2017-18 HSC Section 3 Green Book

TABLE II. Post-Tonsillectomy Hemorrhage Rates in Subjects Receiving Perioperative Ketorolac and Unexposed Controls.

Study

Year

Population

Design

Timing

Toradol

Toradol (%)

Control

Control (%)

Agrawal 10

1999

Pediatric

R

Intra

5/213

2.3

3/97

3.1

Bailey 1

1997 1995

Adult

P

Post

7/37

18.9

3/43 2/60

7.0 3.3

Gallagher 8 Gallagher 8 Gunter 11 Judkins 4 Judkins 4 Keidan 12 Romsing 13

Pediatric

R

Intra

7/107

6.5

1995

Adult

R

Intra

10/62

16.1

0/29

0.0

1995 1996

Pediatric Pediatric

P

Post

6/41 4/36

14.6 11.1

5/39

12.8

R

Intra

11/200

5.5

1996

Adult

R

Intra

6/22

27.3

1/53

1.9

2004 1998

Pediatric Pediatric

P P

Pre

0/25 8/40

0.0

0/32 5/20

0.0

Pre/post

20.0

25.0

Rusy 14

1995

Pediatric

P

Pre

0/25

0.0

0/25

0.0

Splinter 9 Sutter 15

1996 1995

Pediatric Pediatric

P P

Pre

5/35 0/45

14.3

0/29 0/42

0.0 0.0

Post

0.0

All children

35/567

6.2

26/544

4.8

All adults

23/121 19/211

19.0

4/125

3.2 5.3

Pediatric prospective

9.0

10/187

Pediatric retrospective

16/356

4.5

16/357

4.5

All studies

58/688

8.4

30/669

4.5

Data for 12 groups from 10 studies are reported. Individual-level data from each study were pooled for analysis in the specified subgroups. P 5 prospective, randomized, controlled study; Pre/Intra/Post 5 ketorolac given before/during/after surgery; R 5 retrospective case-control study.

are insufficiently powered to detect this difference. The summary RR for the 10 studies is 1.39. If this is accu- rate, then ketorolac may have a 39% increased risk of post-tonsillectomy hemorrhage. To detect this difference with 95% confidence given a control hemorrhage rate of

5%, which is supported in the pooled data, would require a prospective, randomized, controlled study of approxi- mately 1,500 patients in each arm. Future studies need to be carefully designed, controlled, and reported to avoid the risks of bias identified in prior efforts.

Fig. 2. Relative risk (RR) of post-tonsillectomy hemorrhage in subjects receiving perioperative ketorolac compared to unexposed controls. Twelve groups from 10 studies are included for meta-analysis. RR values are depicted with black closed diamonds (RR), black lines (95% CI), and grey boxes (weighting for the overall meta-analysis. Black open diamonds indicate summary RR and 95% CIs for subgroup meta- analysis of the pediatric and adult studies, and for all studies pooled together. Higher RR indicates increased risk of post-tonsillectomy hemorrhage with ketorolac use. Follow-up indicates the number of days reported. C 5 cold; CI 5 confidence interval; E 5 electrocautery; L 5 laser; NR 5 not reported; P 5 prospective, randomized, controlled study; pre/intra/post 5 ketorolac given before/during/after surgery; R 5 retrospective case-control study. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

Laryngoscope 124: August 2014

Chan and Parikh: Ketorolac and Tonsillectomy

255