HSC Section 3 - Trauma, Critical Care and Sleep Medicine

V ENOUS THROMBOEMBOLISM IN OTOLARYNGOLOGY

FIGURE 3. Forest plots comparing chemoprophylaxis to no chemoprophylaxis in all otolaryngology–head and neck surgery patients in terms of (A) venous thromboembolism (I 2 5 0%) and (B) bleeding complications (I 2 5 0%). 95% CI, confidence interval. [Color figure can be viewed at wileyonlinelibrary.com]

chemoprophylaxis (73% decrease). It is also the only study in which the information on hemorrhage in patients with cancer or who underwent free flap reconstruction was available, and showed an incidence of 8.1% in cancer or free flap cases, and 1.1% in other patients, although this is confounded by nonhomogeneous use of thrombo- prophylaxis. Heterogeneity was not a concern in this anal- ysis and, therefore, subgroup analyses and meta- regression was not performed. Using a funnel plot and Duval and Tweedie’s trim and fill, there was evidence of publication bias with 3 potential missing studies (Figure 4B), however, a classic Fail Safe N demonstrated that 243 missing studies would be required to significantly change our findings. DISCUSSION The CHEST guidelines stratify risk of VTE into very low ( < 0.5%), low (1.5%), moderate (3%), and high (6%). 1 The CHEST guidelines recommend no prophylaxis for very low-risk patients, and mechanical prophylaxis for low-risk patients. However, based on our data, most stud- ies do not report the specific prophylaxis regimen used. In the only study in which it was specifically stated that none of the patients received chemoprophylaxis, 14 the authors mention that the overwhelming majority of their patients were treated with mechanical prophylaxis. In this study of 2016 patients, the incidence of VTE in patients with SCC/free flap was 2.3% and 0.5% in others. 14 How- ever, the true incidence of VTE in OHNS and patients with SCC/free flap might be higher when taking into account that many of the patients identified in the studies from our systematic review may have received mechani- cal prophylaxis with sequential compressive devices. Sequential compressive devices were used in a significant proportion of patients (6, 8, 13). Therefore, we can con- clude that we did not encounter studies evaluating the

chemoprophylaxis regimen reported was unfractionated heparin, low-molecular weight heparin (LMWH), or fon- daparinux for the length of stay. The chemoprophylaxis regimen used by Clayburgh et al 8 was unfractionated hep- arin or LMWH for the duration of stay, and all patients had sequential compressive devices. For Gavriel et al, 13 the authors mention that sequential compressive devices were used on “most” patients, and the chemoprophylaxis regimen was not specified. The addition of chemoprophy- laxis did not result in a decreased VTE incidence (OR, 0.86; 95% CI, 0.48–1.52; I 2 , 0%), but an increased risk of bleeding (OR, 3.78; 95% CI, 2.20–6.48; I 2 , 6%). The het- erogeneity in these studies was minimal, as demonstrated by the low I 2 , despite some difference between the stud- ies. Using a funnel plot and Duval and Tweedie’s trim and fill, there was no evidence of publication bias in both of these analyses. The studies in which the outcome data were available on patients with head and neck cancer or who underwent free flap reconstruction are shown in the forest plot in Figure 4A. The overall OR for squamous cell carcinoma (SCC)/flap cases was 6.3 (95% CI, 3.8–10.5), with an I 2 of 54%. The incidence of VTE in patients with head and neck cancer or who underwent free flap reconstruction when compared with other patient populations was 0.9% versus 0.1%. There was only 1 study by Bahl et al 6 in which the information on VTE risk reduction with che- moprophylaxis in patients with SCC or who underwent free flap reconstruction when compared with other patients was available. This study showed an incidence of VTE of 7.7% in cancer/free-flap cases without chemopro- phylaxis, and 2.1% in SCC/free-flap cases with Patients with head and neck cancer with or without free flap use

HEAD & NECK—DOI 10.1002/HED JUNE 2017

169