Chest radiograph interpretation sparks disagreement over pediatric ARDS diagnosis
Click Here to Manage Email Alerts
Key takeaways:
- Agreement on a pediatric acute respiratory distress syndrome diagnosis was common using 2015 criteria.
- Most disagreements occurred because of differing chest radiograph interpretations.
Use of the 2015 Pediatric Acute Lung Injury Consensus Conference, or PALICC, criteria for diagnosing pediatric acute respiratory distress syndrome mostly resulted in agreement between two physicians, according to results published in CHEST.
However, when disagreements occurred, the primary reason for this conflict was due to differing chest radiograph interpretations, according to researchers.
“In this study, we found that the interrater reliability of the 2015 PALICC criteria for moderate to severe [pediatric acute respiratory distress syndrome] was substantial,” Layne Silver, MD, physician at Northwell Health, and colleagues wrote. “Diagnostic disagreements were most commonly caused by differences in chest radiograph interpretations.”
In a retrospective review, two pediatric ICU physicians assessed 191 hospitalized pediatric patients (admitted between 2016 and 2021) with acute hypoxic respiratory failure receiving invasive mechanical ventilation using the 2015 PALICC criteria for a moderate to severe pediatric ARDS (PARDS) diagnosis in order to determine interrater reliability of the criteria.
After deciding if each patient met the PALICC definition of the condition and rating their diagnostic confidence, the two physicians ended up having 37 diagnostic disagreements but “substantial” interrater reliability (Gwet’s agreement coefficient, 0.74; 95% CI, 0.65-0.83), according to researchers.
The most frequent reason for diagnostic disagreement was different interpretations of chest radiographs (56.8%), followed by ambiguity in origin of pulmonary edema (37.8%) and uncertainty over the presence of significant differences between a patient’s current vs. baseline state (27%).
Notably, three patient factors significantly increased the odds for disagreement between the two physicians in univariate logistic regression, with the highest odds for this outcome found during assessment of those who underwent cardiothoracic surgery during admission (OR = 4.9; 95% CI, 1.6-15)
The remaining patient factors that heightened the odds for disagreement included chronic ventilation (OR = 4.66; 95% CI, 2.16-10.08) and a primary cardiac admission diagnosis (OR = 3.36; 95% CI, 1.18-9.53).
Researchers observed that 73% of the time, the physicians reported moderate to high diagnostic confidence. The presence of cardiac disease or chronic respiratory failure in patients reduced confidence in the given diagnosis, with the only exception being the presence of a primary respiratory diagnosis.
“Additional guidance on diagnosing PARDS in these specific subgroups, and instructive parameters for interpreting radiographic imaging, would be beneficial to improve the identification of these patients, which would allow further understanding of how treatment interventions impact patient outcomes in future research,” Silver and colleagues wrote.
This study brings light to factors that may cause disagreements among pediatric physicians diagnosing PARDS, but it is important to remember that the clinical environment for these diagnoses is not replicated in this research, Robinder G. Khemani, MD, MsCI, interim vice chair of research for the department of anesthesiology and critical care medicine at Children’s Hospital Los Angeles, and Nadir Yehya, MD, MSCE, attending physician in the pediatric sepsis program and the division of critical care medicine at Children’s Hospital of Philadelphia, wrote in an accompanying editorial.
“Like all diagnostic decisions in medicine ... and in particular all critical illness syndromes, there is rarely a true gold standard that definitively rules in or rules out a disease or condition,” Khemani and Yehya wrote. “Therefore, we must review all the available clinical data and use clinical judgment when making diagnostic decisions related to PARDS, specifically to rule out other causes. Sometimes definitions do not need to be perfect — just useful.”