SYSTEMS CHANGE TO REDUCE TOBACCO USE IN IN CLIN INICAL SETTINGS THE NUTS AND BOLTS • DONNA WARNER, MA, MBA • INDEPENDENT CONSULTANT FORMER MANAGING PARTNER, MULTI-STATE COLLABORATIVE FOR HEALTH SYSTEMS CHANGE AND CESSATION DIRECTOR, MASSACHUSETTS TOBACCO CESSATION AND PREVENTION PROGRAM
TOPICS Why are clinical systems to address tobacco use important? What do we mean by ‘systems change’ in the clinical setting? Show me a good ‘system’! How do we know it works? Forces be with us 2/3/2016
PRESENTATION BASED ON TW TWO HEALTH SYSTEMS AND PUBLISHED RESULTS: • Atrius Health, Massachusetts • Thad Schilling MD, formerly with Atrius Health, currently with Reliant Medical Group, Worcester MA • LSU Public Hospitals, Louisiana • Sarah Moody-Thomas PhD, Charles L. Brown, Jr., MD Professor of Health Promotion Director, Tobacco Control Initiative, LSU Health New Orleans, School of Public Health • Lead Researcher • Thomas Land PhD, Director, Office of Data Management and Outcomes Assessment, Commissioner’s Office, MA Department of Public Health • AND, experience of many colleagues around the country working to improve systems in healthcare to reduce tobacco use
WHY CLINICAL SYSTEMS? • Unparalleled Reach - 70% Tobacco Users -including hard-to-reach and specific populations • Feasible to Implement • Powerful Incentives • Benefits to Patients and Population Health • Cost Savings to Healthcare Facilities
WHAT IS IS A SYSTEM? • Not an individual clinician Multiple Components burden • CEO, CFO, Clinical Leadership • Focus on reaching • Integrated clinical and EHR workflow • Defined staff or team roles and sustained high levels of training: It’s your job! identification and brief • Feedback reports by team and site intervention • Performance measurement reported • Electronic referrals for to the “corporation” more intensive treatment
SHOW ME A GOOD ‘SYSTEM’!
WHAT IS IS A GOOD SYSTEM? • A good system is one that, when implemented fully, is easier to do than not do. Trying harder to make the existing system work better may not be enough
THE A ATRIUS HEALTH SYSTEM • HVMA serves 350,000 Patients in 17 Clinical Sites • Uses EPIC • Known as a “high performance” system • Operating under alternative quality payment contracts • An MD project leader assigned to implement the Tobacco Intervention Project • Tobacco Intervention System was later expanded to 5 additional healthcare organizations ATRIUS umbrella, serving 770,000 patients.
BEFORE QI: “IN SHORT, WE HAD NO SYSTEM -WIDE INTERVENTION” • Attempted to mine existing EPIC data to establish a QI goal • Data showed 84% - ask? • Didn’t know what workflow produced these “ Our Chie ief Medical l Officer results identifie id ied sm smokin ing • No record of the assessments or interventions cessation as ce as an an im important • No record of cessation prescriptions initia in itiative to o im improve th the • Only knew for sure the ½ of 1% referred to quitline; only loose affiliation with state overall ll heal alth of of ou our quitline pop opula lation. ” • No academic detailing, no supports to assist clinicians
The intervention must: Atrius Health • Be evidence-based, iterative, and consistent through the system • Integrated into daily practice. • Support clinician intervention at point- of-care. • Involve all members of the clinical team. • Leverage community resources (like the QuitWorks — the state’s quitline referral program).
AFTER: WORKFLOW IM IMPLEMENTED IN IN 12 OF 17 CLINICS Changes in 4 areas : Teaming Approach with Decision Support • clinical pharmacy Integrated into EHREHR • information technology • patient health education and marketing • team training New relationship with Quitline • fully electronic referrals • customized feedback report on patients referred
HOW THE WORK WAS ORGANIZED • Clinical workflow features a teaming approach • Roles defined for medical assistants, visit clinicians (usually an MD), and advanced practice clinicians. • Workflow includes scripts and prompts for each step in the intervention built into the EHR. • Real-time data entry in the EHR is required: the identification of smokers and interventions has to be recorded as they occur during the patient visit. • Measurement and feedback monthly to clinical teams
BY FEEDBACK, WE MEAN AN ACTIVE R REVIEW PROCESS M onitor performance data monthly by clinic, by team, and by team member on the “ask,” “assess,” and “assist” steps . • Provide monthly feedback reports • Review performance with each department, clinician and medical assistant on each measure • Review patient cases with missing assessments • Identify staff for coaching to improve assessment and identification rates. • For staff with lower scores, conduct periodic refreshers on the EHR screens that contain the smoking assessment and intervention prompts It is this active review process with each team, not simply providing reports, that is critical to performance improvement.
Helping Patients Quit: New Research on Clinical Tobacco Interventions August 9, 2012 HOW DO WE KNOW IT IT WORKS? WHAT WAS DONE? HOW DID PATIENTS FARE? OBTAIN AND ANALYZE ENCOUNTER RECORDS — MILLIONS OF THEM!
WHAT WAS DONE? GROUPING 17 SIT ITES BY ROLL-OUT DATE Measure: Total number Identification Rates by Site of identifications in Jan 2008 - Oct 2010 proportion to the total number of visits 1 Data Source: 17 sites 0.75 Population : 310,577 adult % of Visits with Tobacco Use Identification patients 0.5 Green Sites : Identification rate over 80% in Jan 2008 Orange Sites : Identification 0.25 rate increased significantly between January 2008 and October 2010 Blue Sites: Identification 0 J Jul- J Jul- J Jul- rate did not change an- 08 an- 09 an- 10 08 09 10 significantly between BOS BTR BUR CAM CHE CND COP FAU KEN LMA MFD PBY QCY January 2008 and SOM WEL WRX WR October 2010
HOW DID ID PATIENTS FARE? FEWER SELF-REPORTED SMOKERS Total Current Smokers at Most Recent Screening (n = 75,129, 1st screening <7/1/2008) 11500 # Current Smokers 11000 10500 10000 Jun-09 Jun-10 Oct-08 Dec-08 Feb-09 Oct-09 Dec-09 Feb-10 Oct-10 Apr-09 Aug-09 Apr-10 Aug-10 Population: 75,129 adults who were screened prior to 7/1/2008 Data Sources: 15 Sites (not including BTR and CHE) Tobacco use status for population determined by most recent status prior to the beginning of each quarter.
TO SU SUM IT IT UP • Decreases in smoking prevalence were 40% greater at sites that achieved systems change (13.6% vs. 9.7%, p<.01). • Likelihood of quitting increased by 2.6% per occurrence of brief intervention . (p<0.05, 95% CI: 0.1% – 4.6%) • For patients with a recent history of current smoking, the likelihood of an office visit for smoking-related diagnoses decreased by 4.3% on an annualized basis in sites where systems change occurred (p<0.05, 95% CI: 0.5% – 8.1%). It’s the System Itself That Works
IS IS TH THIS A ONE-TRICK PONY? NOT! “
LS LSU TOBACCO CONTROL IN INITIATIVE (2 (2000 TO 2012) • Louisiana State University (LSU) 10- hospital public safety-net system--one Sarah Moody-Thomas PhD of our country’s largest Professor LSU School of Public Health, New Orleans Director, Tobacco Control Initiative, • With a diverse, low-income, primary “The [Tobacco Control Initiative} data care population: Mean age 49 years; 71% collection allowed us, for the first time, to female; 53% AA; 66% uninsured understand the prevalence of tobacco use among our primary care patients, while analyses of EHR data showed improved rates of intervention and quit rate”. -- S. Moody- Effect of systems change in Thomas Louisiana similar to Atrius Health
LS LSU CLINICAL AND EHR WORKFLOW
WANT TO KNOW MORE? Effect of Systems Change and use of The Effect of Systematic Clinical Electronic Health Records on Quit Rates Interventions with Cigarette Smokers on Among Tobacco Users in a Public Quit Status and the Rates of Smoking- Hospital System Related Primary Care Office Visits. Sarah Moody-Thomas, Laura Nasuti, Yong Yi, Michael Land TG, Rigotti NA, Levy DE, Schilling T, Warner D, Li D. Celestin, Jr, Ronald Horswell, and Thomas G. W (2012); PLoS ONE 7(7): e41649. Land. American Journal of Public Health: April 2015, doi:10.1371/journal.pone.0041649 Vol. 105, No. S2, pp. e1-e7. doi: 10.2105/AJPH.2014.302274
LIMITATIONS AND PRACTICAL LESSONS LEARNED
LI LIMITATIONS Timeline for LSU Public Hospitals Project • Length of time to ramp- up to steady state • Leadership essential • Measurement essential • Access to large patient encounter data sets
LESSONS LEARNED FROM ATRIUS HEALTH Team-based intervention • Quitline Referrals Doubled Physician involvement is critical but minimal After e-Referral Launch • Medical Assistants trained successfully to begin 5A’s intervention 150 125 Operations 100 No. of Referrals • Make the right thing to do the easy thing to do” 75 50 25 Performance feedback 0 October November December January February March • Real time, usable, public/transparent, operator level critical Atrius Referrals 2009-2010 (Before Launch) Atrius Referrals 2010-2011 (After Launch) Community partnerships • Close collaboration with Massachusetts Quitline helped us define our vision
Recommend
More recommend
