Costs of a clinical pathway with point‐of‐care testing during influenza epidemic in a Dutch hospital

Abstract Our study aim was to determine how a new clinical pathway, including PCR‐based influenza point‐of‐care test (POCT), influences the hospitalisation costs of patients suspected of influenza presenting at the emergency department of a Dutch hospital during two consecutive influenza epidemics (2016‐2017 and 2017‐2018). Compared to mean costs per patient of €3661 in 2016‐2017, the implementation of this new clinical pathway with influenza POCT in 2017 was associated with mean costs per influenza‐positive patient of €2495 in 2017‐2018 (P = .3). Our study suggests favourable economic results regarding a new clinical pathway with influenza POCT, reflecting a more efficient care of patients suspected of influenza presenting at the emergency department.


| INTRODUC TI ON
Seasonal influenza epidemics cause substantial morbidity and lead to excess hospitalisations and mortality, especially in the elderly. 1 A sudden increase in the number of patients requiring hospital care for severe acute respiratory infections (SARI), especially pneumonia as complication of influenza virus infection, may pose a significant burden for hospitals in managing bed and staff capacity. 2 Whether a SARI patient is tested for influenza virus infection is the decision of the individual attending physician in most hospitals and mainly relies on laboratory-based PCR testing with a turnaround time of 24-48 hours.
More frequent testing and timely diagnosis of influenza may better guide isolations and improve patient flow through the hospital and thereby contribute to a more efficient management of patients. 3,4 In influenza season 2017-2018, the Jeroen Bosch Hospital (JBH) in 's-Hertogenbosch, the Netherlands, implemented a PCR-based point-of-care test (POCT) for influenza virus type A and type B and respiratory syncytial virus (RSV) for all patients presenting with SARI at the Emergency Department (ED). 5 Furthermore, a temporary ward dedicated specifically to care of influenza-positive patients was established. The POCT (Cobas Liat Assay) has high sensitivity and specificity for influenza virus types A and B, and RSV, and can be performed by non-laboratory personnel at the ED. 3 The aim of this study was to determine how a new clinical pathway, including POCT, influences the hospitalisation costs of patients suspected of influenza presenting at the ED.

Abstract
Our study aim was to determine how a new clinical pathway, including PCR-based

| Clinical data for cost estimates
Retrospective data were collected using the electronic patient records. The dataset included number of requested influenza tests (LDT/POCT), admissions to hospital (ward/ ICU), isolated patients, treatment for patients suspected of influenza (antibiotics/antivirals), and median length of hospital stay (LOS). Costs were calculated from the start of the epidemic using a bottom-up approach following Dutch guidelines for economic evaluations. 6 Test costs were retrieved from JBH, ED consultations and hospitalisation costs were based on the national cost manual for economic evaluations, 7 isolation costs were taken from literature, 8 and medication costs were taken from the National Health Care Institute website. 9 Hospitalisation costs were calculated by multiplying recorded units of used healthcare resource with corresponding unit prices (Table S1). The maximum isolation duration, additional diagnostics and type and duration of antibiotic and/or antiviral treatment were based on recent literature (footnotes Table 1). All costs were expressed in 2018 euros.

| RE SULTS
The

| PUBLIC ATION E THIC S
The Dutch Medical Research Involving Human Subjects Act did not apply to this study, because anonymous data were used and there were no interventions other than routine clinical care. A waiver for full medi-

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets generated during and/or analysed during the current study are not publicly available yet, due to legal concerns and ongoing additional research. Data can be made available for peer review on reasonable request through contacting the corresponding author.