"ESCMID SHEA Training Course in Hospital Epidemiology Surveillance State"
ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Overview Important issues SURVEILLANCE Computer screening State of the Art Computer algorithms Loreen A. Herwaldt, MD Process Surveillance Process & incidence surveillance + intervention Designing A Surveillance System Focus on events that Can be prevented Occur frequently Cause serious morbidity Increase mortality Increase length of stay Are costly to treat Components of Surveillance Components of Surveillance Definitions Definitions Written Case finding and data collection Applied consistently Data entry or tabulation Imprecise definitions can lead to Data analysis incorrect conclusions Data interpretation Data reporting 1 ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Surveillance Case-finding Methods Computerized Screening at Computerized Screening the UIHC Advantages Positive cultures Automatic Positive laboratory findings Use data in available databases CSF + gram stain, WBC > 5, protein > 45, glucose < 40 Provides data with little effort after C. difficile toxin test programming is completed RSV antigen test Disadvantages Rotavirus Accuracy of the data in other data bases Combination of diagnostic tests cannot be assumed CXR and respiratory culture/24 hrs Necessary data may not be available in Cultures from > 2 body sites/24 hrs computer databases Computerized Screening for SSI at the UIHC Obtain list of all patients who had an operation T - 35 days Patients selected for screening if: Wound cultures obtained > 2 days post op and/or Antibiotics begun > 4 days postop or antibiotics continued > 4 days post op ICD9 code for post-op infection ICP Time UIHC Computerized Surveillance Total: 20.4 Adult in-patient units 7 hrs/week hrs/week BSI C difficile infections Three ICPs: SSI 6.4 hrs/week 6.8 hrs/week/ICP Women & Children’s 7 hrs/week (17%) Service (estimated) APIC Staffing Research Project 14 hrs/week/ICP estimated time for (35%) surveillance/epidemiologic investigations 2 ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Epidemiology of SSI Surveillance Algorithm for Occurring after Hospital Discharge Identifying SSI Sensitivity (%) Positive Predictive Results Value (%) Proce- Routine Abx Dx Abx Abx Dx Abx 111 SSI occurred dure code + Dx code + Dx after discharge code code 70 SSI were CABG 59 91 54 93 36 84 36 diagnosed and treated entirely outside of the C-sect 38 84 78 97 37 67 38 hospital In-pt Post-discharge Breast 33 94 70 96 33 79 33 Sands, JID, 1996 DS Yokoe, et al EID 2004:101924-30 Percent of Patients Meeting Sensitivity & PVP of Criteria for Criteria Identifying SSI after Hip Replacement Procedure Abx Diagnosis Both Sensitivity (%) PVP (%) exposure code Routine surv 80 100 CABG 19.1 4.6 19.6 > 48 h abx 50 1.21 > 7 d abx 15 6.76 C-section 12.7 6.4 14.1 Dc dx code 5 10.29 Readmitted 95 5.98 Breast 6.7 2.0 6.8 Abx readmiss 95 7.8 Dc dx code 80 64.78 DS Yokoe, et al EID 2004:101924-30 M Bolen, et al. IDSA abstract 1019; 2006, Toronto Sensitivity & PVP of Criteria for Identifying SSI after Hip Replacement Sensitivity (%) PVP (%) Dc dx code 85 49.43 >7 d abx or 100 5.87 readmitted >7 d abx or 100 7.33 abx w/ readmis >7 d abx or dc 90 27.31 dx code M Bolen, et al. IDSA abstract 1019; 2006, Toronto 3 ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Data Mining (MedMined) Process Surveillance Sensitivity Specificity Time Monitoring practices that directly or HW DM .86 (.76- .98 (.98- 2 h/10000 indirectly contribute to a health outcome .96) .99) Using data to improve process and HW MRR .92 (.84-1) .99 (.98-1) 1.5 possibly the outcome * FTE/10000 Adjunct to surveillance for nosocomial ICU DM 1.0 .99 -- infection ICU MRR 1.0 1.0* -- Use as a surrogate for outcome ICU NNIS Surv .61* .94* -- assessment ICU NNIS Retro .94 .68* -- SE Brossette, et al. Am J Clin Pathol 2006;125:34-9 Examples of Process Process Surveillance Surveillance The proportion of: Process should be closely Perioperative antibiotic doses given in associated with the designated the 1 hour period before the incision outcome Staff who follow UTI prevention Urinary catheter associated with protocol urinary tract infection HCW who receive influenza vaccine Central line associated with bloodstream infection 4 ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Process Surveillance Overall Hand Hygiene Compliance Percentage over the iNICQ Project UTI 100% 90% 74% 85% 82% 80% 64% 70% Number of cases/10,000 Percent compliance with UTI 59% 60% patient days prevention protocol 50% 40% 20 100 30% 20% 80 10% 15 0% Baseline Jun-05 Jul-Sep 2005 Oct-Dec 2005 Jan-Mar 2006 60 Compliance 10 40 UTI Vermont Oxford Network iNICQ project: Improve 5 Hand Hygiene 20 Initiate NICU hand hygiene education 0 0 Monitor hand hygiene by direct observation 1994 1995 1996 1997 1998 Present themed project with posters and props Continue improvement above baseline Year Planned re-education to maintain gains and improve outcome Improved Hand Hygiene with Hand Hygiene Program: Nosocomial Hospital-wide Program Infection & MRSA Transmission Rates Educational campaign Observed hand ↓ nosocomial hygiene compliance infection rate from Introduced bedside 16.9% → 9.9% portable dispensers (p=0.04). of alcohol-based Reduced new MRSA hand disinfectant in acquisition, MRSA 1995. infection rates, & Improved compliance nosocomial MRSA w/ hand disinfection. BSI. Overall, 48% → 66% (p<0.001). Pittet D, et al. Lancet 2000;356:1307-1312. Pittet D, et al. Lancet 2000;356:1307-1312. Central Venous Catheter-Related Bloodstream Infection Rates with NNIS Benchmarks and Interventions IHI Central Line Bundle Elements 25 M ICU Hand Hygiene Education Chloroprep 20 Coated Cath Rate CVCBSI per 1000 CVC-Days Hand hygiene 15 Maximal sterile barrier precautions Chlorhexidine skin antisepsis 10 Optimal catheter site selection, with subclavian vein as the preferred site 5 for non-tunneled catheters in adults 0 Daily assessment of line necessity; May-02 May-03 May-04 May-05 Jul-01 Aug-01 Sep-01 Oct-01 Nov-01 Dec-01 Jan-02 Feb-02 Mar-02 Apr-02 Jun-02 Jul-02 Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Jan-03 Feb-03 Mar-03 Apr-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03 Jan-04 Feb-04 Mar-04 Apr-04 Jun-04 Jul-04 Aug-04 Sep-04 Oct-04 Nov-04 Dec-04 Jan-05 Feb-05 Mar-05 Apr-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05 Jan-06 Feb-06 Mar-06 Apr-06 promptly remove unnecessary lines Rate Mean 10%ile 25%ile Median 75%ile 90%ile 5 ESCMID/SHEA Training Course in Hospital Epidemiology Surveillance: State of the Art Loreen A. Herwaldt, MD Percent of Months with No CVC- Days Between CVC-BSI: Adult BSI ICUs 70 70 60 60 50 50 40 SICU 40 SICU 30 MICU 30 MICU CVICU CVICU 20 20 10 10 0 0 Before After Before After The Median CVC-BSI rate decreased in each adult ICU Estimated Cost Savings from CVC-BSI Intervention CVICU Before: 20 BSI/21 mns = 0.95/mn After: 24 BSI/39 mns = 0.62/mn 37 (expect) – 24 (observe) = 13 (prevent) 13 x $28,000 = $365,400 MICU 92 (expect) – 42 (observe) = 50 (prevent) 50 x $28,000 = $1,400,000 SICU 84 (expect) – 63 (observe) = 21 (prevent) 21 x $28,000 = $588,000 Total: ~84 CVC-BSI prevented; saved = $3,164,000 6