Soluble ST2 as a prognostic marker in community-acquired pneumonia

Masato Watanabe,Hajime Takizawa,Masaki Tamura,Akira Nakajima,Daisuke Kurai,Haruyuki Ishii,Saori Takata,Keitaro Nakamoto,Erei Sohara,Koujirou Honda,Masuo Nakamura,Toshiya Inui,Hiroo Wada,Hajime Goto
DOI: https://doi.org/10.1016/j.jinf.2015.02.004
Abstract:Objectives: Community-acquired pneumonia (CAP) is associated with high mortality when initial treatment fails. Early identification of these patients allows physicians to modify treatments earlier, increasing survival. Methods: Ninety-one hospitalized patients with CAP were studied. Serum soluble ST2 levels were measured at diagnosis and at 3, 7, and 14 days (days 0, 3, 7, and 14) after the initiation of antimicrobial treatment. The predictive value of all-cause in-hospital mortality and the additive effect of soluble ST2 on the pneumonia severity index (PSI) were evaluated. Results: In univariate analysis, high serum levels of soluble ST2 at days 0, 3, 7, and 14 were predictive of death (hazard ratios: 3.1, 10.0, 12.0, and 22.6, respectively). In multivariate analysis, a combination of soluble ST2 at day 3 (above 2700 pg/ml) and PSI were predictive of death with higher accuracy than PSI alone (net reclassification improvement, 0.44; integrated discrimination improvement, 0.17; P = 0.001 for both). Specifically, simultaneous presence of high soluble ST2 (day 3) and a PSI of 5 was suggestive of higher mortality risk than a PSI of 5 alone (mortality 78% vs. 39%, respectively). Conclusions: Soluble ST2 is prognostic indicator of CAP and can add to the predictive value of the PSI.
What problem does this paper attempt to address?