Managing Rising-Risk Patients in a Value-Based World

Transcription

1 Managing Rising-Risk Patients in a Value-Based World Katie Nelson, MA Ed Rafalski, PhD, FACHE Sandra Bailey, FACHE Razvan Marinescu, MD, FACHE Session 6A Monday 3/16 10:00 AM 11:30 AM Session 6B Monday 3/16 2:15 PM 3:45 PM

2 Agenda Learning objectives How did we get here? Congregational Health Network Participatory Hot Spotting Memphis Model Familiar Faces Methodology and Familiar Faces demographics and statistics Results Rising Risk Methodology Goals Bios and Contact Information Bibliography Appendix 2

3 Learning Objectives Gain knowledge about addressing health issues in high-risk patient populations plagued by chronic co-morbid disease in a value-based care world. Learn about implementing a patient-centric population health management program for a high-risk population that decreases health system costs. 3

4 4 HOW DID WE GET HERE?

5 Segment Patients by Risk Level The most successful population health managers don t build one model for one population. They segment patients by risk level to manage three distinct populations, each requiring different goals, resources, and care delivery models. Source: Advisory Board Playbook for Population Health Building the High-Performance Care Management Network 5

6 How Did We Get Here? Congregational Health Network (2013) In 2006, Methodist Le Bonheur Healthcare (MLH) created the Congregational Health Network which works closely with clergy in the most under-served zip codes of the city to improve access to care and overall health status of the population. CHN locations in and around Memphis, TN 6

7 Participatory Hot Spotting The Mid-South has some of the highest prevalence of chronic disease: heart disease, stroke, lung disease, cancer, diabetes, and asthma. In an effort to identify ways to improve the health of its community, MLH used geocoding technology to identify hot-spots of healthcare utilization. The goal was to identify geographic areas of focus on which to direct hospital resources in a targeted effort to improve the health of the neediest communities. The result was the identification of zip code in South Memphis. Patients from had the highest utilization of MLH emergency departments (EDs) as well as the highest consumption of hospital charitable care. Source: T. Cutts, E. Rafalski, C. Grant and R. Marinescu, "Utilization of Hot Spotting to Identify Community Needs and Coordinate Care for High-Cost Patients in Memphis, TN," Journal of Geographic Information System, Vol. 6 No. 1, 2014, pp doi: /jgis IP and OP visits & variable cost by block group for zip codes: 38109, 38126, 38106, 38132,

8 Participatory Hot Spotting (continued) These techniques were coupled with the community health needs assessment process at MLH and qualitative, participatory research findings captured in collaboration with church and other community partners. The methodology, which we call participatory hot spotting, is based upon the Camden Model, which leverages hot spotting to assess and prioritize community need in the provision of charity care, but adds a participatory, qualitative layer, represented by the CHN. 8

9 Focus Area Memphis Model Spatial analysis was employed to evaluate hospital-based inpatient and outpatient utilization and define costs of charity care for the health system by area of residence. The top ten zip codes accounted for 56% of total system charity care costs. The hot spot of utilization and cost was found to be South Memphis in one zip code, In 2010, in this zip code, IP volume accounts for 9% of visits, while representing almost 65% of total cost has a high percentage of under-served persons and has only one FQHC safety net clinic, serving roughly the 49,000 residents in the zip comprises 14% of the total Memphis population and is 97% African American. Managing patients at the neighborhood level should have the biggest impact on the charity care population. These findings were combined with grassroots intelligence that enabled a partnership with clergy and community members and Cigna Healthcare to better coordinate care in a place-based population health management strategy. 9

10 40% of all Charity care is in the blue, so are more than half of CHN congregations. T or

11 Overview Of MLH Top 10 ED Patients In ZIP (2010) Patients 2010 Visits (MHS and Christ Community Health Loop CHN Church in Visits* Age MUH combined) in Area in Area Area Hospital Patient # MHS Patient #2 25 y y Bloomfield Baptist MHS Patient #3 23 Y Y Mt. PisgahM.B.C MHS Patient #4 22 y y Maranatha Faith MHS Patient #5 21 y y Mt. Vernon Baptist MUH Patient #6 16 y y Mt. Vernon Baptist MHS Patient #7 12 y y Bloomfield Baptist MUH Patient # MUH Patient #9 10 y y Rising Sun MHS Patient #10 9 y y Mt. Vernon Baptist MHS Patients Main reason for ED visits Comorbidities Patient #1 Pain Yes Depression Patient #2 Pain Yes Depression Patient #3 Alcohol intox No Mental ilness Patient #4 COPD related Yes No Patient #5 Suicidal ideations No Depression/Bipolar Patient #6 Back pain Yes Depression/Bipolar Patient #7 CHF/Chest pain Yes No Patient #8 Chronic Pain Yes Mental ilness Patient #9 Sore throat Yes No Patient #10 Dizziness Yes No *Three-year visit trend shows only the main location for the visits, if visits where at more than one hospital. Data Source: MLH Ascent Mental/Psych Story Essential Service Needed Homeless Self pay until 2011, Medicare since Homeless?/Polysubstance abuse Painkiller request Stopped taking medication BCBS, 2010 self pay, 2011 TennCare The Healing Center The Healing Center The Healing Center CCHS on Third The Healing Center The Healing Center CCHS on Third The Healing Center; CCHS CCHS on Third CCHS on Third 11

12 12 FAMILIAR FACES

13 Overview As a first step in addressing the health needs of 38109, MLH launched an on-going, innovative community health navigator pilot program Familiar Faces in January The pilot program provides additional, non-clinical support to the most frequent users of MLH EDs and tests the impact of navigator intervention on improving health behaviors and appropriate healthcare utilization among the Familiar Faces. The most common diagnoses for our Familiar Faces (FF) a group of about 100 patients are heart failure, COPD, diabetes, hypertension and chronic kidney disease. Patients with chronic kidney disease and hypertension have the highest hospital encounter rates in this cohort. Additionally, this group has a shockingly high 2013 allcause readmission rate approaching 60% at MLH hospitals 13

14 Definition Pilot Cohort: Familiar Faces Patients originating from zip code with 11 or more ED visits in 12 month period (May 2012 April 2013) Excludes pediatric patients 14

15 Goals Improve chronic disease self-management. Redirect patients to the most appropriate point of care thereby reducing ED encounters, IP readmissions and, as needed, increasing primary and specialty care physician visits. Test hypothesis that socio-economic and social factors are hindering the Familiar Faces ability to manage their health and chronic disease and more important, the provision of a trusting and knowledgeable navigator can actually improve the health outcomes of these patients. 15

16 Navigator Provides non-clinical support to overcome the socio-economic barriers to good personal health and chronic disease management. This support ranges from: scheduling appropriate physician appointments arranging transportation to and from appointments securing a warm meal or groceries getting prescriptions filled, financial aid for prescriptions and more Partners with community churches in this effort to further involve community stakeholders and engage community resources. 16

17 How it Works ~100 of the most frequent users of MLH EDs are assigned to the Familiar Faces navigator. When a patient in the Familiar Faces (FF) program has an encounter at a MLH hospital, the electronic medical record (EMR) sends a notification to the navigator. The navigator meets the patient in the ED or in the hospital if he/she is admitted. The navigator is responsible for building a relationship based on trust with the patient. The goal is to create a partnership between the navigator and the patient, identify the underlying causes for frequent ED use and developing an action plan to change the individual s health behaviors. 17

18 38109 Household Income Average household income in is just over half the national average Demographic Characteristics Shelby County USA 2010 Total Pop 46, , ,745, Total Pop 46, , ,861, Total Pop 46, , ,322,277 % Change % 3.0% 3.3% Average HH Income $38,687 $64,934 $69,637 Data Source: Truven Health Analytics, Demographics Expert 10/28/

19 38109 Demographics Income and Race Distribution 1 in households makes less than $15K per year; one of the poorest zip codes in Shelby County. $75-100K 6% $50-75K 14% Over $100K 5% $25-50K 32% Income Distribution <$15K 25% $15-25K 18% Pop Hispanic 1% All Others 1% White Non- Hispanic 2% 96% of is African American. Black Non- Hispanic 96% Data Source: Truven Health Analytics, Demographics Expert 10/28/

20 May 12 Sep MLH Visits 26,313 patients living in choose MLH; Equivalent to 57% of total population Data Source: Ascent,; May Sep 2013 I/P Visits O/P Visits (includes ED visits) ED Visits MUH 1,888 10,662 4,947 MSH 3,639 31,089 25,311 MNH MGH 496 2,771 1,161 MOBH MFH MECH 84 N/A N/A Le Bon ,101 3,761 TOTAL 6,681 55,512 35,898 20

21 38109 MLH Payer Mix TennCare is largest payer in MLH patients originating from Payor # of Visits UniquePts (MRN) T ENNCARE 26,061 10,735 MEDICARE 14,254 4,936 PRIVAT E PAY 9,477 5,330 CONTRACT/MANAGED CARE 8,376 4,221 BLUE CROSS 2,657 1,454 COMMERCIAL LIABILIT Y WORKMAN'S COMPENSATION MEDICAID MEDASSIST CHAMPUS SPONSOR AGENCY METHODIST STAFF SERVICES Data Source: Ascent, Costflex May Sep

22 38109 Familiar Faces Age and Race Distribution The largest age group in Familiar Faces is 30-39; 97% are African American Age Distribution Familiar Faces # of Patients Race Distribution Familiar Faces # of Patients < > Black 3 White Data Source: Ascent, Costflex May Sep 2013 Note: Age distribution patient total is greater than 97 because several patients moved between groups during the specified analysis period (May 12- Sept 13) 22

23 38109 Familiar Faces Visits by Patient Type Data Source: Ascent, May Sep Familiar Faces visit our EDs on average about once per month Familiar Faces Visit Totals by Patient Type 372 I/P # of Visits 1910 O/P I/P Familiar Faces Patient Type Detail Maternity I/P Newborn AmbSurg # of Visits 1757 ED O/P 106 O/P 33 Recur O/P 23

24 38109 Familiar Faces ED Visits Frequency 46/97 Familiar Faces visited an MLH ED more than 12 times from May 2012 Sep 2013 Three patients have been seen in a MLH ED over 75 times from May 2012 Sep 2013 Two of the three exceeded 90 ED visits during the period. One of whom insured by Medicare, the other uninsured. Data Source: Ascent, May Sep Familiar Faces By ED Visit Frequency # of ED Visits May '12-Sep '13 12 >12/Period >25/Period >50/Period >75/Period

25 38109 Familiar Faces Payer Mix The largest payer group is TennCare, covering 66 lives # of Visits by Payer Familiar Faces TENNCARE UHCCOMMUNITYPLA TENNCARE BLUECARE TENNCARE SELECT MEDICARE UHC DUAL COMPLE MEDICARE AMERICHOICE MEDICARE PART A ONLY MEDICARE WINDSOR HMO MEDICARE A & B COVERAGE DISC 60% UNINSURED DISC LOCATED IN SELFPAY FIRSTSOURCE/DISABILITY Private Pay{Sybase Blank SRC/STRATEGIC RESOURCE C ***GENERIC PLAN*** LIABILITY / GENERIC HOSPICE MANAGED CARE UNITED HEALTHCARE / CIGNA /FLEXCARE BCBS /OTHER STATES CIGNA /PPO MEDICAID ILLINOIS MEDICAID SOUTH CAROLINA ***GENERIC PLAN*** MEDICAID TEXAS TENNCARE MEDICARE OTHER COMMERCIAL MEDICAID # of Lives = 66 # of Lives = 30 # of Lives = 29 # of Lives = 10 # of Lives = 4 Data Source: Ascent, May Sep

26 38109 Familiar Faces Costs Total Visits Total Charges Total Cost Tot Act Reimb 2,282 $14,577,584 $3,439,109 $3,424, Familiar Faces account for $3.4M Total Cost. Top 15 highest cost patients contribute 50% of total group cost; 8 of which are TennCare patients. Medicare reimbursement floats this group. Majority of cost comes from the IP volume. Data Source: Ascent, Costflex May Sep

27 Comorbidities and Readmissions 97 Familiar Faces (FFs) from Zip Code were identified using Ascent Report Timeline: May 2012 September 2013 MRNs of the 97 Familiar Faces were used to extract All Encounters Data with up to 40 Secondary Diagnosis Codes from CERNER using Power Insight Reporting Data was processed to represent one row per encounter 3 Major Categories: ALL (n=2250), IP (n=351) and ED (n=1618) High Risk Comorbid Conditions: Identified from research articles & papers published in PubMed-NCBI and Elsevier Conducted analysis on: ALL (n=2250), IP (n=351) and ED (n=1618) encounters Readmissions Analysis (FFs): Conducted on 325 records out of 351 IP encounters after exclusions Readmissions Analysis (Entire Population): Conducted on entire Methodist Population (4 adult facilities) taking exclusions into account for the same timeline 27

28 Total Encounters: 97 Familiar Faces Encounters for the FF cohort cluster around ED and IP, with the majority in the ED. Comorbidities analysis is conducted on 3 groups: 1. All Encounters = 2, ED encounters = 1, IP encounters = 351 Readmissions analysis is conducted on inpatient encounters after the hospice and same day readmissions were excluded. Encounter Type Encounters (#) Encounters (%) Emergency (ED) 1, % Inpatient (IP) % Observation % Outpatient % Labor & Delivery % ED Alternative- LB % RPRecurring % Ambulatory Surgery % Recurring Outpt % FPRecurring % Newborn % NUCMED % Skilled Nursing % Sleep Disorders % Grand Total 2, % 28

29 Comorbidities Distribution by Primary Discharge Diagnosis (All Encounters) 50% of the encounters had between 1 and 5 comorbid conditions associated with the principal diagnosis. 2 diagnosis codes associated with Chronic Kidney Disease stood out as the top 2 categories in the Other category. The maximum numbers of high-risk comorbidities were noticed in the case of those who were discharged with a Primary Dx of Chronic Lung Disease (CLD), Heart Failure, Diabetes and Chronic Kidney Disease. 83% of the encounters with a primary (principal) diagnosis code of Heart Failure (HF) had at least 2 or more highrisk comorbidities. # of Comorbidities Primary Diagnosis Total Other ,720 CLD HF Diabetes Sickle PNE ESRD HTN GED AMI CAD Drug Stroke Alcohol HIV Total # 1, ,250 Total % 49.1% 18.9% 14.7% 8.1% 4.9% 2.6% 1.1% 0.5% 0.1% 23 highlighted by the Orange Box means that there were 23 encounters in which patients discharged with a primary (principal) diagnosis code of HF had 4 comorbidities. 29

30 Comorbid Conditions as a Secondary Diagnosis Affecting the Familiar Faces Population The top 5 comorbid conditions are present in more than 70% of the encounters, regardless of patient type. These patients have more than 1 high risk comorbid condition regardless of their principal diagnosis. The Ratio of Comorbidities to Encounters is as follows: All Encounters = 1.16 Inpatient (IP) = 2.95 Emergency (ED) = 0.79 This indicates IP encounters had an average of 2.95 comorbidities. Timeline: May 2012 September 2013 Comorbid Condition (Secondary Dx) All Encounters IP Encounters ED Encounters (n = 2,250) % (n = 351) (n = 1,618) Hypertension % Diabetes Mellitus % Heart Failure % Chronic Lung Disease % End-Stage Renal Disease % Gastroesophageal Reflux Disease % Chronic Obstructive Pulmonary Disease % Coronary Artery Disease % Drug Dependence % Obesity - Familial, Morbid, Severe % 68 5 Human Immunodeficiency Virus % 11 6 Sickle Cell Disease % 0 12 Peripheral Vascular (Arterial) Disease % 1 8 Alcohol Dependence % 9 2 Acute Myocardial Infarction 8 0.3% 3 5 Failure To Thrive 6 0.2% 4 0 Stroke (Ischemic & Hemorrhagic) 3 0.1% 3 0 Peptic Acid Disease 3 0.1% 2 1 Peptic Ulcer Disease 3 0.1% 3 0 Dementia 1 0.0% 1 0 Total High Risk Comorbidities (Secondary DXs) 2, % 1,036 1,273 Though we had only 2250 All encounters, the Total number of High Risk Comorbidities came out to be 2611 which indicates that some encounters had multiple diagnosis codes related to the same disease condition. How many times is a comorbid condition present as a secondary dx code in the population, regardless of the principal dx. 30

31 Familiar Faces 30-Day Readmission Rate (IP) FF population has a higher readmit rate than the system population. The highest readmission rate burden is for HF patients. The 30-day Medicare HF Readmissions Rate is reduced by 3.15% at Methodist South Hospital.when the encounters of the 97 FFs were excluded from the entire Methodist Healthcare Memphis Hospitals (MHMH) population. This clearly shows the disproportionate impact of FF readmission rates on the system average. MHMH = Methodist Healthcare Memphis Hospitals or System or 4 adult facilities EP = Entire Methodist Population (4 adult facilities MUH, MSH, MNH and MGH) FFs = 97 Familiar Faces identified from Zip Code EP FFs = Entire Methodist Population minus encounters of the 97 Familiar Faces identified from Zip Code Burden = EP (EP-FFs) May 2012 September 2013 Financial Class = All Payer Readmission Cause FFs EP EP-FFs Impact AMI 66.67% 11.84% 11.76% 0.08% HF 57.53% 20.99% 20.16% 0.83% Pneumonia 37.50% 15.21% 15.14% 0.07% All Cause 52.38% 12.90% 12.72% 0.18% Financial Class = Medicare Readmission Cause FFs EP EP-FFs Impact AMI % 14.78% 14.69% 0.09% HF 19.91% 21.60% 20.89% 0.71% Pneumonia 33.33% 15.47% 15.43% 0.04% All Cause 58.06% 15.90% 15.68% 0.22% 31

32 Familiar Faces Program Is a Success Goals: Improve patient health behaviors and chronic disease management skills through community health navigator intervention and support Navigate patient to the most appropriate point of care Decrease preventable hospital encounters or readmissions Decrease hospital cost of patient care Results Community Health Navigator the Intervention is working Program began January out of 94 Familiar Faces touched 28 out of 94 Familiar Faces followed Per Patient Cost: down 43% as of Nov YTD 2014 Hospital Encounters (inpatient, outpatient, ED) significant decrease: Aug YTD 2014 average monthly IP encounters down to 10.3 compared to 19 per month in 2013 (for entire Pilot Cohort) Aug YTD 2014 average monthly ED encounters down to 76.6 compared to 99 per month in 2013 (for entire Pilot Cohort) 32

33 Familiar Faces Report Card YTD November 2014 Through November 2014 total costs per patient in the program have decreased by 43% compared to the baseline. Hospital utilization has decreased most significantly in the ED (23% on average for the cohort). Interestingly, the average length of inpatient stay has stayed flat. It may suggest that admitted patients are truly in need of hospital based care and are not being admitted as a result of poor management of chronic disease which presumably would require a shorter hospital stay Familiar Faces Report Card Baseline JAN 14 FEB 14 MAR 14 APR 14 MAY 14 JUN 14 JUL 14 AUG 14 SEP 14 OCT 14 NOV YTD Progress # of Patients in Familiar Faces (first day of month/year) N/A # Expired Patients (last day of month/year) N/A # of Patients w ith MLH Encounter N/A ENCOUNTER DATA Monthly Avg YTD Avg IP Visits/Month (total = 228) ALOS (Total Days=1072 ) OP Visits/Month (non-ed) (total=94) ED Visits/Month (total =1193) All Visits/Month (total =1515) FINANCIAL DATA YTD Total Total Charges (Data updated each month for 2014 cells) $9,868,763 $808,476 $601,172 $673,937 $472,541 $587,292 $321,079 $481,483 $472,449 $432,410 $366,484 $380,142 $5,597,466 Total Cost (Data updated each month for 2014 cells) $2,619,457 $225,509 $156,299 $152,896 $115,637 $141,073 $61,651 $98,929 $105,650 $90,479 $74,977 $72,059 $1,295,159 Cost/Patient (DMAP Metric) $2,416 $2,783 $1,928 $1,762 $1,330 $1,629 $715 $919 $1,214 $1,040 $862 $828 $1,369 % Cost Savings/Patient compared to 2013 Baseline 15% -20% -27% -45% -33% -70% -62% -50% -57% -64% -66% -43% *Progress: Calculated based on YTD data; Red indicates unintended direction, green indicates intended direction IP Visits/ Month - Baseline = 2013 Monthly Average; YTD = 2014 Monthly Average LOS - Baseline = 2013 Monthly Average; YTD = 2014 Monthly Average OP Visits/MonthBaseline = 2013 Monthly Average; YTD = 2014 Monthly Average ED Visits/Month - Baseline = 2013 Monthly Average; YTD = 2014 Monthly Average All Visits/Month - Baseline = 2013 Monthly Average; YTD = 2014 Monthly Average Pediatric patients (5) removed in May and revised in JUNE Only adult Familiar Faces remain. One pt was miscoded with disp code 20. Familiar Face members adjusted to correct. 33

34 Is the MLH Community Navigator Program Working Outside MLH Walls? Are Familiar Faces visiting other Memphis hospitals outside the MLH network? If so How frequently prior to the navigator intervention? How frequently following the navigator intervention? Is the program truly meeting its goals? Or are patients going to other facilities? 34

35 Mid-South ehealth Alliance Provides Answers Mid-South ehealth Alliance (MSeHA) patient level database of hospital encounters for Memphis hospitals that excludes financial indicators, ie. hospital cost. Allows us to answer the question, Are we meeting our goals when accounting for patient activity outside MLH? Hospital encounter data for Familiar Faces extracted from MSeHA for CY2014. Average cost per ED or IP encounter computed at patient level for Familiar Faces (using MLH financial data). Average cost per encounter at the patient level applied to visits outside MLH during

36 MSeHA Results 17 Familiar Faces were found to have at least one encounter outside MLH network in CY2014 (Source: MSeHA Sep 2014): All encounters were ED visits, except one IP visit Only 7 patients were found to have an encounter outside the MLH network after the navigator intervention: The program appears to be increasing patient loyalty to MLH 36

37 Intervention Has Positive Impact In and Out of MLH Network 2013 Aug YTD 2014 Conservative Estimate: program has decreased per patient cost by 35.2% (Aug 2014 YTD). Navigator intervention is positively impacting patient behavior in and out of MLH network. Baseline data for 2013 excluded MSeHA cost estimates for Familiar Faces making the above cost savings a conservative estimate actual savings are likely to be significantly higher. MLH Cost $2,619,457 $1,057,643 Non-MLH Cost $40,017 TOTAL Cost $1,097,660 MLH per patient Cost $2,416 $1,540 Total per patient Cost (MLH + Non-MLH) $1,565 Non-MLH Cost Pre-Intervention $31,751 Non-MLH Cost Post-Intervention $8,266 Non-MLH Cost Total $40,017 Non-MLH Encounters Pre-Intervention 104 Non-MLH Encounters Post-Intervention 28 Non-MLH Encounters Total

38 Contribution Factor Used to Identify Rising Risk Patients Chronic Kidney Disease and Hypertension patients have highest utilization rates. Principal Discharge Diagnosis Encounters - IP & OP (Unique Patients) Contribution Factor The contribution factor can help us identify the focus of our intervention within the familiar faces population. Initial results show these two comorbidities are the most likely disease states to result in frequent ED and IP visits. Chronic Kidney Disease 35 (4) 8.75 Hypertension 83 (10) 8.3 Diabetes 59 (11) 5.36 HF 74 (18) 4.11 COPD 47 (16) 2.94 AMI 3 (3) 1 PNE 8 (8) 1 STK 2 (2) 1 Other 1252 (94) Total 1563 (94) Contribution Factor = Ratio of the number of encounters to the number of unique patients. Data Source: Power Insight 2013 Discharges Note: 1) 311 encounters out of 1563 have a principal diagnosis of something other than "Other". 2) These 311 encounters came from 39 unique patients. Of these: 1 patient was discharged with 5 different principal diagnosis codes, 3 patients with 4 different diagnosis codes, 6 patients with 3 different diagnosis codes. 38

39 Familiar Faces Executive Summary The 97 FFs (people with encounters >= 11) identified from contributed 2,250 encounters (Inpatient = 351 (15.60%), Emergency=1,618 (71.91%)) between the time period of: May 2012 September The largest age group in Familiar Faces is 30-39; 97% are African American Familiar Faces visit our EDs on average about once per month. 46/97 Familiar Faces visited an MLH ED more than 12 times from May 2012 Sep The largest payer group is TennCare, covering 66 lives. 50% of the encounters had between 1 and 5 comorbid conditions associated with the principal diagnosis. The top 5 comorbid conditions (secondary dx) are present in more than 70% of the encounters, regardless of patient type. The maximum numbers of high-risk comorbidities were noticed in the case of those who were discharged with a Primary (Principal) Diagnosis Code of: Chronic Lung Disease (CLD), Heart Failure, Diabetes Mellitus (DM) and Chronic Kidney Disease. 39

40 Familiar Faces Executive Summary 31 % of the encounters with a primary (principal) diagnosis code of Chronic Lung Disease (CLD) had at least 2 or more high-risk comorbidities. 37% of the encounters with a primary (principal) diagnosis code of Diabetes Mellitus (DM) had at least 2 or more high-risk comorbidities. 83% of the encounters with a primary (principal) diagnosis code of Heart Failure (HF) had at least 2 or more high-risk comorbidities. The 30-day Medicare HF Readmissions Rate is reduced by 3.15% at Methodist South Hospital when the encounters of the 97 FFs were excluded from the entire Methodist Healthcare Memphis Hospitals (MHMH) population Program started in January As a result of the intervention, per patient cost: is down 43% as of Nov YTD The contribution factor shows that two co-morbidities, Chronic Kidney Disease and Hypertension, are the most likely disease states to result in frequent ED and IP visits. 40

41 41 RISING RISK

42 Rising Risk Results suggest the community navigator pilot program is working. However, we now want to expand the program to a larger population of patients in those that are at risk of becoming a Familiar Face. We believe that the next step is to extend and strengthen our support for these high/rising risk patients by: 1) identifying patients with the potential of becoming a Familiar Face by expanding navigation support to include patients who are referred to home health services for ongoing care post hospital stay and 2) working more closely with patients who are registered members of the Congregational Health Network. 42

43 Congregational Health Network The Congregational Health Network is a partnership between Methodist Le Bonheur Healthcare and participating congregational and community health organizations in Memphis designed to improve access to healthcare and health status. These covenantal partnerships have grown to over 599 church partners, 10 navigators, trained congregational liaisons and more than 20,000 registered members. Of the 5,886 patients (members of CHN congregations) who have had hospital encounters, 891 (15%) live in the zip code. Of these patients, 30% have had readmissions within 30 days of discharge. 43

44 How We Plan To Make It Work Form a collaborative relationship between the community navigator, the existing home health liaisons and the Congregational Health Network navigators, in order to meet the social and medical needs of these patients. Additional support outside the hospital walls will allow participating patients to be cared for in their homes with wrap-around support from a trusted navigator who can also seek and provide additional services as required. A community navigator would work closely with home health agencies who are assigned to home health patients in to reinforce the care of those who are identified as Rising Risk patients. The navigator will become a member of that patient s existing home health care team and provide the social and personal support, similar to that provided to the Familiar Faces. This navigator will visit the patient in the home and the combination of home health and individual navigation will provide a more comprehensive wrap-around outpatient care network. The navigator can also supplement the care which is sometimes important to enabling the rising risk patient to actively engage in managing their chronic conditions. 44

45 Rising Risk Identification/Methods To identify the Risking Risk (R 2 ) cohort we begin by analyzing the emergency department (ED) encounters over an 18-month period. We then isolate adult-only patients that originate from 38109, our hot spot zip code. Using Pareto analysis, we then segregate those patients that are considered a high utilization cohort, >=6 ED encounters over an 18-month period. We remove those patients who expired during the study period and further refine the cohort to isolate Familiar Faces (FFs) from those who are R 2, >=6 and <=10 ED encounters. Once the R 2 cohort has been identified, we stratify the group by encounter classification: ED, Inpatient (IP) and Observation (OBS). We remove patients diagnosed with HIV for patient confidentiality and proceed to score the three groups based on their comorbidities, utilization patterns and acuity/readmission/length of stay (LOS). 45

46 Goals Building upon our success in the Familiar Faces effort to date, we propose to expand our reach to begin to prevent or delay patients from becoming a Familiar Face to Methodist - we will use essentially the same model employed with trusted navigation provided to Rising Risk patients. Our metrics for success will be three-fold: 1) Reduction in ED visits, hospital admissions & cost. 2) Increasing the days between visits to an acute care provider. 3) Management of chronic disease through successful use of tele-health products and personal support by a navigator by employing a case/control cohort prospective study design. 46

47 Methodist Healthcare Rising Risk Methodology Total Encounters* through the Emergency Departments (Timeline: January 2013 June 2014) *The encounters may remain as Emergency (ED) or may change to Inpatient (IP) or Observation (OBS) or Outpatient (OP) at the time of discharge. In other words, we are looking at all the encounters whose entry point to MLH is through the Emergency Department. Total ED/IP/OBS Encounters through the Emergency Departments at the Adult Facilities** **MUH, MSH, MNH, MGH, MFH, and MOBH Total Encounters originating from ZIP Code Excluded 93.8% of patients had 5 or fewer encounters 1. Identification of the final cohort for Rising Risk analysis 6.2% of patients had 6 or more encounters Excluded Expired: Died during an ED/IP/OBS encounters 6 or more Encounters Alive Excluded Cases/Intervention: Cohort 1 (97 FFs) Encounters from non-familiar Faces Excluded Control/Cohort 2 FF: (>= 11 encounters) 2014 Copyright Methodist Healthcare Memphis, TN Cohort 1 RR: (6 10 encounters) 47

48 Methodist Healthcare Rising Risk Methodology Cohort 1 RR: (6 10 encounters) Total ED encounters that remained ED encounters at the time of discharge Total ED encounters whose status changed to IP encounters at the time of discharge Total ED encounters whose status changed to OBS encounters at the time of discharge ENCOUNTER CLASSIFICATION METHODOLOGY Patients may have had multiple encounters i.e., ED encounter that remained an ED encounter at the time of discharge, ED encounter that got re-classified as an IP encounter or as an OBS encounter at the discharge. 1. Identification of the final cohort for Rising Risk analysis (continued) In order to avoid over-counting of the patients, we created the following criteria: 1) If a patient had both ED and IP encounters then the final patient classification = IP 2) If a patient had both ED and OBS encounters then the final patient classification = OBS 3) If a patient had all the encounter types then the final patient classification = IP 4) If a patient belongs to only one category the final patient classification remains the same ED Patients IP Patients OBS Patients Exclusion: HIV/AIDS Patients RAW ED Cohort for Risk Analysis RAW IP Cohort for Risk Analysis RAW OBS Cohort for Risk Analysis 48

49 Methodist Healthcare Rising Risk Methodology 2. Scoring methodology for the identification of Rising Risk Population) 49

50 Methodist Healthcare Rising Risk Methodology 2. Scoring methodology for the identification of Rising Risk Population) Once the Rising Risk population is determined through the above methodology, the contribution factor presented earlier can help us identify patients with the comorbidities that are the most likely disease states to result in frequent visits and thus enable us to maximize our intervention within this cohort. 50

51 Epilogue Secure funding from our payer partner Replicate cohort in case control study with University of Memphis and University of Tennessee Create a health disparities coordination council (a multidisciplinary cross-functional team) to supercharge this work 51

52 52 QUESTIONS

53 Bios and Contact Information Katie Nelson, MA Katie Nelson is the senior analyst in Business Development at University of Minnesota Health a partnership between the University of Minnesota Physicians and Fairview Health Care. In this role, she is responsible for market analytics and decision support for University of Minnesota Health Cardiovascular, Oncology, and Women s and Children s Services in addition to other academic programs of distinction. In her previous role in Business Development, Planning and Research at Methodist Le Bonheur Healthcare in Memphis, TN, she was responsible for performing environmental analyses for business development opportunities and supporting MLH physician alignment. She also supported the market research team and was involved in patient experience survey implementation, analysis and process improvement for MLH adult hospitals. Ms. Nelson has a background in strategic planning, data analysis, market research, survey methodology, quantitative and qualitative methods, and GIS. She received a BA in political science from the University of Wisconsin - Madison and a MA in political science from the University of Memphis. Sandra Bailey, FACHE Sandra.Bailey-DeLeeuw@mlh.org Sandra Bailey is Vice President of Senior Care Services and CEO of Methodist Extended Care Hospital. She is responsible for system initiatives in readmission, community health, and care transitions (Case Management). Ms. Bailey rejoined Methodist in 2002 as Administrator of the Methodist Healthcare-managed UT Bowld Hospital after a oneyear tenure as Chief Operating Officer at Emergency Coverage Corporation, headquartered in Knoxville Tennessee. She initially joined Methodist Healthcare as the Administrator of Methodist Healthcare - Brownsville Hospital from Prior to this time, she served in various healthcare leadership roles in both east and west Tennessee hospital systems. Ms. Bailey holds a Bachelor of Arts degree in Audiology from the University of Tennessee and a Master of Science Degree in Health Systems Engineering from the Georgia Institute of Technology. She is currently taking coursework toward a PhD in Public Health at the University of Memphis. Bailey is a member of the American College of Healthcare Executives and a former National Board member of the Healthcare Information and Management Systems Society. She is a graduate of Leadership Memphis, Chief Volunteer Officer of the Greater Memphis YMCA, an active member of Collierville United Methodist Church and actively involved in various community service groups. 53

54 Bios and Contact Information Razvan Marinescu, MD, MHA, FACHE Razvan Marinescu is currently Director of Planning and Business Development for Methodist Le Bonheur Healthcare (MLH) in Memphis, TN. In this role he is responsible for strategic planning and business development at MLH. Most recently he was in an Administrator role for a large multi-specialty hospital-owned physician practice at Memorial University Medical Center in Savannah, GA. Dr. Marinescu has a background in patient care, data analytics, strategic planning, physician practice administration, hospital operations, decision support, GIS, and business development. He has also served as an examiner for the state of Tennessee Baldridge program, the Tennessee Center for Performance Excellence. After graduating from Carol Davila University of Medicine and Pharmacy in Romania with a Doctor of Medicine degree, Dr. Marinescu received a Master s degree in Health Administration from Ohio University. He received his Six Sigma Green Belt certification from the Methodist Quality Institute at Methodist Le Bonheur Healthcare, and his Lean Six Sigma Black Belt certification from the Memphis Lean Six Sigma Institute. Dr. Marinescu is a Fellow of the American College of Healthcare Executives. Edward M. Rafalski, PhD, MPH, FACHE Ed.Rafalski@mlh.org Edward Rafalski is currently Senior Vice President of Strategic Planning & Marketing for Methodist Le Bonheur Healthcare, Memphis TN. In this role he is responsible for all planning, business development, market research, marketing, web services, public relations and communications at MLH. Most recently he was Vice President of Marketing for Alexian Brothers Hospital Network, Arlington Heights, IL. Dr. Rafalski has an extensive background in strategic planning, data analytics, decision support, GIS, business development, marketing, public relations, group purchasing and managed care contracting. He has also served as an executive liaison for emergency department and children s hospital operations. After graduating from the University of Chicago with a bachelor s degree in public policy studies, Dr. Rafalski received a master s degree in public health from Yale University School of Medicine. He received his Ph.D. in public health sciences from the Division of Health Policy and Administration at the University of Illinois, School of Public Health where he has taught as Clinical Assistant Professor. His health services research and teaching interests include: the effects of market economics on health care services, healthcare decision support, quantitative methods, health disparities, marketing and strategic management. He recently joined the faculty of The University of Tennessee Health Science Center, College of Medicine in the Department of Preventive Medicine as Adjunct Associate Professor and The University of Memphis School of Public Health as Affiliate Research Professor where he is teaching managerial epidemiology. Dr. Rafalski is a Fellow of the American College of Healthcare Executives and Six Sigma Black Belt. 54

55 Bibliography T. Cutts, E. Rafalski, C. Grant and R. Marinescu, "Utilization of Hot Spotting to Identify Community Needs and Coordinate Care for High- Cost Patients in Memphis, TN," Journal of Geographic Information System, Vol. 6 No. 1, 2014, pp doi: /jgis Advisory Board; Playbook for Population Health Building the High-Performance Care Management Network Bending the Cost Curve and Improving Healthcare Quality in One of America s Poorest Cities, Department of Health and Human Services, Best Practices for Health Systems in the Field. Jeff Brenner, CEO Camden Coalition; Kelly Craig, Director of Care Management Initiatives; Aaron Truchil, Data Analyst Planning and Managing Community Health Through Hotspotting, 2012 Annual SHSMD Conference Presentation, Teresa Cutts, PhD and Razvan Marinescu, M.D., MHA, FACHE Successful Strategies for Managing Community Health and Charity Care, 2013 ACHE Congress on Healthcare Leadership, Razvan Marinescu, MD, MHA, FACHE, Associate Director, Strategic Planning and Business Development, Methodist Le Bonheur Healthcare, Memphis, TN; Ed Rafalski, PhD, MPH, FACHE, Senior Vice President, Strategic Planning and Marketing, Methodist Le Bonheur Healthcare, Memphis, TN "The Memphis Congregational Health Network Model: Grounding ARHAP Theory When Religion and Health Align: Mobilizing Religious Health Assets for Transformation, edited by James R. Cochrane, Barbara Schmid, and Teresa Cutts. Pietermaritzburg: Cluster Publications, pp , Gawande, Atul - The Hotspotters Agency for Healthcare Research and Quality (AHRQ) Innovation Profile: Church-Health System Partnership Facilitates Transitions from Hospital to Home for Urban, Low-Income African Americans, Reducing Mortality, Utilization, and Costs. Teresa Cutts, Bobby Baker and Gary Gunderson, Chief Innovators. 55

56 56 APPENDIX

57 Comorbid Conditions Acute Myocardial Infarction (AMI) Heart Failure (HF) Pneumonia (PN) Septicemia Respiratory Failure Chronic Lung Disease (CLD) CMS Readmission Focus Chronic Obstructive Pulmonary Disease (COPD) (Asthma is a subset) Coronary Artery Disease (CAD) Stroke (STK) or Cerebrovascular Accident (CVA) Peptic Acid Disease (PAD) Peptic Ulcer Disease (PUD) Gastroesophageal Reflux Disease (GERD) Hypertension (HTN) Diabetes Mellitus (DM) Human Immunodeficiency Virus (HIV) End-Stage Renal Disease (ESRD) Alcohol Dependence Drug Dependence Dementia Frailty Failure to Thrive Obesity - Familial, Severe, Morbid Peripheral Vascular (Arterial) Disease (PVD) Sickle Cell Disease (SCD) 57 57

58 List of Disease Conditions for Rising Risk Comorbidities Identification Cardiac Arrhythmia Acute Myocardial Infarction (AMI) Heart Failure (HF) Pneumonia (PNE) Septicemia Respiratory Failure Asthma COPD Chronic Bronchitis Stroke (Ischemic & Hemorrhagic) Diabetes Hypertension Coronary Artery Disease Chronic Kidney Disease Kidney Problems Renal Failure Schizophrenia Depression Lupus Cardiomyopathy (Non-ischemic) Cardiomyopathy (Ischemic) Hypothyroidism Liver Disease Sickle Cell Disease Frailty Failure To Thrive Obesity Alcohol Abuse Drug Abuse Peripheral Vascular Disorders Valvular Disease Pulmonary Circulation Disorders Paralysis Other Neurological Disorders Peptic Ulcer Disease excluding bleeding GERD Lymphoma Metastatic Cancer Solid Tumor without Metastasis Rheumatoid Arthritis/collagen Coagulopathy Fluid and Electrolyte Disorders Blood Loss Anemia Deficiency Anemia Pre-Diabetes Pre-Hypertension 58