Performance of Harmonic devices in surgical oncology: an umbrella review of the evidence

Cheng, Hang; Clymer, Jeffrey W.; Sadeghirad, Behnam; Ferko, Nicole C.; Cameron, Chris G.; Amaral, Joseph F.

doi:10.1186/s12957-017-1298-x

World Journal of Surgical Oncology

Table 3 Overview of the results of included reviews comparing HS use to CT in oncologic surgeries

From: Performance of Harmonic devices in surgical oncology: an umbrella review of the evidence

Review (year)	Population	Effect size (95% CI)	P value for difference	No. of participant (HS/CT)	Heterogeneity (I²)	Publication bias††	GRADE Certainty in evidence‡
Operative time (min)
Sun (2015) [24]	Gastric cancer	MD − 24.5 (− 46.0 to − 3.0)	0.026	199/198	95%	Asymmetric funnel plot	Low^{1, 2, 3}
Cheng (2015) [12]		MD − 27.5 (− 42.2 to − 12.8)	< 0.001	399/382	91%	Symmetric funnel plot	Moderate¹
Chen (2014) [23]		MD − 27.1 (−45.2 to − 9.1)	0.003	172/168	91%	Symmetric funnel plot	Very low^{1, 2, 4}
Huang (2015) [13]	Breast cancer	MD − 1.4 (− 4.2 to 1.4) †	0.85	333/327	74%	NS Egger’s and Begg’s tests	Very low^{1, 2, 5}
Cheng (2016) [21]		MD − 5.1 (− 11.0 to 0.8)	0.09	390/391	83%	Not assessed	Moderate^{1, 6}
Currie (2012) [22]		MD 1.7 (− 3.8 to 7.3) †	0.81	125/120	42%	Not assessed	Low^{1, 2}
Ren (2015)[14]	Oral, head, and neck cancer	MD − 29.3 (− 44.3 to − 4.3)	< 0.001	201/205	92%	Symmetric funnel plot	Moderate^{1, 7}
Tou (2011) [30]	Colon cancer	MD − 26.2 (− 62.0 to 9.6) *	0.15	94/92	87%	Not assessed	Low^{1, 2, 8}
Intraoperative blood loss (mL)
Sun (2015) [24]	Gastric cancer	MD − 137.5 (− 224.9 to − 50.2)	0.002	195/196	91%	Asymmetric funnel plot	Low^{1, 2, 3}
Cheng (2015) [12]		MD − 93.2 (− 125.3 to − 61.0)	< 0.001	349/336	86%	Symmetric funnel plot	Moderate¹
Chen (2014) [23]		MD − 106.3 (− 151.0 to − 61.7)	< 0.001	172/168	93%	Symmetric funnel plot	Very low^{1, 2, 4}
Huang (2015) [13]	Breast cancer	MD − 87.5 (− 130.1 to − 45.0)†	< 0.001	226/237	92%	NS Egger’s and Begg’s tests	Low^{1, 5}
Cheng (2016) [21]		MD − 87.5 (− 137.1 to − 38.0)	< 0.001	323/321	99%	Not assessed	Moderate¹
Currie (2012) [22]		MD − 127.4 (− 227.5 to − 27.3)†	0.013	126/137	91%	Not assessed	Very Low^{1, 2, 9}
Ren (2015) [14]	Oral, head, and neck cancer	MD − 141.1 (− 315.0 to 6.4)	0.112	153/151	100%	Symmetric funnel plot	Moderate ^{2, 10}
Tou (2011) [30]	Colon cancer	MD − 42.1 (− 62.0 to − 21.2)	< 0.001	94/92	0.0%	Not assessed	Moderate²
Drainage volume (mL)
Sun (2015) [24]	Gastric cancer	MD − 292.3 (− 708.3 to 123.7)	0.168	148/145	77%	Asymmetric funnel plot	Low^{1, 2, 3}
Cheng (2015) [12]		MD − 138.8 (− 177.6 to − 100.1)	< 0.001	375/359	94%	Symmetric funnel plot	Moderate¹
Chen (2014) [23]		MD − 74.6 (− 95.2 to − 54.0)	< 0.001	69/69	84%	Symmetric funnel plot	Very low^{1, 2, 4}
Huang (2015) [13]	Breast cancer	MD − 211.6 (− 353.9 to − 69.2)†	0.004	258/269	91%	NS Egger’s and Begg’s tests	Moderate¹
Cheng (2016) [21]		MD − 42.1 (− 65.9 to − 18.9)	< 0.001	127/129	87%	Not assessed	Low^{1, 2}
Currie (2012) [22]		MD − 141.5 (− 335.9 to 53.0)†	0.154	138/149	81%	Not assessed	Low^{1, 2}
Ren (2015) [14]	Oral, head, and neck cancer	MD − 64.9 (− 110.4 to − 19.3)	0.005	191/195	97%	Symmetric funnel plot	Low^{1, 2}
Duration of hospitalization (days)
Sun (2015) [24]	Gastric cancer	MD − 2.1 (− 4.0 to − 0.2)	0.027	50/50	0.0%	Asymmetric funnel plot	Moderate²
Cheng (2015) [12]		MD − 0.6 (− 2.5 to 1.2)	0.509	81/81	65%	Symmetric funnel plot	Low^{1, 2}
Chen (2014) [23]		MD − 3.2 (− 6.3 to − 0.1)	0.040	20/20	–	Symmetric funnel plot	-**
Cheng (2016) [21]	Breast cancer	MD − 1.4 (− 2.4 to − 0.4)	0.007	184/186	98%	Not assessed	Low^{1, 2}
Ren (2015) [14]	Oral, head, and neck cancer	MD − 0.21 (− 0.48 to 0.07)	0.142	79/81	0.0%	Symmetric funnel plot	Moderate²
Tou (2011) [30]	Colon cancer	MD − 0.42 (− 0.84 to 0.00)	0.051	94/92	0.0%	Not assessed	Moderate²
Overall perioperative complications
Cheng (2015) [12]	Gastric cancer	RR 0.58 (0.3 to 1.0)	0.059	235/229	12.0%	Symmetric funnel plot	High
Chen (2014) [23]	Gastric cancer	RR 0.75 (0.4 to 1.3)	0.276	126/121	0.0%	Symmetric funnel plot	Moderate²
Huang (2015) [13]	Breast cancer	RR 0.38 (0.2 to 0.6)	0.01	199/209	23.0%	NS Egger’s and Begg’s tests	High
Cheng (2016) [21]		RR 0.5 (0.3 to 0.8)	0.002	NR	0.0%	Not assessed	Moderate²
Currie (2012) [22]		OR 1.6 (0.7 to 3.7)	0.3	NR	35.0%	Not assessed	Very low^{1, 2, 9}
Tou (2011) [30]	Colon cancer	RR 1.28 (0.7 to 2.3)	0.395	106/103	0.0%	Not assessed	Moderate²
Seroma development
Huang (2015) [13]	Breast cancer	RR 0.5 (0.3 to 0.7)	< 0.001	82/125	0.0%	NS Egger’s and Begg’s tests	Moderate²
Cheng (2016) [21]		RR 0.5 (0.4 to 0.7)	< 0.001	410/411	25.0%	Not assessed	High
Currie (2012) [22]		OR: 0.8 (0.4 to 1.4)	0.368	45/49	0.0%	Not assessed	Low^{2, 9}

HS Harmonic devices, CT conventional techniques, MD mean difference, SMD standardized mean difference, RR risk ratio, OR odds ratio, NR not reported, NS non-significant
*For the comparison of monopolar electrocautery scissors and ultrasonic coagulating shears
**Only one study in this category
†Original SRs reported SMD. Mean differences were calculated using data provided in forest plots of published SRs
††An asymmetric funnel plot or significant Egger’s or Begg’s test indicates the possibility of publication bias
‡ GRADE Working Group grades of evidence:
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect;
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different;
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect;
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹ Given the substantial heterogeneity in the pooled estimate, we rated down for inconsistency
² For continuous outcomes, GRADE guideline suggests downgrading for sample size less than 400
³ We decided not to rate down for publication bias as Cochrane suggests tests for funnel plot asymmetry should be used only when there are at least ten studies included in the meta-analysis
⁴ The quality of RCTs was assessed using Jadad scale, and their scores were located at the low level, mainly due to the absence of randomization details
⁵ The quality of RCTs was assessed using Jadad scale. We decided not to rate down as four out of seven RCTs were categorized as high quality
⁶ We decided not to rate down for risk of bias as only 1 out of 12 RCTs were considered high risk of bias
⁷ We decided not to rate down for risk of bias as four out of seven RCTs were identified as being of high or moderate quality
⁸ We decided not to rate down for risk of bias as only one out of six included RCTs were considered high risk of bias
⁹ We decided to rate down for risk of bias as four out of six included RCTs were with a high risk of bias
¹⁰ Although effect estimates and their 95% Cis from RCTs did not overlap, we decided not to rate down for inconsistency as all had the same direction and I² for authors sensitivity analysis is zero

Back to article page

ISSN: 1477-7819

Contact us

General enquiries: journalsubmissions@springernature.com