2220. Gut Microbiota and Clinical Predictors of Carbapenem-Resistant Enterobacteriaceae (CRE) Carriage Among Patients at the Time of Admission to a Long-Term Acute Care Hospital (LTACH)
Session: Poster Abstract Session: Microbiome: GI
Saturday, October 29, 2016
Room: Poster Hall

Background: CRE infections have high mortality rates and are difficult to treat due to multidrug resistance. CRE colonization of the gastrointestinal tract precedes infection and contributes to cross-transmission in health care facilities. In this study we investigated the gut microbiota and clinical predictors of CRE colonization at the time of LTACH admission.

Methods: Stool or rectal swab samples were collected from LTACH patients ≤ 4 days of admission. CRE status was determined by cultivation and duplex qPCR of blaKPC/blaNDM or history. The V4 region of the 16S rRNA gene was amplified from sample DNA and sequenced with an Illumina MiSeq. Sequences were processed and analyzed using mothur with operational taxonomic units (OTUs) based on 0.02 level sequence difference. Medical records were reviewed. Construction of prediction models proceeded in 3 phases: 1) variable selection by microarray analysis; 2) graphical examination of data to generate dichotomous dummy values for possible predictors, followed by backwards stepwise logistic regression to identify independent predictors; and 3) construction of a unit-weighted risk scale. The risk scale was tested on a calibration sample then applied to a validation sample to test its robustness.

Results: 158 subjects participated: 71 (33 CRE+) in the calibration sample and 87 (29 CRE+) in the validation sample. The gut microbiota of CRE+ and CRE– patients was not significantly different based on inverse Simpson diversity index (Mann-Whitney test) or analysis of molecular variance (AMOVA) on θYC distances, a metric that takes into account relative abundances of both shared and non-shared OTUs. A unit-weighted risk scale comprised of 6 independent clinical (Charlson score, decubitus ulcer, mental status) and OTU variables (OTU12 [Proteus], OTU47 [Streptococcus], and absence of OTU75 [Clostridium XlVa]) was more strongly associated with CRE colonization status in both the calibration and validation samples than a scale comprised of either clinical or OTU variables alone (Table).  

Conclusion: A unit-weighted risk scale that comprised clinical and gut microbiota factors predicted CRE colonization at the time of LTACH admission better than a risk scale that included clinical or gut microbiota variables alone.

Christine M. Bassis, PhD1, Anna Seekatz, PhD1, Louis Fogg, PhD2, Karen Lolans, BS2, Nicholas M. Moore, MS3, Koh Okamoto, MD4, Yoona Rhee, MD, ScM2, Laura Bardowski, MSN5, Pamela Bell, BS6, Efrain Salazar, BS7, Thelma Dangana, MBBS, MS2, Galina Sidimirova, BSN8, Robert A. Weinstein, MD9, Michael Y. Lin, MD, MPH2, Vincent B. Young, MD, PhD, FIDSA10, Mary K. Hayden, MD, FIDSA, FSHEA11 and For the CDC Prevention Epicenters Program, .12, (1)Internal Medicine/Infectious Diseases, University of Michigan, Ann Arbor, MI, (2)Rush University Medical Center, Chicago, IL, (3)Medical Laboratory Science, Rush University Medical Center, Chicago, IL, (4)University of Tokyo Hospital, Tokyo, Japan, (5)Section of Infectious Disease, Rush University Medical Center, Chicago, IL, (6)Pathology, Rush University Medical Center, Chicago, IL, (7)Rush University, , College of Health Sciences, chicago, IL, (8)Infection Control, Rush Oak Park Hospital, Oak Park, IL, (9)Cook County Health and Hospitals System, Chicago, IL, (10)Internal Medicine-Infectious Diseases, University of Michigan, Ann Arbor, MI, (11)Internal Medicine (Infectious Diseases) and Pathology, Rush University Medical Center, Chicago, IL, (12)CDC, Atlanta, GA


C. M. Bassis, None

A. Seekatz, None

L. Fogg, None

K. Lolans, None

N. M. Moore, None

K. Okamoto, None

Y. Rhee, None

L. Bardowski, None

P. Bell, None

E. Salazar, None

T. Dangana, None

G. Sidimirova, None

R. A. Weinstein, None

M. Y. Lin, None

V. B. Young, None

M. K. Hayden, None

F. T. CDC Prevention Epicenters Program, None

See more of: Microbiome: GI
See more of: Poster Abstract Session

Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.