Measuring the composite incidence of ICU-acquired multidrug resistant organisms (MDROs) and Clostridium difficile electronically: adapting surveillance algorithms from the NHSN MDRO and C difficile modules

Background: Validated electronic measures of aggregate hospital antimicrobial resistance are needed to evaluate the impact of antimicrobial stewardship interventions. Data from surveillance algorithms in CDC’s National Healthcare Safety Network’s (NHSN) MDRO and Clostridium difficile infection (CDI) modules may be adaptable for this purpose.

Objective: 1. Create a composite measure of ICU-acquired MDRO and CDI incidence using data generated by applying NHSN’s surveillance algorithms to electronic microbiological and administrative data in an urban teaching hospital.  2. Compare incidence rates derived thereby with rates derived via manual chart review, which allows for confirmation of patient location at the putative time of MDRO or C difficile acquisition.

Methods: NHSN’s MDRO and CDI module surveillance algorithms were applied to all inpatients during April 1-September 30, 2010 to identify non-duplicate C difficile toxin assays and non-duplicate, non-surveillance clinical isolates of MRSA, VRE, multidrug-resistant Acinetobacter and ceftazidime-resistant Klebsiella pneumoniae (specified by NHSN’s MDRO module); and imipenem-resistant Pseudomonas aeruginosa and ceftazidime-resistant E coli and Enterobacter sp (not included in the MDRO module).  Composite incidence rates for our 8 ICUs were created by summing the incidence of these pathogens (per 1000 patient-days).  Rates obtained when ICU acquisition was defined by ICU residence on the day of 1^st specimen collection (NHSN rule) and by patient location 3 calendar days before the day of 1^st specimen collection (local rule) were compared.  

Results: There were 750 numerator events (694 non-duplicate MDRO and 56 CDI) over 6 months; 579 were community- and 171 were hospital-acquired. By NHSN definitions, 101 were ICU-acquired, yielding a composite incidence of 7 (range: 0.9-17.9) per 1000 patient days for all ICUs. Rates did not differ significantly when ICU acquisition was defined by the NHSN versus local rule (Table).  

Conclusions: A composite measure of MDRO and CDI incidence can feasibly be derived by adapting NHSN’s MDRO and CDI module surveillance algorithms, though the clinical import of this measure needs to be defined.  ICU-specific rates obtained by using the NHSN rule for ICU acquisition differed minimally from those derived by defining location of pathogen acquisition through more rigorous manual record review.

Comparison of Aggregate Rates in ICUs [events/1000 patient days]
ICU type	Local Rule	NHSN rule	Abs Diff	Proportional difference %	p value
Medical	17.2	17.9	-0.7	-3.9%	0.84
Surgical	9.4	10.2	-0.8	-7.8%	0.85
Burns	13.6	13.6	0	0	1.0
Trauma	13.9	13.2	0.7	5.3%	0.88
Neurosurgical	2.3	2.3	0	0	*
Pediatric	1.2	2.4	-1.2	-50%	*
Neonatal	0.2	0.2	0	0	*
CCU	0	0.9	-0.9	-100%	*
Total	6.7	7	-0.3	-4.3%	0.78

*=not calculated