2017-18 HSC Section 3 Green Book

Jackson et al

Figure 1. PRISMA flowchart demonstrating study selection. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analyses.

0.77, –0.28; P \ .0001). There was no heterogeneity among these trials ( I 2 = 7%, P = .37; Figure 2 ). Operative time. Only 6 of the included studies had sufficient data available for the outcome of operative time. 15,16,19,21,22 Data were extracted from 6 studies with 174 tumors. Operative time was significantly lower in patients under- going preoperative embolization than patients undergoing surgical excision alone (SMD = –0.46; 95% CI: –0.77, –0.14; P = .004). There was significant heterogeneity among these trials ( I 2 = 58%, P = .03; Figure 3 ).

the studies. No studies described the strengths and weak- nesses of the data source. Adequate description was not pro- vided regarding how treatment effects were measured or how to avoid classification bias. There was no description of con- founding in any study. General interpretation of the results in the context of current evidence was presented in all studies.

Primary Analysis

Estimated blood loss. Of the 22 studies evaluated, only 15 had a control group. Of these, another 3 studies 4,7,11 did not have suf- ficient data available to calculate standard deviation and there- fore were not included in the analysis. Data were extracted from 12 studies with 295 tumors. The EBL was significantly lower in patients undergoing preoperative embolization than patients undergoing surgical excision alone (SMD = –0.52; 95% CI: –

Subgroup Analysis

Embolization method. The majority of studies with data avail- able for analysis of EBL indicated transarterial embolization

196