Impact of a Clinical Pathway and Rapid Direct Influenza Polymerase Chain Reaction Test Introduction on Appropriate Testing and Treatment Among Non-hospitalized Children with Influenza-Like Illness

Background: Influenza-like illness (ILI) symptoms in children are nonspecific, making diagnosis of influenza challenging. A clinical pathway was introduced in the emergency department (ED) of a large pediatric hospital to effectively evaluate high risk patients with ILI. The next year, a new highly accurate, rapid polymerase chain reaction (PCR) test was introduced within the pathway. We aimed to measure the impact of the pathway and new PCR test on rates of appropriate influenza testing and treatment for children with ILI.

Methods: We conducted a retrospective cohort study of non-hospitalized children≤18 years presenting to the ED with ILI over 3 respiratory viral seasons (2012 to 2015). We used an interrupted time series and logistic regression to measure trends in appropriate testing and treatment rates for all children while accounting for high risk status (age<1 year, presence of chronic conditions or immunosuppression) and adjusting for pertinent covariates. Appropriate testing was defined as testing of all high risk children. Appropriate treatment was defined as oseltamivir administration for high risk children with a positive or unknown test result, and no treatment of low risk children.

Results: Among 18012 children with ILI, 4.4% were tested, 3.4% received oseltamivir and 45.5% met high risk criteria. Most tested and treated children were high risk (70.9% and 77.9%). Appropriate testing increased from 54.9% to 55.1% to 58.5% (p<.01). Appropriate treatment increased from 55% to 55.2% to 58.9% (p<.01). In multivariable analysis, the likelihood of appropriate testing was increased with pathway introduction (aOR 1.08, 95% CI 1.01-1.15) and private insurance (aOR 1.10, 95% CI 1.02-1.18), and decreased in males (aOR 0.82, 95% CI 0.77-0.87) and black patients (aOR 0.83, 95% CI 0.76-0.90). The likelihood of appropriate treatment was increased with new PCR test introduction (aOR 1.15, 95% CI 1.08-1.22) and private insurance (aOR 1.10, 95% CI 1.02-1.19), and decreased in males (aOR 0.82, 95% CI 0.77-0.87) and black patients (aOR 0.81, 95% CI 0.74-0.88).

Conclusion: Clinical pathway and rapid PCR test introduction may increase appropriate influenza testing and treatment. However, results also suggest potential sociodemographic disparities.