Antibiotic Review Kit - A New Approach to Antimicrobial Stewardship

1. Antibiotic Review Kit - Hospital (ARK-hospital) Grand Round, RSCH March 22nd 2017

2. The ARK-Hospital Team Nationally Locally The antimicrobial stewardship team Sally Curtis Sam Lippett Vikesh Gudka Martin Llewelyn Jasmin Islam Lizzie Cross Will Hamilton Catherine Sargent Dan Agranoff Nelson Barbon

3. ARK-Hospital A new approach to antibiotic prescribing for medical patients Improve antibiotic prescription ‘review and revise’ Stop antibiotics in patients who don’t need them To reduce use across the hospital To protect patients from antibiotic resistance Achieve quality targets

4. Antimicrobial resistance…. The end of modern medicine…? Defined modern medicine as the era of Germ Theory Thomas Wooton in ‘Bad Medicine – Doctors doing harm since Hippocrates’ “ when medicine stopped killing more people than it cured” Joseph Lister - 1865

5. Germ theory, antisepsis and antibiotics have given us: AMR certainly a blow to modern medicine • Successful surgery • Cancer chemotherapy • Immunotherapies Antimicrobial resistance by 2050 threatens2: Antibiotic resistance causes1: • A global population reduction of 11-444 million people • A reduction in the world economy of 0.1-3.1% 2. Taylor J, et al. Estimating the economic costs of antimicrobial resistance: model and results. 2015. http://www.rand.org/ randeurope/research/projects/antimicrobial-resistance-costs.html • WHO Antimicrobial Resistance Global Report on Surveillance 2014

6. Outline • Antimicrobial resistance and antimicrobial use • Where we are in NHS Hospitals. • Where should we / could we get to? • How ARK-Hospital will help us get there

7. Prof Marc Bonten, Utrecht; Twitter, March 7 2017 ‘The English Miracle’1 Introduction of targets Financial penalties • Substantial changes in antibiotic use • Cephalosporins and Quinolones  • β-lactam/β-lactamase inhibitors  Cooke J, et al. Longitudinal trends and cross-sectional analysis of English national hospital antibacterial use over 5 years (2008–13): working towards hospital prescribing quality measures JAC 2015

8. The challenge of antibiotic resistant GNRs in the UK • >3x increase in E . coli bacteraemia since 2004 • Currently 36,000 cases / year Rates of resistance to key agents… Gentamicin Ciprofloxacin Co-amoxiclav Piperacillin – tazobactam ….climbing alarmingly 46% of E. coli bacteraemia isolates now reported as co-amoxiclav resistant English surveillance programme for antimicrobial utilisation and resistance (ESPAUR) report 2015

9. Local data confirm this applies to us Some key rates of antibiotic resistance at this trust in 2016 • If you don’t believe predictions of a world without antibiotics by 2050 • Imagine working in your hospital without these antibiotics by 2020 At what rate of resistance would these drugs cease to be reliable clinically selected treatments?

10. So what is antimicrobial stewardship? • Term coined by Dale Gerding in 1996 who hoped that: “…the long-term effects of antimicrobial selection, dosage, and duration of treatment on resistance development should be a part of every antimicrobial treatment decision”1 • According to the Society for Healthcare Epidemiology of America ”a set of coordinated strategies to improve the use of antimicrobial medications with the goal to enhance patient health outcomes, reduce antibiotic resistance, and decrease unnecessary costs” 1. McGowan, JE Jr,; Gerding D. "Does antibiotic restriction prevent resistance?" New Horizon. 4 (3): 370–6.

11. A time-line of stewardship in the UK £ £ £

12. Hospital antibiotic use in England Total and broad-spectrum antibiotic prescribing NHS England 2010 - 2014 Consumption of systemic antibacterials in the hospital sector in Europe 2014 UK https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/477962/ESPAUR_Report_2015.pdf(accessed August 2016) http://ecdc.europa.eu/en/healthtopics/antimicrobial_resistance/esac-net-database/Pages/Antimicrobial-consumption-rates-by-country.aspx (accessed August 2016) Courtesy of Dr Kieran Hand

13. Antibiotic Use at BSUH This trust 8% more than average for acute trusts 12.5% increase since 2013-14 Change in antibiotic use 2015-16 vs 2013-14 Amount of antibiotic use 2015-16 (DDD/ Occupied Bed Day

14. The five 2016-17 AMR CQUIN goals • Submission of consumption data for years: 2014/15 to 2016/17 • 90% (by Q4) of antibiotic prescriptions reviewed within 72 hours • Reduction of 1% or more in total antibiotic consumption • Reduction of 1% or more in carbapenem use • Reduction of 1% or more in piperacillin-tazobactam use     

15. How much could antibiotic use be cut? • “By 2020, significant outcomes will include: • Reduction of inappropriate antibiotic use • By 50% in outpatient settings • By 20% in inpatient settings” “we will cut inappropriate prescribing in the UK by half by 2020…” https://www.cdc.gov/features/antibioticuse/

16. Acute medical patients under an infection physician received less antibiotic treatment; 173 vs 282 DOT/100 admissions (~39% less) Allowing for case mix ID managed patients were • Less likely to receive an antibiotic (OR=0.25 (95% CI 0.07 to 0.84),p=0.03) • More likely to receive a shorter course (RR=0.71 (95% CI 0.54 to 0.93), p=0.01) With no differences in treatment failure or mortality But longer hospital stay 2 (2-6) vs 4 (3-6) days

17. How can we reduce antibiotic overuse • An everyday scene in primary care… frugalnurse.com

18. Patients presenting to their GP with symptoms of RTI judged not to need immediate antibiotics • Randomised to • Prescription • post-dated prescription • Reconsultation • Delayed prescription • No-prescription

19. Patients presenting to their GP with symptoms of RTI judged not to need immediate antibiotics • Randomised to • Prescription • post-dated prescription • Reconsultation • Delayed prescription • No-prescription • 40% reduced antibiotic use • No detectable harm • Reduced ‘belief’ in antibiotics

20. But in hospital patients are at greater risk from infection Especially need for prompt recognition and treatment of sepsis Give IV antibiotics in the first hour according to trust policy The Surviving Sepsis Campaign Guidelines 2012: Update for Emergency Physicians. Jones AE and Puskarich MA Annals of Emergency Medicine 2014.

22. In reality ’focus’ rarely means stop >95% of ‘Review and Revise’ Decisions are to continue antibiotic treatment With thanks to Dr Kieran Hand, Southampton

23. Start ‘smart’ without ‘focus’ my actually increase antibiotic overuse

24. A typical risk – benefit balance for stopping antibiotics Continue unless I can justify stopping =

25. What’s the evidence that less antibiotic exposure reduces risk of resistance?

26. Antibiotic consumption correlates with antibiotic resistance Log odds of penicillin resistance in Strep. pneumoniae Defined Daily Doses of beta-lactam antibiotics / 1000 population Bronzwae S et al 2002 Emerg Infect Dis

27. Antibiotics also place individual patients at risk of resistant infection

28. Antibiotics also place individual patients at risk of resistant infection • Explored relationship between prior antibiotic use and antibiotic resistance • Reviewed 24 studies • 19 Observational • 5 RCTs • >27,000 participants • Urinary Infections • Respiratory Tract infections

29. Forest plots of included studies Substantial Lasting Impact on risk of resistant infection Respiratory Tract Infection Urinary Infection

30. Giving less antibiotics reduces this risk • Chastre J et al JAMA 2003 • 401 patients with VAP on 51 French ICUs • Randomized to 8 vs 15 days antibiotic treatment • Short course patients had • equivalent outcomes on every safety measure • half as much antibiotic exposure • Lower risk of resistant re-infection • Singh N et al Am J RespirCrit Care Med 2000 • 81 patients treated for suspected VAP randomised to Review and Revise at 3 days or standard course • Review and Revise group had less antibiotics and lower risk of resistance / super-infection (15 vs 35% p = 0.017).

31. Giving less antibiotic probably improves clinical outcome Effect on mortality of antibiotic de-escalation Overall 56% (95% CI 34-70%) REDUCED risk of mortality with antibiotic de-escalation strategies

32. Do we even know how long a course should be? There is a lack of evidence that recommended durations are superior to antibiotic-sparing approaches* *w/exception of otitis media, Hoberman A et al New Eng J Med. 2016;375:2446-2456 Recommended course durations have fallen

33. Prescriber behaviour ‘etiquette’ Identified team behaviours resulting in poor antibiotic use • Seniors dominate the decision making and regard themselves as autonomous • There is a culture of non-interference among juniors and colleagues • Unwritten rules of etiquette prevent prescribers reassessing others decisions

34. What is ARK-hospital? A 5-year applied research programme funded by NIHR The overarching aim of ARK is to reduce the incidence of serious infections caused by antibiotic-resistant bacteria in the future, through substantially and safely reducing antibiotic use in hospitals now • Underlying hypotheses: • In hospitals, most antibiotics are started appropriately; but there is reluctance to stop them once started • Short durations of antibiotic treatment are sufficient to treat most genuine bacterial infections in hospitals • Clinical review will identify those whose condition has not improved who need to continue taking them

35. What is ARK-hospital? A 5-year applied research programme funded by NIHR The overarching aim of ARK is to reduce the incidence of serious infections caused by antibiotic-resistant bacteria in the future, through substantially and safely reducing antibiotic use in hospitals now • The programme applies: • Complex behaviour change approaches successful in primary care • Grace-INTRO (Internet Training for reducing Antibiotic use) • STAR (Stemming the tide of antibiotic resistance) Educational programme • To ‘Review and Revise’ decisions taken in secondary care: target behaviour being to discontinue antibiotics

36. What is ARK-hospital? A 5-year applied research programme funded by NIHR • March 2016 – pre-trial work began • qualitative studies in service users and healthcare professionals viz. antibiotic review/duration in hospitals • co-design of complex intervention for healthcare professionals and inpatients /carers to optimise "review&revise" in hospitals • Liaison with specialist societies • Network Development • Spring 2017 – single site feasibility trial of intervention • Autumn 2017 – 1-3 site pilot impact trial • Winter 2017 – 2019 – full impact trial. Cost effectiveness analysis

37. ARK-hospital provides • Information for prescribers about Review and Revise decision making • A decision aid which • acknowledges that when antibiotics are started the diagnosis usually isn’t certain • Are you prescribing for a probable diagnosis of infection or a possible risk from infection? • Encourages prescribers to review daily taking a “stop or justify continue”approach • Leads to a senior clinician’s finalised antibiotic prescription • Information for nurses and pharmacists about ways to support Review and Revise • Information for patients who have had their antibiotic treatment stopped • Structure for Team Meetings and Monitoring to support Review and Revise

38. The key resource is a web-page / app

39. ARK-hospital provides • Education for prescribers about Review and Revise decision making • Why over prescribing antibiotics is hazardous • Why stopping more often is safe • Endorsed by specialist societies • Fulfils mandatory training requirements for stewardship • Will replace parts of existing mandatory training

40. The ARK decision aid • Initial antibiotic prescriptions categorized by diagnostic confidence • Probable diagnosis of infection: Infection is the most likely diagnosis but diagnosis and treatment still needs to be reviewed • Possible risk from infection: Infection is not the most likely diagnosis but you want to use antibiotics as a precaution In both cases prescriptions are time limited (to 48-72 hours) In both cases daily review and revise must happen to establish whether antibiotics can be….

41. At review and revise antibiotics will be • Stopped • Continuation justified • and a finalised antibiotic prescription written. A finalised antibiotic prescription is written after investigation, observation and senior or specialist review. A decision has been made about a final choice of agent, route and duration of antibiotic. Senior here means ST3 or above. Finalised does not necessarily mean microbiologically proven.

42. A finalised antibiotic prescription May be based on a microbiologically proven diagnosis e.g. A patient with microbiologically confirmed Streptococcus pneumoniae pneumonia is prescribed five days amoxicillin to complete a 7-day course of treatment. May be based on a clinically finalised diagnosis e.g. A patient with radiologically confirmed infective spondylodiscitis is prescribed a six-week course of ceftriaxone as out-patient parenteral antibiotic therapy, followed up in the orthopaedic clinic.

43. For example - 1 A 72-year old male presents with a three day history of worsening cough productive of brown sputum, dyspnoea and chills. On examination there is bronchial breathing at the left base. Blood investigations show neutrophilia and a CRP of 320. Chest x-ray shows left lower lobe consolidation. CURB-65 score is 3. You diagnose severe community acquired pneumonia and prescribe intravenous amoxicillin and oral doxycycline according to local policy.

44. For example - 1 A 72-year old male presents with a three day history of worsening cough productive of brown sputum, dyspnoea and chills. On examination there is bronchial breathing at the left base. Blood investigations show neutrophilia and a CRP of 320. Chest x-ray shows left lower lobe consolidation. CURB-65 score is 3. You diagnose severe community acquired pneumonia and prescribe intravenous amoxicillin and oral doxycycline according to local policy. This prescription would be categorised as Probable diagnosis of infection The patient has an organ-based syndrome of bacterial infection, It’s not a Finalised Prescription because you should send samples for microbiology and monitor progress clinically and with laboratory tests. Review and Revise will allow a Finalised Prescription to be written. For instance, the doxycycline may be stopped and the amoxicillin may be converted to oral form.

45. For example 2 A 64-year old female presents with two days of increasing DIB. She has a background of COPD and CCF. On examination she is tachypnoeic and tachycardic, requiring supplemental oxygen to maintain her saturations. She has bilateral wheeze and crepitations. Her bloods show a normal neutrophil count, a CRP of 45 and an acute kidney injury. Chest x-ray shows bilateral pulmonary infiltrates. Your impression is that this is an acute exacerbation of her COPD and CCF possibly triggered by a viral infection. However, she is quite unwell and you are concerned there could be an element of bacterial infection. Therefore, in addition to treating her bronchospasm and cardiac failure you prescribe intravenous Co-amoxiclav.

46. For example 2 A 64-year old female presents with two days of increasing DIB. She has a background of COPD and CCF. On examination she is tachypnoeic and tachycardic, requiring supplemental oxygen to maintain her saturations. She has bilateral wheeze and crepitations. Her bloods show a normal neutrophil count, a CRP of 45 and an acute kidney injury. Chest x-ray shows bilateral pulmonary infiltrates. Your impression is that this is an acute exacerbation of her COPD and CCF possibly triggered by a viral infection. However, she is quite unwell and you are concerned there could be an element of bacterial infection. Therefore, in addition to treating her bronchospasm and cardiac failure you prescribe intravenous Co-amoxiclav. This prescription would be categorised as Possible risk from infection. You are prescribing antibiotics as a precaution. It’s likely that, as early as the next day, a review and revise decision will allow her antibiotics to be stopped based on her treatment response (e.g. resolving breathlessness) and the results of investigations (e.g. her CRP remaining low and the throat swab result).

47. ARK decision aid Actions Probable diagnosis of infection • Observations and investigations should be put in place to find out: • The causative organism (samples sent for diagnostic microbiology). • The need for source control (e.g. imaging studies). • Clinical response (e.g. clinical observations, biomarkers). • More patient historymay become available from previous hospital records. • Daily senior review should record: • If there is still a probable risk of infection and so the antibiotic prescription needs to continue. • Whether a finalised antibiotic prescription can be made. • Whether the antibiotics can be stopped.

48. ARK decision aid Actions Possible risk from infection • Observations and investigations should be put in place to decide on: • A non-infective alternative diagnosis, e.g. pulmonary embolism, congestive cardiac failure, cerebrovascular accident, renal stones. • The absence of infection, e.g. clinical observations, biomarkers, urine microscopy, imaging studies, negative specific microbiology (blood culture). • Daily senior review should EITHER: • confirm that antibiotics can now be stopped OR • document why antibiotics need to be continued (for example, if infection now seems probable)

49. Categories Flow Diagram

50. Lizzie’s slides