OPPORTUNITIES FOR DATA INTEGRATION AND BEST PRACTICE INTERVENTIONS TO IMPROVE CLINICAL AND FINANCIAL OUTCOMES

OPPORTUNITIES FOR DATA INTEGRATION AND BEST PRACTICE INTERVENTIONS TO IMPROVE CLINICAL AND FINANCIAL OUTCOMES Elizabeth S Miller, MPA November 2014 President BPKMiller Associates 1 2 3 HEDIS DATA IMPROVEMENT: Improve HEDIS Administrative Rates Optimize Medicaid HEDIS Pregnancy Measures QI INTERVENTION BEST PRACTICES: Provider Sourced Reminder Program Office Staff Incentives DATA INTEGRATION IMPROVING POPULATION HEALTH MANAGEMENT Member Contact Information Disparity Index Tool Integrating Clinical and Non-Clinical 1 1

1 HEDIS DATA IMPROVEMENT 2 IMPROVE PREVENTIVE HEALTH SCREENING RATES BY IDENTIFICATION OF SERVICES RECEIVED AND NOT REFLECTED IN HEDIS RATES Opportunity: Are you Optimizing administrative HEDIS rates? Seeing competitors in a shared network reporting higher rates? Hearing contracted providers reporting that their members are receiving services that are not reflected in the rate reports? Noting staff reporting that medical record review in the provider s office confirms care? Premise: Physicians treat their patients same regardless of payer, but HEDIS rates vary Improvement in HEDIS rates critical for Medicare Stars rating, Medicaid state-specific bonus/sanction, and business retention and growth Accurate HEDIS rates are important for assessing quality of care, ensuring the data is credible and usable by provider groups, and in contracting and physician incentive programs. Only medical record review accurately identifies where care is rendered. Matching to Claims and HEDIS reports can identify where discrepancy lies. Especially for administrative rates, where HEDIS specs do not allow for MRR to supplement claims information, the impact on more complete and accurate reporting can not be understated. 3 2

IMPROVE PREVENTIVE HEALTH SCREENING RATES BY IDENTIFICATION OF SERVICES RECEIVED AND NOT REFLECTED IN HEDIS RATES Action: Obtain and abstract medical records for administrative records, according to HEDIS specifications Identify areas of opportunity to capture data that is currently being missed, e.g., Member didn t receive service within HEDIS time requirements Office bill didn t include required codes or didn t reflect service Payer claims system didn t capture code HEDIS reporting system didn t reflect service Address the opportunities Focused Medical Record Review will: Confirm service was provided and within the requirements of HEDIS specifications, e.g., date, age, etc. Identify any measure-specific exclusions removing the member from the HEDIS report Provide the information required to verify: Correct coding on bill from provider office of service rendered Complete capture of information required in MCO claims system Accurate HEDIS report reflecting the care provided 4 IMPROVE PREVENTIVE HEALTH SCREENING RATES BY IDENTIFICATION OF SERVICES RECEIVED AND NOT REFLECTED IN HEDIS RATES Results: Improved HEDIS rates Improved future claims-based reporting by ensuring that Client Claim system accurately and completely captures information required for HEDIS reporting Education of providers in proper coding/billing requirements Improved knowledge of providers of Client s expectation that Members will receive all appropriate preventive services, and understanding of the HEDIS specifications as they relate to this reporting Compliance by health plan with state and federal requirements relative to improvement of preventive health rates Closer collaboration with Physician Groups in improving rates by providing accurate complete reports on preventive care 5 3

USING A LAB CODE NOT IN THE HEDIS PREGNANCY Measure Specifications To Help Improve The HEDIS Pregnancy Measure Rates (and reduce NICU stays) Opportunity: Need for early identification of pregnant members. Premise: Earlier identification allows for risk assessment, provision of preventive health services and case management to reduce likelihood of delivering a premature or atrisk baby that requires NICU care. Why this is important: Prenatal care is a highly visible indicator of quality of life for the Medicaid membership, and state sanctions/ incentives reflect that priority Estimated if only 1 high risk pregnancy out of the pregnancies newly identified each month with this innovative data improvement is THEN provided high risk case management and AVOIDS a NICU stay for that baby that results in an improved outcome for mom and baby, AND a savings of at least $100,000 per month. Premature infants have a greater likelihood of having continued medical problems If one birth per month can be impacted to prevent a NICU stay, the savings would also result in subsequent reduced medical care in the first years of life and beyond Positive impact and reduced costs for the mother as well, improved quality of life for child, mother, family. 7 6 HEDIS PREGNANCY MEASURE PROJECT SUMMARY Health plan identified a best practice from the Center for Health Care Strategies to increase the rate of identifying pregnant members earlier in their pregnancy Health plan and health plan s operations development department replicated this documented best practice using the pregnancy lab panel code (80055) to identify more members for the pregnancy risk assessment team at health plan to outreach and assess risk so that care management services could be provided to reduce the risk to the mother and baby Health plan identified nearly 1,000 more pregnant members each month compared to prior identification approach for risk assessment by health plans pregnancy outreach program using this best practice data improvement 8 HEDIS Measures impacted: Timeliness of Prenatal Care Frequency of Prenatal Care PostPartum Care 7 4

2 QUALITY IMPROVEMENT BEST PRACTICES 8 Health Plans Partnering with Physicians and Provider Groups to Outreach to Members (rather than only outreach from the Health Plan) is a part of the Future of Healthcare Quality Improvement 9 5

SUPPORT PROVIDER COMMUNICATION WITH PATIENTS Action: Coordinated education program to PCP s patients in need of targeted services via Letter (signed by PCP & Health Plan) and Call (using voice of PCP) Focus on large practices in need of rate improvement that require support in identifying patients and services Provide to PCP member-and measure-specific reports to be used at time of office visit Coordinated Provider Outreach will: Focus on members in need of specific services Utilize a different approach to member communication that is more effective than mass, impersonal mailings Support the physician in improving the patient/physician relationship 10 SUPPORT PROVIDER COMMUNICATION WITH PATIENTS Results: Improvement in HEDIS rate compliance for those physicians/practices participating in the program Enhanced patient/physician relationship based on the specificity of the letter to the member and the direct involvement of the physician in the reminder Education of practice/physician in the HEDIS measurement requirements, including time and condition sensitive information Improved member satisfaction with health plan and PCP Compliance of health plan with any applicable state and federal requirements relative to improvement of preventive health rates 11 6

ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF Is it more efficient and effective to outreach to 200,000 members, or to 40 doctor offices who care for those 200,000 members? A health plan in Tennessee analyzed Medicaid members monthly EPSDT screening rates and found rates ranging from 48% to 54% *AHIP Innovations in Medicaid Managed Care Report 12 ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF Health Plan offered parents and guardians $5 Wal-Mart gift card for receiving EPSDT services 13 7

ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF To get the $5 gift card the Members had to: Make the appointment Get transportation to the appointment Keep the appointment Remember the postcard from the TN health plan when they went to the appointment Get their doctor s signature on the postcard from the TN health plan to prove they had the screening visit Mail the doctor-signed postcard to the TN health plan AND THEN wait for the mail to receive the $5 gift card 14 ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF Health Plan also offered $5 Wal-Mart gift card to Office Staff in PCP offices of same members offered $5 gift card incentive. Recognized that office staff and office managers are key to opening door to improved HEDIS rates and member satisfaction. 15 8

ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF Office Staff manage and coordinate schedules with patients and doctors Patients listen to them and talk with them more than anyone from any health plan Office Staff at network doctor offices Called members who had not received recommended EPSDT screenings Placed EPSDT reminders for doctors in members charts and completed and sent in EPSDT forms to TN health plan Received gift card incentive after they completed and submitted member-specific data collection forms to health plan Incentive Difference: Office Staff received a $5 gift card for each member s completed EPSDT screening visit 16 ACTIVATING MEMBERS BY INCENTING THEIR DOCTOR S OFFICE STAFF Results: Member gift card incentive resulted in 770 visits Doctor office staff incentive resulted in 23,814 visits TN health plan s EPSDT visit rate increased from 54% to 73% 17 9

3 DATA INTEGRATION IMPROVING POPULATION HEALTH MANAGEMENT 18 UPDATED MEMBER CONTACT INFORMATION Threat Member contact info incorrect or outdated Members missed and mistakenly thought to be uninterested, or at least not contacted Care coordination not provided to members in need Lower quality care compliance (HEDIS, Stars), complications resulting in hospitalization, higher MLR Opportunity Member contact info corrected and updated with non-health plan data Increase % of members successfully contacted Improved % of engaged members to provide care coordination Improve care coordination and support services for members, reduce MLR, improve satisfaction with health plan 19 10

IMPACT CULTURAL DISPARITY IN QUALITY CARE DELIVERY By Improving Cultural Relevance of Member & Provider Outreach Threat Disparity is quantity that separates a group from a reference point on a particular measure of health, expressed in terms of a rate, proportion, mean, or other quantitative measure. (Healthy People 2010) Outreach efforts not culturally sensitive or relevant; missed opportunities to effectively outreach and communicate with members. Changes in HEDIS or Star rates (or lack of changes/trend) may mask changes at race, ethnicity, or primary language levels of population groups in a health plan/aco membership Cultural communication or perception factors unique to race, ethnicity, or primary language of members. Conclusions regarding member receptivity to outreach efforts is potentially misread, since may not be wrong message to activate members, but cultural context of message off the mark Opportunity Tool measures both absolute and relative disparity, and variance from quality goals to determine largest gaps in care and area(s) where interventions may have most impact on HEDIS scores. Disparity: HEDIS rate performance difference between racial/ethnic groups using bestperforming group as point of reference Quality Variance: difference between HEDIS target rates and rates of each racial/ethnic group using state-mandated rate, or HEDIS percentile goal rate, as benchmark. Quality performance provides context for group disparities Tracking both disparities and quality variance over time provides a view to impact of quality improvements on disparities between racial/ethnic groups, and to ensure that disparities are not improved/reduced at cost of reducing quality of care. 20 DISPARITY INDEX AND QUALITY VARIANCE Example: Cervical Cancer Screening Ethnicity Group REL-Specific HEDIS Rate REL Population (members) Black or African 90.12% 30,750 American White 86.48% 67,422 Hispanic 85.96% 17,856 American Indian 82.01% 567 & Alaskan Native Unknown 83.55% 15,461 Some Other Race 81.41% 7,585 Asian 75.82% 6,150 21 11

DISPARITY INDEX TOOL Helps Plans Prioritize Quality Improvement Project Goals Monroe Plan for Medical Care (New York): measure and report reducing neonatal intensive care unit admissions in 2005 from 11% to 4.9% among babies of African-American teen mothers. Molina Healthcare (Michigan): increase childhood immunization rate for African Americans from 38.3% in 2004 to 58.4% in 2006. Blue Cross of California: increase African Americans' use of its Asthma personalized education program from 0% to 15% in eight non-chain pharmacies. UPMC for You (Pennsylvania): decline in low-birth weight deliveries for African-American women in Braddock County from >20% in 2004 to 0% in 2005, and decline from >20% to 8% among all women in Braddock. L.A. Care Health Plan (California): enhance delivery of pharmacy services to racial and ethnic minorities by providing language labeling in 10 most common languages of its members to 500 pharmacies. 22 INTEGRATING CLINICAL AND NON-CLINICAL DATA FOR A MORE COMPLETE PICTURE OF YOUR POPULATION 23 12

Clinical and Non-Clinical Sources Holistic Info from Health Plans/ ACOs to Physicians and Members Analysis Feed Summary Analysis to Vendor Partners 24 The Picture of Your Members/Patients/Customers Is Incomplete Clinical and Non- Clinical Sources Lots of, But Missing Links NHS UK video: Better Information Means Better Care http://www.youtube.com/watch%3fv%3d3pkb65rwzvi&sa=u&ei=7wxyu_spcca2yatdp4kgbg&ved=0cbwqtwiwaa&usg=afqjcneha4inciej96edunpm16rbx9qixg 25 13

Member Enrollment data: Race, Ethnicity, Primary language only available for Medicaid, Medicare Member Health Risk Assessments, Satisfaction Surveys: Available on growing number of members, but only partially leveraged, not integrated with other sources Provider networks data: Contact information, languages spoken, credentialing status data continually needs to be updated Utilization Management data: Only Partially Leveraged, needs IT Integration with from Other Sources Case Management data: Detailed data, not fully integrated with data from other sources Pharmacy Claims: Most timely and complete data, but more valuable when combined with other data Professional and Institutional Healthcare Delivery System claims: Claims data from delivery systems/ hospitals, but other services and medical record information not provided on claims Lab data: Hospital, physician practice labs not all captured Clinical and Non- Clinical Sources 26 Health Management Vendors: Includes Behavioral Health entities, yet Proprietary Info, HIPAA, IT Integration issues State/local Health Departments data: Only partially leveraged, varying levels of bureaucracy IT Integration Member Electronic Records: Available at limited # of physician offices, HIPAA, IT Integration Issues Clinical and Non- Clinical Sources Additional data sources to enhance the patient picture: Food/OTC pharmacy purchases Internet providers - Member s health-related information seeking data Medical device usage data Geomedicine Alert sent to member Don t eat there! http://www.zeit.de/datenschutz/malte-spitz-vorratsdaten Twitter crime analytics 27 14

Clinical and Non-Clinical Sources Holistic Info from Health Plans/ ACOs to Physicians and Members Analysis Feed Summary Analysis to Vendor Partners 28 Holistic Info from Health Plans/ ACOs to Physicians and Members Perspective Demographically and clinically comprehensive Likelihood to respond to engagement with culturally-relevant information and language preferences Within and across: Physician practices Populations Lines of business Benefits Holistic info from health plans/ ACOs to physicians and members Enables more accurate and customized clinical outreach and management approach for members, AND information sharing with physicians Utilizes information gathered on each member to quantify their level of engagement, combined with both clinical and social risk elements that provide level of risk, AND nuances about receptivity to outreach and activation response history to engagement from health plan Allows for sharing of this information to physicians to give them a similar holistic perspective on their patients clinical status and activity Supports physicians assessment of their practice performance and identify opportunities to improve by patient as well as across the practice, by disease or condition. 29 15

Results Holistic Info from Health Plans/ ACOs to Physicians and Members Case Management Programs focused on coordination of services and care needs of members Risk Information and Status via portal for physicians and health management partners, empowering them with information focused on coordination services and care needs of members Risk and activation-adjusted member engagement programs for members to obtain preventive health services, screenings, and to continue medical management of existing conditions 30 Analysis Predictive Clinical and Social Risk Identify Opportunities for Caregiver Support Proactive Non-Clinical Risk Management Race, Ethnicity, and Language Disparity with Quality Variance Drivers 31 16

Predictive Clinical and Social Risk Analysis Healthcare decisions or behavior are not made in a sterile environment Predictive modeling forecasting accuracy varies from 0%-35%, with some higher exceptions for specific populations (e.g., Medicaid) In context of other information/data points, helps to prioritize outreach, contact, BUT NOT an answer unto itself. While this predictive risk metric may help prioritize member to outreach/or type of care coordination/action, how much it helps is secondary to the completeness, accuracy, and the accurate integration of your other clinical and demographic data sources. 32 Identify Opportunities for Caregiver Support Analysis Threat Members need care management but lack support network Health plan does not know member has minimal support network Impact to members on independence, self-care, behavioral health Care management efforts undermined by lack of social or relative support network Feedback loop of health status decline, higher MLR Opportunity Identify member s relatives within specific proximity (e.g., 20 miles) Assess risk based on member s support by relatives measured by geographic proximity Care management outreach prioritization and approach dynamically adjusted based on member support network risk Improve care coordination with support services for members, assess impact on MLR and satisfaction with health plan 33 17

Proactive Non-Clinical Risk Management Analysis Threat Members who have experiences not measured in clinical data, but impact health and are documented in public records Health plan does not know member has high risk behaviors if not diagnosed or documented in claims data Behaviors may impact member s care plan, family members, community Care Management efforts undermined by incomplete picture of the member s comprehensive personal risk behavior Lack of feedback loop of health status decline, resulting in missed opportunities to help member Opportunity Integrate public record data with member s clinical history Develop more comprehensive risk assessment, prioritize care and outreach Identify additional resources otherwise not known to be needed using only claims data Develop holistic risk summary based on Care Management outreach prioritization and approach Improve care coordination and support services for members, as well as caregivers/family members 34 Race, Ethnicity, & Language Disparity w/ Quality Variance Drivers Analysis Knowing the Disparity Rates of your membership will help to focus and to direct messaging in a more accurate, impactful, and cost effective manner. This includes: RACE ETHNICITY LANGUAGE Locate Geographic pockets with Disparities for direct member outreach Identify Barriers to getting needed medical care based on these Disparities and address same in communications Tailor written and telephonic communications, both language and content, to the identified Disparities 35 18

Clinical and Non-Clinical Sources Holistic Info from Health Plans/ ACOs to Physicians and Members Analysis Feed Summary Analysis to Vendor Partners 36 Feed Summary Analysis to Vendor Partners Member Health Risk Assessments, Satisfaction Surveys Provider Networks contractual performance Utilization Management outcomes and performance Case Management performance and outcomes Pharmacy outcomes Professional and Institutional Healthcare Delivery Systems contractual performance Laboratory Test Result outcomes Health Management Vendor performance, contractual and clinical outcomes State/Local Health Departments for population profile and outcomes 37 19

Recap/Questions? 1. Non-HEDIS Lab Code to ID Pregnancies Earlier http://www.chcs.org/usr_doc/towardimprovingbirthoutcomes.pdf 2. Member address corrections to improve member outreach engagement Lizzy Feliciano, Director, LexisNexis Leonides.Feliciano@lexisnexis.com 3. Disparity and Quality Variance Information http://www.chcs.org/?s=racial+%26+ethnic+health+disparities 4. Doctor office staff incentive https://www.ahip.org/innovations-in-medicaid-managed-care-report/ 5. Integrating for a Better Picture of Your Population German cell phone tracking map Twitter crime analytics NHS UK video: Better Information Means Better Care 38 20