Prospective validation of a proposed severity-of-illness score in patients with tuberculosis requiring intensive care unit admission

We previously retrospectively validated a 6-point severity-of-illness score aimed at identifying patients at risk of dying of TB in ICU. The parameters used were septic shock, HIV with a CD₄&lt;200/mm³, impaired renal function, PaO₂:FiO₂&lt;200, diffuse parenchymal infiltrates and no concomitant TB treatment on admission. The aim of the present study was to prospectively validate and potentially refine the scoring system. All patients admitted to ICU with confirmed TB over a 2-year period were included. A planned <i>post-hoc</i> analysis was performed to assess the significance of various parameters and potentially derive a simplified score with an equally/higher predictive power. 78 patients (40 males, 36.1+/-14.4 y) were admitted with TB (pulm, n=76; extrapulm, n=2). 41 (52.6%) died. The 6-point scores of non-survivors were higher (3.5+/-1.3 vs 2.7+/-1.2; p=0.01) than survivors. A score ≥3 was associated with higher mortality compared to &lt;3 (63.3% vs 34.5%; OR 3.3; 95%CI, 1.3-8.7; p=0.02). <i>Post-hoc</i>, a PF&lt;200 failed to predict mortality, whereas other causes of immunosuppression did. A revised 5-point score (septic shock, any immune suppressed state, impaired renal function, diffuse parenchymal infiltrates and no concomitant TB treatment) was retrospectively applied. The 5-point scores of non-survivors were higher than survivors (3.3+/-1.2 vs 2.7+/-1.1; p&lt;0.001). A score ≥3 was associated with higher mortality compared to &lt;3 (31.0% vs 65.3%; OR 5.8; 95%CI, 2.2-15.8; p&lt;0.001). Although the 6-point score identified patients at higher risk of dying, we were able to derive and retrospectively validate a simplified 5-point score with superior predictive power.