1.1

Traditional tonsillectomy techniques

Traditional tonsillectomy techniques using “cold” steel (e.g. scalpel, snare, guillotine) are infrequently used in the United States due to a high rate of intra-operative blood loss . Monopolar electrosurgical devices supplanted “cold” techniques, resulting in decreased intra-operative blood loss ; however, there are clinicians who feel these devices might lead to increased post-operative pain due to the high level of continuous energy utilized.

1.2

Bipolar radiofrequency ablation

Bipolar radiofrequency ablation (Coblation, Arthrocare, Inc., Austin, TX) is a form of high-frequency electrosurgery in which radiofrequency energy passes through a conductive medium between two active electrodes, producing a plasma field. The resultant charge-carrying ions have sufficient energy to break organic molecular bonds resulting in disintegration of tissue. The bipolar probe or wand is used to accurately remove tonsil and adenoid tissue, with less heat generation (40–60 °C) than traditional monopolar electrosurgical devices and minimal damage to surrounding structures . Although this advanced technology is purported to lead to a less painful recovery, a Cochrane review on this subject concluded that current evidence supporting the benefits of bipolar radiofrequency ablation over other techniques is lacking .

1.3

Pulsed-electron avalanche knife

The pulsed-electron avalanche knife (PEAK) PlasmaBlade device (Medtronic Inc., Minneapolis, MN) is a soft tissue dissection instrument that uses very brief, high-frequency pulses of radiofrequency (RF) energy to induce electrical plasma along the edge of a thin (12.5 um), 99.5% insulated electrode . Due to the low duty cycle from RF pulsing and proprietary thermal protection shield (TPS) insulation technology, the device uses less total energy and, like bipolar radiofrequency ablation, operates at significantly lower temperatures (40–170 °C) than traditional electrocautery (200–350 °C), presumably allowing dissection of tonsil and adenoid tissue and hemostatic control with less collateral damage than traditional electrocautery .

1.4

Purpose of present study

The purpose of the present study was to compare post-operative pain and complications in children undergoing adenotonsillectomy with bipolar radiofrequency ablation vs. pulsed-electron avalanche knife technology. The primary null hypothesis was that adenotonsillectomy would be equally painful with pulsed-electron avalanche knife vs. bipolar radiofrequency ablation. The study also was designed to compare pulsed-electron avalanche knife surgery vs. bipolar radiofrequency ablation with respect to 1) number of post-operative analgesic medication doses (narcotic and non-narcotic), and 2) post-operative hemorrhage rate. The type, frequency and the amount of analgesic medication taken by each patient also served as an indirect indication of the amount post-operative pain they experienced.

2.1

Enrollment mechanism

Children were consecutively enrolled until the quota for the study was reached. Two experimental arms were utilized: pulsed-electron avalanche knife adenotonsillectomy and bipolar radiofrequency ablation adenotonsillectomy. The study included 100 total subjects (50 in each arm). The first 25 patients enrolled were assigned to undergo bipolar radiofrequency ablation; the second group of 25 patients was assigned to undergo pulsed-electron avalanche knife surgery; the third group of 25 patients was assigned to undergo bipolar radiofrequency ablation; the fourth and final group of 25 patients was assigned to undergo pulsed-electron avalanche knife surgery.

2.2

Surgical procedure techniques

All procedures were performed in the outpatient surgical center affiliated with the Center for Pediatric ENT – Head and Neck Surgery, and all procedures were performed by one of three senior fellowship-trained pediatric otolaryngologists (ZS, DJK, DLM). All three surgeons performed both techniques; the decision of which technique to use was pre-determined based on the assignment scheme (see description in previous paragraph). In this practice, bipolar radiofrequency ablation had been performed for many years as the primary mechanism of adenotonsillectomy. In an attempt to minimize “learning curve” bias, each surgeon performed as many pulsed-electron avalanche knife adenotonsillectomies as possible in the 6 months prior to initiation of the study (over 20 cases for each surgeon). General anesthesia with orotracheal intubation was identical for all patients. All tonsillectomies were extra-capsular. As per American Academy of Otolaryngology tonsillectomy guidelines , no peri-operative antibiotics were given, no local anesthetic infiltration was used, and every patient received a single IV dose of dexamethasone during surgery .

Bipolar radiofrequency ablation procedures were performed with the Procise XP wand (Arthrocare Corp., Sunnyvale, CA), using an ablation setting of 6 or 7 and a coagulation setting of 3 to 5 for tonsillectomy and an ablation setting of 9 and a coagulation setting of 5 for adenoidectomy. PEAK procedures were performed with the PlasmaBlade TnA device (Medtronic, Inc., Minneapolis, MN) on a coagulation setting of 4 for tonsillectomy and 8 for adenoidectomy (coagulation was used for dissection and hemostasis). Patients were discharged home the same day as the surgery, and were encouraged to use over-the-counter ibuprofen every 6 h as-needed, acetaminophen every 4 h as-needed, and were given a prescription for acetaminophen and hydrocodone, dosed based upon weight, to be used as-needed for severe pain unresponsive to over-the-counter analgesics.

2.3

Data collection

One of two trained operating room registered nurses collected subject data prior to same-day discharge (day #0), and then made daily telephone calls to parents or legal guardians from post-operative day #1 through post-operative day #14. One nurse was required to contact the assigned patient’s parents or legal guardians throughout the entire 14-day postoperative period. Parents or legal guardians were specifically asked the following items: 1) to quantify maximal pain that day per a validated 11-point proxy-evaluated pediatric pain scale (0 = no pain, 10 = intense pain) (Children’s and Infants’ Postoperative Pain Scale, CHIPPS) ; 2) whether or not any doses of narcotic and non-narcotic medication were needed that day, and the number of doses; 3) whether or not there was any bleeding that day, and how it was managed (e.g. conservatively at home; in the ENT’s office; in the emergency room; in the hospital overnight with observation; or in the operating room). Primary bleeding was defined as bleeding on the day of surgery, and secondary bleeding as hemorrhage that began on post-operative day #1 or later.

2.4

Statistical analysis

One goal of the study was to compare the amount of post-operative pain between pulsed-electron avalanche knife surgery and bipolar radiofrequency ablation. The null hypothesis was that the two population means would be equal. A mean difference of 3 points on the pain scale was considered to be clinically relevant, since the pain scale we utilized has shown that scores between 0 and 3 are obtained in pain-free post-operative situations, whereas scores of 4 of higher are obtained in painful post-operative situations . Prior research has shown that the CHIPPS score was 3.0 in pain-free post-surgical pediatric situations, vs. 5.7 in painful situations (e.g. difference of 2.7 points) . This effect was selected as the smallest difference that would be important to detect, in that any smaller effect would likely not be of clinical or substantive significance. The criterion for significance was determined by a two-tailed t-test with significance (alpha) of 0.050. With a proposed sample size of 45 subjects in each group, the study was calculated to have a power of 80.4% to yield a statistically significant result. (We were planning to recruit 100 subjects, but prepared our power analysis for 90 subjects, assuming a potential 10% drop-out rate.)

Another goal of the study was to estimate the mean difference between the two populations. On average, a study of this design would enable us to report the mean difference in pain scale scores with a precision (95.0% confidence level) of plus/minus 2.09 points. For example, an observed difference of 3.0 would be reported with a 95.0% confidence interval of 0.91 to 5.09. The precision estimated here is the median precision and will vary as a function of standard deviation. Median pain scores between the two study populations were compared with a two-sample Wilcoxon rank-sum (Mann–Whitney U) test.

2.1

Enrollment mechanism

Children were consecutively enrolled until the quota for the study was reached. Two experimental arms were utilized: pulsed-electron avalanche knife adenotonsillectomy and bipolar radiofrequency ablation adenotonsillectomy. The study included 100 total subjects (50 in each arm). The first 25 patients enrolled were assigned to undergo bipolar radiofrequency ablation; the second group of 25 patients was assigned to undergo pulsed-electron avalanche knife surgery; the third group of 25 patients was assigned to undergo bipolar radiofrequency ablation; the fourth and final group of 25 patients was assigned to undergo pulsed-electron avalanche knife surgery.

2.2

Surgical procedure techniques

All procedures were performed in the outpatient surgical center affiliated with the Center for Pediatric ENT – Head and Neck Surgery, and all procedures were performed by one of three senior fellowship-trained pediatric otolaryngologists (ZS, DJK, DLM). All three surgeons performed both techniques; the decision of which technique to use was pre-determined based on the assignment scheme (see description in previous paragraph). In this practice, bipolar radiofrequency ablation had been performed for many years as the primary mechanism of adenotonsillectomy. In an attempt to minimize “learning curve” bias, each surgeon performed as many pulsed-electron avalanche knife adenotonsillectomies as possible in the 6 months prior to initiation of the study (over 20 cases for each surgeon). General anesthesia with orotracheal intubation was identical for all patients. All tonsillectomies were extra-capsular. As per American Academy of Otolaryngology tonsillectomy guidelines , no peri-operative antibiotics were given, no local anesthetic infiltration was used, and every patient received a single IV dose of dexamethasone during surgery .

Bipolar radiofrequency ablation procedures were performed with the Procise XP wand (Arthrocare Corp., Sunnyvale, CA), using an ablation setting of 6 or 7 and a coagulation setting of 3 to 5 for tonsillectomy and an ablation setting of 9 and a coagulation setting of 5 for adenoidectomy. PEAK procedures were performed with the PlasmaBlade TnA device (Medtronic, Inc., Minneapolis, MN) on a coagulation setting of 4 for tonsillectomy and 8 for adenoidectomy (coagulation was used for dissection and hemostasis). Patients were discharged home the same day as the surgery, and were encouraged to use over-the-counter ibuprofen every 6 h as-needed, acetaminophen every 4 h as-needed, and were given a prescription for acetaminophen and hydrocodone, dosed based upon weight, to be used as-needed for severe pain unresponsive to over-the-counter analgesics.

2.3

Data collection

One of two trained operating room registered nurses collected subject data prior to same-day discharge (day #0), and then made daily telephone calls to parents or legal guardians from post-operative day #1 through post-operative day #14. One nurse was required to contact the assigned patient’s parents or legal guardians throughout the entire 14-day postoperative period. Parents or legal guardians were specifically asked the following items: 1) to quantify maximal pain that day per a validated 11-point proxy-evaluated pediatric pain scale (0 = no pain, 10 = intense pain) (Children’s and Infants’ Postoperative Pain Scale, CHIPPS) ; 2) whether or not any doses of narcotic and non-narcotic medication were needed that day, and the number of doses; 3) whether or not there was any bleeding that day, and how it was managed (e.g. conservatively at home; in the ENT’s office; in the emergency room; in the hospital overnight with observation; or in the operating room). Primary bleeding was defined as bleeding on the day of surgery, and secondary bleeding as hemorrhage that began on post-operative day #1 or later.

2.4

Statistical analysis

One goal of the study was to compare the amount of post-operative pain between pulsed-electron avalanche knife surgery and bipolar radiofrequency ablation. The null hypothesis was that the two population means would be equal. A mean difference of 3 points on the pain scale was considered to be clinically relevant, since the pain scale we utilized has shown that scores between 0 and 3 are obtained in pain-free post-operative situations, whereas scores of 4 of higher are obtained in painful post-operative situations . Prior research has shown that the CHIPPS score was 3.0 in pain-free post-surgical pediatric situations, vs. 5.7 in painful situations (e.g. difference of 2.7 points) . This effect was selected as the smallest difference that would be important to detect, in that any smaller effect would likely not be of clinical or substantive significance. The criterion for significance was determined by a two-tailed t-test with significance (alpha) of 0.050. With a proposed sample size of 45 subjects in each group, the study was calculated to have a power of 80.4% to yield a statistically significant result. (We were planning to recruit 100 subjects, but prepared our power analysis for 90 subjects, assuming a potential 10% drop-out rate.)

Another goal of the study was to estimate the mean difference between the two populations. On average, a study of this design would enable us to report the mean difference in pain scale scores with a precision (95.0% confidence level) of plus/minus 2.09 points. For example, an observed difference of 3.0 would be reported with a 95.0% confidence interval of 0.91 to 5.09. The precision estimated here is the median precision and will vary as a function of standard deviation. Median pain scores between the two study populations were compared with a two-sample Wilcoxon rank-sum (Mann–Whitney U) test.