochrane Library were performed for the period of 1990 to June 1, 2005, using relevant search terms. A manual check of all references in accepted studies was also performed.

Study Selection  Only comparative studies (including randomized and nonrandomized control trials) of ultrasonic surgical instrumentation with nonultrasonic instrumentation were accepted. Procedures of interest included the following: colorectal surgery, gynecologic surgery, head and neck surgery, solid organ surgery, vessel harvesting, cholecystectomy, hemorrhoidectomy, mastectomy, and Nissen fundoplication.

Data Extraction  Two investigators reviewed each study: the first investigator extracted all relevant data, and consensus of each extraction was performed by a second investigator to verify the data. Data were then entered into a database and quality checked for accuracy.

Data Synthesis  Fifty-one primary studies that examined 4902 patients were included in this systematic review, of which 24 were randomized trials and 27 were nonrandomized studies. Comparative meta-analyses for blood loss, surgery time, and hospital length of stay were performed using a random-effects model and stratified by surgery type. Heterogeneity was tested using Q statistics. Statistical significance was defined as P < .05.

Conclusion  Meta-analysis of outcomes comparing ultrasonic with conventional nonultrasonic surgical instrumentation demonstrates significant improvement of several perioperative outcomes in procedure-specific settings when ultrasonic instrumentation is used.

With the continued increased dependence of surgeons on energy-based surgical techniques, patient safety continues to be a significant concern. Ultrasonic surgical instrumentation was introduced into clinical practice approximately a decade ago in an effort to minimize the risks of traditional electrosurgical technologies, local peripheral energy damage, and the rare, but potentially devastating, complications associated with monopolar energy. During this time, multiple in vivo and in vitro animal experiments have demonstrated improved safety with ultrasonic instrumentation, which includes devices such as ultrasonically activated scalpel and scissors or shears.

Unlike traditional monopolar energy, which has been in clinical practice for a century, ultrasonic instrumentation denatures protein by means of ultrasonic vibration at a frequency of 55 500 Hz with a vibratory excursion of 50 to 100 μm. The vibration transfers mechanical energy to the tissue, resulting in simultaneous cutting and coagulation. As with all energy technologies, the precise mechanism of action remains unclear, but the effect produces a coagulum of denatured protein and blood clot that occludes adjacent blood vessels and reduces bleeding.¹ Vibration of the scalpel blade does not generate as much heat as monopolar cautery or laser cautery, and the vibration in potential spaces results in cavitation, which may facilitate tissue dissection.² Photomicrographs of tissue incised using an ultrasonic scalpel demonstrate an approximately 1-mm zone of denatured tissue around the ultrasonic incision similar to the lateral energy dispersion demonstrated with ultrasonic instrumentation in porcine models.³ This finding compares favorably with the 0.24- to 15.0-mm range of thermal energy damage associated with electrocautery. Although transmural necrosis to nearby structures can occur even with the ultrasonic scalpel if care is not taken during the use of this device, the risk of collateral tissue injury would seem to be decreased with ultrasonic instrumentation.^2,4

This systematic review of the literature and meta-analysis on ultrasonic instrumentation is intended to compare the efficacy and safety of ultrasonic energy with nonultrasonic traditional surgical techniques, including monopolar or bipolar cautery, clips, stapling, and conventional clamp or tie techniques. General morbidities, such as blood loss, total surgery time, length of hospital stay, and outcomes of specific surgical procedures, were assessed.

Procedures for this review followed established best methods for the evolving science of systematic review research.^5,6 A written prospective protocol defined the objectives, search criteria, study selection criteria, data elements of interest, and plans for analysis.

The English-language literature was searched from 1990 to June 1, 2005. Electronic searches of the MEDLINE (PubMed) database, Current Contents, and the Cochrane library were performed using the following keywords: harmonic scalpel or UltraCision or ultrasonic scalpel or ultrasonically activated scalpel or ultrasonic dissection or ultrasonic dissector or ultrasonic coagulator or harmonic shears or ultrasonic shears or ultrasonic scissors or LaparoSonic or coagulating shears or Autosonix or SonoSurg. A manual reference check of all accepted manuscripts and recent reviews was also performed to supplement the electronic searches and identify additional potentially relevant studies.

Studies had to meet the following inclusion criteria: the study had to be a randomized control trial (RCT) or a nonrandomized control trial (nRCT) that compared ultrasonic surgical instrumentation with nonultrasonic instrumentation in at least 10 patients per treatment group undergoing surgery and had to have reported efficacy and/or safety outcomes. All surgical procedures were of interest, except tonsillectomy, neurosurgery, and nephroureterectomy.

A data extraction form was designed from protocol-defined data elements. These elements were extracted and confirmed by 2 researchers. Differences were resolved before data entry. Data were then checked for accuracy and exported for analysis. The primary efficacy and use measures included conversions from laparoscopic to open procedures, length of surgery time, blood loss, and length of hospital stay. Safety outcomes sought were complications associated with blood loss, pain, tissue and organ injury, and other procedure-specific perioperative complications.

Basic descriptive statistics were used to summarize overall and procedure-specific study, patient, and treatment data. Efficacy, safety, and use outcomes were summarized for ultrasonic and nonultrasonic instrumentation groups by surgical procedure. Nonultrasonic instrumentation included electrocautery (monopolar and bipolar), stapling, clips, and clamp-and-tie techniques. Comparative meta-analyses of selected efficacy and safety outcomes were performed using empirical Bayes random-effects meta-analyses between ultrasonic instrumentation and alternative nonultrasonic instrumentation for the following variables: blood loss, operating room or surgery duration, and lengths of hospital stay. Other outcomes (number of patients who required conversions, additional operations, transfusions, perioperative deaths, and perioperative complications) were of interest, but data were sparse. The results of the meta-analyses of continuous variables were expressed as mean differences and as odds ratios for binary outcomes with 95% confidence intervals. Meta-analyses of the binary variables were conducted on the log odds ratios to satisfy the assumption of normality of effect sizes. All analyses were stratified by surgery type. Because of the lack of reporting of effect size dispersion, not all data could be included in the meta-analyses.

Statistical calculations were performed using SAS statistical software, version 8.1 (SAS Institute Inc, Cary, North Carolina), and SPSS statistical software, version 14.0 (SPSS Inc, Chicago, Illinois). Statistical significance was defined as P < .05. The Q statistic (Cochrane Q) was used to determine the presence of significant heterogeneity among studies.

The search yielded 581 citations, of which 543 full articles were retrieved for further screening. Ultimately, 54 studies met all inclusion criteria, of which 3 were determined to be kinship studies (studies that involved the same or overlapping patient populations). Thus, 51 primary studies with 111 treatment groups and 4902 patients were accepted into this systematic review (Table 1). Most of the studies were published after 1998 and were from North America and Europe. Most were from single centers (48 studies [94%]) and teaching or urban institutions (41 studies [80%]). Twenty-four studies were RCTs and 27 were nRCTs. All eligible studies were rated for level of evidence based on criteria adapted from recommendations from the Centre for Evidence-Based Medicine.⁷ The studies were evenly distributed among level 1 (18 studies), level 2 (16 studies), and level 3 (17 studies) evidence, indicating that most data in this review came from RCTs or prospective comparative studies. Procedure categories included colorectal surgery (5 studies); gynecologic surgery (6 studies); head and neck surgery (8 studies); solid organ surgery, including liver, pancreas, and adrenal glands (6 studies); vessel harvesting (9 studies); cholecystectomy (4 studies); hemorrhoidectomy (7 studies); mastectomy (2 studies); and Nissen fundoplication (4 studies) (Table 2). A list of all studies included in this review is provided in the eBox.

Box Section Ref ID

Patient demographic characteristics are summarized in Table 3. Mean ages were similar among treatment groups. In most surgical procedure groups, most patients were male. The exceptions were for head and neck surgery (female, 72%), Nissen fundoplication (female, 50%), mastectomy (female, 100%), and gynecologic procedures (female, 100%).

The proportion of patients who underwent laparoscopic or minimally invasive surgery varied by surgical procedure. All cholecystectomy and Nissen fundoplication procedures were performed laparoscopically. For the colorectal surgical procedures, laparoscopic or minimally invasive surgery procedures accounted for more than 50% in both the ultrasonic and nonultrasonic groups. Conventional or open techniques were used in most head and neck and solid organ procedures. In gynecologic surgical procedures, there were twice as many patients undergoing laparoscopic or minimally invasive surgery procedures with ultrasonic instrumentation than nonultrasonic instrumentation (88.3% and 43.5%, respectively). This difference can be explained by a single study that compared laparoscopic hysterectomy using ultrasonic instrumentation (48 patients) with conventional hysterectomy using nonultrasonic instrumentation (96 patients).⁸

From the available data, the efficacy and use outcomes for ultrasonic and nonultrasonic instrumentation are summarized by surgical procedure type in Table 4. The most commonly reported outcomes across all surgical procedures were length of operating room or surgery time, blood loss, length of stay, and additional operations. The number of patients who required conversion to an open procedure was reported in several studies in which procedures were performed laparoscopically, whereas other outcomes such as drainage output and/or drainage duration were rarely reported. Meta-analyses that compared ultrasonic instrumentation with nonultrasonic instrumentation for blood loss, operating room or surgery time, and hospital length of stay are presented in Table 5. Trends and statistically significant findings for efficacy and use outcomes are described for each procedure category.

Most studies in this review involved general surgery procedures (36 studies). Of these, 10 (28%) were from North America, 12 (33%) were from Europe, and 14 (39%) were from other countries, including Asian-Pacific countries (11 studies), India (2 studies), and Israel (1 study). These studies included 3028 patients, and 17 (47%) were RCTs.

Mean blood loss seemed to be less for most general surgery procedures with ultrasonic instrumentation, the exceptions being colorectal and Nissen fundoplication procedures (Table 4). In a single RCT that involved 200 patients undergoing a laparoscopic cholecystectomy procedure,⁹ a small but statistically significant difference in blood loss was detected, favoring ultrasonic instrumentation. A trend in favor of ultrasonic instrumentation for blood loss was also observed in patients undergoing hemorrhoidectomy and solid organ procedures; however, no statistical significance was reached. Meta-analysis of blood loss was not possible for mastectomy, colorectal, and head and neck procedures.

All general surgery studies reported operating room or surgery time. The average surgery time seemed to be shorter in the ultrasonic instrumentation group with the exception of mastectomy (Table 4). A significant difference in surgery time in favor of ultrasonic instrumentation was seen with cholecystectomy and head and neck surgery. All cholecystectomy studies in this analysis were RCTs conducted in Europe, and all were performed laparoscopically. Head and neck surgery studies mostly involved patients undergoing thyroidectomy procedures. In the 2 modified radical mastectomy studies,^10,11 a shorter surgery time was noted with nonultrasonic instrumentation use; however, the results were not significant. In solid organ surgery, 4 nRCTs that evaluated open hepatectomy procedures were included in the meta-analysis.^12-15 No significant difference in surgery time was detected. Similarly, in 3 studies of laparoscopic Nissen fundoplication procedures, no significant difference in surgery time was detected between the ultrasonic and nonultrasonic groups (Table 5).

Hospital length of stay was reported for all general surgery procedures except mastectomy and solid organ procedures. No significant difference in hospital stay was detected for any of the surgical procedures except for cholecystectomy, for which a small but significant difference was observed with ultrasonic instrumentation in a single study.⁹ In a single study of Nissen fundoplication,¹⁶ the length of stay was similar in both the ultrasonic and nonultrasonic instrumentation groups.

With regard to other clinical outcomes of interest, fewer additional operations were reported in the ultrasonic group for all surgical procedures except for colorectal procedures. The 2 additional operations in the ultrasonic group were (1) in a patient undergoing transanal endoscopic microsurgery and simultaneous right hemicolectomy and (2) in a patient undergoing a hand-assisted laparoscopic total colectomy procedure for postoperative bleeding.^17,18

Fewer conversions from laparoscopic to open procedures were required with ultrasonic instrumentation use except in solid organ procedures. Drainage output, duration of drainage, or both were reported for mastectomy, head and neck, and solid organ procedures and were reported less frequently with ultrasonic instrumentation use.

Four studies in this category were hysterectomy procedures, and the fifth study¹⁹ involved 48 patients undergoing an ovarian drilling procedure. All 5 hysterectomy studies^8,20-23 were included in the meta-analysis. All hysterectomies were performed laparoscopically or were laparoscopically assisted, with the exception of 1 study,²⁰ which was an open total abdominal hysterectomy. The types of hysterectomies performed included vaginal hysterectomy, supracervical hysterectomy, and abdominal hysterectomy.^8,20-23 With the exception of the study performed by Grassie and Makii,²⁰ no significant differences were seen in blood loss, surgery time, or length of stay in the individual studies. Grassie and Makii demonstrated a statistically significant reduction in blood loss (P = .05) in patients undergoing open total abdominal hysterectomy and bilateral salpingo-oophorectomy using ultrasonic instrumentation. In addition, Gyr and colleagues⁸ compared laparoscopic total hysterectomy using ultrasonic instrumentation with total abdominal hysterectomy using conventional instrumentation. When this study was excluded from meta-analysis, the results remained similar, favoring ultrasonic instrumentation for blood loss and hospital stay with the smaller difference and no significance reached.

Most vessel-harvesting procedures were radial artery vessel harvesting for coronary artery bypass graft surgery. In 2 studies,^24,25 the vessel harvested was the internal thoracic or internal mammary artery. Blood loss was not reported for these procedures, and length of surgery time was slightly less in the nonultrasonic instrumentation group; however, data on surgery time did not permit meta-analysis. In a single study,²⁵ a shorter hospital length of stay was observed when ultrasonic instrumentation was used; however, no significant difference was detected.

Safety outcomes commonly reported for each of the surgical procedures are summarized in Table 6. Surgical complications varied highly across studies and by each surgical procedure. In most instances, only a few treatment arms contributed to the data, and the number of patients with complications was small. Therefore, meta-analysis of these outcomes was not performed. Bleeding was the most commonly reported complication across all surgical procedures.

For all studies and surgical procedures that contributed data, fewer patients seem to have bleeding complications in the ultrasonic group. The exception was in colorectal surgical procedures, for which 2 of 49 patients (4.1%) in the ultrasonic group had bleeding complications, compared with only 3 of 166 patients (1.8%) in the nonultrasonic group. Thromboses, bowel leaks, and fistulas all occurred less often in the ultrasonic instrumentation groups, and stenosis complications were similar in both groups. Drainage or seroma in patients undergoing cholecystectomy, head and neck, or mastectomy surgical procedures also occurred less often with ultrasonic instrumentation use. Nerve, organ, and tissue injuries were reported for only a few procedure types. All occurred less often in the ultrasonic group, as did vessel stenosis. Vessel injury, however, was similar for both the ultrasonic and nonultrasonic groups. Overall, ultrasonic surgical instrumentation use seems to be associated with fewer complications across all studies that report these outcomes; however, data are sparse.

This systematic review of the literature and meta-analysis were performed to examine efficacy, efficiency, and safety outcomes in patients undergoing surgical procedures in which ultrasonic instrumentation was used. The results are based on 51 comparative studies, including 24 randomized trials, examining almost 5000 patients in various geographic locations.

Overall, the findings of this review and meta-analysis demonstrate that ultrasonic instrumentation is as safe and effective and, in some procedures, produces more favorable outcomes than nonultrasonic instrumentation. With numerous options available, surgeons can select energy-based surgical instrumentation with regard to the versatility, reliability, and limitations of available technology and the technical requirements of the intended procedure. Instruments that use thermal dissection are easy to use, reliably provide local hemostasis, and allow versatility. Surgeons can alternate between cutting and coagulation modes during a procedure with relative ease. However, the electric current used to create the heat for cutting and coagulation can cause nonselective local thermal damage, leading to inadvertent tissue necrosis.²⁶ Bipolar electrosurgery may minimize nonselective tissue effects compared with monopolar electrosurgery.

Existing ultrasonic instrumentation can be versatile and exists in multiple configurations, including shears, hooks, and probes that can be used to grasp, dissect, coagulate, and cut tissue. The mechanism of vessel coagulation with ultrasonic instrumentation is similar to electrosurgery, coapting with a denatured protein coagulum. However, ultrasonic instrumentation works at much lower temperatures than radiofrequency energy, minimizing local thermal injury while maintaining the ability to coagulate small vessels and minimize bleeding.

Ultrasonic instrumentation is used frequently in gynecologic, urologic, and general surgery procedures. As this review of the literature illustrates, ultrasonic instrumentation has been used successfully in head and neck procedures, vessel harvesting, hemorrhoidectomy, mastectomy, and antireflux surgery. Ultrasonic instrumentation has even been successfully used in the ablation of malignant hepatic metastases for palliative care in patients with cancer.²⁷ It has proved useful in the special situation of a surgical patient with a pacemaker or cardioverter defibrillator, for whom bipolar electrical cautery can lead to device dysfunction.²⁸

Meta-analyses have notable limitations. Specifically, the heterogeneity of the surgical procedures reported ultrasonic and nonultrasonic instrumentation use and the limited amount of level 1 data in the published literature related to energy sources. Although inclusion of only RCTs enables one to assemble the most rigorous evidence, there is often a paucity of clinical trials. Faced with assembling the best available evidence in the absence of numerous and rigorous clinical trials, it becomes important to include nonrandomized comparative studies as well. In this review, studies are split almost equally between RCTs and nRCTs; however, only approximately 40% of the patients were enrolled in the RCTs. Variability in surgical procedures within each surgery type contributed to significant heterogeneity observed in some meta-analytic results and precluded meta-analyses for other outcomes. Therefore, it is difficult to make conclusive statements about the results of the meta-analysis in some categories. Furthermore, some meta-analysis results are based on a limited number of studies and, therefore, should be interpreted with caution.

Last, cost data comparing ultrasonic instrumentation with electrocautery or other conventional methods (eg, clips, staples, and sutures) were rarely reported. When reported, the data were highly variable across studies and were for the most part study and institution specific relative to geographic location, procedure performed, reimbursement structure, and actual costs and/or charges that were reported. Studies documenting and reporting cost-effectiveness are needed and will become increasingly important as we continue to examine a broader application of ultrasonic instrumentation.

Despite these limitations, the data presented in this literature review and meta-analysis support the use of ultrasonic instrumentation in a wide variety of surgical procedures. Furthermore, the analysis performed in the present study demonstrates improvement in multiple surgical outcome variables in a procedure-specific setting.

In conclusion, the results of this meta-analysis comparing ultrasonic with nonultrasonic instrumentation for tissue dissection and vessel ligation reveal that ultrasonic instrumentation is safe and effective in various surgical procedures and provides several procedure-specific advantages in perioperative outcomes. The expanded use of ultrasonic instrumentation in general or minimally invasive surgery, colorectal surgery, gynecologic surgery, plastic and reconstructive surgery, and otolaryngology is supported by the outcomes from this systematic review. A decrease in operative blood loss combined with less local thermal damage than traditional electrosurgical instrumentation may prove to be of particular benefit in selected surgical procedures.

Correspondence: Jaime Landman, MD, Department of Urology, Columbia University School of Medicine, 161 Fort Washington Ave, Room 1111, New York, NY 10032 (landmanj@yahoo.com).

Accepted for Publication: April 27, 2007.

Author Contributions:Study concept and design: Matthews, Nalysnyk, Estok, Fahrbach, Banel, and Landman. Acquisition of data: Nalysnyk, Estok, Fahrbach, Banel, and Linz. Analysis and interpretation of data: Matthews, Nalysnyk, Estok, Fahrbach, and Banel. Drafting of the manuscript: Matthews, Nalysnyk, Estok, Fahrbach, and Landman. Critical revision of the manuscript for important intellectual content: Matthews, Estok, Banel, and Linz. Statistical analysis: Fahrbach. Administrative, technical, and material support: Nalysnyk, Estok, Banel, Linz, and Landman. Study supervision: Matthews, Nalysnyk, Estok, Banel, and Landman.

Financial Disclosure: None reported.

Funding/Support: This study was supported by Ethicon Endo-Surgery.

Additional Information: The eBox is available at http://www.archsurg.com.