Study reference

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

Bias due to inadequate blinding of participants to treatment allocation?³

Bias due to inadequate blinding of care providers to treatment allocation?³

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

Bias due to selective outcome reporting on basis of the results?⁴

Bias due to loss to follow-up?⁵

Bias due to violation of

intention to treat analysis?⁶

Coblation vs. Cold dissection

Omrani et al, 2012

Random number table

Unclear * concealment not described

Unlikely

Unlikely

Unlikely

Unlikely

Unclear * loss to follow-up not described. Secondary hemorrhage is passively measured (patients need to call), so not very reliable

Unlikely

+ Same surgeon with much experience in both techniques.

-not all outcome measures are given in absolute numbers; in text some numbers are incorrectly rounded. Overall nog very well structured and written.

Paramasivan et al, 2012

Randomly selected

Unclear* concealment not described

Unlikely* not blinded, but bias is not likely

Unlikely

Unlikely (blinded)

Unlikely

Unclear* not reported

Unclear

Not very well written, outcomes are grouped without clear explanation

Shapiro et al, 2007

Not described

Unlikely

Unclear* patients were not blinded, unclear if this caused bias

Unlikely

Unclear

Likely** stratified data without explanation why, data not given in numbers but only in figures, definitions of outcome measures unclear, time in operation room is not reported in results section.

Unclear

Unclear

Laser vs. Cold dissection

Ishlah et al, 2005

Random, not described

Unclear * not described

Unlikely ** blinding of participants is not stated, however this is not likely to cause bias (objective outcome measures)

Unlikely* operator is not blinded, but risk of bias is unlikely

Unclear * not stated whether outcome assessors are blinded

Unlikely

Unlikely * patients who did not come for follow up were contacted, and interviewed by telephone

Unlikely

Not very well written English, outcomes reported as means (range), no SD, no critical appraisal of own study. Overall methodology weak.

Kothar et ali, 2002

Random by computer

Unclear * not described

Unlikely

Unlikely * operator is not blinded, but since outcome measurement is done by others, risk of bias is unlikely.

Unlikely

Unlikely

Unclear * loss to follow-up is not described

Unlikely * not described, but since intervention was surgery which was performed right after randomisation risk of bias is low.

Some outcomes not reported in absolute numbers, which is less transparent. Also, in and exclusion criteria were not described, way of VAS recording was not clearly described.

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison/ control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Coblation vs. Cold dissection

Omran et ali, 2012

Type of study:

RCT

Setting:

Hospital

Country:

Iran

Source of funding:

None reported

Inclusion criteria: chronic recurrent tonsillitis, OSAS

Exclusion criteria: history of peritonsillar abscess, ongoing analgesic use for medical conditions, bleeding disorders.

N total at baseline: 94

Intervention: 47

Control: 47

Important prognostic factors²:

age ± SD:

I: 11.2

C: 11.8

Groups comparable at baseline? unclear

AthroCare 2 assisted Evac-70 coblator wand

No antibiotics, topical or local anaesthesia, or other medications were administered before surgery. Aesthetic approach was the same for all patients: IV fentanyl, propofol 2mg/kg, atracurium 0.5mg/kg following intubation. Maintenance achieved by mixture of nitrous oxide, oxygen, propofol.

Steel cold dissection

Length of follow-up:

Not described

Loss-to-follow-up:

Intervention:

N=2

Control:

N=1

Reasons: unknown

Incomplete outcome data:

Not described.

Pain (VAS)

No absolute outcomes: exclusion of this outcome measure

Intraoperative blood loss (volume of blood in suction bottle ml)

I: 103.4±28.7

C:   161.5±11.8

P<0.001

Primary hemorrhage (bleeding <24h after surgery, patients were asked to call if occurred, n(%))

I: 1 (2.12)

C: 2 (4.26)

p>0.05

Secondary hemorrhage (bleeding >24h after surgery, patients were asked to call if occurred, n(%))

I: 1 (2.12)

C: 5 (10.64)

p>0.05

Time to return to normal diet (questionnaire, mean days)

I: 6.27±1.07

C: 9.25±1.30

P<0.001

Time to return to normal activities (questionnaire)

I: 7.63±1.16

C: 11.70±1.68

P<0.001

Paramasivan et al, 2012

Type of study:

RCT

Setting:

Hospital

Country:

India

Source of funding:

Not reported

Inclusion criteria: age 5-12years, tonsillar and adenoid hypertrophy causing obstructive sleep apnoea

Exclusion criteria: septic tonsils, other general contraindications for (A)TE.

N total at baseline: 100

Intervention: 50

Control: 50

Important prognostic factors²:

age ± SD:

I: 11.2

C: 11.8

Groups comparable at baseline? unclear

AthroCare 2 assisted Evac-70 coblator wand

Adenoid was curetted using St. Clair Thomson adenoid curette and bleeding arrested using postnasal pack, tonsillectomy was performed by blunt dissection and bleeding arrested with ligatures.

Length of follow-up:

6h and 4days postoperative

Loss-to-follow-up:

Not described

Incomplete outcome data:

Not described.

Pain (Wong Bakers FACES pain rating + questionnaire)

Day 1 (high pain score 5-8)

I: 2/50 (4%)

C: 12/50 (24%)

Day 4

I: 1/50 (2%)

C: 8/50 (16%)

p-value: <0.001

Intraoperative blood loss (volume of blood in suction bottle ml)

I: 7.4

C: 56.3

P<0.001

Primary hemorrhage (bleeding <24h after surgery, n)

I: 1/50 (2%)

C: 1/50 (2%)

Secondary hemorrhage (bleeding >24h after surgery, n)

I: 0/50

C: 0/50

Shapiro et al, 2007

Type of study:

RCT

Setting:

Hospital

Country:

USA

Source of funding: received materials from industry.

Inclusion criteria:

Age 2-16, undergoing outpatient adenotonsillectomy

Exclusion criteria:

Severe comorbidities: systemic disease, known bleeding diathesis, craniofacial disorders, chromosomal abnormalities, motor/developmental delays.

N total at baseline: 47

Intervention: 24

Control: 23

Important prognostic factors²:

Age mean

I: 7.39

C: 6.1

Sex:

I: 43.5% M

C: 65.2% M

Groups comparable at baseline?

Coblatie using EVAC T&A handpiece.

Anaesthesia and recovery room techniques were standardized

Cold-dissection tonsillectomy was performed using curved Metzenbaum scissors to enter peritonsillar space, blunt dissection to remove the tonsil from superior to inferior, and wire snare to divide the inferior pole. Hemostasis with bipolar cautery.

Cold dissection adenotonsillectomy with adenoid curette, and hemostasis with suction Bovie cautery.

Anaesthesia and recovery room techniques were standardized

Length of follow-up:

12 months

Loss-to-follow-up:

None

Incomplete outcome data:

Intervention:

N=1

Reasons (describe): incomplete postoperative data

Control:

N=-

Postoperative pain (Wong-Baker Faces pain scale, for 14 days): no difference (no absolute numbers given)

P=0.296

Intraoperative bleeding (ml blood loss):

Higher in control than the intervention group (no absolute numbers given)

P<0.001

Primary hemorrhage, n(%)during hospital stay

Not reported

secondary hemorrhage, (n)

I: 1 (treated with operative control of bleeding without blood transfusion)

C: 0

Time before resumption of normal diet (average day)

I: 4

C: 3

P=0.982

Time before resumption of normal activities (average day)

I: 2-3

C: 1

P=0.631

Adverse effects; no unscheduled telephone calls, physician office visits or dehydration admissions were recorded

Indicatie wordt niet vermeld

Laser vs. Cold dissection

Ishlah, 2005

Type of study: RCT

Setting: hospital

Country: Malaysia

Source of funding: not stated

Inclusion criteria:

-

Exclusion criteria:

-

N total at baseline: 60

Intervention: 30

Control: 30

Important prognostic factors²:

age (range):

I:15.3 (6-49)

C:16.8 (6-42)

Sex:

I: 60% M

C: 56% M

Indication for tonsillectomy

Chr/rec. Tonsillitis (n, %)

I: 24 (80%)

C: 26 (86.7%)

Groups comparable at baseline? Yes

Laser tonsillectomy using the Sharplan model 1080S, power set between 10-15 watts and laser was used to separate the tonsil tissue from tonsillar bed.

Hemostasis was secured by suture ligation or diathermy

Intramuscular and oral Tramal was a standard pain control regimen used in all patients, given every 8 hours. Patients were discharged 1 or 2 days after operation.

Conventional dissection technique using blunt dissector to dissect tonsillar tissue from tonsillar bed and tonsil snare at the tissue attachment in the inferior pole.

Hemostasis was secured by suture ligation or diathermy

Intramuscular and oral Tramal was a standard pain control regimen used in all patients, given every 8 hours. Patients were discharged 1 or 2 days after operation.

Length of follow-up: 2 weeks and 1 month after surgery

Loss-to-follow-up:

Not stated

Incomplete outcome data:

Not stated

I vs. C

Intraoperative blood loss (measured weighing tonsil swab and the amount in the suction bottle), ml (range)

94.8 (19-256) vs. 151.7 (16-654) p=0.014

Pain scores (VAS, from day 1 to 2 weeks) mean

24.93 vs. 29.17 p=0.43

Readmitted to hospital 2 vs. 2. For: postoperative fever, halitosis and persistent blood-stained saliva. All patients were treated conservatively.

Kothari et al, 2002

Type of study:

RCT

Setting:

Hospital

Country:

UK

Source of funding: grant from Oxford Regional Health Authority

Inclusion criteria: -

Exclusion criteria :-

N total at baseline:151

Intervention:79

Control:72

Important prognostic factors:

Mean age :

I: 15

C:16.5

Sex:

I: % M

C: % M

Groups comparable at baseline?

Describe intervention (treatment/procedure/test):

KTP 532 (laserscope) laser at 10 W continuous beam for initial dissection and a defocused beam for haemostasis. Laser was delivered via 0.6mm Endostat^TM fibre

Standard postoperative care: discharge with dispersible analgesia and mouth washes (10% hydrogen peroxide after food and Difflam mouthwash before food) sufficient for 10 days.

Describe control (treatment/procedure/test):

Standard cold dissection and snare technique. Haemostasis was achieved with bipolar diathermy of ligation.

Standard postoperative care: discharge with dispersible analgesia and mouth washes (10% hydrogen peroxide after food and Difflam mouthwash before food) sufficient for 10 days.

Length of follow-up: Reviewed 1 day postoperative and 1 month postoperative.

Loss-to-follow-up:

Not stated

Incomplete outcome data:

Not stated

I vs. C.

Peroperative bleeding (weight of saturated swabs and blood volume in suction bottle), median ml (range):

20 (0-320) vs. 95 (4-400)

P<0.001

Postoperative pain (VAS)

Day of surgery: not significant (p=0.9)

First postoperative day: larger proportion of patients from laser group had moderate to severe pain (p=0.055)

One month: 30% experiencing pain vs 18% (p=0.056)

Time to return to normal diet postoperatively (SF-36) 1=not at all, 5=normal

See figure 2 below

Quality of life (patient comfort and mood (SF-36) 1=silent, 5=cheerful)

See Figure 3 below

Haemorrhage

(a) primary 11.3% (n=9) vs 5.5 (n=4)P=0.3

 (b) secondary (with hospital readmission) 8% (n=6) vs. 4% (n=3) p=0.5

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Coblation vs. Cold dissection

Burton and Doree, 2009

PS., study characteristics and results are extracted from the SR only for studies comparing coblation with cold steel dissection.

SR of  RCT’s comparing the effectiveness of coblation tonsillectomy vs. Other surgical methods

Literature search up to December 2006

A: Anthony, 2006

B: Bäck, 2001

C: Jayasinghe, 2005

E: Philpott, 2005

Study design: RCT

Setting and Country:

A: hospital, UK

B: hospital, Finland

C: tertiary referral centre, UK

E: hospital, UK

Source of funding:

Not stated

Inclusion criteria:

Randomisation of patients (not tonsils),

Adults, Children,

Day-case or inpatient setting, adenoidectomy

Exclusion criteria: emergency tonsillectomy in adults and trials in which participants underwent additional procedures other than adenoidectomy

9 studies included

Important patient characteristics at baseline:

N, age range

A: 27, 3-64

B: 40, 18-65

C: 60, 18-65

E: 92, 18-45

Groups comparable at baseline? yes

Describe intervention (n)

Coblation tonsillectomy (ArthroCare, using Coblator I or II)

A: coblation tonsillectomy (136)

B: bipolar radiofrequency thermal ablation tonsillectomy (18)

C: coblation tonsillectomy (30)

E: coblation tonsillectomy (43)

adenotonsillectomy (17)

Perioperative medication:

A: participants were issued with analgesics for home use (no information type or dosage)

B: premedication (if requested) 10mg oral diazepam and postoperative analgesia as requires (100mg ketoprofen + 500mg acetaminophen and 30mg codeine phosphate)

C: post-operative analgesia was standardized (no information type and dosage)

E: Co-codamol (2 tabs, 4x day) and diclofenac (50mg 3xday) post discharge

Describe  control (n):

Other surgical methods

A: standard cold steel dissection tonsillectomy with clip and tie haemostatis (n=138)

B: cold steel dissection tonsillectomy with packs and diathermy haemostasis (n=19)

C: cold steel dissection tonsillectomy with  packs and diathermy haemostasis (30)

 E: cold dissection tonsillectomy (49) using a standardized dissection technique, method of haemostasis not stated.

End-point of follow-up:

A: 14d

B:  21d

C:  11d

E:  14d

For how many participants were no complete outcome data available?

(intervention/control)

A: 111 (40.5%)

B: 3 (7.5%)

C: 20 (33%)

E: 22 (24%)

I vs C

Outcome measure:

Pain (Validated pain scores  (VAS))

A:adults: VAS 0-4; children: unspecified VAS 0-4, daily pain scores during 14 days

B: 100mm line, daily during 14 days

C: 20 point, daily during 11days, overall mean pain score.

E: 10cm line + multiple choice questions on days 1,3,7 and 14

A: no significant differences.

B: only graphically. No significant differences.

C:  4.43 (sd 1.36) vs  5.14 (sd 1.42) p-value: 0.18

E: no significant differences.

Outcome measure: haemorrhage

Intraoperative Blood loss (suction aspirate, ml):

A: -

B:80 (range5-300) vs 20  (5-100) p= clinically insignificant

C: 19.50(SD 18.7) vs 68.95 (64.95), P = 0.0002*

E: -

(b) Primary hemorrhage, n(%)during hospital stay

A: -

B: 5 (28%) vs 3 (16%) p=0.56

C: -

E: no incidence

(c) secondary hemorrhage, (%)

A: 13.6% vs 2.1% p=0.001

B: 9 (50%) vs 8 (42%) =0.82

C: 1 vs 1

E: 11(31.4%) vs 8 (22.2%) p=0.331

Outcome measure-3

Time before resumption of normal diet

A:mean days (95%CI)

Adults: 2 (2-3) 2 vs (1-3)

Children: 1 (1-2) vs 1 (1-3)

B:-

C: -

E: mean days (sd)

6.59 (3.18) vs. 8.64 (4.47), p-value: 0.032

and/or normal activities

A: -

B: median days (range): 14 (14-27) vs 14 (14-21) p-value: 0.92

C: -

E: (sd) 10.47(4.41) vs 10.20 (4.2) p-value:0.669

Anthony 2006 and Jayasinge 2005 are unpublished studies

Brief description of author’s conclusion:

9 trials. All but two studies were of low quality and therefore meta-analysis was not appropriate. In most studies, when considering most outcomes, there was no significant difference between coblation and other tonsillectomy techniques.

Level of evidence: Overall quality Grade (per study as stated in SR)

A: C

B: C

C: C

E: C

Reasons: >20% loss of follow up, Bäck had quasi randomisation and initial outcome assessment was performed by non-blinded operator

*outcome preoperative hemorrhage in study C (Jayasinghe): student’s T test was used, however not normally distributed so non-parametric test would have been better

Diathermy  versus Cold dissection

Pinder et al, 2011 

Study characteristics and results are extracted from the SR

SR and meta-analysis of RCT’s comparing tonsillectomy by dissection and diathermy

Literature search up to 31 January 2011

A: Kujawski, 1997

B: Nunez, 2000

Study design: RCT

Setting and Country:

-

Source of funding:

-

Inclusion criteria SR: RCT’s of children and adults undergoing TE or ATE by dissection or diathermy techniques.

Exclusion criteria SR:

Trials were tonsils were randomised, where patients underwent additional procedures (except adenoidectomy)

2  studies included

Important patient characteristics at baseline:

N, age range

A: 200 patients, 80 children <7 and 120 >7 yrs

B: 54 children, 3-12 years

Groups comparable at baseline? yes

Describe intervention (n)

Diathermy

A: Binocular microscope and bipolar dissection (100)

B:  diathermy tonsillectomy by monopolar diathermy dissection at 70W and haemostasis at 30W. (24)

Neither protocol included the routine administration of antibiotics

Describe  control:

Dissection

A: dissection by scissors and bipolar haemostasis (100)

B: Cold dissection tonsillectomy was by

Gwynne-Evans dissector with an Eves snare to the lower pole.

Haemostasis was with monopolar diathermy at 30 W. (26)

End-point of follow-up:

Not stated

For how many participants were no complete outcome data available?

 A: 42 patients (20.5%) were excluded from pain analysis due to loss to follow up or antibiotic use. Therefore pain was excluded from review

B: 4 children (7.4%) were excluded (2 withdrew, 2 protocol violations)

Outcome measure:

Outcome measure:

Haemorrhage

Intraoperative hemorrhage

A: suction aspirate, ml : 12 (18) vs 36 (35)

B: swab weight and volume suction aspirate. (after adenoidectomy),15.1 (11.7) vs 33.7 (18.4)

combined: -21.56 [ -27.26, -15.85 ] (favours intervention)

(a) primary hemorrhage <24h postoperative:

No episodes recorded

(b) secondary hemorrhage >24h postoperative. (with/without hospitalisation)

Effect measure: mean difference [95% CI]:

A: 0.38 [0.11 to 1.29]

B: 2.18 [0.22 to 22.03]

Pooled effect (random effects model / fixed effects model):

0.56 [95% CI [0.19 to 1.63] favoring diathermy

Heterogeneity (I²): 1.70.  Z=1.06 (p=0.29)

Outcome measure: pain control

Defined as analgesic requirement (total analgesic doses, mean difference (95%CI)

First 24h

B: -0.60 [-1.36 to 0.16]

P=0.12

First 12 days

B: 7.5 [1.05 to 13.95]

P=0.023

Time before resumption of normal activities

Days to normal activity (95%CI)

B: 7 (5.0-8.0) vs 5.0 (3.0 – 8.0)

Days to normal diet (95%CI)

B: 7.5 (5.0-8.0) vs 50 (3.0-7.0)

Operating time

A: 36.9 vs 35.9 NS

Pain data from Kajuwaski (1997) was excluded because of loss to follow up due to antibiotic use.

Both studies assessed intraoperative bleeding by measuring suction aspirate; nut only Nunez weighed the swabs as well. This illustrated the difficulty of comparing data of hemorrhage between studies.

Brief description of author’s conclusion:

Data from RCT’s to support one method of tonsillectomy over another are currently lacking, particularly when considering haemorrhage rates. The combined data suggest that there is less intraoperative blood loss using diathermy dissection: this may be relevant for certain patients such as small children and infant. No differences were found in either the primary of secondary haemorrhage rates. This may reflect the fact that the studies were insufficiently powerful to pick up small differences in such rates