, based on data from approximately 400 individuals.
Although a mortality risk-prediction score could improve treatment for pulmonary tuberculosis patients, such a score has not been previously reported, wrote Takeshi Osawa, MD, of Fukujuji Hospital, Tokyo, and colleagues.
In a study published in the journal CHEST, the researchers used 252 patients from a previous perspective study of newly diagnosed pulmonary tuberculosis as the development cohort, and recruited 165 additional patients between March 2021 and September 2022.
The primary endpoint was all-cause in-hospital mortality. Based on data from the development group, the researchers found that age 65 years and older and age 80 years and older, hypoxemia, activities of daily living, bilateral pulmonary lesions, lymphocyte count of less than 720 microliters, serum albumin less than 2.86 mg/dL, C-reactive protein (CRP) 3.97 mg/dL or higher, and procalcitonin (PCT) 0.130 ng/mL or higher were predictors of all-cause in hospital mortality.
The researchers used this information to create the disease severity score, known as the AHL score. The AHL included three clinical parameters: activity in daily living (semi-dependent, 1 point; totally dependent, 2 points); hypoxemia (1 point) and lymphocytes (< 720 /mcL, 1 point).
The scoring systems for the three parameters were, respectively, 1 point for semi-dependent and 2 points totally dependent (for activity in daily living), 1 point for presence of hypoxemia, and 1 point for lymphocytes less than 720 per microliter. The researchers stratified the scores into levels of low, intermediate, and high risk, with scores of 0, 1-2, and 3-4, respectively.
All-cause in hospital mortality occurred in 39 (15.5%) and 17 (10.3%) of patients in the developmental and validation cohorts, respectively.
The AHL score effectively predicted mortality, dividing patients into three groups of 1.3% low-risk, 8.9% intermediate risk, and 39.3% high-risk in the validation cohort, with a Harrell’s c-statistic of 0.842.
The corresponding numbers for the development cohort were 0, 13.5%, and 55.8%, with a c-statistic of 0.902.
The findings were limited by several factors, including the lack of data from “smear-negative” patients who were treated as outpatients, and more research is needed to determine the applicability of the AHL score in an outpatient population, the researchers noted. Other limitations included the lack of data on long-term mortality in surviving patients who were discharged, and the reliance on assessments that can be performed only in clinical settings in developed countries, they said.
However, the results support the feasibility of the AHL score in clinical settings to accurately predict mortality in patients with pulmonary TB, and may help optimize treatments for this population, they concluded.
The study received no outside funding. All authors disclosed nonfinancial support in the form of measuring reagents from Fujifilm Wako Pure Chemical Corporation during the study but had no relevant financial conflicts to disclose.