The LDL Challenge: Using Health Information Technology to Drive Clinical Quality Improvement

Transcription

1 The LDL Challenge: Using Health Information Technology to Drive Clinical Quality Improvement Tricia Lee Wilkins, Pharm D, PhD Kathy Reims, MD Cory Sevin, RN, MSN, NP March 11, 2014 Session C4 Financial Disclosures Although participants included grantees of ONC funded demonstration programs, the LDL Challenge was an unfunded quality improvement initiative. Technical Assistance for the LDL Challenge was provided through a sub contract with IHI. 1

2 Learning Objectives: Learning Objectives: Participants will describe quality improvement strategies to achieve outcomes in cholesterol control. Participants will identify how to effectively utilize HIT to support these strategies- e.g. system level changes and population health. Participants will identify practice readiness and strategies to create an environment conducive to HIT driven Quality Improvement. Participants will learn a number of successful approaches to achieve change, including leadership engagement, population health and clinical decision support. 2 Session Overview Overview of the LDL Challenge What data did teams need? How did they make the data actionable? What new actions did they take based on their data? Lessons learned 3 2

3 What Brought You to This Session? Turn to your neighbor, introduce yourself, your role at your home organization and interest in this session. 4 The Beacon Community Program: Where HITECH Comes to Life 17 communities each funded ~$12-16M from yrs to: Build and strengthen health IT infrastructure and exchange capabilities - positioning each community to pursue a new level of sustainable health care quality and efficiency over the coming years. Improvecost, quality, and population health -translating investments in health IT in the short run to measureable improvements in the 3-part aim. Test innovative approachesto performance measurement, technology integration, and care delivery -accelerating evidence generation for new approaches. 3

4 Challenge Background Health information technologies position practices for population health management and rapid cycle quality improvement. Quality improvement challenge for patients with diabetes to demonstrate how data can drive improvement across a practice and at the point of care. Target stubborn clinical outcomes (e.g. LDL) for dramatic improvement. One national multi-payorestimate of LDL control is 60%. 6 What were we Trying to Accomplish? Challenge Goal: Cut gap to goal (60%) in half. Example-If baseline LDL control is 40%, then gap to goal is 20 percentage points. Challenge goal is 50% LDL control. 7 4

5 Theory of Change 1. Ability to break problems down into their sub processes is key for successful resolution. 2. Overarching system level changes are effective ways to address down stream sub processes. 3. Data provides the basis for evidenced based care. 4. Any one can improve. Employing the right methods is paramount-lack of experience does not preclude ability to improve. Theory of Change: Our Approach 1. Ability to break problems down into their sub processes is key for successful resolution. Use EHR to focus on identifying segments of patients for whom targeted interventions are warranted. 2. Overarching system level changes are effective ways to address down stream sub processes. System-level best practices suggested as scripted moves. 3. Data provides the basis for evidenced based care. Baseline, weekly and monthly reporting of measures. Use an intensive, innovative quality improvement approach in a fast-paced, learning environment e.g. daily/weekly PDSA cycles that include scripted moves. 4. Any one can improve. Employing the right methods is paramount- Lack of experience does not preclude ability to improve. Create a supportive learning community with access to subject matter, quality improvement and HIT expertise. 9 5

6 Participating Practices Clinics Providers Patients Site A Site B ,492 Site C Site D ,076 Site E ,530 Total , Monthly Measures Measure 1: Screening Measure (NQF 0063) Number of patients in the denominator who had an LDL- C test performed during the measurement year Measure 2: LDL Control Measure (NQF 0064 modified) Number of patients in the denominator whose most recent LDL-C level performed during the measurement period is < 100 mg/dl Measure 3: Statin Use Measure (de novo) Total number of patients years of age with diabetes (type 1 or type 2) whose most recent LDL-C level performed during the measurement period was equal to or greater than 100 mg/dl, and not prescribed a statin 11 medication 6

7 How did teams approach their data needs? EHR MU Measures IT customization needed Manual processes supplemented Clearly Defined Cohort Population of Focus Outcome Measure Denominator Outcome Measure Numerator 13 7

8 Cohort and Baseline Cohort Definition Cohort: Patients years of age who had a diagnosis of diabetes (type 1 or type 2) during the measurement year. Measurement year: April 1 st of 2012 to March 31 st of 2013 Baseline and Goal: -Baseline: Measured in March of Provided numerical values for clinic goals -Clinical Goal: Cut gap to national goal (60%) in half. Example-If baseline LDL control is 40%, then gap to goal is 20 percentage points. Challenge goal is 50% LDL control. 14 Data Segments Correspond to Patient Subpopulations 15 8

9 Summary of Data Needed Cohort: Patients years of age who had a diagnosis of diabetes (type 1 or type 2) during the measurement year. Measurement year: April 1 st of 2012 to March 31 st of 2013 Subpopulations of that cohort Those with screening due Those with LDL in control Those with LDL not in control Helpful to know if treatment just adjusted or persistently out of control 16 Segments and Drilling Down How did Teams Make the Data Actionable? 17 9

10 Clinic A Spotlight: Weekly and Monthly Data 57.5% 61.7% LDL of <100 mg/dl 65.8% 71.7% Baseline Apr-13 May-13 Jun Baseline April 2013 May 2013 June 2013 Data Segment 1: No LDL Screen in last 12 months Data Segment 2: No StatinLDL > 100 mg/dl N= 120 patients with DM Teams identified themes within subpopulations Providers Clinic site SE status Ethnicity/cultural issues 19 10

11 Drilling Down In addition, teams identified: Which patients were not screened Which patients were not on a statin Which patients were on a statin but still not controlled 20 Data was Actionable Systems Level Scripting Policy Staffing Standing orders Nurse Protocols Improved Access Best Practice Patient Level Adherence issues Concerns about side effects Treatment intensification Cultural conversations Accommodations to meet patient needs 21 11

12 Connecting the Dots Making it Relevant What are you working to improve and how can you think about segmenting your population of focus in order to accelerate system changes? Describe your current data segments What additional data or segmentation would be useful? 22 What Actions did Teams Take? Data Driven Action 23 12

13 Scripted Moves Concept born out of work in Viet Nam with childhood malnutrition by Jerry Sternin Found that if you focused on evidence based focused practices you can accelerate improvement Scripted moves often manifested as change packages -limited number of concrete actionable interventions 24 Improvement Story: Moving the Dots Screening Statin Rx Treatment Adjustment Population Management, Access, Other Standing orders, nursing protocols Protocolto get on right statin Drop in appointments Drop in appointments Test when pts. in clinic (nonfasting) EHRalerts EHRalerts Pre-appt. prep via care managers, huddles, chart prep Scripting calls to patients Standing orders Standing orders Registry functions; daily and weekly dashboard Scripting IVR calls EHRdecision aids RN intensifying statin treatment under protocol. Lunch and learn for pts. Health Fairs/Lunch and learn for pts. LDL/Statin talking points in exam room Testing scripting to engage pts. /pts of various cultures Tracking adherence by pharmacy fills In-clinic phlebotomist EHRalerts + care team members engage with pts. Huddles with individual providers Check adherence/ Fill history through Sure Scripts Patient Education Calling patients with Office of the National reminders Coordinator of for follow up Health Information appointment Technology s Tracking patients in each segment in real time: daily or weekly Modifiedopen access scheduling 25 13

14 Data Walls and Real-time Learning Source: IHI 27 14

15 Results Through August 30, 2013 Note: Reporting through June for Site D.2 Results Through August 30, 2013 Weighted Total Percentages by Measure (N= 12,871) Measure Baseline August Change Annual LDL Screening 75.1% 78.2% 3.1% LDL Control < 100 mg/dl Statin Measure LDL > 100 mg/dl NOT on Statin 37.7% 40.4% 2.7% 37.8% 28.5% -9.3% Measure 1: Screening Measure (NQF 0063) Number of patients in the denominator who had an LDL-C test performed during the measurement year Measure 2: LDL Control Measure (NQF 0064 age range modified) Number of patients in the denominator whose most recent LDL-C level performed during the measurement period is < 100 mg/dl Measure 3: Statin Use Measure (de novo) Total number of patients years of age with diabetes (type 1 or type 2) whose most recent LDL-C level performed during the measurement period was equal to or greater than 100 mg/dl, and not prescribed a statin medication 15

16 Theory of Change: Lessons Learned 1. Ability to break problems down into their sub processes is key for successful resolution. Not easily done!!! The data teams needed was not readily available but integral to team success. Data for subpopulations is needed to identify opportunities for system change 2. Overarching system level changes are effective ways to address down stream sub processes. Practice transformation and information flow sets the stage for improvement. Rapid improvement requires change at the system level -Population management 3. Data provides the basis for evidenced based care. Employing small tests of change and scaling Using data to support rapid iterative testing. Small changes were helpful to increase buy-in and acceptance of changes Communication about changes and progress was essential 4. Any one can improve. Employing the right methods is paramount-lack of experience does not preclude ability to improve. Success not dependent on prior QI experience but QI skills were needed for success Leadership buy-in was critical- including executive, clinical and IT leadership was best Thank You ONC / Beacon Community Program/ Office of the Chief Medical/Regional Extension Program Joseph Bormel, Medical Officer Christine Castle, Summer Intern Lisa-Nicole Danehy, HIT Vanguard Coordinator Janhavi Kirtane Fritz, Beacon Director Amy Helwig, Medical Officer Jesse James, Senior Medical Officer Kevin Larsen, Meaningful Use Director Farzad Mostashari, National Coordinator Kerri Petrin, Project Officer Jacob Reider, Chief Medical Officer Julia Skapik, Fellow Institute for Health Care Improvement Roger Chaufournier, Quality Improvement Faculty Lindsay Hunt, Project Management Kathy Reims, Quality Improvement Faculty Puneet Sahni, IT Reporting Lead Cory Sevin, Project Lead Booz Allen Hamilton Eunice Choi, Project Assistance Giao Phung, Project Assistance Rupali Sardesai, Project Management Subject Matter Experts Anne Camp, Endocrinology Jerry Osheroff, CDS/Practice Transformation Judith Schaefer, Patient Engagement Julie A. Schmittdiel, Quality Improvement 16

17 Additional Resources ONC Resources: Continuous Quality Improvement (CQI) Strategies to Optimize your Practice mentprimer_feb2014.pdf ONC Resources: Continue Quality Improvement beyond EHR Implementation IHI Resources: How to Improve: ementhowtoimprove.aspx 32 For more information please contact Tricia Lee Wilkins 33 17