Examining the Use of Chest Radiographs in Complicated Pneumonia Requiring Thoracotomy

To promote high-value care, lower health care expenditures, and avoid potentially unnecessary exposure to radiation, recent studies have focused on opportunities to safely decrease routine imaging practices in children for a variety of clinical indications. For instance, studies have challenged the need for daily chest radiographs (CXR) in stable intubated patients1–3 and routine CXR following chest tube removal in asymptomatic patients.4,5 However, there is a paucity of data on the use and value of routine CXR in children with chest tubes for complicated community-acquired pneumonia (CAP). In this issue of Hospital Pediatrics, Polich et al6 and Cameron et al7 present findings from 2 complementary studies evaluating nationwide variations in CXR use for children with complicated CAP following chest tube placement for parapneumonic effusion or empyema.The first, by Polich et al,6 is a cross-sectional study that estimated frequency of CXR obtained postchest tube placement across 35 tertiary care pediatric hospitals between 2019 and 2023. The primary outcome was the proportion of days with at least 1 CXR ordered postchest tube placement until discharge. Hospitals were grouped into quartiles based on CXR frequency to explore differences across quartiles in clinical outcomes, including chest tube-related complications, hospital length of stay, and readmissions. The second study, by Cameron et al,7 was a short clinician survey to pediatric hospital medicine representatives from 49 hospitals to gather insights on current institutional policies and practices regarding chest imaging for children with complicated CAP.The main findings from both studies are that the landscape of postthoracostomy CXR policies and practices across institutions are wide and varying. The adjusted proportion of daily CXR ordered postchest tube placement ranged from 30% to 91% across hospitals. They also found that 78% of total variation was attributable to differences among hospitals. Only half of survey respondents (n = 17) noted having a clinical pathway or order set to guide decision-making for children with complicated pneumonia, likely explaining some of this variation. Identifying this degree of variation highlights uncertainty about the optimal approach, likely stemming from the absence of robust, high-quality evidence to guide best practices. This leaves decisions to local culture and individual clinician preferences rather than by evidence-based guidelines. Variation has important implications: it risks unnecessary testing in some settings, underuse in others, and ultimately inequities in the care children receive. Identifying these practice differences provides an opportunity to target studies toward this area of uncertainty to promote more consistent, evidence-based care across institutions.The overall adjusted median frequency of postchest tube CXR obtainment was 0.68 (IQR 0.55–0.81), meaning that more than two-thirds of the days after chest tube placement, children received at least 1 CXR. After grouping hospitals into quartiles based on CXR frequency, there were no differences in outcomes representing potential chest tube-related complications, such as postprocedural pneumothorax, chest tube dislodgement, repeat chest tube placement, or video-assisted thoracoscopic surgery (VATS). Overall, rates of all these events were low, keeping in mind that administrative data rely on the accuracy of diagnosis and procedure codes, and, therefore, may underestimate actual rates. There were also no statistical differences seen across quartiles in postthoracostomy days, length of stay, 30-day readmissions, or overall costs.Of note, children that were intubated were excluded because of the possibility that CXR would be obtained with higher frequency to confirm endotracheal tube placement. Therefore, study conclusions may not extend to the subset of patients who are most critically ill or those potentially at higher risk of postthoracostomy complications, for whom routine CXR may be indicated. Although several recent studies challenge the need for routine daily CXR for stable mechanically ventilated children, a multicenter study by Quasney et al,1 evaluating the usefulness of routine morning CXR, found that the likelihood of an intervention triggered by findings from a daily routine CXR was significantly higher in children with at least 1 device (including chest tube) compared with no devices (38% vs 19%).The use of frequent CXR is contrary to the national guideline from the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America, which does not recommend daily radiographs in children who are clinically stable after thoracostomy or VATS.8 However, this guideline points out the level of evidence supporting this recommendation is “low quality,” emphasizing the importance of research in this area to better guide clinical practice.8 Moreover, only 9% of survey respondents reported finding routine CXR helpful in their treatment of patients. Two major caveats were that no survey respondents were surgeons or interventional radiologists, the ones typically placing and managing the chest tubes, and only 1 hospitalist was surveyed per hospital, which may not represent the perspective of other hospitalists at these institutions. Still, the low percentage of hospitalists finding value in routine CXR may point to the rarity with which, in their experience, routine CXR affect daily patient treatment. Given the lack of differences in chest tube-related complications or clinical outcomes seen in this study between higher and lower CXR-using hospitals, there may be opportunities to move toward fewer CXR with targeted deimplementation. Like many things in pediatrics, this may be a scenario in which less is more, and there are opportunities to improve costs and risk of unnecessary harm to children by doing less. Although these institutional-level results are encouraging, more evidence is needed before implementing or deimplementing practice changes. Specifically, we need to better understand which children and under which circumstances imaging adds value beyond clinical assessment and monitoring of chest tube output.An additional question that deserves further investigation is whether chest ultrasonography (US) can replace CXR as a serial imaging modality for children who do warrant closer, potentially daily monitoring for chest tube position and development of complications. Chest US is now the preferred method for evaluation and diagnosis of parapneumonic effusions given its high specificity, greater sensitivity compared with CXR and comparable sensitivity to computed tomography (CT) scans, and benefit of being radiation free.8,9 Both studies found that clinicians are using US for complicated CAP. However, only 41% of patients had more than 1 US per encounter, and only 2 sites in the survey reported using US for monitoring, suggesting that CXR are still the default modality for serial imaging postthoracostomy. With the favorable diagnostic, safety, and cost characteristics of US compared with CXR, more efforts should be made to promote US use in this population. Additionally, US can be portable, quick, and cost effective, particularly when performed at the bedside by a trained clinician. Only 1 of 28 survey respondents reported hospitalists performing chest US. Although performing point-of-care US is fairly routine for many intensivists and clinicians in adult hospitals, this is not yet a widely adopted practice in many pediatric hospitals or pediatric training programs.10 Although there are barriers to implementation, there is an opportunity to better educate and train more pediatric hospitalists on the use of US, at least for this indication, and set up a system so that pediatric hospitalists can bill for this service. This would empower hospitalists to take more ownership on monitoring effusions and evaluating for complications. Although the use of chest US excels at visualizing the pleural space and is likely adequate for assessing chest tube position, it does still have important limitations that are worth noting, including its inability to adequately assess extensive cavitations, bronchopleural fistula, or vascular complications, findings which are also difficult to visualize on CXR and require CT scans for full characterization.11 Given the shift from fee-for-service to bundled diagnosis-related groups based reimbursements and financial challenges across many pediatric hospitals, reducing imaging costs, either by decreasing use or switching to lower cost methods, is important to consider.These studies, while focused on a relatively small subset of children with CAP, underscore a broader issue regarding the need to deliver high-value care for all children with pneumonia. Beyond the overuse of CXR, antibiotics, and ancillary diagnostics (eg, complete blood cell counts) for children with uncomplicated CAP, opportunities for greater value lie in reducing unnecessary admissions for children able to be treated and monitored outpatient as well as finding ways to shorten length of stay for those ultimately needing hospitalization. These improvements matter not only in terms of health care costs and resource use, but also in the day-to-day disruptions they impose on children and families. Although process improvement efforts have aimed to reduce length of stay and subsequent hospitalization costs for CAP, there has not been substantial change over the last decade in length of stay for children hospitalized with pneumonia,12 which remains one of the most common and costly diagnoses in pediatrics.13,14 To achieve this, clinicians and systems must think creatively—for example, by implementing home oxygen programs, reconsidering routine use of intravenous fluids or antibiotics, and adopting rapid oxygen weaning protocols, early intravenous-to-oral antibiotic transitions, and timely cessation of intravenous fluids to challenge children to try oral intake. Ultimately, well-designed studies are needed that do not just ask whether we can deliver a service but whether that service meaningfully improves outcomes for children and their families from diagnosis, to treatment, to monitoring. In the end, improving value in pediatric pneumonia treatment is not from the tests that we order but rather from the hospital days that we prevent.