Are Pathways an Untapped Opportunity to Generate Knowledge and Improve Outcomes for Croup? Free

In 2 studies featured in this month’s issue of Hospital Pediatrics, Lefchak et al¹  and Bergmann et al²  examine the presence, characteristics, and impact of croup pathways at 38 children’s hospitals participating in the Pediatric Health Information System database. In recent years, pathways have become a growing part of inpatient clinical practice. Given the significant investment involved with writing and implementing pathways, it is important to understand if and how pathways improve patient outcomes. Lefchak et al and Bergmann et al attempted to answer this question in relation to croup. The authors first examined differences in pathways across hospitals and described variations in pathway content. Next, the authors compared outcomes for croup subjects cared for at pathway hospitals versus nonpathway hospitals and found no difference in the adjusted odds ratio of admission or mean cost. It is important to unpack these findings given the potential impact of pathways on the practice of pediatric hospital medicine. With this in mind, we present 2 issues for consideration. First, we discuss the role of pathways to standardize practice and generate new learning in the absence of strong evidence to guide care. Next, we explore how we can measure pathways’ impact on care and outcomes.

Though the terms are often used interchangeably and there is some overlap, in this commentary, we distinguish between care pathways and practice guidelines. Clinical practice guidelines (CPGs) are evidence-based recommendations developed by expert panels or professional organizations. They are systematically developed using a rigorous process that involves reviewing and analyzing the available scientific literature, synthesizing the evidence, and formulating recommendations for diagnosing, treating, and managing specific conditions or diseases. Clinical pathways are developed for a local context by multidisciplinary teams and provide an algorithmic approach to patient care by outlining the indicated diagnostic tests, treatments, medications, and interventions. Importantly, clinical pathways can be a tool for implementing recommendations from CPGs because pathways facilitate adaptation to the local context and make evidence rapidly accessible to clinicians at the point of care.

The authors’ objectives to present a full, detailed depiction of the current state of croup pathways and to evaluate the link between pathways and outcomes are commendable. A clear strength of the studies was the complete and thorough description of pathway content across hospitals. It was notable that the authors report imperfect agreement between reviewers for some pathway recommendations (eg, racemic epinephrine [RE] doses recommended before admission), likely signaling that these recommendations are not clear to end users. Pathway recommendations around hospital discharge criteria were not reported by the authors and only 45% of pathways made recommendations regarding repeat dexamethasone dosing. This is important given prior studies have demonstrated variability in steroid-dosing practices and an association between dexamethasone dosing and utilization outcomes such as length of stay.³  Although the authors’ goal of evaluating outcomes associated with pathways is well-intentioned, we would contend that the mere presence of pathways does not change outcomes. Therefore, the primary limitation of the study by Lefchak et al, which the authors acknowledge, was the lack of measures of implementation or adoption. Additionally, the authors looked at outcomes associated with components of the pathways where recommendations differed by site (admission criteria), making it especially hard to demonstrate a change in outcomes.

Although half of the Pediatric Health Information System hospitals in the first study had croup pathways, no croup CPG has been published by a pediatric professional society in the United States to recommend best practices for the inpatient management of croup. There is no strong evidence to guide admission decisions or in-hospital-specific management of croup.^4,5  With this lack of evidence, it is perhaps not surprising that Bergmann et al found variation across pathways’ content and recommendations. More surprisingly, despite evidence that it is a poor predictor of the need for further interventions,⁶  the pathways were strikingly similar in their recommendation for admission based on the number of RE doses administered. Pathways varied in the number of doses recommended before admission (the majority recommended admission for ≥2 RE doses) and criteria for the administration of RE (30% of pathways recommended RE for any degree of stridor). This begs the question: should pathways encourage admission based on RE dosing when it may be prescribed without clear indication and does not predict the need for further treatments once hospitalized? Using RE doses, the pathways likely recommend admission for patients who will derive limited or no benefit from hospitalization and could be safely discharged from the emergency department. This may explain the consistent finding that many children admitted for croup require no additional interventions following admission,^7–10  and why, in adjusted models, Lefchak et al found that pathways had no significant impact on admission rates and costs.

We propose that the persistence of moderate or severe symptoms after steroid onset would provide a better decision point for admission, and a more meaningful measure of pathway impact (vs. admission rate) would be the need for further interventions following admission. More research is needed to determine which patients are likely to require additional treatments to inform a CPG for the management of croup and decrease the variability of institutional pathway recommendations. While awaiting definitive evidence for children hospitalized with moderate to severe croup, pathways serve 2 purposes. First, to promote treatments where strong evidence exists (eg, dexamethasone) and advise against non–evidence-based tests such as neck films and viral panels. Second, with care pathways in place, data on croup management can be systematically collected, pooled, and analyzed to generate valuable insights about the effectiveness of various management strategies. These findings can then be used to both refine pathways and grow the evidence base for croup guidelines. Important areas for standardization of care and systematic study in croup include admission criteria, indications for RE, indications for repeat steroid dosing, steroid dosing intervals, and length of in-hospital observation.¹¹ 

Unclear from these studies is how institutions encouraged clinicians to use croup pathways. There are no measures of adoption or descriptions of implementation strategies used at each institution. Absent these measures, it is not possible to conclude that the pathway recommendations were ineffective at improving outcomes. Rather, the pathway recommendations may not have been adopted by clinicians. This highlights the need for implementation efforts to accompany the pathways. Implementation science is a rapidly growing field of study to promote the adoption, integration, and sustainability of evidence-based practices in routine clinical practice. Proctor et al suggested a multistage framework for the adoption of evidence-based practice: access, adoption, implementation, and assessment.^12,13  The croup studies discussed here are limited in scope to the access phase because all that can be inferred from the available data are that providers at these sites have access to pathways.

Once access to pathways is established, the next step is to consider the adoption of pathways. Namely, are clinicians at institutions with pathways using them to inform their clinical decisions? In the absence of adjustment, the authors found that hospitals with pathways recommending admission after only 1 RE dose had lower admission rates than those that recommended 2 or 3, and the trend persisted after adjustment, though was no longer statistically significant. This may suggest that providers at these institutions have not adopted the pathways because it is difficult to imagine a scenario in which admission rates are lower at institutions recommending admission after 1 dose of RE compared with hospitals with pathways recommending admission after >1 dose of RE. Because of the limitations of their data, the authors were not able to provide any information related to adoption.

Another important consideration is how the pathways are implemented. Prior quality improvement work in croup suggests that a pathway may decrease hospital admission rates for croup. However, although the authors of that work suggested that the pathway was the primary driver of this improvement, in their paper, they discuss the use of additional implementation strategies that accompanied the pathway.¹⁴  Similarly, in studies of bronchiolitis, authors found that clinical pathways alone did not lead to sustained changes in practice. However, when pathways were implemented with higher reliability strategies, such as order sets and best practice alerts, their effectiveness increased.¹⁵  Without any evidence that implementation work was done by the institutions that had pathways and no measures of adoption, it is hard to draw conclusions about the impact/effectiveness of pathways or specific pathway recommendations.

This study provides a complete view of croup pathway content and examines outcomes. Future work is needed to refine evidence-based management of inpatient croup and guide admission decisions. As seen in these studies, when pathways vary in their recommendations, it is difficult to demonstrate a meaningful impact on outcomes. In the absence of strong evidence, the use of pathways represents an untapped potential to standardize care across institutions, study both adoption and patient outcomes, and add to our evidence base. It is important to remember that, even with strong evidence, pathways are still just a starting place and must be accompanied by implementation efforts to realize benefits for patients and health care systems.