2017-18 HSC Section 4 Green Book

Aesthetic Surgery Journal 35(3)

Figure 1. Hematoma Forest plot.

Figure 2. Seroma Forest plot.

Table 3. Postoperative Seroma Pooled Relative Risk

Table 4. 24-hour Postoperative Drainage Differences of Means

Study

Relative Risk

95% Confidence Interval

Percent Weight

Study

Standardized Mean Difference

95% Confidence Interval

Percent Weight

Kamer and Nguyen 6

0.14

0.0177

1.150

51.85

Hester et al. 8

− 0.78

− 1.21

− 0.35

40.2

Hester et al. 8

0.2

0.0098

4.052

18.51

− 1.28

− 1.63

− 0.93

Hester

59.8

Hester et al. 9

et al. 9

1.5

0.4410

5.101

29.62

0.55

0.2313

1.334

Pooled

0.55

− 1.35

− 0.81

Pooled

Heterogeneity chi-squared = 3.18 (d.f. = 1), P = .075. Test of SMD = 0: z = 7.81, P < .001.

Heterogeneity chi-squared = 4.60 (d.f. = 2), P = .100. Test of RR = 1: z = 1.31, P = .189.

Oliver et al. conducted the fi rst trial of fi brin glue in rhyti- dectomy, but only measured the median 24-hour drainage. 4 They found a signi fi cant reduction in volume of the drainage (10 mL vs 30 mL), but they only reported the interquartile range and not the standard deviation, which is needed for comparison of means tests. Consequently, this trial was not included in any of the analyses. Fezza et al. found a non-signi fi cant reduction in hemato- ma, but found an interesting 13.3 minute reduction in oper- ating time. 5 This is likely due to less time spent obtaining hemostasis. As an editorial side note, the authors strongly recommend absolute attentiveness to hemostasis, and these glues cannot be considered a replacement for meticulous hemostasis. Marchac and Greensmith found what they considered evidence against the ef fi cacy of fi brin glue in fi nding drainage rates that were statistically different at 26 mL and 33.5 mL, but with results that they did not consider clinically signi fi cant. 3 They concluded that fi brin glue was not cost- effective compared to drains alone. They also did not provide the standard deviation, so their drainage rates were not included in this analysis. Using drains after rhytidectomy has never been shown to decrease the rate of hematoma, although it is arguable that their use may allow an earlier and more con fi dent

patients. Zoumalan and Rizk found that only 0.4% of their patients who received fi brin glue developed hematomas, compared to 3.4% of patients who did not receive fi brin glue. 15 Marchac and Sandor found similar results in retro- spectively examining 200 of their fi brin glue rhytidectomy patients. 16 However, Jones and Grover found no bene fi t of fi brin glue in a review of 910 of their patients. 1 These con fl icting retrospective studies called for the randomized, controlled trials that we examined in this meta-analysis. A previous meta-analysis by Por et al. found no statisti- cal difference in the incidence of hematoma development in patients who received fi brin glue. 10 The authors only included two fi brin glue trials (Marchac and Greensmith 3 and Oliver et al. 4 ) and one on platelet-rich plasma trial. They found a trend towards reduction in ecchymoses and wound-drainage formation, but pooled analyses in their Forest plots crossed the line of no effect, showing no signi fi - cance. Since then, multiple studies have been published with large enough sample sizes to call for a new meta- analysis. In this project, Forest plots were chosen, as in other meta-analyses, including the Por study. A Forest plot is a method for graphical display used to demonstrate the relative effects of treatments in multiple quantitative scien- ti fi c studies addressing the same question. This makes them perfect for meta-analyses such as this one.

130