Program Schedule

Assessment of Automated Surveillance Strategies to Identify Infectious Complications Following Implanted Cardiac Device Procedures

Session: Poster Abstract Session: HAI Surveillance and Public Reporting
Friday, October 10, 2014
Room: The Pennsylvania Convention Center: IDExpo Hall BC

Infectious complication rates for implanted cardiac devices (ICD) have varied from <1% to >5%.  The optimum approach for surveillance of these infections is unclear. The objective of this study was to create an automated surveillance tool for infectious complications after ICD procedures using ICD-9 and microbiology data. 


All patients undergoing ICD procedures at Duke University Hospital from 1/1/2005-12/31/2011 were identified using ICD-9 procedure codes.  Potential infection-related complications among the cohort were noted using specific ICD-9 diagnosis codes and microbiology data for the 365 days following the index procedure.  All potential cases identified from microbiology data and a subset identified using ICD-9 codes were reviewed for infection by expert reviewers.  A subset of procedures without associated microbiology data or ICD-9 codes was also reviewed.  Sensitivity, specificity, positive predictive values, and negative predictive values for specific queries were calculated. 


A total of 6,097 patients had 7,137 procedures during the 7-year timeframe.  A total of 1,686 patients with potential infectious complications were identified: 174 met criteria from both ICD-9 and microbiology data; 14 met only microbiology criteria; and 1,498 met only ICD-9 criteria.  We reviewed 558 cases, including all 188 microbiology cases, 250 randomly selected cases with ICD-9 criteria only, and 120 with neither microbiology nor ICD-9 codes.  Of the 558 procedures reviewed, 71 unique infections were identified.  Query test characteristics are shown in Table 1.  Only 10 in 250 reviewed cases with ICD-9 codes but without microbiology were true infectious complications.  


Our surveillance tool using both microbiology and ICD-9 data was sensitive and specific for infectious complications following EP procedures.  Further modifications of our query (e.g., type of ICD-9 code, time to infection) may further improve the performance of our automated surveillance.

Table 1.  Test performance of surveillance strategies  

Query Result




Positive Predictive Value

Negative Predictive Value

Micro + / ICD-9 +






Micro - / ICD-9 +






Micro + / ICD-9 -






Joel C. Boggan, MD, MPH1,2, Arthur W. Baker, MD3, Kristen V. Dicks, MD4, Michael J. Durkin, MD4, Sarah S. Lewis, MD4, Rebekah W. Moehring, MD, MPH4, Luke F. Chen, MBBS, MPH, CIC, FRACP4, Lauren Knelson, MSPH5 and Deverick Anderson, MD, MPH4, (1)Medicine, Durham Veterans Affairs Medical Center, Durham, NC, (2)Department of Medicine, Duke University Medical Center, Durham, NC, (3)Duke University Medical Center, Durham, NC, (4)Division of Infectious Diseases, Duke University Medical Center, Durham, NC, (5)Duke University CDC Prevention Epicenter Program, Durham, NC


J. C. Boggan, None

A. W. Baker, None

K. V. Dicks, None

M. J. Durkin, None

S. S. Lewis, None

R. W. Moehring, None

L. F. Chen, None

L. Knelson, None

D. Anderson, None

Findings in the abstracts are embargoed until 12:01 a.m. EDT, Oct. 8th with the exception of research findings presented at the IDWeek press conferences.

Sponsoring Societies:

© 2014, All Rights Reserved.

Follow IDWeek