Layout 1


ISDS Annual Conference Proceedings 2012. This is an Open Access article distributed under the terms of the Creative Commons Attribution-
Noncommercial 3.0 Unported License (http://creativecommons.org/licenses/by-nc/3.0/), permitting all non-commercial use, distribution, and
reproduction in any medium, provided the original work is properly cited.

ISDS 2012 Conference Abstracts

Enhanced Influenza Surveillance using Telephone Triage
Data in the VA ESSENCE Biosurveillance System
Cynthia A. Lucero-Obusan*1, Carla A. Winston1, Patricia L. Schirmer1, Gina Oda1,
Anoshiravan Mostaghimi1 and Mark Holodniy1, 2

1Department of Veterans Affairs, Office of Public Health, Washington, DC, USA; 2Division of Infectious Diseases & Geographic
Medicine, Stanford University, Stanford, CA, USA

Objective
To evaluate the utility and timeliness of telephone triage (TT) for

influenza surveillance in the Department of Veterans Affairs (VA).

Introduction
Telephone triage is a relatively new data source available to bio-

surveillance systems.1-2 Because early detection and warning is a high
priority, many biosurveillance systems have begun to collect and an-
alyze data from non-traditional sources [absenteeism records, over-
the-counter drug sales, electronic laboratory reporting, internet
searches (e.g. Google Flu Trends) and TT]. These sources may pro-
vide disease activity alerts earlier than conventional sources. Little is
known about whether VA telephone program influenza data corre-
lates with established influenza biosurveillance.

Methods
Veterans phoning VA’s TT system, and those admitted or seen at a

VA facility with influenza or influenza-like-illness (ILI) diagnosis
were included in this analysis. Influenza-specific ICD-9-CM coded
emergency department (ED) and urgent care (UC) visits, hospital-
izations, TT calls, and ILI outpatient visits were analyzed covering
2010-2011 and 2011-2012 influenza seasons (July 11, 2010-April 14,
2012). Data came from 80 VA Medical Centers and over 500 outpa-
tient clinics with complete reporting data for the time period of in-
terest. We calculated Spearman rank-order coefficients, 95%
confidence intervals and p-values using Fisher’s z transformation to
describe correlation between TT data and other influenza healthcare
measures. For comparison of time trends, we plotted data for hospi-
talizations, ED/UC visits and outpatient ILI syndrome visits against
TT encounters. We applied ESSENCE detection algorithms to iden-
tify high-level alerts for influenza activity. ESSENCE aberration de-
tection was restricted to the 2011-2012 season because limited
historical TT and outpatient data from 2009-2010 was available to
accurately predict aberrancy in the 2010-2011 season. We then cal-
culated the peak measure of healthcare utilization during both in-
fluenza seasons (2010-2011 and 2011-2012) for each data source and
compared timing of peaks and alerts between TT and other healthcare
encounters to assess maximum healthcare system usage and timeli-
ness of surveillance.

Results
There were 7,044 influenza-coded calls, 564 hospitalizations,

1,849 emergency/urgent visits, and 416,613 ILI-coded outpatient vis-
its. Spearman rank correlation coefficients were calculated for in-
fluenza-coded calls with hospitalizations (0.77); ED/UC visits (0.85);
and ILI-outpatient visits (0.88), respectively (P< 0.0001 for all cor-
relations). Peak influenza activity occurred on the same week or
within 1 week across all settings for both seasons. For the 2011-2012
season, TT alerted with increased influenza activity before all other
settings.

Conclusions
Data from VA telephone care correlates well with other VA data

sources for influenza activity. TT may serve to augment these exist-
ing clinical data sources and provide earlier alerts of influenza activ-
ity. As a national health care system with a large patient population,
VA could provide a robust early-warning system for influenza if on-
going biosurveillance activities are combined with TT data. Addi-
tional analyses are needed to understand and correlate TT with
healthcare utilization and severity of illness.

Keywords
Surveillance; Influenza; Telephone triage; Veterans

References

1. Yih WK, Teates KS, Abrams A, Kleinman K, Kulldorff M, Pinner R,
Harmon R, Wang S, Platt R: Telephone triage service data for detec-
tion of influenza-like illness. PLoS One 2009, 4(4):e5260.

2. van Dijk A, McGuinness D, Rolland E, Moore KM: Can Telehealth
Ontario respiratory call volume be used as a proxy for emergency de-
partment respiratory visit surveillance by public health? CJEM 2008,
10(1):18-24.

*Cynthia A. Lucero-Obusan
E-mail: cynthia.lucero@va.gov

Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org * 5(1):e103, 2013