A meta-analysis-informed diagnostic stratification tool for invasive pulmonary aspergillosis in severe fever with thrombocytopenia syndrome: systematic review, meta-analysis, and single-center cohort-based assessment.

BACKGROUND

This study aimed to develop and assess a meta-analysis-informed diagnostic stratification tool for invasive pulmonary aspergillosis (IPA) in patients with severe fever with thrombocytopenia syndrome (SFTS).

METHODS

A systematic review and meta-analysis were conducted to identify bedside clinical factors associated with IPA/SFTS-associated pulmonary aspergillosis (SAPA) in patients with SFTS. Because individual participant data were unavailable, pooled study-level odds ratios were translated into a pragmatic three-variable diagnostic stratification score rather than a conventional individual-level prediction model. The tool was assessed in a separate single-center retrospective cohort of 220 patients with SFTS, among whom 63 (28.6%) met the clinically ascertained IPA/SAPA endpoint.

RESULTS

Six studies involving 2,342 patients with SFTS, including 407 patients with IPA/SAPA, were included in the quantitative meta-analysis. In the primary random-effects analyses, neurological symptoms showed the clearest pooled association with increased IPA/SAPA risk, whereas ICU admission and corticosteroid use showed positive but less precise pooled directions of effect. The final three-variable score assigned 9 points for neurological symptoms, 9 points for ICU admission, and 6 points for corticosteroid use, yielding a total score range of 0-24 points. In the single-center cohort-based assessment, the diagnostic stratification tool showed encouraging discrimination, with an AUC of 0.910 (95% CI, 0.869-0.951) and a Brier score of 0.097 after cohort-based logistic recalibration. The observed IPA/SAPA incidence increased stepwise across low-, medium-, and high-risk groups: 3.7% (4/108), 22.8% (13/57), and 83.6% (46/55), respectively.

CONCLUSION

This meta-analysis-informed diagnostic stratification tool showed encouraging cohort-based risk separation in a single-center retrospective assessment. It may help prioritize fungal diagnostic work-up and surveillance intensity in patients with SFTS and support more targeted diagnostic resource allocation and antifungal stewardship. However, it should not be used as a stand-alone basis for antifungal treatment decisions or as a broadly generalizable individual-level prediction model. External validation, local recalibration, and prospective multicenter evaluation are required before clinical implementation.

SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=5115813, identifier CRD420251115813.