Clinical Associate Professor (Affiliated) Stanford University School of Medicine Valley Children's Healthcare Fresno, California, United States
Background: National bronchiolitis guidelines recommend the reduction of two practices associated with medical overuse: (1) respiratory virus panel (RVP) testing in otherwise healthy children and (2) continuous pulse oximetry monitor use (cSpO2) when not receiving supplemental oxygen. Objective: To evaluate the association between RVP testing and cSpO2 overuse in children with bronchiolitis and to evaluate for any spillover effect of cSpO2 deimplementation to RVP testing rates. Design/Methods: This multicenter cross-sectional study was conducted concurrent with baseline (12/2021-6/2022) and active deimplementation phases (12/2022-4/2023) of the Eliminating Monitor Overuse Trial, a hybrid effectiveness-deimplementation trial of strategies to reduce cSpO2 in children with bronchiolitis. Active deimplementation included education and audit and feedback, with or without clinical decision support within the electronic medical record. Study population included a convenience sample of patients aged 2-23 months with bronchiolitis not receiving supplemental oxygen.
The primary exposure was RVP testing status (i.e. no RVP vs RVP negative vs RVP positive vs RVP performed and results unknown). The primary outcome was cSpO2 overuse. Multivariable conditional logistic regression using hospital-level cSpO2 quartile as strata and covariates including baseline vs active deimplementation phase, time off of oxygen, age combined with preterm birth, presence of a feeding tube, neurological impairment, and overnight observation were included. A chi squared test was performed to evaluate for change in RVP testing rates before vs after cSpO2 deimplementation strategy roll-out. Results: The sample included 4341 patient observations from 36 US and Canadian hospitals in the Pediatric Research in Inpatient Settings Network. 12% had no RVP, 54% had RVP positive, 23% had RVP negative, and 11% had RVP unknown. Multivariable conditional logistic regression found a borderline association between RVP testing and cSpO2 overuse, with RVP positive being associated with higher odds of cSpO2 overuse compared to no RVP testing (43% overuse for RVP positive vs 32% overuse for no RVP; aOR 1.26; 95% CI .995-1.60; P=.055; Table 1). RVP testing rates did not decrease over time following cSpO2 deimplementation (P=.429; Figure 1).
Conclusion(s): cSpO2 overuse was more prevalent when RVP testing was performed and positive, consistent with other work demonstrating that initial overuse has the potential to lead to an overuse cascade. In contrast, strategies which led to successful deimplementation of cSpO2 did not lead to decreased RVP testing.
