An Update on Stewardship Measurement in Hospitals: Programs and An#bio#c Use

Transcription

1 Na#onal Center for Emerging and Zoono#c Infec#ous Diseases An Update on Stewardship Measurement in Hospitals: Programs and An#bio#c Use CAPT Arjun Srinivasan, MD Division of Healthcare Quality Promo#on Centers for Disease Control and Preven#on

2 Disclosures No financial disclosures.

3 NHSN Annual Facility Survey Must be completed by each facility that reports data to NHSN. Almost all acute care hospitals are currently repor;ng data. The survey is completed in the 1 st quarter of each year to reflect the previous year. The 2017 survey is currently being completed. CDC added ques;ons about implementa;on of the CDC Core Elements for Hospital An;bio;c Stewardship Programs in 2014.

4 Uptake of Core Elements, 2016 Of the 4,781 acute care hospitals responding to the 2016 NHSN Annual Hospital Survey, 3,063 (64.1%) reported uptake of all 7 core elements Percent of hospitals mee;ng core element 100% 95% 90% 85% 80% 75% 70% 94.8% 96.5% 90.9% 86.2% 88.1% 83.1% 81.0% Leadership Accountability Drug Exper;se Act Track Report Educate Infrastructure Implementa;on

5 Percentage of U.S. acute care hospitals repor#ng uptake of all 7 CDC Core Elements, by facility demographic, Na#onal Healthcare Safety Network, 2016 (N=4,781) 100% 90% 80% 70% 60% 50% 40% 43.0% 58.1% 69.5% 73.9% 46.0% 69.0% 81.5% 58.5% 76.3% 30% 20% 10% 0% Cri;cal access hospital Surgical hospital General acute care hospital Children's hospital 50 beds beds >200 beds Nonteaching Major teaching Facility Type Bed Size Teaching Status

6

7 Number of facili#es responding to NHSN Annual Hospital Surveys : Number and percentage mee;ng all 7 Core Elements 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1, Mee;ng all 7 Not mee;ng all % 48.1% 64.1% n=4,184 n=4,569 n=4,781

8 Respondent Demographics: Characteris#c All Hospitals 4,184 4,569 4,781 Facility Type Children's hospitals General acute care hospitals 3,385 3,537 3,610 Surgical hospitals Cri;cal access hospitals Bed Size >200 beds 1,320 1,361 1, beds 1,571 1,634 1, beds 1,293 1,574 1,758 Teaching Status Major teaching 1,121 1,248 1,486 Non-teaching/undergrad 3,063 3,321 3,295

9 How Accurate Is The NHSN Stewardship Survey? The NHSN survey goes to the facility administrator for NHSN- almost always in infec;on preven;onist in hospitals. The survey instruc;ons indicate that members of the stewardship team should be asked for input on the stewardship ques;ons. We don t know how ofen that happens or how the results might differ if we sent the survey directly to the stewardship team.

10 How Accurate Is The NHSN Stewardship Survey? CDC partnered with Vizient to try and assess the validity of the stewardship ques;ons on the NHSN survey. In 2016, Vizient added the NHSN stewardship survey ques;ons to their annual stewardship survey, which went to hospitals par;cipa;ng in the Vizient stewardship list-serv. The Vizient survey goes directly to a member of the stewardship team at the hospital.

11 Vizient and NHSN Survey Comparison 189 of the 211 hospitals comple;ng the Vizient survey were successfully matched to NHSN hospitals. 83% of Vizient respondents were pharmacists, but who completed the survey varied by bed size: Hospitals with <250 beds: 88% directors of pharmacy comple;ng Hospitals with >250 beds: 77% clinical pharmacists comple;ng Overall, about 50% of Vizient respondents indicated they had helped complete the NHSN survey. Reported implementa;on of all 7 elements in this subset of 189 hospitals was 58.2% in the Vizient survey compared to 54.5% in NHSN.

12

13

14 NHSN and Vizient Survey Comparison Overall, there was preky good agreement, both on individual core elements and on implementa;on of all 7 elements. Reported implementa;on was higher when pharmacists completed the survey- likely reflects more awareness of specific stewardship ac;vi;es. Only half of Vizient respondents had helped with the NHSN survey. It s possible that other stewardship staff at these hospitals did have input into the NHSN survey, but this seems like an area for improvement.

15 NHSN and Vizient Survey Comparison Differences in reported implementa;on between pharmacy directors and clinical pharmacists could reflect either more familiarity with stewardship efforts by clinical pharmacists or simply reflect less stewardship implementa;on at smaller hospitals. Surveys in smaller hospitals were much more likely to be completed by pharmacy directors and smaller hospitals have less stewardship implementa;on.

16 What s Next for the Stewardship Survey? As compliance with the core elements increases, we need to revise the survey to get a beker understanding of what exactly stewardship programs are doing. How can we use the survey data, combined with an;bio;c use, resistance and C. difficile data to understand what factors might be associated with more effec;ve stewardship programs? We have asked many experts for their thoughts on this.

17 What Have People Told Us to Learn More About? The level of physician and pharmacist effort commiked to stewardship. How many hours per week do people spend on stewardship? How much of their ;me is supported for stewardship? What types of an;bio;cs are people focusing their interven;ons on? Where does the stewardship program fit in the hospital repor;ng structure? Did someone from the stewardship team par;cipate in filling out the stewardship ques;ons?

18 Revisions to the Stewardship Ques#ons We are currently revising the survey based on the input we ve received along with informa;on from many hospitals that have done some pilot tes;ng. The revised survey will have to be approved by the Office of Management and Budget and will be implemented in 2019 to cover prac;ces in calendar year 2018.

19 Percentage of U.S. acute care hospitals repor#ng uptake of all 7 Core Elements, by facility type, % 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% % 69.5% 58.1% 53.1% 53.2% 50.0% 44.6% 43.0% 45.4% 33.6% 26.3% 19.6% General acute care hospitals Cri;cal access hospitals Children's hospitals Surgical hospitals

20 Cri#cal Access Hospitals Are Not Spared the Problems of Resistance and C. difficile Most cri;cal access hospitals have had to treat infec;ons with Methicillin Resistant S. aureus (MRSA) and Clostridium difficile. A growing number are now encountering infec;ons caused by highly resistant gram-nega;ve pathogens as well. 2015: 1.7% of CAHs reported a healthcare associated infec;on with a highly resistant gram nega;ve pathogen. 2016: 2.9% of CAHs reported a healthcare associated infec;on with a highly resistant gram nega;ve pathogen.

21 Cri#cal Access Hospitals Use An#bio#cs About As Much As Larger Hospitals Mean rates (DOT/1,000 days present) Mean Rates of Total An;bio;c Use beds beds >50 beds

22 Barriers to Implemen#ng Stewardship Programs in Cri#cal Access Hospitals Has not been a priority, perhaps due to (mis)percep;on that an;bio;c overuse and complica;ons of an;bio;c use are limited to larger hospitals. Limited access to infec;ous disease clinicians who have historically been champions for and led an;bio;c stewardship efforts. Previous recommenda;ons for stewardship have ignored the reali;es of cri;cal access hospitals (e.g. assemble a team led by an ID clinician and ID pharmacist).

23 Suppor#ng Implementa#on An#bio#c Stewardship in Cri#cal Access Hospitals Discussions with staff in cri;cal access hospitals indicated that specific guidance on how to implement the core elements in CAHs would be helpful. CDC embarked on an effort to do just that. Efforts informed at all stages by discussions with and guidance from staff working in cri;cal access hospitals. Sugges;ons from CAHs for CAHs

24 Implemen#ng Core Elements in CAHs Input from CAHs in several states (in green below) Developed in collabora;on with: Federal Office of Rural Health Policy, The American Hospital Associa;on, The Pew Charitable Trusts

25 Some Key Lessons Learned and Tips from CAHs Bedside nurses can play a key role in an;bio;c stewardship in CAHs. CDC and American Nurses Associa;on recently released a white paper on the role of nurses in hospital stewardship programs: hkp:// WhitePaper Strong rela;onships between staff members ofen makes performance improvement efforts easier to implement and more successful in CAHsonce CAHs decide to do it, they will succeed. Small prescribing staff makes individual feedback and one on one educa;on a realis;c, and very effec;ve, op;on. Design mul;-faceted improvement efforts: C. difficile+stewardship+sepsis

26 Incen#vizing Stewardship Implementa#on in Cri#cal Access Hospitals Medicare Beneficiary Quality Improvement Program (MBQIP) is seeking to expand stewardship programs in CAHs. 99% of CAHs par;cipate in MBQIP Hospitals are supported by state Flex grantees Implementa;on of the CDC Core Elements is now a Core/Required Element of MBQIP ( ).

27 National Healthcare Safety Network Antibiotic Use Option q Captures electronic data on antibiotics administered, along with admission/discharge/transfer data. q Calculates rates of administration for use: By facilities to monitor interventions on single units or facility wide To collect aggregate information on antibiotic use at a regional and national level Eventually, to create antibiotic use benchmarks.

28 Facility-level AU repor#ng 548 facili;es have reported at least one month of data 500 facili;es have reported at least one month of data in Number of facili;es data as of December 1 st

29 Demographics of AU reporters As of December 1 st 2017, AU data has been submiked from: 81 different loca;on types 46 states, the District of Columbia (DC), and the Armed Forces Europe (AE) Characteris#c Number of facili#es Facility type General acute care 373 Veteran's Affairs 83 Cri;cal access 46 Military 22 Children's 7 Other* 13 Bed-size Category Small: <50 beds 97 Medium: beds 205 Large: >200 beds 242 Teaching status No medical school affilia;on 187 Undergraduate teaching hospital 69 Graduate teaching hospital 120 Major teaching hospital 166 *Surgical (3), oncology (2), orthopedic (2), rehab (2), women s (2), women and children s (1), long-term acute care (1)

30 Percentage of NHSN facili#es submieng to AU Op#on, by state Data as of November 2017

31 Standardized Antibiotic Administration Ratio (SAAR) q CDC s 1st attempt at developing a risk-adjusted benchmarking measure for hospital antibiotic use. q SAAR expresses observed antibiotic use compared to predicted use. Predicted use is calculated with risk adjusted models q CDC working with many partners to develop the SAAR measure to try and make it most useful for stewardship.

32 Standardized An#bio#c Administra#on Ra#o (SAAR) Categories Broad spectrum agents predominantly used for hospital-onset/multi-drug resistant bacteria. Broad spectrum agents predominantly used for community-acquired infection. Anti-MRSA agents. Agents predominantly used for surgical site infection prophylaxis. All antibacterial agents. SAAR categories are calculated separately for different patient care locations ICU and Ward Pediatrics and Adult

33 Adult Medical, Medical/Surgical, and Surgical Intensive Care Units, SAAR distribu;ons, 2016 (n=275) SAAR Broad Spectrum Hospital Onset Agents Broad Spectrum Community Onset Agents An;-MRSA Agents Surgical Site Infec;on Prophylaxis Agents All An;bacterial Agents

34 Adults ICUs: 2015 vs SAAR distribu;ons Adult Medical, Medical/Surgical, Surgical Intensive Care Units: SAAR DISTRIBUTIONS 2015 (n=138) 2016 (n=275) Q1 Median Q3 Q1 Median Q3 Broad Spectrum Hospital Onset Agents Broad Spectrum Community Onset Agents An;-MRSA Agents Surgical Site Infec;on Prophylaxis Agents All An;bacterial Agents

35 Key Ques#ons About the SAAR Does the SAAR help point to loca;ons and/or agents where there are meaningful opportuni;es to improve an;bio;c use? How would addi;onal data for risk-adjustment impact the SAAR? Will the SAAR values change if an;bio;c use is improved?

36 Comments on the SAAR From One Hospital The SAAR has pointed us to agents and loca;ons to priori;ze further inves;ga;ons. Knowing that our use is higher than others, rather than just thinking that it is, helps us when we talk to providers and has pushed us to do more. It would help to know the distribu;ons of SAAR values. We d like a SAAR for NICUs!

37 Using NHSN AU Data to Focus Stewardship Efforts Courtesy of Eddie Stenehjem

38 Using the SAAR to Evaluate Stewardship Facility-level Standardized Antimicrobial Administration Ratios (SAAR), Livorsi DJ, et al. Using the SAAR to monitor the influence of antimicrobial stewardship activities. Poster presented at: IDWeek 2016; October 2016; New Orleans, LA.

39 Assessing the SAAR Risk Adjustment: Facility Level vs. Pa#ent Level Data Collabora;on with Kaiser of Southern California looking at risk adjusted benchmarks using a variety of pa;ent level data (DRG, case-mix, diagnosis codes, WBC, Charlson etc) in addi;on to facility level data. NHSN AU currently only gets facility level data. Comparing these benchmark values (KP ra;o) to SAARs. Slides courtesy of Kalvin Yu, Jason Jones, Liz Moisan

40

41

42 Next Steps for the SAAR Measure More work on exploring and refining risk adjustment. More work on assessing applica;on of the SAAR for stewardship. Duke An;microbial Stewardship Outreach Network working on a project to Enroll a group of hospitals in NHSN AU Implement or expand stewardship core elements Assess the impact on SAAR measures

43 Using NHSN AU Data: Poten#al Impact of SEP-1 Measure The Centers for Medicare and Medicaid Services (CMS) began requiring hospitals par;cipa;ng in the Inpa;ent Quality Repor;ng Program to implement a sepsis management bundle (SEP-1), effec;ve October Concerns raised about poten;al increases in broad spectrum and an;-mrsa an;bio;cs. We looked at NHSN AU rates across two 12-month ;me periods: Pre-SEP1: October 2014 September 2015 Post-SEP1: October 2015 September 2016 Compared VA and non-va hospitals since VA hospitals did not have to implement SEP-1.

44 CDC Preliminary Exploratory Sepsis Analysis: Results Total number of loca#ons No. (%) of loca#ons with a sta#s#cally significant increase in use of BSHO agentsᵠ No. (%) of loca#ons with a sta#s#cally significant increase in use of An#-MRSA agents* VA loca;ons Non-VA loca;ons VA loca;ons Non-VA loca;ons VA loca;ons Non-VA loca;ons ICUs (adult med, med/surg, surg) (21%) 7 (23%) 9 (17%) 5 (16%) Wards (adult med, med/surg, surg) (22%) 20 (34%)** 14 (14%) 12 (20%)** ᵠBSHO agents are broad spectrum agents used predominantly for hospital-onset or mul;-drug resistant infec;ons and include: Amikacin, Aztreonam, Cefepime, Cefazidime, Cefazidime/Avibactam, Cefolozane/Tazobactam, Colis;methate, Doripenem, Gentamicin, Imipenem/Cilasta;n, Meropenem, Piperacillin, Piperacillin/Tazobactam, Polymyxin B, Ticarcillin/Clavulanate, Tigecycline, Tobramycin *An;-MRSA agents include: Cefaroline, Dalbavancin, Daptomycin, Linezolid, Ori;vancin, Quinupris;n/Dalfopris;n, Tedizolid, Telavancin, IV Vancomycin ** Differences not sta;s;cally significant, p=0.10 for BSHO and p=0.30 for an;-mrsa agents

45 Conclusions Measurement remains a key focus for CDC hospital stewardship work. What gets measured gets done Measurement is important on several fronts: Program implementa;on An;bio;c use Appropriate an;bio;c use We value your sugges;ons on what we can do to advance measurement in stewardship.