Low-Value Clinical Practices in Pediatric Trauma Care

Abstract

Importance: Reducing low-value care has the potential to improve patient experiences and outcomes and decrease the unnecessary use of health care resources. Research suggests that low-value practices (ie, the potential for harm exceeds the potential for benefit) in adult trauma care are frequent and subject to interhospital variation; evidence on low-value practices in pediatric trauma care is lacking.

Objective: To estimate the incidence of low-value practices in pediatric trauma care and evaluate interhospital practice variation.

Design, setting, and participants: A retrospective multicenter cohort study in a Canadian provincial trauma system was conducted. Children younger than 16 years admitted to any of the 59 provincial trauma centers from April 1, 2016, to March 31, 2022, were included.

Main outcomes and measures: Low-value practices were identified from systematic reviews of clinical practice guidelines on pediatric trauma. The frequencies of low-value practices were evaluated by estimating incidence proportions and cases per 1000 admissions (low if ≤10% and ≤10 cases, moderate if >10% or >10 cases, and high if >10% and >10 cases) were identified. Interhospital variation with intraclass correlation coefficients (ICCs) were assessed (low if <5%, moderate if 5%-20%, and high if >20%).

Results: A total of 10 711 children were included (mean [SD] age, 7.4 [4.9] years; 6645 [62%] boys). Nineteen low-value practices on imaging, fluid resuscitation, hospital/intensive care unit admission, specialist consultation, deep vein thrombosis prophylaxis, and surgical management of solid organ injuries were identified. Of these, 14 (74%) could be evaluated using trauma registry data. Five practices had moderate to high frequencies and interhospital variation: head computed tomography in low-risk children (7.1%; 33 per 1000 admissions; ICC, 8.6%), pretransfer computed tomography in children with a clear indication for transfer (67.6%; 4 per 1000 admissions; ICC, 5.7%), neurosurgical consultation in children without clinically important intracranial lesions (11.6%; 13 per 1000 admissions; ICC, 15.8%), hospital admission in isolated mild traumatic brain injury (38.8%; 98 per 1000 admissions; ICC, 12.4%), and hospital admission in isolated minor blunt abdominal trauma (10%; 5 per 1000 admissions; ICC, 31%).

Conclusions and relevance: In this cohort study, low-value practices appeared to be frequent and subject to interhospital variation. These practices may represent priority targets for deimplementation interventions, particularly as they can be measured using routinely collected data.

Figure 1.. Relative (Incidence Proportions) and Absolute (Case Volumes) Frequencies of Low-Value Practices in Pediatric Trauma Admissions, 2016-2022

CT indicates computed tomography; DVT, deep vein thrombosis; ICU, intensive care unit; PECARN, Pediatric Emergency Care Applied Research Network; and TBI, traumatic brain injury. ^aApplies to level III/IV referral centers. ^bApplies to pediatric trauma centers and level I/II adult trauma centers. ^cConsultation in the emergency department for neurosurgical centers (level I /II) and transfer to neurotrauma trauma centers for level III/IV centers.

Figure 2.. Summary of Relative and Absolute Frequencies and Interhospital Variation of Low-Value Practices in Pediatric Trauma Admissions, 2016-2022

CT indicates computed tomography; DVT, deep vein thrombosis; ICC, intraclass correlation coefficient; ICU intensive care unit; PECARN, Pediatric Emergency Care Applied Research Network; and TBI, traumatic brain injury. ^aApplies to level III/IV referral centers. ^bApplies to pediatric trauma centers and level I/II adult trauma centers. ^cConsultation in the emergency department for neurosurgical centers (level I/ II) and transfer to neurotrauma trauma centers for level III/IV centers.

Figure 3.. Interhospital Variation in Low-Value Practices for Pediatric Trauma Admissions, 2016-2022

CT indicates computed tomography; DVT, deep vein thrombosis; ICC, intraclass correlation coefficient; ICU, intensive care unit; NA, not applicable; NE, not estimated; PECARN, Pediatric Emergency Care Applied Research Network; PTC, pediatric trauma center; and TBI, traumatic brain injury. ^aInterpreted as low (<5%), moderate (5%-20%), and high (>20%). ^bCould not be estimated due to low sample sizes but considered to be low as numbers were low in all centers.