Flow diagram of the reviewed studies. RCTs indicates randomized controlled trials.
Meta-analysis comparing outcomes of tonsillectomies performed using vessel sealing systems (VSS) with tonsillectomies performed using the conventional technique (cold steel and/or electrocautery). CI indicates confidence interval.
Meta-analysis comparing outcomes of tonsillectomies performed using Harmonic Scalpel (HS) with tonsillectomies performed using the conventional technique (cold steel and/or electrocautery). CI indicates confidence interval.
Meta-analysis comparing outcomes of tonsillectomies performed using Coblation with tonsillectomies performed using the conventional technique (cold steel and/or electrocautery). CI indicates confidence interval.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Alexiou VG, Salazar-Salvia MS, Jervis PN, Falagas ME. Modern Technology–Assisted vs Conventional Tonsillectomy: A Meta-analysis of Randomized Controlled Trials. Arch Otolaryngol Head Neck Surg. 2011;137(6):558–570. doi:10.1001/archoto.2011.93
To systematically review evidence regarding modern technology–assisted tonsillectomy pertaining to operative time, intraoperative and postoperative bleeding, postoperative pain, and other outcomes.
A systematic search for randomized controlled trials comparing total tonsillectomies performed using vessel sealing systems (VSS), Harmonic Scalpel (HS), or radiofrequency ablation (ie, Coblation) with the conventional technique of cold steel and/or electrocautery dissection (CS/EC). Estimation of odds ratios and 95% confidence intervals (CIs), weighted mean differences (WMD), or standardized mean difference (SMD), as appropriate.
Thirty-three randomized controlled trials studying a total of 3139 patients were included in this meta-analysis.
Main Outcome Measures
Operative time, perioperative and postoperative bleeding, and postoperative pain.
For the VSS group compared with the CS/EC group, operative time was significantly shorter (WMD), −4.09 minutes; 95% CI, −7.43 to −0.75 minutes; 760 patients), perioperative bleeding was significantly less (SMD, −1.67; −2.80 to −0.53; 355 patients), and postoperative bleeding was significantly less (odds ratio, 0.28; 0.13 to 0.61; 792 patients). Pain on the first and seventh postoperative days was significantly less in the VSS group (SMD, −1.73; 95% CI, −3.07 to −0.39; 740 patients; and SMD, −1.46; −2.35 to −0.57; 684 patients; respectively). For the HS group compared with the CS/EC group, the only studied outcome that differed significantly was perioperative bleeding, which was significantly less in the HS group (WMD −37.71 mL; 95% CI, −52.98 to −22.43 mL; 535 cases). No difference was noted between the Coblation and CS/EC groups for any of the studied outcomes.
For tonsillectomies, the Coblation and HS techniques do not provide any significant advantage compared with CS/EC. Synthesis of the limited and heterogeneous data regarding VSSs showed a significant benefit in all studied outcomes.
Tonsillectomy is one of the oldest and most commonly performed otorhinolaryngologic procedures.1 The advent of antimicrobial therapy2 and the establishment of specific surgical indications3 have substantially decreased the number of tonsillectomies. In 2003 and 2004, a total of 50 531 patients underwent tonsillectomy within English National Health Service trusts, of whom 49 765 (98.5%) were elective admissions,4 with a significant health care expenditure aggravated by the cost of postoperative morbidity.5 Modern tonsillectomy is a safe procedure. The conventional technique is cold steel and/or electrocautery dissection (CS/EC). However, postoperative morbidity in terms of pain, bleeding, and return to normal activity and diet is notable. Thus, a variety of techniques and approaches have been tested over the years. Yet, no definite consensus has been reached regarding the optimal technique with the lowest morbidity rates.
Recent advances in surgical instrumental technology have introduced energy-based devices that are able to simultaneously dissect tissue and seal vessels.The patented energy-based vessel sealing systems (VSS) designed for tonsillectomy use different technologies to provide a similar dissection-ligation effect. The LigaSure Vessel Sealing System (LS) (Valleylab, Boulder, Colorado) is a hemostatic electrosurgical device that consists of a handpiece with a ratcheted scissors mechanism that grasps and compresses the tissue and an electrosurgical generator that senses the density of the tissue bundle, automatically adjusting the amount of energy to deliver the optimal amount to denature collagen and elastin within the vessel wall and connective tissue.6 The Thermal Welding System (TWS) (Starion Instruments Corp, Saratoga, California), instead of using electric current, uses heat and pressure to simultaneously fuse and divide soft tissue and blood vessels. The manufacturer claims that it produces less heat and minimizes collateral tissue damage. It consists of a simple electrically resistant heating wire driven by low-voltage direct current.7 The third device of this type is the BiClamp (BC) (Erbe Elektromedizin GmbH, Tübingen, Germany), which combines bipolar coagulation with high-frequency current modulation automatically regulated and dosed according to tissue impedance. The manufacturer claims that it results in pulse-and-pause duration that offers gentle coagulation, avoiding carbonization. The coagulation induces vessel wall swelling and sealing of the lumen.8
The Harmonic Scalpel (HS) (Ethicon Endo-Surgery Inc, Cincinnati, Ohio), another device that has been used to safely perform tonsillectomies, uses ultrasonic energy to vibrate its blade at 55 kHz, providing simultaneous cutting and coagulation of the tissue. This vibration transfers energy to the tissue and leads to superficial denaturation and coagulation of protein by heating the tissue. The temperature of the surrounding tissue reaches 80°C. The manufacturer claims that this procedure results in precise cutting with minimal thermal damage.9
Finally, radiofrequency ablation (ie, cold ablation) is a technology that has been used successfully in various surgical specialties.10 Bipolar radiofrequency ablation, referred to by the trade name Coblation (ArthroCare Corp, Sunnyvale, California), was designed for head and neck surgery; it ablates and coagulates soft tissue by generating a field of ionized sodium molecules. The ionized plasma layer between the device tip and the tissue is produced by a radiofrequency current that passes through a medium of normal saline at a temperature of 40°C to 70°C. This process breaks molecular bonds and produces a melting tissue effect. The manufacturer claims that the lower temperature provides important benefits, such as improved precision cutting and minimal thermal damage in tissue.11
Many studies have been published regarding the use of all these surgical techniques and devices. However, the reported results are, in many cases, conflicting and controversial. Thus, we sought to systematically review and synthesize the available evidence regarding modern technology–assisted total tonsillectomy pertaining to operative time, intraoperative and postoperative bleeding, postoperative pain, and other clinical outcomes. We then compared them with the conventional technique of CS/EC by performing a systematic review and meta-analysis of randomized controlled trials (RCTs).
This meta-analysis was conducted according to the guidelines issued by the Quality of Reporting of Meta-analyses conference.12 To identify relevant RCTs, we systematically searched PubMed for articles dated through July 14, 2010, and the Cochrane Central Register of Controlled Trials for the key word tonsillectomy using only articles published after January 1, 1990. Furthermore, we reviewed the references of the included RCTs to identify additional resources. We did not seek abstracts of conference proceedings.
Two of the authors (V.G.A. and M.S.S.-S.) independently performed literature searches to locate potentially eligible reports. All RCTs comparing total tonsillectomies performed using VSS (ie, LS, TWS, and BC), HS, or Coblation with tonsillectomies performed using the conventional technique of CS/EC and reporting on operative time, intraoperative and postoperative bleeding, postoperative pain, and other clinical outcomes were considered for inclusion in this meta-analysis. The RCTs reporting on subtotal (ie, intracapsular) tonsillectomies were excluded. Thus, techniques such as microdebrider and laser-assisted tonsillectomy were excluded. Furthermore, we excluded case series reporting on fewer than 10 patients. Finally, we excluded RCTs pertaining to the use of Argon Plasma Coagulation (Erbe Elektromedizin GmbH), a technology that has been used for tonsillectomy after successfully having been used in endoscopic procedures.13 Argon Plasma Coagulation–assisted tonsillectomy never has been a widely used technique and has been almost abandoned today.
Two reviewers (V.G.A. and M.S.S.-S.) independently collected from all eligible articles the first author, year of publication, country of origin, age range of study population, number of patients enrolled, tonsillectomy techniques compared, and studied outcomes. Detailed data regarding operative time, intraoperative and postoperative bleeding, postoperative pain, and other clinical outcomes were tabulated. Also, we individually assessed randomization, generation of random numbers, details of the double-blinding procedure, information regarding withdrawals, and concealment of allocation to evaluate the methodologic quality of each RCT according to a modified Jadad score.14 One point was awarded for the specification of each criterion; the maximum score that a study could achieve was 5.
The outcome measures for this meta-analysis were operative time, intraoperative and postoperative bleeding, and postoperative pain. Postoperative bleeding was defined as any primary and secondary bleeding reported by the authors of an RCT.
Statistical analyses were performed using Review Manager (RevMan, version 5.0.24, for Linux; Nordic Cochrane Center, Cochrane Collaboration, Copenhagen, Denmark). The heterogeneity among the RCTs was assessed with the I2 statistic and a χ2 test; P <.10 was defined to note statistical significance in the analysis of heterogeneity. Publication bias was assessed according to funnel plot asymmetry. Continuous outcomes were analyzed using weighted mean difference (WMD) or standardized mean difference (SMD) if different scales were used to calculate the same outcome15 and 95% confidence intervals (CIs). The SMD is necessary to standardize the results of the studies to a uniform scale before they can be combined; it expresses the size of the intervention effect in each study relative to the study variability.15 Pooled odds outcomes of this meta-analysis were calculated by using a fixed-effects model or the DerSimonian-Laird random-effects model if statistically significant heterogeneity was noted.
In Figure 1, we present a flow diagram describing the selection process followed to identify the pool of RCTs included in the meta-analysis. The PubMed search yielded 3929 potentially relevant articles; the search using the Cochrane Central Register of Controlled Trials did not reveal any additional relevant RCTs. An additional 6 articles were retrieved using the references cited in the retrieved articles. In total, 33 articles16-48 fulfilled the inclusion criteria for this meta-analysis.
In Table 1, we summarize the main characteristics of the studies included in this meta-analysis: 33 RCTs studying a total of 3139 patients (median, 68; range, 20-316). Seven studies compared tonsillectomies performed using VSS (ie, 4, 2, and 1 studies for TWS, LS, and BC, respectively) with tonsillectomies performed using CS/CS. Eleven and 16 studies compared conventional tonsillectomy using CS/EC with tonsillectomies performed using HS and Coblation, respectively. Eleven studies were conducted among children only and 8 among adults only. Fourteen studies had a mixed-age population. Five studies randomized tonsils instead of patients. Regarding quality score, 16 of 33 RCTs achieved a score of 3, 3 studies achieved a score of 4, and 1 study achieved the maximum score of 5. The rest of the included studies had a low score of 1 or 2 points.
In Table 2A and Table 2B, we present various clinical outcomes summarized by our study. Data regarding operative time, perioperative and postoperative bleeding, and postoperative pain were reported by most of the included studies; thus, a meta-analysis was feasible. Return to normal diet and activity, patient satisfaction, need for analgesics, tonsillar fossa healing, postoperative nausea and vomiting, need for antibiotics, and other complications, such as hematoma and taste changes, were reported only by certain studies.
In Figure 2, we present a meta-analysis comparing outcomes of tonsillectomies performed using VSS with tonsillectomies performed using the conventional technique (ie, CS/EC). Operative time was significantly shorter in the VSS group (WMD, −4.09 minutes; 95% CI, −7.43 to −0.75 minutes; 760 patients). Perioperative bleeding was noted to be less using VSS compared with the conventional technique (SMD, −1.67; 95% CI, −2.80 to −0.53; 355 patients). Moreover, VSS yielded significantly less postoperative bleeding compared with the conventional technique (odds ratio [OR], 0.28; 95% CI, 0.13 to 0.61; 792 patients). Pain on the first and seventh postoperative days was significantly less in the VSS group (SMD, −1.73; 95% CI, −3.07 to −0.39; 740 patients; and SMD, −1.46; −2.35 to −0.57; 684 patients, respectively).
In Figure 3, we present the meta-analysis comparing outcomes of tonsillectomies performed using HS with tonsillectomies performed using CS/EC. No significant difference was found between the compared groups regarding operative time (WMD, −0.10 minutes; 95% CI, −6.26 to 6.05 minutes; 655 cases), postoperative bleeding (OR, 0.78; 95% CI, 0.50 to 1.23; 1473 cases), and averaged postoperative pain (SMD, −0.38; 95% CI, −1.20 to 0.43; 517 cases). However, perioperative bleeding was significantly less in the HS group (WMD, −37.71 mL; 95% CI, −52.98 to −22.43 mL; 535 cases).
In Figure 4, we present the meta-analysis comparing outcomes of tonsillectomies performed using Coblation with tonsillectomies performed using the conventional technique. No significant difference was found between the compared groups for any of the studied outcomes: operative time (WMD, −0.35 minutes; 95% CI, −2.84 to 2.13 minutes; 406 cases), perioperative bleeding (WMD, −4.22 mL; 95% CI, −14.39 to 5.95 mL; 292 cases), postoperative bleeding (OR, 0.99; 95% CI, 0.58 to 1.69; 1092 cases), and postoperative pain (SMD, −1.56; 95% CI, −3.48 to 0.35; 313 cases).
The main finding of this meta-analysis of RCTs is that the Coblation and HS tonsillectomy techniques that have been used during the past decade in an attempt to decrease postoperative morbidity in terms of pain and bleeding do not provide any significant advantage over the conventional CS/EC technique. The only outcome that differed significantly in tonsillectomies performed using HS compared with those using CS/EC was perioperative bleeding. However, it is commonly accepted that this variable does not have any measurable clinical significance. Regarding postoperative bleeding and pain in tonsillectomies performed using HS, the meta-analysis plot (Figure 3) identified 2 outlier studies19,26 reporting significant benefits that were not verified by the total OR and SMD, respectively. However, none of the 16 RCTs comparing Coblation with CS/EC showed significant benefits regarding postoperative bleeding. This finding was verified by the meta-analysis, including a total of 1092 patients; no benefit of Coblation over CS/EC was observed regarding the most important clinical outcome of postoperative bleeding. Regarding other outcomes, the pooled population was considerably smaller: 406, 292, and 303 patients for operative time, perioperative bleeding, and postoperative pain, respectively. Thus, results should be considered cautiously. In all outcomes, HS and Coblation were at least equivalent to the conventional technique. However, equivalence may not be enough to justify a change in the current clinical practice.
The meta-analysis of the RCTs reporting the use of VSS showed a statistically significant difference in favor of this technique compared with CS/EC for all studied outcomes: operative time, perioperative bleeding, and, most important, postoperative bleeding and pain. Of interest, none of the included VSS studies was highly powered enough to show statistical significance in the clinical outcome of postoperative bleeding. Synthesis of the available evidence from 792 patients included in 7 RCTs showed that the meta-analysis has higher statistical power to detect an effect on the rare but clinically important outcome of postoperative bleeding. Furthermore, 4 of the 5 studies18-22 that evaluated pain in the first postoperative day showed that it was significantly less for the VSS group. One study17 did not show any significant difference, and another16 showed that pain was significantly greater in the VSS group. Combining all the available evidence helps resolve such controversies and reach a safer conclusion regarding postoperative pain.
This systematic review has included 4 studies of TWS,16,18,20,21 2 studies of LS,19,22 and 1 study of BC17 that have not equally addressed all the outcomes included in the meta-analysis. Thus, splitting the VSS RCTs into subgroups would not permit us to draw any meaningful conclusions. When further evidence from ongoing or future studies of these devices is available, it would be useful to perform subanalyses of the different VSS devices, as well as a comparative analysis. Vessel sealing systems constitute a new technique; therefore, one should consider many issues before trying to implement it in everyday clinical practice: cost, training issues, the learning curve for novice surgeons and additional morbidity associated with the procedure, and, most important, limited evidence to support a change from the use of CS/EC, the tonsillectomy procedure that has proven safe for several decades.
Regarding the quality score, 20 of 33 RCTs achieved a score of 3 or higher, which is generally acceptable for inclusion in a meta-analysis. Only 4 studies18,26,45,46 had a low score of 1 and accounted for less than 12% of the weight of the meta-analysis. This finding may be attributed to the nature of this surgical intervention that does not permit double-blinding for most of the studied outcomes. Thus, we chose to maintain a lower threshold for inclusion and to include the studies with weak evidence. Other methodologic issues regarding the included RCTs pertain mainly to the data collection methods used (ie, questionnaires, daily diaries, and visual analog scales) and short follow-up periods. These issues may have introduced a degree of bias in our analysis that needs to be acknowledged.
The results of this study should be considered in view of several limitations. The included studies have significant methodologic heterogeneity, and various scales were used to measure postoperative pain and perioperative bleeding. To overcome this limitation, the SMD was calculated if different scales were used to measure the same outcome.14 The SMD is appropriate for depicting significant differences but does not provide any meaningful quantification. In general, for continuous outcomes (eg, pain scores, blood loss, and operative time), the noted statistical heterogeneity was significant. The random-effects model was used to deal with this issue. However, the clinical diversity of the included studies was not strong. The absence of outliers, especially for the VSS analysis, may indicate a certain degree of publication bias and may be attributed to the fact that many of the included studies were sponsored by the manufacturers of the relevant devices. However, we did not note any significant funnel plot asymmetry for postoperative bleeding in any of the pooled analyses. In all cases, the well-described selective publication of RCTs49 and the “file drawer effect”50 also should be acknowledged when interpreting the results of this meta-analysis. Moreover, the quality score of the included RCTs is relatively low, which may be attributed to the nature of this surgical intervention that does not permit double-blinding for most of the studied outcomes. Finally, as previously mentioned, the VSS analysis pooled studies of 3 different devices; manufacturers often claim that their products have unique characteristics that make them more efficient than those of their competitors. The available evidence that directly compares different VSS devices in tonsillectomy51 and in other surgical procedures8 shows that these devices are, in fact, similar or equivalent in terms of surgical effectiveness and clinical outcomes. Still, the results of this analysis should be interpreted with caution.
In conclusion, despite its limitations, this meta-analysis provides evidence that the use of Coblation and HS for tonsillectomy is equivalent to the use of the conventional CS/ES technique. Surgeon experience, training, and preferences, as well as cost-effectiveness criteria, should be considered. However, statistical synthesis of the limited available data regarding VSS showed a significant benefit in all studied outcomes. Well-designed and well-performed RCTs are warranted to further investigate the effectiveness of VSS techniques for tonsillectomy.
Correspondence: Vangelis G. Alexiou, MD, MSc, Northampton House, Flat P14, Wellington St, Northampton, England (email@example.com).
Submitted for Publication: January 9, 2011; final revision received February 23, 2011; accepted March 21, 2011.
Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Alexiou. Acquisition of data: Alexiou and Salazar-Salvia. Analysis and interpretation of data: Alexiou, Salazar-Salvia, Jervis, and Falagas. Drafting of the manuscript: Alexiou and Salazar-Salvia. Critical revision of the manuscript for important intellectual content: Alexiou, Jervis, and Falagas. Statistical analysis: Alexiou. Administrative, technical, and material support: Salazar-Salvia. Study supervision: Jervis and Falagas.
Financial Disclosure: None reported.
Additional Information: Accepted for oral presentation at the First Congress of the Confederation of the European Otorhinolaryngology–Head and Neck Surgery; July 2-6, 2011; Barcelona, Spain.
Create a personal account or sign in to: