Abstract
BACKGROUND: Repeated chest X-rays serve as an essential screening tool to identify and describe new or stable (i.e., unchanged) lung abnormalities suggestive of pulmonary tuberculosis (TB) disease. The time for which a patient’s chest X-ray has not demonstrated appreciable change prior to treatment, or pretreatment chest X-ray stability duration, has been considered clinically useful in distinguishing inactive from active disease at four or 6 months. This relationship, however, has not been previously quantified. METHODS: This study relied on retrospective medical record review to assess the relationship of documented pretreatment chest X-ray stability duration thresholds relative to four and 6 months with a future clinical or culture-confirmed (Class 3) diagnosis of pulmonary TB disease. Multivariable logistic regression quantified this association among 146 patients who were evaluated and started on treatment for pulmonary TB disease in the San Diego County tuberculosis clinic between May 2012 and March 2017. RESULTS: After adjusting for age and Class B1 TB, Pulmonary status, a CXR stability duration of 4 months or more was not significantly associated with a Class 3 pulmonary TB diagnosis (adjusted odds ratio [AOR], 0.830; 95% confidence interval [CI], 0.198–3.48). Results were similar for the 6-month cut-point after adjusting for age and Class B1 Pulmonary status (AOR, 0.970; 95% CI, 0.304–3.10). Compared with less than 4 months, CXR stability durations of four to 6 months (AOR, 0.778; 95% CI, 0.156–3.89) and greater than 6 months (AOR, 0.875; 95% CI, 0.187–4.10) were also not significantly associated with a Class 3 TB diagnosis after adjusting for covariates. CONCLUSION: Repeated chest X-rays remain a valuable tool for clinicians identifying and describing new or unchanged lung abnormalities suggestive of pulmonary TB disease. This study found no statistically significant association between pretreatment chest X-ray stability duration and subsequent TB disease diagnosis, with a wide range of estimates compatible with the data, suggesting the stability duration cut points relative to four and 6 months may not be as informative as previously understood. DISCLOSURES: All authors: No reported disclosures.