How to Initiate and Sustain Operational Excellence in Healthcare Delivery: Evidence from Multiple Field Experiments Aravind Chandrasekaran PhD Peter Ward PhD Fisher College of Business Ohio State University June 2017 Institute of Medicine (IOM) Report 1999 Almost 20 years later.. People die from the care they receive more often than the illness that brought them to the hospital in the first place Makary et al. 2016 BMJ Examples of Preventable Medical Errors (IOM) Wrong site procedures Incorrect prescriptions Healthcare Associated Infections Central Line Infections Handoffs 2 1
Why does this happen? Consider simple handoffs and discharge instructions delivered 3 Discharge process (say after a surgical procedure) HOSPITAL STAY OUTPATIENT CARE Teaching related to discharge begins at various points during stay One directional conversation rushed due to topical coverage Outpatient staff may use different terms and vary instructions Care for life Patient Arrives for Surgery (T = 0) Surgery (T=2 days) Multiple RNs, PAs, Doctors, Residents/ Fellows, Social Workers, Psychologists emphasize different aspects of discharge Multiple teaching/ technology aids (with errors) Discharge from Hospital (T=7days) Lack of Handoff between inpatient and outpatient teams regarding the patient Labs and Other wellness Appointments (T= 9 days) [48 hrs. after discharge] 30-days after discharge 1 in 3 patients get readmitted back to the hospital Several forms of variation introduced by the care provider 4 2
Patient demographics Variations can also come from the patients Education level Language/ Culture Learning styles (e.g. visual vs. narrative) Support structure (e.g. family members) Technology readiness (e.g. use of internet, online systems) 5 Other Healthcare Delivery Challenges Hierarchy Process is too complex to standardize Limited empowerment Initiatives fail to sustain 6 3
What is the solution? Developing Patient Centric Healthcare Delivery systems that are highly standardized yet have in built processes to allow customization for each patient 7 Evidence from Two Experiments Site The Ohio State Wexner Mayo Clinic Medical Center Context Kidney Transplant Outpatient Setting Discharge Process Years 2013 2016 2016 ongoing Intervention Standardized Discharge Daily Huddles Process Type of Experiment Quasi Experiment Randomized Control Trial Results 8% reduction in 30 day readmission and 6% increase in HCAHPS scores Preliminary results better access and improved screening outcomes 8 4
Experiment 1 Research Question What is the effect of caregiver designed and implemented Patient Centric Standardized Discharge Work (PSDW *) with patient input on care delivery outcomes? PSDW set of activities standardized in terms of content, sequence, timing, and outcome but flexible enough to allow patient specific customization Other Collaborators: Susan Moffatt Bruce MD, Todd Pesavento MD, Mary Lou Hauenstien RN, Gopesh Anand PhD 9 Study Design 3 year grounded work at OSUWMC Field work: 40 caregivers, 102 transplant patients T 1 = 6 months Phase 1: Pre Study Work (Q1 2014 Q2 2014) T 2 = 9 months Phase 2: Intervention (Q3 2014 Q1 2015) T 2 = 12 months Phase 3: Post Implementation (Q2 2015 Q1 2016) 1. Pre Study Work Map Current Discharge process Collect Patient Data (100+ patients) Collect Nurses/Coordinator data Observe Discharge Process Shadow Nurses and Talk to Patients 3. Patient Input for SDW Collect patient input on the newly developed SDW 2. SDW Workshop Develop SDW by the team Synchronize all teaching kits and discharge planning with the standards) 4. Implement SDW Train all the nurses CI of SDW 5. Post SDW (Q2 2015 +) Collect patient and caregiver data Observe the effectiveness of CI 6 10 5
Phase 1: Pre Implementation Shadowing caregivers and observations on the discharge process Instructions on water intake Evidenced Based Standards Transplant recipients must drink at least 3 liters of water every day. Failure to adhere to this standard can result in increased creatine levels. More creatine levels results in increased blood pressure resulting in readmission. Scenario 1: Drink a lot of water Scenario 2: Drink 2 liters of water Scenario 3: Drink a lot of fluids Scenario 4: Drink six 16.9 ounce bottled water 11 Pre Study* Surveyed 87 kidney transplant recipients and matched them with their medical records Conducted focus groups with kidney transplant recipients and families to understand the current state of the discharge process Results Risks of getting readmitted is 102% higher for a one unit increase in anxiety levels one week following discharge * Journal of Surgical Research (October 2016) Lack of standardized discharge work and lower levels of empathetic care delivery increases patient anxiety upon discharge (explains 27% of variation in anxiety) 12 6
Phase 2: Implementation Caregivers developed PSDW through six (3 hour) workshops Content, Sequence, Timing and Outcome Components Content Instructions prior to discharge based on evidenced based standards (Before our study: 90+, Now = 20+) Sequence Sequence of instructions (Before our study: Varied, Now = one standard sequence agreed upon) Timing How long and when to start instructions (Before our study: Anytime, Now = 24 hrs, RASS =0, First meal) Output Assessing patient compliance (Before our study: Varied, Now = Smart phrases and Teach back) 13 14 7
Major Topics Pre PSDW Major Topics Post PSDW Infection Prevention * Symptoms of Rejection * Labs * Dental Care Eye Care Vaccinations * Sports and Recreation Activities Lifestyle changes Going back to Work * Treatment of Complications * When to Call Medications Vital Signs * Gardening OTC Meds * Follow up Appointments * Fluid Intake * Activity Progressions * (e.g. Pregnancy, Sexual Activity, Special Activity) Pet Care Holiday Schedules for Labs Wound Care * Going out to Public Places Lifting Instructions * Emergency Contact (e.g. 911, primary care) * Smoking and Drinking Use of EMR/Patient Portal * Multiple topics Part I (During Hospital Stay) by Inpatient Nurses Starts when the follow: 24 hrs. post transplant and no later than 32 hours, RASS Score = 0, family present and patient tolerate first meal Infection Prevention * Vital Signs and Symptoms * Labs * When to Call Medications Vital Signs * Fluid Intake * Emergency Contact (e.g. 911, primary care) * Part II (After Discharge) by Outpatient Nurses Starts within 48 hours after discharge and continues for about 3 months (face to face as well as over the phone) Medications Follow up Appointments Activities (e.g. pregnancy, sexual, social, gardening) Vaccinations Eye and Dental Care Lifestyle Change Treatment of Complications Post op care (Wound Care) * Emergency Contact (e.g. 911, primary care) * OTC Meds Going back to work, Public Places Pet Care Exercising & Sports Activities * 15 Phase 3: Post Implementation Hospital Stay Outpatient Clinic Patient Arrives for Surgery (T = 0) Transplant Surgery (T = 2 days) Teaching begins 24 hours after surgery (Patient must tolerate meal to demonstrate attentiveness) Teach-back (Conversational) using multiple aids to reinforces learning Patient discharged from hospital (T = 6) Outpatient nurse calls patents (T = 8) 48 hrs. after discharge First lab appointments (T = 14 days) 7 days after discharge 30-days after Discharge 50% Reduction in topics (Most important for immediate well being are taught) Handoff between inpatient and outpatient teams (Outpatient nurse sees patient before discharge) Everything done in outpatient unit is standardized and communicated with inpatient unit Instruction standardized and consistent across teaching aids (cheat sheets, flip charts etc) Bi weekly Problem solving huddles Weekly Problem solving huddles Weekly huddles between inpatient and outpatient managers with nursing director 16 8
Phase 3: Inpatient/Outpatient huddles Bi weekly Huddles (Inpatient) Weekly Huddles (Outpatient) 17 Empirical Validation (Post Implementation) Difference in Difference Methodology Readmission Control Treatment Effect of Intervention Pre During Post Control Group Heart and Liver Transplant process Same hospital unit, share technologies Discharge process were streamlined with kidney transplant prior to study Influenced by any changes at OSUWMC Patients face similar issues post discharge Units are spatially separated to avoid any spillovers 18 9
Empirical Validation (May 2015 April 2016) Treatment Group Kidney Transplants Heart Transplants Control Group Liver Transplants Volume 571 62 103 Age 50.27 years 58.59 years 54.74 years Length of Stay 7.84 days 24.58 days 16.25 days 30 day 36.60% 32. 46.6% Readmissions 30 day Mortality 0.7% 3. 1.9% Gender 40% Female Female 30% Female Readmissions and other demographics are comparable to national average Minimal nursing and physician turnovers during the study period 19 Benefits Probability of Readmission.2.25.3.35.4.45 Likelihood of readmission postimplementation is 0.32 times lower than control group and pre and during implementations Pre-Implementation Control Group Time Period Treatment Group Post-Implementation Cost Benefits Data on direct (room and board, pharmacy, drugs etc) and indirect costs (salaries & benefits, building, labs) Total cost / Readmission = $14, 804 9 readmissions avoided (Savings) = $ 133236 20 10
HCAHPS Quality Ratings VBP Weight FY 2016 Outcome 40% Process 10% Efficiency VBP Weight FY 2017 Efficiency and Cost Reduction, HCAHPS PROPOPSED VBP Weight FY 2018 Efficiency and Cost Reduction, HCAHPS HCAHPS Safety, 20% Clinical Care Process, 5% Clinical Care Outcomes, Safety, Clinical Care Outcomes, HCAHPS Scores are a part of reimbursements for Medicare and Medicaid Patients, Source: CMS.gov Access to HCAHPS unit level data for treatment (Unit A) and control (Unit B) groups 21 HCAHPS Post Implementation Overall Patient Satisfaction (%) Pre-Implementation.64.66.68.7.72 Control Group Time Period Treatment Group Post-Implementation HCAHPS scores increased by 6% for treatment group while it is decreased by 7% for control group 22 11
Key Takeaways Showing that the problem exists is a good approach to drive change People doing the work must develop their work standards Sustaining requires middle managers to initiate and sustain huddles Requires Cultural Transformation (Not just tools!) 23 Experiment 2 Research Question How does physician engagement in daily huddles improve team effectiveness and outpatient performance? Other Collaborators: Dave Rushlow MD, Marc Tummerman MD, Brad Staats PhD and Nilay Shah PhD 24 12
Background Mayo Model of Outpatient Care (MMOC) deployed in 2014 with mixed results Element of MMOC Team Based Care Lack of proper buy in among physicians (Not studied in literature) Poor team effectiveness and mixed effects on screening performance 25 Study Design Pre Work Observations (April 2016 Aug 2016) Study Design and Recruiting Teams (Sep 2016 April 2017) Study Workshop with treatment group (April 2017) Post observations (May 2017 Oct 2017) 26 13
Pre Work Observations Visited multiple outpatient sites and semi structured interviews with care team Mixed huddle participation and compliance Correlated with team climate and screening outcomes 27 Correlational Observations Pod 3 huddled most frequently with physician mostly present while Pod 2 huddled very ad hoc 28 14
Study Design and Recruiting Multiple sites from WI and MN regions to participate Each team physician, nursing and admin 30 teams expressed interest divided into treatment and control arms 29 April 12 2017 Workshop for Treatment arm Each team with an assigned coach Training on content and process of huddling 30 15
Huddle Workshop Communicate and Assign Accountabilities 1. Status and Flow Match resources to patients Create plan and schedule 2. Occurrence Tracking Compare the plan to actual Identify and capture issues 3. Problem Solving Show scientific thinking Experiment using PDSA Status & Flow Occurrence Tracking Problem Solving Huddle Managing Daily Improvement System Match resources to patients Create plan and schedule Compare the plan to actual Identify and capture issues Scientific thinking Experiment using PDSA Communicate Assign accountabilities 31 Post Observations Bi weekly coaching by the assigned coaches Onsite visits 30 day & 100 day follow up by coaches Pre survey on team climate, psychological safety and trust for all 30 teams (response rate 70% no sig. difference across the arms) Track outcomes after 6 months for both the arms 32 16
Post Observations Results pending on the overall effectiveness (Nov 2017) Preliminary qualitative insights are positive Nice process story by tracking progress over time 33 Questions Thank you! chandrasekaran.24@osu.edu Ward.1@osu.edu 34 17