Background: Invasive fungal diseases (IFDs) are crucial causes of morbidity and mortality among febrile neutropenia (FN) patients with haematological malignancies. With the improved awareness of IFD associated disease burden and the clinical benefit from early antifungal therapy, the empirical antifungal strategy, recommended by the majority of international IFD related guidelines, has been regarded as the main antifungal treatment either on initial presentation or after other potential causative bacteria have been treated. But the downside of this strategy is the overtreatment of patients meeting the above criteria but who do not have an IFD, leading to an increase in adverse events, prolonged hospitalizations and elevated usage of antifungal drugs. In recent, diagnostic driven antifungal strategy has been evaluated as an alternative therapeutic strategy in these patients. Therefore, the comparison of both the efficacy and safety between diagnostic driven antifungal strategy and empirical strategy is explored in our meta-analysis.Methods: PubMed, Embase, Cochrane Library, Web of Science, and another two Chinese database- CNKI and WanFang- were searched to identify randomized controlled trials (RCTs)published from January 2000 to July 2018 comparing empirical versus diagnostic driven antifungal strategies in IFD among patients with hematologic malignancies. Two reviewers performed the quality assessment and extracted data independently. The primary outcome was all-cause mortality. Secondary outcomes included IFD-related mortality, incidence rate of invasive fungal infections, and adverse events. Data analysis was performed using Stata15.0 software. The measurement of effect was done by calculating the Risk Ratio (RR) and 95% Confidence Interval (CI) using a fixed-effect Mantel-Haentszel method. We assessed for heterogeneity via calculating chi-square test and I2 test, and the following meta-regression and sensitivity analysis were applied to figure out heterogeneous sources for further subgroup analysis.Results: Four RCTs involving 1026 patients were included (Table1). Meta-analysis showed that diagnostic-driven antifungal strategy did not reduce all-cause mortality in IFD patients with hematologic malignancies compared with empiric strategy (P=0.638), RR=1.14 (95% CI: 0.78-1.65, Fig.1); No statistical difference was found between diagnostic drive antifungal strategy and empirical strategy in either IFD related mortality(P=0.593), RR=1.24 (95% CI: 0.57-2.68, Fig.2), or IFD incidence rate(P=0.158), RR=2.015(95% CI:0.76-5.33, Fig.3). But, subgroup analysis showed, in patients with acute leukaemia, IFD incidence rate was increased with diagnostic driven antifungal strategy (P=0.002, RR=3.75(95% CI:1.649-8.541, Fig.4). In terms of safety, the four RCTs included cannot be combined and analyzed because of the inconsistent effects (Table2). As reported in Yuan, et, al study, two patients in the empirical group and 1 patient in the preemptive group quit treatment due to adverse reactions to voriconazole; As showed in Cordonnier, et, al paper, Creatinine clearance decreased significantly during the study period in the empirical group and preemptive group. Although the mean decreased value± Standard Deviation(SD) was larger in the empirical treatment group than in the preemptive treatment group ( -8.7±20.8 vs -5.8±27.2 ), the difference was not significant, SAEs occurred in similar proportions among patients in the 2 groups; In Hobart, et, al research, no significant difference was observed in creatinine, bilirubin, AST, and ALT between diagnostic driven and empirical groups.Conclusion: No statistical differences were found between diagnostic driven antifungal strategy and empirical antifungal strategy in all-cause mortality, IFD-related mortality, and IFD incidence rate. But for acute leukaemia among haematological malignancies, the diagnostic-driven antifungal strategy may increase the risk of developing IFD, suggesting adoption of a more selective diagnostic driven antifungal strategy in hematological patients without acute leukaemia. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.