Variance in Clostridium difficile surveillance reporting: implications for quality improvement outcomes DAVESON KL 1,2, WILSON C 1, MENZIES A 3. 1 DEPARTMENT OF INFECTIOUS DISEASES, CANBERRA HOSPITAL AND HEALTH SERVICES, CANBERRA, ACT, 2 AUSTRALIAN NATIONAL UNIVERSITY, CANBERRA, ACT, 3 INFECTION PREVENTION AND CONTROL UNIT, CANBERRA HOSPITAL AND HEALTH SERVICES, CANBERRA, ACT
Clinical Context - Clostridium difficile infection (CDI) Important cause of health-care associated (HCA) diarrhoea
Important outcome indicator of infection prevention and control (IPC) programs antimicrobial stewardship (AMS) program Attractive benchmarking tool locally and nationally
Clostridium difficile infection (CDI) drivers http://www.slideshare.net/carlaston1/is-clostridium-difficile-still-a-problem-in-uk-hospitals. CDC 2012
Australian Commission on Safety and Quality in Health Care Implementation Guide for Surveillance of Clostridium difficile Infection 2013
Extended Surveillance?
Hospital Identified Hospitalidentified Other (HIO) ED + OPD Hospital Identified Community associated
Surveillance Options - Required Hospital-Identified (HI) intended to apply to hospitals in jurisdictions across Australia as an indicator of the hospital burden of CDI disease Advantages Easier to collect Easier for administration to understand Independent of clinician review Disadvantages Little granularity in expected drivers?usefulness for other QA programs Adds confusion if performing extended surveillance
Surveillance Options Extended Surveillance Additional to hospital-identified should be considered by hospitals, organisations and jurisdictions according to priority, local risk assessment and the capacity to categorise cases correctly Advantages Able to determine change in CDI drivers over time Able to be used by varying QA programs to inform quality improvement interventions Disadvantages More complex definitions than HI-CDI More time consuming for surveillance programs Easier to make cross institutional comparisons (CAVEAT) Consider High burden of disease Complex drivers present Other e.g targeted surveillance
CDI Control & Antimicrobial Stewardship (AMS) Majority of Interventions were restrictive effects in certain subsets of patients Faezel et al. JAC 2014; 69: 1749.
How should we measure CDI to show AMS-based quality improvements?
Methods Case ascertainment: prospectively, collected CDI surveillance database Definitions: cases classed by HI and extended surveillance definitions Time period: 2012-2014 Sensitivity Analysis: changes in HCA-CDI and HI-CDI rates based on a hypothetical quality improvement program that reduced HCA-CDI rates by 50%.
Results Difference in Case Definition HI-CDI:HCA-CDI 3:2 HI-CDI (n) HCA- HCFO (n) HCA-CO (n) CA (n) Indeter (n) HIO (n) 2012 210 134 (64%) 4 (2%) 68 (32%) 1 (<1%) 3 (1%) 2013 160 99(62%) 3 (2%) 54 (34%) 2 (1%) 2 (1%) 2014 189 118 (62%) 11 (6%) 48 (25%) 4 (2%) 8 (4%)
210 (138,68) 160 (102,54) 189 (129,48) Total(HCA,CA) 3->10%
AMS Impact Depends on type of intervention Community versus hospital based Hospital based interventions likely to make the most impact on HCA-HCFO Sensitivity Analysis
Cases Sensitivity Analysis 100 80 60 40 20 0 Rate Comparison Before and After Hypothetical Intervention 20% 40% 60% 80% HCA:HI-CDI HICDI Rate Before Intervention HCA Rate Before Intervention HICDI Rate After Intervention HCA Rate After Intervention
Reported QI Reductions Reported Objective Improvements 60% 50% 40% 30% 20% 10% 0% 20% 40% 60% 80% HCA:HICDI Incidence HICDI Difference HCA Difference
20 18 16 14 12 10 8 6 4 2 0 Case Overestimation HICDI:HCA Public group C-D??? Principal Referral 10% 20% 30% 40% 50% 60% 70% 80% HCA-HICDI Rate
How do we progress the choice together: institution/jurisdiction/nationally? Why is important?
One Health do we care? Differences important for surveillance from a public health point of view We (will) need to assess and address community and hospital drivers Every program needs to consistently demonstrate its utility