In July 2009, New York State (NYS) began using the National Healthcare Safety Network (NHSN) LabID Event module to report facility-wide Clostridium difficile (C. diff) at all NYS hospitals as part of the mandatory public reporting law. NYS staff performs annual on-site audits to ensure the accuracy of the data submitted by hospitals.
Objective:
Analyze the accuracy of the overall number of C. diff events reported to the NHSN. Determine effect of data entry errors on the case status (CO – community onset; CO-HCFA – community onset-healthcare facility associated; HO – hospital onset) assigned to these events.
Methods:
179 NYS hospitals entered 2009 facility-wide C. diff LabID Event data into the NHSN. NYS staff audited a sample of data from 93 (52%) of these facilities for accuracy and completeness.
Results:
(Table 1) In the 3365 charts that were examined, reviewers indentified an additional 235 infections that should have been reported to the NHSN, an underreporting of 7.0% in the sample. Over reporting was indentified in 63 (1.9%) events. Discrepancies were identified in 259 (8.6%) specimen dates, 99 (3.3%) admission dates, and 213 (8.2%) last discharge dates.
Table 1: C. diff Event – Data Entry Discrepancies |
|||
|
# of Differences |
# of Events Reviewed |
% Difference |
Specimen Date |
259 |
3026 |
8.6 |
Admission Date |
99 |
3008 |
3.3 |
Last Discharge Date |
213 |
2609 |
8.2 |
|
|
|
|
Underreported |
235 |
3365 |
7.0 |
Over reported |
63 |
3365 |
1.9 |
(Table 2) The overall case status match was 96.9% (2991/3088). An additional 50 (1.7%) events, previously classified as CO, were changed to CO-HCFA or HO events after audit. Conversely, 19 (0.6%) events, previously classified as CO-HCFA or HO, were changed to CO events after audit.
Table 2: C. diff Event – Case Status Match |
|||
Hospital |
Reviewer |
||
|
CO |
CO-HCFA |
HO |
CO |
799 (25.9%) |
24 |
26 |
CO-HCFA |
6 |
482 (15.6%) |
15 |
HO |
13 |
13 |
1710 (55.4%) |
Conclusions:
The audits revealed a small amount of underreporting caused mainly by misunderstanding of the reporting requirements, or miscommunication between the laboratory and the Infection Prevention staff.
Despite all C. diff LabID Event data being manually entered into the NHSN, NYS hospitals are very accurate with data entry. There were a total of 571 (6.6%) date discrepancies identified that may have affected case status assignment in NHSN; this resulted in 97 (3.1%) changes in case status among the events reviewed.
Moving forward, increased data accuracy could be accomplished by allowing facilities to import their data; electronic surveillance may also eliminate most of the underreporting and will be important as mandatory reporting requirements continue to increase.