ED Throughput Improvements Dr. David Allard, Chief Medical Information Officer Dr. Bruce Muma, Chief Medical Officer, Henry Ford Physician Network
Problem Statement The Henry Ford Accountable Care Organization (HF ACO) is participating in the CMS Next Generation Model which offers significant upside/downside financial risk for total cost of care and measurable quality of care The primary challenge (problem) is to reduce non value added medical expense while simultaneously improving preventive care and chronic disease care. The secondary challenge is to organize the beneficiary data into a valid registry within the EMR/EDW to facilitate population analytics and develop real time interventions.
Problem Statement Analysis of data sources led the HF ACO to focus on three hot spots of value improvement opportunity: High Risk Populations The top 30% based on future risk predictive models using claims and EMR data. The waste is driven by discontinuity and inadequate clinical/social supports Variation in Medical Decision Making Focus on high cost medical decisions (global). The waste is driven by variation in risk aversion and inconsistent application of standard protocols Post Acute Care Focus on SNF expenses. The waste is driven by inconsistent adherence to communication protocols, variability in care planning, and absence of provider oversight.
Design and implementation The ACO is governed by a Board of Trustees and managed by a team of dedicated Population Health division staff who are accountable to the board for achieving performance targets The Population Health staff developed a comprehensive work plan in late 2015 with a focus on reducing total cost of care, improving selected quality measures, engaging providers and patients, communication and building analytic tools with dashboards to drive continued success. The population was analyzed using a projected beneficiary attribution model derived from both EMR and EDW data elements. This model allowed us to estimate financial and utilization performance in real time and identify hot spot opportunities.
Design and implementation Specific interventions were developed for each of the hot spots utilizing the ACO registry and creating patient lists for targeted interventions. The following interventions were developed and launched: High Risk Populations Top 5%: Comprehensive Care Centers Top 30%: Universal Case Management Program Variation in Medical Decision Making Hospital Admission Decision in ED: EDS Program Specialty Referral Decision: Referring Wisely Program Targeted treatment/testing: Choosing Wisely BPA Program Post Acute Care SNF length of stay/transition pathways: PAC Surveillance
EDS Program Objective: provide alternative pathways for ED physicians to avoid hospitalization of patients not meeting IQ criteria for admission Scope: all HF ACO, HAP MA and HAP commercial patients presenting to the HFH/HFWBH/Fairlane ED Intervention: EDS navigator resides in the ED and is equipped with tools to orchestrate OPD care pathways (~15 pathways) Technology: EMR track board identifies targeted patients for EDS navigator EMR flow sheets to track intervention/impact Patient scheduling software to facilitate appointments/treatments EDW database to track utilization impact and potential harm
Design and implementation (how we used technology) ACO Registries were established for HAP Medicare Advantage and NextGen ACO. Patient lists are received from the insurer, matched with HFHS MRN and then automatically moved into Epic using a datalink process on an ongoing basis CMS sends periodic updates on enrollees in the program which sometimes need to get back dated to January or sometimes only affect membership going forward Enrollees are compared to known patients with a matching algorithm using demographics.
Registry population CMS update files can be additions or removal updates which must be back dated to the beginning of the year Expired patients are removed from reports as of the date of death Updates are moved up to the EMR to keep registries current Transactional patient data fed to EDW Registries updates based on upload CMS ACO Update files received by EDW Files matched to MRN with weighted algorithm Patient updates migrated to EMR
Registries used to alert users of Status
ED Trackboard Alerts
Creating actions
How was HIT utilized 2 Registries used to drive improvements in CMS hcc capture Multi provider schedules shows the number of conditions that need refreshed Best Practice Advisories automatically fire for the providers and identify the specific conditions that need to be addressed
Dashboard Reports are available for various populations by Provider, Department and Care Team member In addition to PCPs, Ambulatory Case Managers, Population Health Coaches and Diabetes Educators can follow their own panels of patients
Risk Scoring Tools are available to identify patients that may need more intensive management Chronic Disease Registries allow CareTeam members to focus on specific cohorts of patients this example is CHF patients and is stratified by risk score
Allignment of clinical data with registries allows comprehensive dashboard reporting
Self Serve Access to Near Real Time Data Mart On the fly queries (one day delay in data) Allows identification of patients and immediate bulk actions
EDS Results 29% of all ED visits at participant hospitals by target patient population were reviewed by EDS Navigator To date 235 admission avoided (IPD and OBS) which is equivalent to an avoidance rate of 4.9%. To date 141 additional patients had interventions which provided direct benefit (potential avoidance of future admission or other additional costs associated with delays).
All ED Navigator Impact on Disposition (08/13/2016 09/02/2017) IPD avoided IPD downgraded to observation Observation avoided 25 20 Number of Cases 15 10 16 19 13 12 10 5 0 10 9 13 13 9 6 4 5 5 3 3 1 3 1 3 2 2 2 1 1 1 1 1 1 Aug 2016 Sep 2016 Oct 2016 Nov 2016 Dec 2016 Jan 2017 Feb 2017 Mar 2017 Apr 2017 May 2017 Jun 2017 Jul 2017 Aug 2017
All ED Population (08/01/2016 08/01/2017) HF ACO Patients HFACO Patients HAP Commercial Patients Number of Patients 1800 1600 1400 1200 1000 800 600 400 200 0 421 321 307 271 260 275 677 143 498 472 447 485 448 362 261 270 245 237 401 492 551 475 518 606 156 475 159 206 242 225 211 Aug 2016 Sep 2016 Oct 2016 Nov 2016 Dec 2016 Jan 2017 Feb 2017 Mar 2017 Apr 2017 May 2017 Jun 2017 Jul 2017 All ED Arrival Trends (08/13/2016 08/31/2017) Average number of Patients per hour 3.0 2.5 2.0 1.5 1.0 0.5 0.0 2.7 2.7 2.5 2.6 2.6 2.5 2.3 2.2 1.8 1.9 1.8 1.6 1.3 1.3 1.1 0.9 0.8 0.8 0.7 0.6 0.5 0.5 0.5 0.4 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Admits/1000 (YTD) 400 390 380 370 360 392 384 392 391 384 374 369 366 365 364 362 361 388 364 380 373 374 368 350 340 336 330 320 310 300
All ED Navigator Facilitated Interventions: Impacted Disposition Scheduled Diagnostic Test (blank) Facilitated direct admission to a Skilled Nursing Facility Referred Patient to Community Paramedicine Referred to Diabetes Care Center DME PACE 1 1 1 4 3 3 3 15 20 25 28 0 10 20 30 40 50 60 Count of Interventions 37 43 51 Percentage of Patients with Revisits to Any ED within 72 hours (where Navigator Impacted Disposition OBS & IPD Avoided) No Revisits to ED 86% Revisits to ED 14%
Comprehensive Care Centers Objective: Provide a more effective care model for high risk patients to reduce total cost of care, hospitalizations, ED visits and improve health and quality of live. Scope: All HF ACO, HAP MA and HAP commercial patients living within the 10 mile radius of the 2 sites (Taylor and DNW) Intervention: Enroll targets patients in the program which offers comprehensive support with 24/7 availability of providers, case managers, virtual behavioral health services, pharmacy support, home care support, longer visits, management of disease exacerbations in clinic (vs ED/OBS).
Comprehensive Care Centers (CCC) Technology: Ongoing surveillance of EDW to identify high risk patients and initiate review and recruitment processes (Optum One) EMR embedded flow sheets and intake processes for CCC staff Patient portal services to promote communication with patient/caregiver Virtual video services EDW database and dashboard to track utilization impact and potential harm
CCC: an at risk population
CCC Results 20% reductions in ED usage and Admission rates
CCC Results Overall 4% reduction in PMPM charges
PAC Surveillance (PACS) Objective: Conduct ongoing review of care plans for target patients who are admitted to a SNF and provide support for transitional care back to patient centered medical home Scope: all HF ACO, HAP MA and HAP commercial patients admitted to facilities who belong to the HF preferred SNF network Intervention: PACS case managers conduct outreach (phone, onsite visits, chart review) in real time for patients admitted to the SNF
PAC Surveillance Technology: EDW registry to identify target patients HIE tracking system to identify target patients admitted to selected SNF s EDS track board to identify potential candidates for direct to SNF transfer Flow sheet tools to document intervention and report to PCP EDW analytic tools to track length of stay and admission rates.
PAC Graceful transfer to SNF
PAC: Facilitating PCP followup
SNF Admits/1000 (YTD) PACS Results With real time monitoring of clinical events, SNF usage and length of stay can be closely monitored and optimized 300 250 200 150 100 205 277 263 249 236 228 222 217 212 209 207 205 193 191 201 205 204 199 171 50 0 Average Length of Stay (Preferred SNF vs. Non Preferred SNF) 25.0 20.0 15.0 10.0 5.0 0.0 Jan 2016 Feb 2016 Mar 2016 Apr 2016 May 2016 Jun 2016 Jul 2016 Aug 2016 Sep 2016 Non Preferred 14.0 16.1 13.8 14.3 13.9 14.8 16.7 14.3 15.0 15.7 16.4 15.7 15.2 13.8 16.2 16.0 14.2 14.9 15.0 Preferred 12.2 14.7 17.3 16.5 14.0 16.6 16.3 13.6 16.0 13.4 13.4 13.4 14.9 13.9 19.5 15.2 14.4 14.3 Oct 2016 Nov 2016 Dec 2016 Jan 2017 Feb 2017 Mar 2017 Apr 2017 May 2017 Jun 2017 Jul 2017