829 Evaluating the Impact of Mandatory Reporting on Central Line-Associated Bloodstream Infection Rates Reported to the National Healthcare Safety Network

Sunday, March 21, 2010
Grand Hall (Hyatt Regency Atlanta)
Shailendra N. Banerjee, PhD , CDC, Atlanta, GA
Teresa C. Horan, MPH , Centers for Disease Control and Prevention, Atlanta, GA
Daniel A. Pollock, MD , CDC, Atlanta, GA
Scott K. Fridkin, MD , CDC, Atlanta, GA
Jonathan R. Edwards, MStat , CDC, Atlanta, GA
Background: Central line-associated bloodstream infection (CLABSI) is one of the leading causes of morbidity and mortality in the United States.  With many states mandating hospitals to report CLABSI and sharing the data with the public, there is concern that under-reporting may occur.

Objective: To evaluate the impact of mandatory CLABSI reporting on CLABSI rates among NHSN hospitals during 2006-2008.

Methods: CLABSI rates per 1,000 central line-days were calculated for the following intensive care unit (ICU) types: burn, medical cardiac, cardiothoracic, pediatric cardiothoracic, medical (major teaching), medical (all others) [MICU-oth], medical/surgical (major teaching), medical/surgical (all others), pediatric medical/surgical (major teaching), pediatric medical/surgical (all others), neurologic, neurosurgical, surgical and trauma.  Using knowledge of states’ CLABSI mandates during 2006-2008, we stratified the hospitals’ data into 3 reporting periods for the rate calculations.  The voluntary period (V) contained data for hospitals reporting only in months without a mandate and the mandatory period (M) contained data from hospitals that reported only under a mandate (never in V).  The period VM contained hospitals reporting data in both voluntary and mandatory periods.  Total time period of observations ranged from 1-9 months.  Distributions of these rates were compared between V and VM, V and M, VM and M by non-parametric tests.   

Results: Comparing VM and M, the distributions for CLABSI rates were significantly different only for MICU-oth (p = 0.02), the median of distribution being higher for VM (Table).  Similarly, distributions for CLABSI rates between V and VM were not significantly different (p > 0.05) for all the ICU types except for MICU-oth, the median of distribution being higher for VM (p = 0.04) [CLABSI rate for V = 0.88].  No significant difference was found for any ICU type for the comparisons V and M. Conclusions: Only in MICU-oth did CLABSI rates show any effect associated with state mandates for reporting.  Overall, there is little evidence that reporting under a mandate is associated with significant difference in rates of CLABSI being reported to NHSN, although, it may still be too early to detect such impacts.

Difference in CLABSI rates (infections per 1,000 central line-days) between VM and M by Location Type

VM

M

Location Type

Sample Size

Median

Sample Size

Median

Difference

p-value

MICU-oth

39

1.75

63

0.63

1.12

0.02

MICU-teach

45

2.54

25

2.15

0.39

0.09

Med/surg-teach

40

1.62

53

1.70

-0.08

0.91

Med/Surg-oth

221

0.47

423

0.26

0.21

0.83

Neurosurgical

22

1.80

22

1.78

0.02

0.99

Surgical

58

2.76

57

1.74

1.02

0.10

Med Cardiac

71

0.93

77

1.36

-0.43

0.73

Cardiothoracic

69

0.86

66

0.70

0.16

0.99

      VM = Reporting started in voluntary mode then a mandate to report started

       M = Reporting started under a mandate