eprintid: 37807
rev_number: 25
userid: 9243
dir: disk0/00/03/78/07
datestamp: 2020-07-30 17:41:13
lastmod: 2020-07-30 17:41:13
status_changed: 2020-07-30 17:41:13
type: thesis_degree
metadata_visibility: show
contact_email: saulgsean@gmail.com
eprint_status: archive
creators_name: Saul, Sean/G
creators_email: ses164@pitt.edu
creators_id: ses164
creators_orcid: 0000-0003-0304-4001
title: The Relationship Between the Ordering of a Respiratory Viral Panel and the Presence of Acute Respiratory Illness and Influenza-like Illness Symptoms Reported in the Electronic Medical Record
ispublished: unpub
divisions: sch_gsph_infectiousdiseasesmicrobiology
full_text_status: restricted
keywords: influenza, flu, RVP, respiratory viral panel
abstract: Background: Influenza causes a large burden of hospitalizations in the United States (U.S.) each year. Influenza can lead to viral or bacterial pneumonia, dehydration, ear infections, and sinus infections, while serious complications can include inflammation of the heart (myocarditis), brain (encephalitis) or muscle (myositis, rhabdomyolysis) tissues, and multi-organ failure (for example, respiratory and kidney failure). Efficient mechanisms to accurately and quickly identify Influenza are needed, especially with respect to those with more severe illness.
Methods: Chart reviews were conducted on a random sample of 1,029 patients appearing on a clinical informatics algorithm (CIA) generated list from 12/1/15 to 5/11/16. This list was used for recruitment in the HAIVEN study which is a Centers for Disease Control and Prevention (CDC)–funded, multicenter, test-negative, case-control study to determine the Influenza vaccine effectiveness (VE) against hospitalization. The CIA queried medical record databases of patients who were 18 years of age and older admitted to University of Pittsburgh Medical Center (UPMC) St. Margaret’s Hospital in the previous 3 days using specified terms and diagnosis codes located in admission notes, emergency department notes, chief complaint upon registration, or presence of a respiratory viral panel charge (RVP). Using chart review data, each patient was deemed eligible for the study using 2 CDC descriptive eligibility boxes that aim to identify Acute Respiratory Illness (ARI) and Influenza-like Illness (ILI) using specified symptom-based and diagnosis-based terms. A Kappa test determined agreement between having a term listed on these eligibility boxes and RVP status. Binary and multivariate logistic regression tests were used to characterize the clinical features of those missed by the clinical RVP approach but found by the CDC’s screening criteria and to characterize the clinical features of those missed by the CDC’s screening criteria but found by the clinical RVP approach.  
Results: Of the 1,029 patients reviewed, 290 patients met the eligibility criteria and received an RVP ordered by a physician and 201 met the eligibility criteria but did not have an RVP ordered by a physician. A Kappa test resulted in a weak agreement between the 2 descriptive eligibly boxes and RVPs (kappa=.43). Both RVP status and the CDC’s criteria were statistically significantly associated with fever, chest x-ray, and CT-scan. 
Conclusion: The findings of this study suggest that physicians are ordering RVPs for ILI only moderately well and improvement through standardized ordering criteria may be needed. Using a CIA for recruitment for Influenza and other respiratory diseases studies was beneficial. The regression model further confirmed these findings. A hybrid case definition for inpatient Influenza may be needed.
Public Health Statement: The likelihood of being diagnosed with ARI or ILI by RVP was significantly higher if a patient had fever, chest x-ray, or CT-scan indicated. The likelihood of meeting the CDC’s criteria was significantly higher if a patient had fever, chest x-ray, or lymphocyte count indicated in the EMR.
date: 2020-07-30
date_type: published
pages: 55
institution: University of Pittsburgh
refereed: TRUE
etdcommittee_type: committee_chair
etdcommittee_type: committee_member
etdcommittee_type: committee_member
etdcommittee_name: Frank, Linda/R
etdcommittee_name: Nowalk, Mary/P
etdcommittee_name: Yassin, Mohamed/H
etdcommittee_email: frankie@pitt.edu
etdcommittee_email: nowalkmp@upmc.edu
etdcommittee_email: mhy8@pitt.edu
etdcommittee_id: frankie
etdcommittee_id: TNOWALK
etdcommittee_id: mhy8
etd_defense_date: 2020-02-18
etd_approval_date: 2020-07-30
etd_submission_date: 2019-11-18
etd_release_date: 2020-07-30
etd_access_restriction: 2_year
etd_patent_pending: FALSE
thesis_type: thesis
degree: MPH
citation:   Saul, Sean/G  (2020) The Relationship Between the Ordering of a Respiratory Viral Panel and the Presence of Acute Respiratory Illness and Influenza-like Illness Symptoms Reported in the Electronic Medical Record.  Master's Thesis, University of Pittsburgh.    (Unpublished)  
document_url: http://d-scholarship-dev.library.pitt.edu/37807/2/SaulSG_MPHthesis_April2020.pdf