Partnering With Patients and Families to Champion Deimplementation and Reduce Low-Value Care

Clinical practice guidelines (CPGs) are widely used in medicine to improve quality of care. They have also been implemented as clinical decision support tools to combat medical overuse and the provision of care in the absence of a clear medical indication or when the benefit does not outweigh the risk.1 Designing CPGs based on strong evidence can reduce practice variation, decrease harm, and maximize positive outcomes for patients. Initiatives such as the Choosing Wisely Campaign, which focus on reducing frequently overused low-value tests and treatments, often lead to CPG development, encouraging healthcare providers and organizations to make these recommendations part of routine practice.2In this edition of Hospital Pediatrics, Cooper and colleagues describe the results of a retrospective quasiexperimental study that evaluated the impact of a CPG focused on the management of preseptal cellulitis in hospitalized children.3 The CPG focused on antimicrobial use, blood tests and computed tomography (CT) scans, with an aim to standardize practice, reduce broad spectrum antibiotics, and reduce practice variation. Over the course of 11 years, 236 patients were included, 175 patients before and 61 patients after the CPG was implemented. After CPG implementation, there were reductions observed in all empirical antibiotic categories, including broad spectrum (100% to 66%), dual/triple therapy (47% to 16%), and methicillin-resistant Staphylococcus aureus active agents (86% to 26%). Similar changes were noticed in definitive antibiotic changes. In parallel, there were reductions in complete blood count and blood culture orders from 75% to 57% (P = .014) and 32% to 18% (P = .047), respectively.Although the decrease in broad-spectrum antibiotic use was significant, study limitations include a smaller post-CPG implementation cohort, and limited postdischarge information. Given the small sample size, the logistic regression analysis did not include important covariates such as age or sex. Another limitation is that most patients with preseptal cellulitis are discharged from the emergency department (ED) and do not require admission to hospital. Therefore, the population of patients with preseptal cellulitis included were those with more severe disease requiring hospital admission. It is unclear whether patients with milder disease also received broad spectrum antibiotics and whether this represents another opportunity for antimicrobial stewardship intervention.Interestingly, the authors noted an increase in CT scan rates after CPG implementation (pre-CPG 79% and post-CPG 84%); the reasons for this increase were not discussed. The authors noted that CT scans are “low yield” in orbital infections; however, little guidance was provided to discern which patients would most benefit from CT scans or how imaging findings would change management along the CPG pathway. In the CPG, “difficult eye exam” was an indication to consider CT scan, but in practice most eye examinations are expected to be difficult given that this infection typically affects those younger than age 5 years (median age 4 years in this study), making this indication unhelpful for risk stratification.CT scan with and without contrast is the gold standard for diagnosing severe orbital infections, but its main utility is to identify children who require surgical intervention or who have complications. Nguyen et al. found 69% of children who presented with a severe orbital infection across 6 Canadian children’s hospitals received a CT scan, which is comparable to 74% in a study of 42 US children’s hospitals.4,5 Importantly, these studies included hospitalized children with both preseptal and orbital cellulitis, in which the imaging rate for preseptal cellulitis was likely lower. In the absence of predictors of surgical intervention, such as older age, elevated C-reactive protein and/or white blood cell count, and proptosis, it may be prudent to wait 24 to 48 hours to monitor the child’s clinical trajectory before ordering a CT scan, particularly in preseptal cellulitis.6 The exclusion of children with orbital cellulitis is a limitation in Cooper et al’s study, as the CT imaging rates would presumably be even higher if these patients were included. Although few will argue against the value of deimplementation initiatives focused on antimicrobial stewardship, would CT scans, with their associated risks from ionizing radiation, not warrant the same deimplementation approach?In 2011, the Institute of Medicine published standards for CPG development that included a call for the inclusion of patients and the public.7 Unfortunately, patient involvement in guideline development is variable and the best methods for doing this are unclear.8 Yet, the impact of patient contribution to CPG development can be significant. In a study by Armstrong et al, 2 guideline development groups were convened, 1 with and 1 without patient representatives. The authors found that patient representatives were instrumental in contributing to a culture of “patient-centeredness” by identifying issues that may be overlooked by medical professionals, identifying patient-relevant topics and outcomes and influencing implementation strategies.9CPGs became popular in the 1970s and were initially based on expert consensus. The emergence of evidence-based medicine in the 1990s led to more rigorous approaches to guideline development, with formalized frameworks becoming the norm in the 2000s. Grading of Recommendations Assessment, Development and Evaluation (GRADE) was published in 2008 to unite these frameworks, providing a systematic approach for rating the quality of evidence, and a decision framework for recommendations.10 The growing realization that decisions should also consider nonevidence factors led to the GRADE framework including patient values and preferences in recommendation development. These “preference-sensitive” recommendations are labeled “conditional” or “weak,” meaning that there is a paucity of evidence or the quality of evidence is uncertain, benefits and harms are closely balanced, wide variation in patient preferences is expected and/or no evidence on patient preferences exists (even with strong evidence of benefit).11In pediatrics, there are few high-quality studies demonstrating superiority of 1 treatment over another, leading to conditional recommendations. At the same time, pediatricians, children and youth, and their caregivers report incorporating shared decision-making as a top priority in hospital care.12 These conditional recommendations provide an opportunity to integrate patients’ values and preferences using shared decision-making. Similarly, in cases of rare diseases with small population sizes, the paucity of evidence may allow for increased flexibility in management and greater emphasis on shared decision-making with families to understand their values and priorities.13Shared decision-making tools can encourage patient involvement and are extremely valuable when making decisions based on conditional recommendations and uncertain evidence.10 For example, the PECARN prediction rules for minor head trauma provide little guidance in the choice of home observation versus head CT for children at intermediate risk of clinically important traumatic brain injury (ciTBI). Hess et al randomized parents of children with ciTBI to shared decision-making facilitated by the Head CT Choice decision aid or usual care: parents who used a decision aid had greater knowledge, less decisional conflict, greater physician trust, and greater involvement in CT decision-making. They also had shorter ED lengths of stays and lower healthcare utilization at 7 days without any missed cases of ciTBI in either study arm.14The concepts of risk and harm are central to deimplementation. Studies on antimicrobial stewardship have described how provider perceptions of risk impact antibiotic prescribing behaviors: providers who made decisions based on their perception of the patient remaining sick versus potentially improving on antibiotics were more likely to prescribe. However, this approach assumes antibiotics are harmless and ignores the possibility that antibiotics may contribute to negative patient outcomes. Conversely, providers who weighed decisions in terms of ongoing illness versus harms from antibiotics had lower prescribing rates.15 Antimicrobials are the leading cause of ED visits for adverse drug reactions,16,17 and their use in childhood has been linked to a variety of negative health outcomes including asthma, obesity, and inflammatory bowel disease.18 With up to 50% of antibiotic use considered inappropriate,19 and the clear short- and long-term risks with antibiotic use, antibiotic prescribing is an example of clinical decision-making in which discussions with patients and families around preferences, values and risk-tolerance are paramount. Framing decisions in terms of number needed to treat and number needed to harm can also help provide a more objective evaluation of a potential treatment’s benefits and harms.Can partnering with patients and families aid in deimplementation? The evidence suggests it can. A recent meta-analysis and systematic review examined the impact of patient engagement on deimplementation interventions for low-value care. Twenty-two studies were included: 9 were randomized clinical trials and 13 were quasi-experimental. The most common clinical context was adult patients making decisions about medical treatments in primary care (largely medication prescribing) and the most frequent interventions were patient-oriented educational materials and tools for shared decision-making. Six studies focused on pediatric patients and their caregivers. Deimplementation interventions that engaged patients within the patient–clinician interaction led to significant reductions in low-value care ranging from 26% to 39%.20 Although promising, future research should focus on the best way to engage pediatric patients and families in deimplementation efforts in different care settings. In addition, new research techniques, such as the Desirability of Outcome Ranking, may be an opportunity for patient values to be proactively incorporated into research studies and interpretation of findings rather than considered after the fact.21In this era of patient-centered practice, involving our patients in deimplementation initiatives and engaging them in shared decision-making could be a powerful way to optimize outcomes and reduce low-value care.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.