1. Point of Care Diagnostics for Hospital Acquired Infections (HAIs): Health Economics Perspective Detection & Identification of Infectious Agents (DIIA) Innovation Platform Professor Ala Szczepura Warwick Medical School, University of Warwick, Coventry, UK Dr Leeza Osipenko Honorary Fellow, Warwick Medical School

2. MRSA UK Context: In 2007 : >23,000 cases Deaths: 460-1,314 Division of Health Sciences, Warwick Medical School, UK

3. MRSA: Patient problem/ Population 1 MRSA accounts for up to ~7,300 Life Years (LYs) lost, a total mortality cost of ~£198 million per annum in England and Wales No. MRSA hospital cases is difficult to estimate (only bacteraemia figures are available) – Falling rates will decrease economic burden National hospital cost data are lacking. –Based on literature value (~ £4,798/case), MRSA cases add ~£114 million p.a. to hospital costs in England Diagnosis/screening, incremental hospital & mortality costs total £381m p.a. Cost-effectiveness evidence is poor for existing tests. Economic impact of other controls (e.g. hygiene) vs. screening programmes not yet evidenced 1. Based on 2007 data Division of Health Sciences, Warwick Medical School, UK

4. MRSA: POC test requirements  Time to result <5 min  Available at point of care  Allows testing using an easily obtainable sample  Differences in the molecular characteristics of emerging strains will necessitate new tests to detect these strains Centre for Evidence in Ethnicity, Health & Diversity (CEEHD) Division of Health Sciences, Warwick Medical School, UK

5. Economic Modelling: Current tests vs. ideal new test Key parameters: Changes in number of false negatives/false positives Increase in number of true positives detected Reduction in number of infections transmitted & associated hospital costs (savings) and mortality costs (major savings). Decreased time to result Sensitivity analysis –Varying assumed transmission rate –Varying national number of diagnostic tests Division of Health Sciences, Warwick Medical School, UK

6. Evidence gaps - MRSA model  Number of tests: National number of hospital tests for suspected MRSA is unknown No data on what % of suspected cases test positive  Transmission: Transmission of MRSA infection in the hospital setting assumed to be between 0.3 and 0.7. Narrower range is unavailable.  Colonised patients: Transmission rate (colonisation) is unknown, so 0.3 is assumed % of colonised patients who go on to develop an infection is unknown (only broad range of 20-60%). Division of Health Sciences, Warwick Medical School, UK

7. Outcomes: Annual saving if POCT is used for all MRSA screening & diagnosis 1. Transmission of infection assumed to fall to ~zero with ‘ideal new test 2. Data missing on rate of colonisation of other patients & likelihood of moving from colonised to infected state 3. Only quantified extra colonised cases detected ; Monetary value could not be placed on these No. patients with suspected Infection (Diagnostic tests) Total Annual Cost (Screening + Diagnosis) Current testsNew POC test (TAT from 48 hrs to 20 mins) 1 Culture screenScreen New POC test (screen & diagnostic) Unit test cost = Culture cost New POC test (screen & diagnostic) Unit test cost = PCR cost 111,250£83.2 m £415.9 m- £1.8 to - £149.8 m+ £182.9 to + £330.0 m 500,000£86.0 m £418.7 m- £296 to - £914 m- £581.0 to + £414.2 m Division of Health Sciences, Warwick Medical School, UK

8. MRSA: Conclusions  The ‘ideal new test’ could be highly money-saving, at least in the first few years of use. Overall savings will depend on the new test’s unit cost and the annual number of diagnostic (non-screening) tests performed  If price of the new test is equal to PCR, then POCT can be money- saving, assuming a higher rate of diagnostic testing (~500,000 p.a.) and a high transmission rate (0.7).  If price of the new test is equal to Culture test, then annual saving will be up to £914 million assuming 500,000 tests p.a.  At lower rates (~111,250 tests p.a.) annual saving will be up to £149.8 m p.a.  Under all scenarios, moving to a new MRSA test for diagnosis will detect additional infections Division of Health Sciences, Warwick Medical School, UK

9. C difficile infection (CDI) UK Context: In 2007: ~ 42,700 cases Deaths: 4,056 Division of Health Sciences, Warwick Medical School, UK

10. CDI: Patient problem/ Population 1 CDI may account for up to ~34,800 Life Years (LYs) lost resulting in a total mortality cost of ~£944 million per annum in England and Wales¹. National hospital cost data are lacking. –Based on literature value (~ £4,300/case) CDI cases add £202 million p.a. to hospital costs, including readmissions Diagnosis/screening, incremental hospital & mortality costs total £1.15 billion p.a. Cost-effectiveness evidence on current tests is not available. – No robust economic evaluations exist on screening/ step-testing strategies to rapidly identify 90% of samples which are negative 1. Based on 2007 data Division of Health Sciences, Warwick Medical School, UK

11. CDI: POC test requirements  Time to result <5 min  Sensitive & specific for C. difficile toxins  Available at point of care  Identifies C. difficile disease, not only carriage  Requires less technical expertise  Easily accessible Centre for Evidence in Ethnicity, Health & Diversity (CEEHD) Division of Health Sciences, Warwick Medical School, UK

12. Evidence gaps – CDI model  Number of tests: National number of hospital tests for suspected CDI is unknown No data on what % of suspected cases test positive  Transmission: Transmission rate of CDI in the hospital setting is unknown, assumed to be between 0.3 and 0.7. Division of Health Sciences, Warwick Medical School, UK

13. Outcomes: Annual cost impact if POCT used for all CDI diagnosis (varying transmission rates) Numbers tested nationally Transmission rate per false negative Incremental Cost (Elisa vs. Culture) Extra cases detected (per annum) Incremental Cost (Elisa vs. Assay0 Extra cases detected (per annum) 427,0000.3+ £ 3.2 m7,686+ £ 1.8 m4,270 0.5- £16.6 m- £ 9.2 m 0.7- £36.4 m- £20.2 m Division of Health Sciences, Warwick Medical School, UK - Reduction in annual cost + Increase in annual cost

14.  An ‘ideal’ new POC test can be money-saving, but only if the infection transmission rate is ≥ 0.33  Moving to an ideal new POC test will detect extra cases:  7,686 p.a. if the test replaces culture &  4,270 p.a. if the test replaces assay. CDI: Conclusions Division of Health Sciences, Warwick Medical School, UK

15. ESBL UK Context: In 2007: ~ 30,000 cases Deaths: 275-550 pa Division of Health Sciences, Warwick Medical School, UK

16. ESBL: Patient problem/ Population ESBL appears to be directly linked to fewer deaths than MRSA & far fewer than CDI. Life Years (LYs) lost cannot be estimated, but paediatric cases will increase LYs lost National hospital cost data are lacking. However, literature suggests hospital cost of ~£8,225 per ESBL case Diagnostic costs & impact on national hospital costs cannot be quantified Only very few & poor quality cost-effectiveness studies exist There is some evidence that mortality & hospital cost burden could be reduced through more rapid and efficient diagnosis Division of Health Sciences, Warwick Medical School, UK

17. ESBL: Technical performance of current tests  Clinical laboratories vary widely in ability to identify ESBLs  Testing is a two step process: - screening for resistance with an indicator cephalosporin - confirmatory test for resistant isolates  Conventional methods (based on agar diffusion) better than automated systems  Isolates may carry more than 1 enzyme  Of the indicator cephalosporins, few have good sensitivity & specificity. Centre for Evidence in Ethnicity, Health & Diversity (CEEHD) Division of Health Sciences, Warwick Medical School, UK

18. ESBL: POC test requirements  Time to result <30 min  Sensitive & specific  Able to identify species & relevant genes for species  Requires low level of technical expertise  Needs to adapt to changing gene adaptation of enzymes Centre for Evidence in Ethnicity, Health & Diversity (CEEHD) Division of Health Sciences, Warwick Medical School, UK

19. Evidence gaps - ESBL model  Number of cases in the UK: Number of ESBL cases diagnosed nationally is not known - ~5,500 cases of ESBL E. coli bacteraemia reported in the UK (excluding Scotland) in 2007. - Total of ~30,000 UK infections suggested due to ESBL-producing E. coli p.a. 2  Number of deaths: Number of deaths linked to ESBL is unknown If mortality is principally linked to blood stream infections, applying a mortality rate of ~5% - 10% to 5,500 ESBL bacteraemia cases as per above would indicate ~275-550 deaths per annum.  Number of tests: National number of hospital tests for ESBL is unknown No data on what % of suspected cases test positive  Transmission: Infection transmission rate is unknown 1. Antibiotic Resistant ESBL-Producing E.coli Strain Spreading in UK. Available from: http://www.medicalnewstoday.com/articles/83453.php http://www.medicalnewstoday.com/articles/83453.php Division of Health Sciences, Warwick Medical School, UK

20. HAIs: Overall Conclusions Significant savings can be achieved by reducing HAIs but… the size of saving will depend on the context Investment in POCT must ensure unit test cost (including staff, capital & consumables) is similar to current lab tests, unless significant clinical benefits and/or cost saving changes to clinical pathway can be demonstrated There is a need to develop more sophisticated micro-economic simulation models to study the impact of new POC test implementation Emerging competitor: non-test technologies should be monitored… for example, improved decolonisation, vaccines etc. Division of Health Sciences, Warwick Medical School, UK