IPBronch Review
🩺 Clinical Context
In interventional pulmonology (IP), the diagnostic workup of undiagnosed pleural effusions remains a cornerstone of our practice. While thoracic ultrasound (TUS) has become the standard of care for guiding pleural biopsies, the integration of advanced imaging modalities—such as ultrasonic elastography—aims to improve the sensitivity of tissue acquisition by identifying areas of increased tissue stiffness (often correlating with malignancy). This correspondence highlights a critical semantic and methodological tension in how we report the success of these interventions: the distinction between “diagnostic yield” (the immediate utility of the procedure) and “diagnostic accuracy” (the long-term confirmation of the diagnosis).
📊 Methodological Strengths & Weaknesses
- Strengths: The original study by Deng et al. is a multicenter, randomized trial, which provides a higher level of evidence than retrospective cohorts. The use of a 12-month follow-up period to establish a “gold standard” for benign disease is a rigorous approach to minimizing false negatives.
- Weaknesses: The primary critique centers on the conflation of diagnostic yield and diagnostic accuracy.
- Misclassification: By defining “diagnostic yield” as (True Positives + True Negatives) / Total Cases, the authors are actually reporting diagnostic accuracy.
- Clinical Utility vs. Retrospective Validation: True diagnostic yield should reflect the procedure’s ability to provide a definitive, actionable diagnosis at the time of the procedure. By including cases that required 12 months of follow-up to confirm a “true negative,” the authors inflate the perceived immediate utility of the elastography-guided biopsy.
- Bias: This methodological choice obscures the “nondiagnostic” rate of the procedure, which is the metric most relevant to an IP physician deciding whether to proceed with a more invasive diagnostic thoracoscopy (VATS) or medical thoracoscopy.
💡 Takeaway for Fellows
- Define Your Metrics: When reading literature on new diagnostic tools (EBUS, navigational bronchoscopy, or pleural biopsy), look closely at how the authors define “yield.” If they include long-term follow-up in their calculation, they are reporting accuracy, not yield.
- The “Actionability” Test: Always ask: “Did this procedure provide a definitive answer that changed management today?” If the answer is “no” because you had to wait for clinical follow-up or repeat imaging, the procedure did not have a high diagnostic yield, regardless of its eventual accuracy.
- Clinical Application: Do not let the “87.83% yield” figure in the Deng et al. study mislead you into thinking that nearly 9 out of 10 patients will have a definitive diagnosis immediately post-procedure. A significant portion of that percentage is derived from retrospective confirmation. When counseling patients, manage expectations regarding the possibility of a non-diagnostic result that necessitates further intervention.
Original Abstract
To the Editor:
I read with great interest the recent publication by Deng et al. [1] titled “Ultrasonic elastography-guided pleural biopsy versus traditional thoracic ultrasound-guided pleural biopsy for the diagnosis of pleural effusion: a multicentre, randomised trial.”
The authors reported that the “diagnostic yield” of ultrasonic elastography-guided pleural biopsy was significantly higher than traditional thoracic ultrasound-guided pleural biopsy: 87.83% versus 76.99% (p=0.032).
The “diagnostic yield” was calculated as the proportion of patients who obtained a subsequent definitive diagnosis (benign or malignant) calculated as: (true positives + true negatives)/total cases. To note, the authors state that the follow-up period to establish true benignity in several cases was at least 12 months.
Ost et al. [2] suggested a methodological framework that lays a foundation for a clear distinction between diagnostic yield and diagnostic accuracy when evaluating a novel diagnostic test in pulmonary medicine. Other authors have used that methodology for evaluating results of diagnostic bronchoscopy [3, 4].
Diagnostic accuracy refers to results deemed diagnostic (positive or negative) that are confirmed to be accurate after a follow-up period.
Diagnostic yield should refer only to the proportion of cases in which the index biopsy provides a pathological result allowing for confident management of the patient without relying on confirmatory follow-up clinical data and/or subsequent additional biopsies.
Deng et al. [1] essentially report diagnostic accuracy estimates rather than true diagnostic yield since the index biopsy was deemed as diagnostic or nondiagnostic based on follow-up clinical data not available at the time of the initial procedure.
A true diagnostic yield would entail only data obtained from the initial biopsy. Hence in the current study the numerator (true positives + true negatives) should include only cases in which the index pleural biopsy provided conclusive specific benign diagnosis, such as sarcoidosis or tuberculosis, or definitive malignant diagnosis, such as metastatic adenocarcinoma or mesothelioma.
While reanalysis of data and establishing the true diagnostic yield of elastography-guided pleural biopsy obtained in the current study is not possible, I believe further studies using the current report as a benchmark for comparison should incorporate these points into consideration.