10/14/2016 Quality Improvement in the Hospital Scott A. Flanders, M.D. Professor of Medicine Associate Chair, Quality and Innovation Director, Hospital Medicine Program Director, Hospital Medicine Safety Consortium University of Michigan 1
10/14/2016 The Focus on Value Multiplier A=APPROPRIATENESS Value=A x Quality / Cost Overview • Need to Improve-External Forces • Common Pitfalls in hospital QI / Case studies – Leaders and champions – Data – Infrastructure – Culture / Socioadaptive elements 2
10/14/2016 Proportion of Value ‐ Based Payments is Increasing Rapidly Goal 1: 30% of Medicare FFS payments are tied to value through alternative payment models by the end of 2016, and 50% by the end of 2018 Goal 2: 85% of all Medicare FFS payments are tied to quality metrics by the end of 2016, and 90% by the end of 2018 CMS Framework 2016 2018 1. FFS w/no link to quality 2. FFS w/link to quality 30% 3. Alternative payment built on FFS 50% 85% 4. Population based payment 90% All Medicare FFS FFS linked to quality Alternative payment models 5 Confidential: not for distribution Physician Payment • Move away from all fee-for-service • Strong focus on measuring and improving quality • Cost control – More risk placed on physicians – But also potential for reward 3
10/14/2016 Hospital Payment • Value Based Purchasing Program – Carrot and Stick; 2% of DRG payments • Readmissions Reduction Program – Stick only; 3% of DRG payments for excess readmits • Hospital-Acquired Conditions Reduction Program – Stick only; 1% of all Medicare payments (not just DRG) Value Based Purchasing $ Millions of Dollars at Stake for Hospitals $ Clinical process (AMI, CHF, Pneumonia, SCIP, healthcare associated infections) Patient experience (HCAHPS) Clinical outcomes (Mortality rates for AMI, CHF, Pneumonia; CLABSI; Patient Safety Indicator 90) Efficiency (Risk-Adjusted spending from 3 days PTA to 30 days post- discharge) 4
10/14/2016 Hospital Payment • Value Based Purchasing Program – $3 million • Readmissions Reduction Program – <$1 million • Hospital-Acquired Conditions Reduction Program – $2 million QI V3.0: Sink or Swim (the burning platform) • Bundled payments / population based payments • Narrow networks – Low premiums are most important to consumers • Increased transparency – Consumer Reports • Delivering high value (appropriate) care: necessary 5
10/14/2016 Optimizing Performance • Target high volume / high cost / high variability – COPD, CHF, afib, CAP, Biliary dz, VTE, Sepsis, THR • Measure your outcomes (cost and quality) • Create guidelines and pathways (key processes) • Integrate into IT systems / identify gaps • Data feedback to providers • Same process for complex / tertiary care – Melanoma, bladder CA / cystectomy, colectomy (IBD) Standardizing Common Processes • Hand-offs • Discharges – Home – Skilled nursing facilities • Bedside rounds • Patient communication 6
10/14/2016 Target Cost and Appropriateness • Over-testing – Troponin, iCa, viral panels – PE CT for low pre-test probability / neg d-dimer – “Repeat ECHO” – MRI use • Over-treatment – UTI, CAP, HCAP, Cellulitis • Under-treatment – EP for afib, PCT “The hospital is the most complex human organization ever devised…” - Peter Drucker 7
10/14/2016 Overview • Need to Improve-External Forces • Common Pitfalls in hospital QI / Case studies – Leaders and champions – Data – Infrastructure – Culture / Socioadaptive elements A National Priority • First attempt to characterize the annual human toll of antibiotic resistance. 8
10/14/2016 Local Interventions The Big 3 Infectious Diagnoses in U.S. Hospitals Ranking at UMHS Urinary Tract Infections #1 Pneumonia #2 Skin and Soft Tissue Infections #3 Gandhi T, et al. ICHE 2009 Improving Antibiotics for UTI • Large AMC, community teaching hospital • Goals: – Evaluate antibiotic use for UTI – Identify inappropriate treatment – Design strategies to improve antibiotic use – Target hospitalists Hartley S, et al. ICHE, 2013 9
10/14/2016 Testing and Treatment for UTI • 60% of patients lack guideline indications for urine culture • Positive urine culture – 40% have UTIs by adjudicated review – 25% of UTIs had inappropriate treatment duration – 65% of asymptomatic bacteriuria was treated – 385 excess antibiotic days at UMHS alone Hartley S, et al. ICHE, 2013 Improving Antibiotic Use • Standardize recommendations for testing • Standardize treatment algorithms • Educate hospitalists • Pharmacist-hospitalist review of urine cultures • Measure the impact IHI Forum, 2013 10
10/14/2016 The UMHS Approach • Multidisciplinary team led by Sarah • Asked hospitalists what might work • Developed tools based on their input Sarah Hartley, M.D. – Pocket Cards – Posters in conference room – Abx app – Pharmacist timeout • Education sessions – Dinner for night docs • Shared data, identified and resolved barriers SHOULD THIS PATIENT BE EVALUATED FOR A URINARY TRACT INFECTION * ? Does the patient have any of the following without alternate explanation ? 1. Urgency, frequency, dysuria 2. Suprapubic pain/tenderness 3. Flank pain or tenderness 4. New onset delirium 5. Fever >100.4 F/Rigors 6. Acute hematuria 7. Increased spasticity or dysreflexia in a spinal cord injury patient 8. > 2 SIRS criteria (T > 38.5 C or < 35 C, HR > 90, RR >20 or PaCO2< 32 mmHg, WBC >12 K/mm 3 or <4 K/mm 3 or > 10% bands) YES NO Send U/A & urine culture Document indication for sending urine culture Do NOT send urine culture Start empiric therapy (see reverse side) * Symptom based screening is not reliable in the following cases: pregnancy, prior to urologic procedures, patients with complex urinary anatomy (i.e., nephrostomy tubes, urinary tract stents, h/o urinary diversion surgery in the past, or renal transplant), patients admitted to the ICU, or neutropenia. Use your clinical judgment for this population. 11
10/14/2016 The Community Teaching Hospital • ID doc who had done research previously led work • Asked hospitalists to watch video of Dr Hartley’s talks • Gave them UMHS pocket cards • Offered “authorship on a paper” for the hospitalist lead who would help make sure everyone watched the video Treatment of Asymptomatic Bacteriuria 100 % ASB Receiving Antibiotics 90 80 70 79 73.8 60 * * 65 62 50 57 Pre 53 40 Post 30 20 10 0 Overall Hospital #1 Hospital #2 * p<0.05 Data modified from original: to make a point! 12
10/14/2016 Key Characteristics of the Champion • Role model for the change – Respected / passionate • Collaborating, commit resources and attention • Communicate throughout work • Have a clear process • Be consistent in behaviors Identifying the Champion / Leader • Passion: ideally they bring the project forward • Subject matter expert • Outstanding clinician / teacher / role model • Effective communicator • Prior improvement experience a plus, but not required • Time for the work 13
10/14/2016 Overview • Need to Improve-External Forces • Common Pitfalls in hospital QI / Case studies – Leaders and champions – Data – Infrastructure – Culture / Socioadaptive elements Total Inpatient HH compliance 14
10/14/2016 MFH Specific Data Barriers and solutions • Closely partnered with IPE personnel to enhance monitoring of HHH Barrier Solution Paucity of Data 2 observers dedicated to hospitalist service Non-specific Data • Unable to ID Hospitalist Shared schedule and pictures of hospitalists • Unit Based observers Covert observers shadowed hospitalists Lack of Awareness Educational sessions, Audit/Feedback, Physician champions Unknown Modes of failure Identified clinical situations with higher failure rates 15
10/14/2016 Hospital Medicine Focused HH Interventions • Educational sessions: – Importance of HH in preventing infections – Review of incidence of MRSA, VRE, CDI rates • Audit/Feedback – Service level-target problem areas – Physician level-target problem areas • Physician Champions • $$$$ (QI Incentive) Overall Hospitalist Hand Hygiene Compliance Overall HHH Compliance 100 90 90 % (190/211) 80 Increased HM Data collection 70 60 50 45 % (28/62) 51 % (14/27) 40 30 HM focused Intervention 20 10 0 Jan - March 2015 Apr-15 Feb- May 2016 System Wide HH Intervention (Clean/Remind/Thank, etc.) 16
10/14/2016 Facilitating Data Collection • Use Data Already Being Collected – Hand hygiene – HCAPS – CAUTI / CLABSI / C.Diff – ED wait times – Readmission rates – Any existing EMR data field • Capture during the new process – 72 hour antibiotic timeout – (challenge is “pre-” data) Facilitating Data Collection • Manual medical record review (if all else fails) – Use standard data audit forms – Make it easy , no judgment; ideally non-healthcare providers can collect data – Samples vs. Consecutive pts • Useful for change over time • High volume conditions / data elements • Group level metrics rather than MD-specific 17
Recommend
More recommend
