RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
***************************************
Review: The authors showed the burden of chronic pulmonary aspergillosis (CPA) among a population with newly diagnosed tuberculosis (TB) at one clinic in South Delhi. They followed up on the population to estimate the changing burden of CPA from TB diagnosis throughout the treatment process. They argued the CPA evaluation at the diagnosis of TB is important.
The studyβs main argument may not be generalizable to other populations in different settings. First, the data was collected from one tuberculosis (TB) clinic in South Delhi, but descriptions of the setting are not presented. It is not clear the burden of chronic pulmonary without TB infection in South Delhi. Whether the clinic was in rural vs. urban, public vs. private, or primary vs. tertiary is also not explained. Without understanding the study setting and population in detail, it is difficult to interpret the study findings in other settings. In addition, the study had a small sample size of individuals co-infected with TB and chronic pulmonary aspergillosis (CPA). For example, only 5% of the recruited population had both lab-confirmed TB and CPA. This limits not only the generalizability but also the precision of the study results. This should be at least acknowledged as a study limitation.
The long-term observation that the prevalence of CPA diagnosis increased over time may mislead the study argument. There were more CPA-positive cases in this study cohort at the last follow-up than at baseline, however, most of the CPA at baseline has been self-resolved while individuals were treated for TB. It is not convincing that early evaluation of CPA at the time of TB diagnosis can be helpful for long-term CPA prevention. It would be interesting to assess the heterogeneity of TB treatment effect (e.g., lung recovery) among individuals who resolved CPA and those who developed CPA during TB treatment. Besides, about 40% of the loss to follow-up is a significant amount missing in a prospective cohort study. The authors could have discussed which direction the lost individuals contributed to a bias based on their baseline data.
The regression analysis is not reliable. The authors mentioned that the logistic regression found the baseline anti-Aspergillus IgG and a follow-up SGRQ score as predictors of CPA development at the end of TB treatment. However, a model construction and its construction process are not presented. Less than half of the individuals with CPA were self-resolved, and it is hard to understand how the initial IgG level could predict CPA at the end of the treatment. The SGRQ score at the second visit may also show significance by random and is hard to interpret from the perspective of public health policy.