2154. How well are we estimating the true burden of acute gastroenteritis? Validation of acute gastroenteritis-related ICD codes in pediatric and adult U.S. populations
Session: Poster Abstract Session: Healthcare Epidemiology: Epidemiologic Methods
Saturday, October 6, 2018
Room: S Poster Hall
Background: Background: International Classification of Diseases (ICD) diagnostic codes from acute gastroenteritis (AGE)-associated medical encounters are used for AGE disease burden estimates, yet the validity of AGE-related ICD codes in both pediatric and adult populations is unknown. We estimated the validity of AGE-related diagnostic codes in these populations using two different multi-regional AGE active surveillance platforms.

Methods: Diagnostic codes, demographic and clinical characteristics, and stool pathogen results from AGE-associated medical encounters were obtained for enrolled children <5 years old from 7 sites in NVSN from 12/1/2011 to 6/30/2016, and for adult Veterans in 4 sites from SUPERNOVA from 12/1/2016 to 2/28/2018. SUPERNOVA also enrolled age- and time-matched non-AGE controls. Using AGE cases from the active surveillance networks, sensitivity and specificity of AGE ICD codes were estimated overall and stratified by age and health care setting using exact binomial tests.

Results: ICD codes were collected from 14,952 enrolled children <5 years old with AGE, and 625 enrolled adults (525 AGE cases and 100 controls). The sensitivity of all-cause AGE codes in children was 54% (9127/14952, 95% confidence interval [CI] 54%-55%), and in adults was 54% (283/525; 95% CI 50%-58%), with a specificity of 100% (100/100; 95% CI 97%-100%). Stratified analyses demonstrated higher sensitivity of all-cause AGE codes in children in the inpatient as compared to outpatient setting: 59% (417/675; 95% CI 57%-61%) versus 53% (934/1827; 95% CI 52%-54%). In adults, this trend was reversed; all-cause AGE codes had a higher sensitivity in the outpatient as compared to the inpatient setting: 72% (50/69; 95% CI 60%-83%), versus 51% (233/456; 95% CI 46%-56%), respectively.

Conclusion: Across two different AGE active surveillance platforms, one enrolling only children and one enrolling only adults, the estimated sensitivity of all-cause AGE ICD codes were similarly low. This suggests that current national estimates for AGE disease burden may be underestimating the true burden of AGE pathogens in the US, and emphasizes the importance of active, prospective surveillance.

Talia Pindyck, MD, MPH1, Aron J. Hall, DVM, MSPH2, Jacqueline Tate, PhD2, Cristina V. Cardemil, MD, MPH2, Anita Kambhampati, MPH3, Mary E. Wikswo, MPH4, Scott Grytdal, MPH2, Daniel C. Payne, PhD, MSPH4, Umesh D. Parashar, MBBS, MPH2 and New Vaccine Surveillance Network & SUPERNOVA Network, (1)Division of Viral Diseases, CDC, Atlanta, GA, (2)Division of Viral Diseases, Centers for Disease Control and Prevention, Atlanta, GA, (3)IHRC, Inc., contracting agency to the Division of Viral Diseases, Centers for Diseases Control and Prevention, Atlanta, GA, Atlanta, GA, (4)Centers for Disease Control and Prevention, Atlanta, GA

Disclosures:

T. Pindyck, None

A. J. Hall, None

J. Tate, None

C. V. Cardemil, None

A. Kambhampati, None

M. E. Wikswo, None

S. Grytdal, None

D. C. Payne, None

U. D. Parashar, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.