New York State Department of Health Mandatory Reporting

New York State Department of Health: Mandatory Reporting of Clostridium difficile via National Healthcare Safety Network Lab. ID 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, Kuang. Nan Xiong, BS, Valerie Haley, MS, Rachel Stricof, MPH, CIC, Carole Van Antwerpen, RN, BSN, CIC Abstract Tables & Results Original abstract not included. Updated data presented here. Table 2 Highlights: Table 1: C. diff Event – Case Status Match 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 (Lab. ID Events). In July 2009, New York State (NYS) mandated reporting of facility-wide inpatient C. diff Lab. ID Events in acute care hospitals. Thirty NYS facilities were also collecting C. diff infection data as part of a prevention collaborative. Infection Surveillance NHSN Lab. ID Event Hospital Onset Facility Associated 1202 (40. 7%) Possibly Facility Associated 138 Not Facility Associated Recurrent Objective Compare Lab. ID case status with infection surveillance definitions requiring chart Results continued review 1. Methods NYS research staff conducted on-site audits, reviewed laboratory data and performed medical record reviews in the 30 facilities reporting both Lab. ID events and clinical infection surveillance data. Lab. ID 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 Mc. Donald et. al. 1 The following definitions were used for this evaluation: NHSN Lab. ID 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 245 42 CO-Healthcare Facility Associated 32 243 (8. 2%) 105 23 Community Onset N = 2953 Recurrent Overall Match 18 16 51 5 667 (22. 6%) 23 28 115 (3. 9%) Overall Match 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 Lab. ID Events [Range 70. 0 – 100. 0% by facility; Median = 95. 7%] ● 73. 9% (1202/1627) of NHSN Hospital Onset Lab. ID 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%) Table 2: C. diff Event – Modified Case Status Match Facility Associated or Possibly Facility Associated Not Facility Associated Recurrent Overall Match NHSN Lab. ID Event N = 2953 Hospital Onset or CO-Healthcare Facility Associated Community Onset Recurrent 1615 (54. 7%) 69 21 350 65 ● 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: 2227 (75. 4%) Infection Surveillance Acknowledgement: GNYHA & UHF C. difficile Prevention Collaborative and participating hospitals Overall Match 667 (22. 6%) 23 28 115 (3. 9%) 2397 (81. 2%) ● 3221 Lab. ID Events reported; 3067 infection events reported; 2953 reported to both ● Of 268 Lab. ID 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 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 Lab. ID 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 Lab. ID data. Ongoing auditing and validation of the data will occur. Reference 1 Mc. Donald LC, Coignard B, Dubberke E, et al. Recommendations for Surveillance of Clostridium difficile-Associated Disease. Infect Control Hosp Epidemiol 2007; 28: 140 -145. Note: The 2007 -2009 New York State Hospital-Acquired Infection reports can be found at: www. nyhealth. gov/nysdoh/hospital/reports/hospital_acquired_infections. The 2010 Report is expected to be released September 2011. No financial disclosures.