Abstract

1
New York State Department of Health Mandatory
Reporting of Clostridium difficile via National
Healthcare Safety Network LabID Event
Evaluating the Validity of the Proxy Measure vs.
Infection Surveillance
New York State Dept. of Health, Hospital Acquired
Infection Reporting
Kathleen A. Gase, MPH, CIC, Johanna B. Lee, MPH,
MA, KuangNan Xiong, BS, Valerie Haley, MS, Rachel
Stricof, MPH, CIC, Carole Van Antwerpen, RN, BSN,
CIC
Abstract
Acknowledgement GNYHA UHF C. difficile
Prevention Collaborative and participating
hospitals
Tables Results
Results continued
Original abstract not included. Updated data
presented here.
Table 2 Highlights ? Overall case status match
improves to 81.2 (2397/2953) when HO/CO-HCFA and
FA/PFA categories are combined Range 54.2
93.6 by facility Median 79.6 ? The NHSN
overestimates the facility incidence (HO
CO-HCFA combined) by 20.4 Range 6.5 45.8 by
facility Median 20.4 Additional
Results ? 3221 LabID Events reported 3067
infection events reported 2953 reported to both
? Of 268 LabID Events reported without
matching infection reported, 156 (58.2) were
reviewed ? 96/152 (61.5) records
reviewed revealed documentation to meet clinical
definition, increasing the sensitivity
of the NHSN criteria. Range 0 30 records
by facility Median 1.5 records ? 52
(54.2) of the unreported cases came from 3 (10)
facilities
Table 1 C. diff Event Case Status Match Table 1 C. diff Event Case Status Match Table 1 C. diff Event Case Status Match Table 1 C. diff Event Case Status Match Table 1 C. diff Event Case Status Match Table 1 C. diff Event Case Status Match
Infection Surveillance NHSN LabID Event NHSN LabID Event NHSN LabID Event NHSN LabID Event N 2953
Hospital Onset CO-Healthcare Facility Associated Community Onset Recurrent Overall Match
Facility Associated 1202 (40.7) 32 18 16
Possibly Facility Associated 138 243 (8.2) 51 5
Not Facility Associated 245 105 667 (22.6) 23
Recurrent 42 23 28 115 (3.9)
Overall Match 2227 (75.4)
Background
In 2009, the Center for Disease Control (CDC)
introduced a module in the National Healthcare
Safety Network (NHSN) allowing reporting of
Clostridium difficile (C. diff) Laboratory
Identified Events (LabID Events). In July 2009,
New York State (NYS) mandated reporting of
facility-wide inpatient C. diff LabID Events in
acute care hospitals. Thirty NYS facilities were
also collecting C. diff infection data as part of
a prevention collaborative.
Objective
Compare LabID case status with infection
surveillance definitions requiring chart review1.
Methods
Table 1 Highlights ? Overall case status match
was 75.4 (2227/2953) Range 51.7 92.2 by
facility Median 74.3 ? 94.8 (1202/1268) of
Facility Associated cases using infection
surveillance criteria were classified as
NHSN Hospital Onset LabID Events Range 70.0
100.0 by facility Median 95.7 ? 73.9
(1202/1627) of NHSN Hospital Onset LabID Events
were classified as Facility Associated cases
using infection surveillance criteria Range 40.0
91.5 by facility Median 78.0 ? Of the
726 total mismatches ? 245 NHSN Hospital
Onset cases were classified as Not Facility
Associated using infection surveillance
criteria, accounting for the largest proportion
of the discrepancies (33.7) Range 0.0 58.3
by facility Median 32.5 ? 105
NHSN CO-HCFA cases were classified as Not
Facility Associated using infection surveillance
criteria (14.5 of discrepancies)
Range 0.0 37.3 by facility Median 10.0
? Combining these two categories (HO/NFA
CO-HCFA/NFA) accounted for 48.2 of mismatches
Range 0.0 79.1 by facility
Median 43.6 ? 180 of 350 (51.4)
HO/NFA CO-HCFA/NFA mismatches are from 6
facilities (20)
NYS research staff conducted on-site audits,
reviewed laboratory data and performed medical
record reviews in the 30 facilities reporting
both LabID events and clinical infection
surveillance data. LabID case status (CO,
CO-HCFA, HO) was assigned by the NHSN and
infection surveillance case status (NFA, PFA, FA)
was assigned by the facilities using modified
clinical case definitions based on
recommendations by McDonald et.al.1 The
following definitions were used for this
evaluation NHSN LabID Event Case Status
CO Community Onset (specimen collected
day 3 of admission and patient was
not discharged from facility within 4 weeks)
CO-HCFA Community Onset-Healthcare
Facility Associated (specimen collected
day 3 of admission and patient was discharged
from facility within 4 weeks) HO
Hospital Onset (specimen collected day 4 of
admission) Recurrent (specimen
collected 2-8 weeks after the last positive
specimen) Infection Surveillance Case
Status NFA Not Facility Associated
(symptom onset day 2 of admission and patient
was not discharged from facility
within previous 4 weeks) PFA Possibly
Facility Associated (symptom onset day 2 of
admission and patient was
discharged from facility within previous 4
weeks) FA Facility Associated
(symptom onset day 3 of admission or within 48
hours after discharge)
Recurrent (specimen collected 2-8 weeks after the
last positive specimen) Prior to the on-site
visit, the two datasets were matched by patient
identifiers. During the on-site visit, NYS staff
verified the data entered into the NHSN. There
was no validation of the hospital assigned case
status data for events that were already matched
this data was being collected by Infection
Preventionists at each facility and the resources
were not available to complete chart reviews for
all of the charts involved.
Conclusions
This preliminary analysis shows wide variation in
results between facilities. Discrepancies may be
due to intensity of surveillance, level of
documentation in medical records, variation in
interpretation of definitions by different
reviewers, or data entry errors. Additional
visits will be made to both very low and very
high match facilities to help understand the
differences. Additional information will be
collected on the 245 HO cases that were reported
as NFA to identify what, if any, additional
information the facility had available in order
to make that classification. The 96 charts that
were identified as incorrectly excluded from the
infection surveillance dataset will also be
investigated further. Preliminary evaluation
indicates that a few outlier facilities account
for the majority of the discrepancies and
mismatches. The LabID Event protocol simplifies
C. difficile surveillance by eliminating the need
for clinical case chart review. For the purpose
of the 2010 public report, NYSDOH will use the
hospital-specific facility-wide NHSN LabID data.
Ongoing auditing and validation of the data will
occur.
Table 2 C. diff Event Modified Case Status Match Table 2 C. diff Event Modified Case Status Match Table 2 C. diff Event Modified Case Status Match Table 2 C. diff Event Modified Case Status Match Table 2 C. diff Event Modified Case Status Match
Infection Surveillance NHSN LabID Event NHSN LabID Event NHSN LabID Event N 2953
Hospital Onset or CO-Healthcare Facility Associated Community Onset Recurrent Overall Match
Facility Associated or Possibly Facility Associated 1615 (54.7) 69 21
Not Facility Associated 350 667 (22.6) 23
Recurrent 65 28 115 (3.9)
Overall Match 2397 (81.2)
Reference
1McDonald LC, Coignard B, Dubberke E, et al.
Recommendations for Surveillance of Clostridium
difficile-Associated Disease. Infect Control Hosp
Epidemiol 2007 28140-145.
Note The 2007-2009 New York State
Hospital-Acquired Infection reports can be found
at www.nyhealth.gov/nysdoh/hospital/reports/hospi
tal_acquired_infections. The 2010 Report is
expected to be released September 2011.
No financial disclosures.