In Search of the Autologous Clip: A Case for Experimental Standardization

Abstract

Background:

In an effort to enable faster and, at times, more challenging surgeries without compromising patient or physician safety, medical device manufacturers have created myriad solutions to vascular ligation through the development of novel tools. The speed of development, FDA approval, and dissemination of these devices into the hands of surgeons often outpaces the ability of investigators to critically evaluate comparative effectiveness of these devices.

Database:

The Medline database was searched for energy-based vessel ligation devices. To remove any perception bias against non-Covidien instruments, critical review was applied only to the devices manufactured by our company.

Conclusions:

We report on the variability present in published results and offer vital metrics for future studies. Standardized testing and reporting for measures of safety and efficacy of these surgical instruments awaits definition from a consensus group.

Introduction

Since the earliest recorded history, achieving and maintaining hemostasis has been important for successful surgery. The use of heat to staunch the flow of blood has been instrumental in furthering the ability of the practicing surgeon to take on more challenging cases. The application of heat in the surgical setting dates back to as early as 3000 BC, but conductive methods of heat transfer can result in extensive collateral tissue damage that overshadows benefits due to the uncontrolled nature of the tissue/energy interaction. A brief synopsis of the surgical energy evolution follows.

Monopolar electrosurgery

In the beginning of the 20th century, alternating electrical current found a number of applications in surgery due in large part to the work of French scientist d'Arsonval. His research demonstrated that alternating current could be passed through tissue without producing any physiologic effects outside of heating, and that if the frequency of alternating current was high enough, the unwanted side effect of muscle stimulation and contraction did not occur.¹

Collaboration between Dr. William Bovie and Dr. Harvey Cushing introduced high frequency monopolar electrosurgery to clinical use. The original Bovie generator delivered alternating current of very high frequency that created heat by ionic agitation (kinetic energy) at the tissue–electric interface. The heat affects proteins in vascular walls and plasma, creating a coagulum. Early monopolar devices did not provide for precise control of energy delivered to tissues, and unwanted collateral tissue damage, including carbonization, was a concern.

Bipolar electrosurgery: introduction

In the 1940s, bipolar electrosurgical instruments were developed. Bipolar instruments allowed surgeons to direct the application of electrosurgical energy to specific tissue held between the electrodes of the instrument, usually in the form of forceps. Bipolar electrosurgery offered potential benefits over monopolar: (1) no return electrode needed to be attached to the patient, (2) effective at lower energy levels, minimizing collateral tissue damage, (3) coagulation in wet fields, and (4) compatible for patients with pacemakers, and implantable defibrillators.

Bipolar electrosurgery: closed-loop control, first improvement

Continuous microprocessor feedback, along with the resultant modulation of thermoelectric output is at the heart of the so-called closed-loop control systems. Closed-loop control systems in electrosurgery sense changes in tissue impedance and adjust the output energy accordingly—much like cruise control on an automobile. Energy output increased in response to increased tissue independence (Ohms law), and real-time power adjustments allowed lower output settings to be used, limiting collateral damage.

Bipolar electrosurgery: vascular ligation, present state

The introduction of a foreign body (suture, clips, etc.) was known to increase the patient's potential for infection,² and the possibility of using closed-loop control to create an autologous clip composed of the patients own proteins for vascular ligation was tremendously compelling. Kennedy et al. were the first to study the use of closed-loop bipolar electrosurgery for vascular ligation.³ Working with the variables of electrical current, bipolar jaw pressure, and energy delivery time, this group developed and subsequently optimized a bipolar energy system that was able to reliably seal vascular structures. They validated the system by measuring acute vascular burst pressure. The generator they developed sensed the characteristics of the tissue being sealed and provided a customized output algorithm for optimized energy delivery. Once the endpoint of the energy delivery algorithm was reached, the generator automatically ceased delivering energy and the user was alerted of this via an audio signal.

Today, surgical hemostasis is often achieved using energy-based device technology that relies upon complex, computer-controlled, impedance-based, automated feedback algorithms. It is essential that new technology be evaluated in a consistent and objective manner that allows the relative effectiveness of competing technologies to be elucidated. Unfortunately, the use of consistent experimental methods for evaluating the safety and efficacy of energy-based devices has been the exception rather than the rule. Considering that every surgeon uses electrosurgery during nearly every operative procedure, this lack of consistency in the evaluation and comparison of electrosurgical devices is something that should be improved upon.

We propose four critical factors that should be considered when evaluating energy-based vascular ligation devices: (1) seal/tissue fusion reliability/repeatability, (2) usability, (3) thermal profile, and (4) cost/economics. This article reviews the results of published literature in which authors have evaluated the effectiveness of energy-based devices, with the purpose of identifying the best methods of determining device safety and efficacy.

Device Evaluation: Need for Standardization

Unfortunately, at present it is impossible to accurately compare the new energy-based ligation devices on the market based on the published literature. Reported data reveal significant variations in experimental design, data stratification, analysis, and collection, making the results difficult to interpret. Interestingly, even for FDA device approval, no standard testing methodology exists.

Of the four critical factors listed above, we intend to examine in detail items 1 and 3 (seal efficacy and thermal profiles). To better understand the contributing factors leading to the reported variations, all device comparison articles that (1) include LigaSure™ instruments, (2) were performed in a bench model or in vivo, and (3) described a metric of acute seal efficacy were used to analyze experimental design, methods, models, and analysis. We searched the Medline database for studies that met the above criteria with the search terms of “LigaSure” OR “bipolar vessel sealer” OR “EBVSS [electronic bipolar vessel sealing system]” “EBVS [electrothermal bipolar vessel sealing]” OR “vessel sealing” OR “vessel ligation” from 1996 to 2009. Table 1 lists the 12 articles that met our search criteria as well as the inclusion criteria.

Table 1.

Publications Selected for Analysis of Device Characterization/Comparison and Sources of Variation

	Methodology reporting
Reference	Instruments, generator, settings	Model	Setting	Vessel type	Sample size	Artery/vein	Vessel size
Kennedy et al.³	Prototype	Porcine	Bench–abattoir	NS	Described	Both	Stratified by small (1–3 mm), medium (3.1–5 mm) & large (5.1–7 mm) as well as arteries and veins
	NS
	NS
Goldstein et al.¹⁵	Partially described	Porcine	Bench–fresh excised	Ureter	Described	NA	NR
	NS
	NS
Carbonell et al.¹⁰	Partially described	Porcine	Bench–fresh excised	Neck or abdomen	Partially described	A	Stratified by small (2–3 mm), medium (4–5 mm) & large (6–7 mm)
	Could be inferred
	Partially described
Harold et al.⁴	Partially described	Porcine	Bench–fresh excised	Neck or abdomen	Described	A	Stratified by small (2–3 mm), medium (4–5 mm) & large (6–7 mm)
	Could be inferred
	Reported
Rajbabu et al.¹⁴	Reported	Porcine	Bench–fresh excised	Carotid	Described	A	Mean diameter of 5.3 mm for all
	Could be inferred
	NS
Lamberton et al.⁶	Reported	Bovine	Bench–frozen/thawed	NS	Described	A	5 mm
	NS
	NS
Newcomb et al.⁵	Reported	Porcine	Bench–fresh excised	Carotid, axillary, renal, iliac	Described	A	Stratified by small (2–3 mm), medium (4–5 mm) & large (6–7 mm)
	Reported
	Reported
Landman et al.⁸	Reported	Porcine	In vivo acute	Renal	Described	Both	Stratified by small (2–3 mm), medium (4–5 mm) & large (6–7 mm)
	Partially described						as well as arteries and veins. Mean diameter is also reported.
	Reported
Richter et al.⁷	Partially described	Porcine	In vivo acute	Renal and Splenic	Described	Both	NS
	NS
	NS
Richter et al.⁷	Partially described	Porcine	In vivo acute	Renal, Splenic, Salpingoovarial, mesenteric	Described	Both	NS
	NS
	NS
Hruby et al.¹²	Reported	Porcine	In vivo acute	Femoral, iliac, renal	Described	Both	Stratified by small (2–3 mm), medium (4–5 mm) & large (6–7 mm) as well
	NS						as arteries and veins. Mean diameter is also reported.
	Reported
Person et al.¹¹	Described	Porcine	In vivo acute	Peripheral and visceral	Described	Both	Mean diameter and SD for each category were reported.
	NR
	NR

NR, not reported; NS, not specified; NA, not applicable; A, artery; ▪=missing information.

Using this data set and exploring the range of reported burst pressure for all devices, the lowest reported mean burst strength is 128 mmHg⁴ and the highest reported burst is 1261 mmHg.⁵ This large discrepancy will be evaluated in the following sections, starting with an overview for sources of variation. A possible explanation for the inconsistency in reported data is the inherent heterogeneity of the methods employed for testing with respect to (1) definitions of failure and success criteria; (2) experimental design; and (3) variables considered in the data analysis (biologic/anatomic, user, device, etc.). A standard experimental methodology and a consistent nomenclature may allow for more judicious evaluation of the safety and efficacy of hemostatic devices.

Methodology

Robust experimental design is essential for achieving repeatable results and controlling sources of variation when possible requires a priori knowledge of the sources. Each column in Table 1 is intended as a mandatory variable that should be described in sufficient detail for the experiment to be replicated. Each column also represents an opportunity to introduce variation or even bias into the data. The instrument type, generator used, and settings selected can impact the outcomes; as many of these instruments employ different technology, comparisons should be made under the conditions of the manufacturer's instructions for use. The animal model selected has predominantly been porcine and the impact of introducing a new species type has not been studied.⁶ Performing sealing experiments in a bench setting may allow some environmental control yet negates the effect that perfusion imparts on the intended use and potential differences in data have not been studied to determine an optimal setting. There has been a remarkable lack of comment in the literature on the subject of whether the vessels sealed were completely or partially skeletonized. For reasons described below, the presence or absence of additional tissue on a vessel may introduce a substantial amount of variability across seals.

Iatrogenic factors may influence results across experiments; however, a more imperceptible source of variability recently described by Sindram et al. (accepted for publication) has been only cursorily addressed in the literature. This variability arises from the anatomic and biomechanical properties innate to arteries arising from functionally distinct vascular beds, as well as the innate differences between arteries and veins.^5,7,8 Failure to control for this biologic variability can result in biased conclusions. Briefly, the findings of Sindram et al. found a significant relationship between the collagen and elastin ratio of functionally different vascular beds directly corresponding to the burst strength of those vascular beds. These results should be replicated in a larger sample; however, this study suggests that a discreet analysis of burst pressure by artery type is essential, and that this variable should be controlled, particularly in device comparison studies to avoid biased conclusions. From a methodology standpoint, the variables that should be controlled and reported include instrument used, hardware setup (generator model, power settings, etc.), animal model, environment/setting (in vivo, bench, etc.), stratification of vessel type (femoral, renal, iliac, carotid, etc.), sample size, distinction between arteries and veins, and explanation/stratification of vessel size.

Burst pressure

A number of authors have compared the hemostatic safety and efficacy of bipolar electrosurgical and ultrasonic vessel sealing devices using the maximum burst pressure of sealed vessels as a surrogate for the efficacy of the seal. Burst testing has been demonstrated by cannulating the open end of a sealed vessel, constricting back flow via suture or an iris clamp, delivering fluid into the lumen behind (upstream of ) the seal, and incorporating a digital manometer in-line with the fluid delivery to provide a pressure reading (Fig. 1). The application of burst testing is not unique to vessel ligation research and dates back to the early 1980s in studies of the strength of vascular anastomoses.⁹

FIG. 1.

Burst testing apparatus: infusion pump, pressure meter, and cannula. The cannula is mounted on an aluminum fixture where the cannula can be fed into the open end of a sealed vessel and an iris clamp used to prevent backflow of saline during fluid delivery.

Interestingly, the minimum criteria established for safety and efficacy was likely established through an engineering mindset where a 3× factor of safety seemed reasonable (and potentially arbitrary), suggesting a “safe range” of 360–400 mmHg (120 mmHg systolic pressure ×3).³ The clinical relevance and significance of this metric has not been established. In some studies where multiple devices are compared, a definitive definition of an adequate seal is frequently omitted.^{3,4, 6–8,10,11} There is often a failure to include basic descriptive statistics such as the standard deviation, rendering comparison of respective mean burst pressures for individual devices futile (Table 2). The end conclusion of device comparison studies is typically that the device with the highest mean burst pressure over multiple seals is the most reliable. A device could fail to make an efficacious seal a number of times yet still have a mean that is high enough to seem safe. Although the mean burst pressure must clearly exceed some critical value, which ideally would be clinically relevant and standardized across all studies, it is necessary that in addition to the mean, the failure/leak rate after one device application (some authors have allowed for reapplication in the event of immediate seal leak/failure) also be reported.

Table 2.

Variable Reporting and Analysis Techniques Regarding Burst Strength from Table 1 Publications

	Burst strength reporting
Reference	Flow rate	Flow mechanism	Success/failure criteria	Analysis	Mean	SD	Range	Failure rates
Kennedy et al.³	Tried to ∼100 mmHg/s for arteries and 10 mmHg/s for veins	NS	Failures reported but criteria not explicit	Typical descriptive statistics, data stratified by dia. ranges, probability of burst <400 mmHg calculated	Reported	Reported	NR	Reported
Goldstein et al.¹⁵	Not a burst pressure study
Carbonell et al.¹⁰	Not controlled	Syringe–manual	NS	Minimal descriptive statistics, data stratified by dia. ranges, basic comparative statistics	Reported	NR	NR	NR
Harold et al.⁴	Not controlled	Syringe–manual	NS	Minimal descriptive statistics, data stratified by dia. ranges, basic comparative statistics	Reported	NR	NR	NR
Rajbabu et al.¹⁴	NA—specimens held at 2 predefined pressures	NS	Failures defined	None	NA	NA	NA	Reported
Lamberton et al.⁶	NA—specimens held at progressively increasing pressures	NS	NS	Typical descriptive statistics, lost resolution due to step increases, basic comparative statistics	Reported	Reported	NR	NR
Newcomb et al.⁵	7.5 cc/s	Automated infusion pump	Failures=burst < 50 mmHg	Minimal descriptive statistics, data stratified by dia. ranges, basic comparative statistics	Reported	NR	NR	Reported
Landman et al.⁸	NR	Pressure syringe	NS^a	Minimal descriptive statistics	Reported	NR	NR	NR
Richter et al.⁷	NS	NS	NS^a	Minimal descriptive statistics	Reported	Reported	NR	NR
Richter et al.⁷	Not a burst pressure study
Hruby et al.¹²	NS	IV Pressure Bag	Artery burst < 300 mmHg, Vein burst < 50 mmHg	Minimal descriptive statistics, data stratified by dia. ranges, basic comparative statistics of vessel characteristics as well as burst pressure	Reported	NR	NR	Reported
Person et al.¹¹	NS	NS	NS	Basic descriptive statistics, data not broken out to control for arteries/veins, basic comparative statistics	Reported	Reported	NR	NR

Burst pressure success/failure not specified.

NS, not specified; NR, not reported; ▪=missing information.

The models and experimental methods utilized for burst pressure testing are inconsistent in the literature (see Table 2). The experimental apparatus has also been inconsistent across studies. Several experimental variables may affect the maximum burst pressure achieved and many of these have not been investigated for their relative contribution to data variability. Most authors have utilized an incompressible fluid, that is, water or saline, to apply pressure to a vessel, which results in a mini aneurysm, and ultimately in a blown seal.^{3–8,10–12} A few authors¹³ used air as opposed to water or saline. Infusion to immediate failure versus infusion to a critical pressure, which is then held over time,¹⁴ or infusion increased incrementally to failure is an additional source of variability noted in the literature.⁶ Infusion rate and infusion method should be standardized and reported.

We categorized all data from the references listed in Table 1 and explored the variability in results that have been reported from these different authors and institutions. Using the LigaSure V instrument as a test case, all burst strength data reported were examined. The average burst strength of this instrument applied to all vessels is 563±266 mmHg.^5,6,11,12 Stratification of the data based on the sources of variability described above changes this number considerably; the mean burst strength of arteries sealed by the LigaSure V is 671±282 mmHg. This difference appears to be a corollary of removing the influence of veins. To investigate this more specifically, we analyzed all data (independent of ligation device) where the investigators tested and reported on both arteries and veins individually.^3,7,8,12 The difference between the mean burst strength of arteries versus veins is 303 mmHg (range −39 mmHg* to 687 mmHg). This outcome provides further compelling evidence not to combine the data for arteries and veins together when performing analyses.

We also explored the influence of the model itself. Using the reported burst strength data for all LigaSure instruments, and only articles where results were reported separately for arteries and veins, two subgroups were created: testing performed on fresh-excised vessels^3–6,10 versus testing performed in situ in the acute preclinical setting.^7,8,12 The mean burst strength of arteries sealed with LigaSure on the bench model is 634±277 mmHg (represented by 218^† data points) versus a mean of 609±139 mmHg for the acute setting (represented by 140 data points). While the mean strength is not drastically different between these two groups, the amount of variability appears greatly reduced in the preclinical setting. We do not believe that this reflects an advantage for preclinical testing; rather, some form of variability must exist, or be more expressed, in the bench setting that is not impacting the preclinical setting as heavily.

Thermal profile

The thermal characteristics of these energy-based devices contribute to the overall safety and usability of each device. Two distinct types of thermal information have been addressed: jaw temperature of an active device and thermal spread (or thermal propagation) of heat along a structure being sealed. These two factors provide important safety data to the user; maximum jaw temperature needs to be considered in laparoscopic settings or in endoscopic/minimally invasive type settings, and thermal spread needs to be considered when operating a device close to critical structures that could be harmed via thermal conduction or convection. Minimizing jaw temperature and thermal spread is typically not as critical as providing efficacious seals; however, a priori knowledge of the device's thermal profile will enable the surgeon to confidently use the device in their environment.

The reporting of thermal metrics has been sparse though nearly every burst strength article has included some type of thermal information. For many of the same biomechanical reasons listed above that contribute to variability in burst strength, these factors also impact the amount of thermal propagation measured. The three governing methods of heat transfer are conduction (direct contact), convection (fluid-based transfer), and radiation (negligible in the presence of the first two). Conduction is the most studied and reported type and is referred to most commonly as thermal spread or lateral thermal spread. The means of evaluating this parameter have been by activating the device on a structure within the jaws and then making a gross hand measurement or using histology methods to determine a more precise measurement based on tissue/cellular damage. Convective heat transfer can play a role in creating adjacent site injury by the actions of steam production or activating in a wet (blood or saline) field.

There is no consensus on how to report thermal characteristics or what to report. Some authors have reported thermal spread as the damaged tissue length from the cut edge (wherever the instrument was used to transect the structure) into healthy tissue,^10,12,15,16 whereas others have reported the length of tissue damage starting from the seal's edge into healthy tissue.^8,11 In one case, the leading pathologist chose one method for thermal spread assessment in an early articles⁸ and then modified the method of measurement in their most recent article without adding a rationale for the modification.¹² Another group chose to use thermistors at a prescribed distance from the outside of the jaw to measure peak temperature but did not follow this with any histological assessment.⁶ The measurement method becomes important when the type of technology used is considered. Devices such as LigaSure, BiClamp^®, and EnSeal^® confine tissue between two jaws during the sealing process, whereas ultrasonic devices have an active blade that is open to the environment during the division process. Arguably, any tissue confined within the jaws of a vessel sealing device is tissue that the user expects to seal and lateral thermal spread should be considered as damage that extends beyond the confines of the jaw (or sealed area).

Table 3 identifies the depth and breadth of thermal spread reporting from the Table 1 sources. Roughly 1/3 of all relevant information regarding thermal spread reporting is missing from these publications. The methods applied and stratification of the data demonstrate a lack of consistency as well as study design considerations. Some of the studies have shown a linear relationship between increasing vessel size and increasing thermal spread,^4,10 and other investigators have reported distinct differences between thermal spread on arteries and veins.^7,8,12 Using the LigaSure V as a test case, we found that the reported thermal spread ranged from as low as 0.4 mm¹¹ to as high as 6.3 mm,¹⁰ depending on the author's methods as well as the model conditions.

Table 3.

Variable Reporting and Analysis Techniques Regarding Thermal Profile Data from Table 1 Publications

	Thermal spread reporting
Reference	Instruments, generator, settings	Model	Setting	Tissue type	Artery/vein	Vessel size	Mean thermal spread	SD thermal spread	Range thermal spread	Method reported	Histology used
Kennedy et al.³	Prototype	Porcine	Bench–abattoir	NS	NA	NA	NR	NR	NR	NA	NA
	NS
	NS
Goldstein et al.¹⁵	Partially described	Porcine	Bench–fresh excised	Ureter	NA	NR	Reported	NR	Reported	Reported	Reported
	NS
	NS
Carbonell et al.¹⁰	Partially described	Porcine	Bench–fresh excised	Neck or abdomen	Both	Reported	Reported	NR	NR	Reported	Reported
	Could be inferred
	Partially described
Harold et al.⁴	Partially described	Porcine	Bench–fresh excised	Neck or abdomen	Both	Reported	Reported	NR	NR	Reported	Reported
	Could be inferred
	Reported
Rajbabu et al.¹⁴	Reported	Porcine	Bench–fresh excised	Carotid	A	Reported	No thermal spread measurements
	Could be inferred
	NS
Lamberton et al.⁶	Reported	Bovine	Bench–frozen/thawed	NS	A	Reported	Performed via thermistor at 2 mm			Reported	NA
	NS
	NS
Newcomb et al.⁵	Reported	Porcine	Bench–fresh excised	Carotid, axillary, renal, iliac	A	Reported	No thermal spread measurements
	Reported
	Reported
Landman et al.⁸	Reported	Porcine	In vivo acute	Renal	Both	Reported	Generalized	NR	NR	Reported	Reported
	Partially described
	Reported
Richter et al.⁷	Partially described	Porcine	In vivo acute	Renal and splenic	Both	NR	Generalized	NR	NR	NS	Reported
	NS
	NS
Richter et al.⁷	Partially described	Porcine	In vivo acute	Renal, splenic, salpingoovarial, mesenteric	Both	Reported	Generalized	NR	NR	Reported	NA
	NS
	NS
Hruby et al.¹²	Reported	Porcine	In vivo acute	Femoral, iliac, renal	Both	Reported	Reported	NR	NR	NS	Reported
	NS
	Reported
Person et al.¹¹	Described	Porcine	In vivo acute	Peripheral and visceral	Both	Reported	Reported	NR	Reported	Reported	Reported
	NR
	NR

NR, not reported; NS, not specified; ▪=missing information.

The other type of thermal information useful for communicating safety is jaw temperature. Knowledge of peak temperature achieved during activation as well as the latent heat capacity of the jaw(s) after activation enables the user to apply an energy-based ligation device in a safe manner for their environment as well as handle the instrument safely after application. While the notion of surface temperature mapping is not new, the application of this technology to energy ligation devices is still in its infancy. In 2003, Campbell et al. performed the first study to use infra-red camera monitoring for evaluating device thermal profiles.¹⁷ Since this inaugural publication we found only three other published works covering the topic.^18–20

The importance of understanding the overall jaw temperature is intimately tied to the usability of the device. The surgeon user should know the maximum temperature typically achieved by the jaws as well as the length of time required for the jaws to safely cool. These two metrics are indicators to the user for safe activation and safe movement postactivation. In the pursuit of recording and reporting the outcomes of thermal imaging, standardization continues to be requisite. Factors that can introduce variability to the data include tissue type used, ex vivo/in vivo setting, number of activations, open/laparoscopic environment, instrument (between manufacturers and within one manufacturer), and amount of tissue within the jaws. We encourage authors to carefully consider their sources of variation and control as many factors as possible in future investigations.

Conclusions

In summary, it is important that surgeons be critical in reviewing data and evaluating electrosurgical devices. The authors chose to use the LigaSure V as a test case for exploring the wide variability present in reporting benchmark performance of energy-based devices; this LigaSure device has the most published data of any LigaSure device. The lack of consistent testing methods and reporting of results makes the comparison of devices difficult. In addition, because there are not defined standards the true safety, efficacy and cost-effectiveness of these devices are equivocal. This review identifies the need for development of a common lexicon and standardized testing and performance evaluation of energy-based devices. The authors believe that a consensus group formed from practicing surgeons and surgical researchers would be the appropriate starting point for defining the appropriate clinical metrics of these important energy-based device parameters.

Footnotes

Disclosure Statement

K. Krugman, K. Martin, and Dr. Cosgriff are employed in the Medical/Clinical Affairs Department and receive salaries from Covidien, Energy-based Devices. Dr. Slakey has received honorarium from Energy-based Devices for prior participation on Surgeon Advisory Committees (Dr. Slakey was not compensated for his participation on this article).

References

d'Arsonval

. Action physiologique des courants alternatifs. C R Seances Soc Biol, 1891; 43:283–286.

Elek

, Conen

. The virulence of Staphylococcus pyogenes for man; a study of the problems of wound infection. Br J Exp Pathol, 1957; 38:573–586.

Kennedy

, Stranahan

, Taylor

, Chandler

. High-burst-strength, feedback-controlled bipolar vessel sealing. Surg Endosc, 1998; 12:876–878.

Harold

, Pollinger

, Matthews

, Kercher

, Sing

, Heniford

. Comparison of ultrasonic energy, bipolar thermal energy, and vascular clips for the hemostasis of small-, medium-, and large-sized arteries. Surg Endosc, 2003; 17:1228–1230.

Newcomb

, Hope

, Schmelzer

, Heath

, Norton

, Lincourt

, Heniford

, Iannitti

. Comparison of blood vessel sealing among new electrosurgical and ultrasonic devices. Surg Endosc, 2009; 23:90–96.

Lamberton

, Hsi

, Jin

, Lindler

, Jellison

, Baldwin

. Prospective comparison of four laparoscopic vessel ligation devices. J Endourol, 2008; 22:2307–2312.

Richter

, Kollmar

, Neunhoeffer

, Schilling

, Menger

, Pistorius

. Differential response of arteries and veins to bipolar vessel sealing: Evaluation of a novel reusable device. J Laparoendosc Adv Surg Tech A, 2006; 16:149–155.

Landman

, Kerbl

, Rehman

, Andreoni

, Humphrey

, Collyer

, Olweny

, Sundaram

, Clayman

. Evaluation of a vessel sealing system, bipolar electrosurgery, harmonic scalpel, titanium clips, endoscopic gastrointestinal anastomosis vascular staples and sutures for arterial and venous ligation in a porcine model. J Urol, 2003; 169:697–700.

von Smitten

, Jiborn

, Ahonen

. Bursting strength of syngeneic aortic vein grafts in the rat. Acta Chir Scand, 1981; 147:115–119.

10.

Carbonell

, Joels

, Kercher

, Matthews

, Sing

, Heniford

. A comparison of laparoscopic bipolar vessel sealing devices in the hemostasis of small-, medium-, and large-sized arteries. J Laparoendosc Adv Surg Tech A, 2003; 13:377–380.

11.

Person

, Vivas

, Ruiz

, Talcott

, Coad

, Wexner

. Comparison of four energy-based vascular sealing and cutting instruments: A porcine model. Surg Endosc, 2008; 22:534–538.

12.

Hruby

, Marruffo

, Durak

, Collins

, Pierorazio

, Humphrey

, Mansukhani

, Landman

. Evaluation of surgical energy devices for vessel sealing and peripheral energy spread in a porcine model. J Urol, 2007; 178:2689–2693.

13.

Higami

, Maruo

, Yamashita

, Shida

, Ogawa

. Histologic and physiologic evaluation of skeletonized internal thoracic artery harvesting with an ultrasonic scalpel. J Thorac Cardiovasc Surg, 2000; 120:1142–1147.

14.

Rajbabu

, Barber

, Choi

, Muir

. To knot or not to knot? Sutureless haemostasis compared to the surgeon's knot. Ann R Coll Surg Engl, 2007; 89:359–362.

15.

Goldstein

, Harold

, Lentzner

, Matthews

, Kercher

, Sing

, Pratt

, Lipford

, Heniford

. Comparison of thermal spread after ureteral ligation with the Laparo-Sonic ultrasonic shears and the Ligasure system. J Laparoendosc Adv Surg Tech A, 2002; 12:61–63.

16.

Harrell

, Kercher

, Heniford

. Energy sources in laparoscopy. Semin Laparosc Surg, 2004; 11:201–209.

17.

Campbell

, Cresswell

, Frank

, Cuschieri

. Real-time thermography during energized vessel sealing and dissection. Surg Endosc, 2003; 17:1640–1645.

18.

Kim

, Chammas

Jr. , Gewehr

, Morihisa

, Caldas

, Hayacibara

, Baptistussi

, Meyer

, Martins

. Temperature safety profile of laparoscopic devices: Harmonic ACE (ACE), Ligasure V (LV), and plasma trisector (PT) Surg Endosc, 2008; 22:1464–1469.

19.

Phillips

, Hruby

, Durak

, Lehman

, Humphrey

, Mansukhani

, Landman

. Tissue response to surgical energy devices. Urology, 2008; 71:744–748.

20.

Song

, Tang

, Campbell

, Cuschieri

. Thermal spread and heat absorbance differences between open and laparoscopic surgeries during energized dissections by electrosurgical instruments. Surg Endosc, 2009; 23:2480–2487.