1. Association of Asian Pacific Community Health Organizations
Collecting social determinants of health data Using PRAPARE to reduce disparities, improve outcomes and transform care
Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences
Presented by:
Tuyen Tran
Association of Asian Pacific Community Health Organizations
April 24, 2017
This project was made possible with funding from:
Collecting Social Determinants of Health Data Using P.R.A.P.A.R.E. to Reduce Disparities, Improve Outcomes, and Transform Care 
© National Association of Community Health Centers, Inc., Association of Asian Pacific Community Health Organizations, Oregon Primary Care Association. PRAPARE and its resources are proprietary information of NACHC and its partners, intended for use by NACHC, its partners, and authorized recipients. Do not publish, copy, or distribute this information in part of whole without written consent from NACHC.

2. PRESENTATION OBJECTIVES
Present AAPCHO’s collaboration in PRAPARE
Discuss the importance of documenting and addressing social determinants of health for improving patient and population health
Describe the PRAPARE protocol and how it was developed and implemented in four different Electronic Health Records at health centers nationally
Learn about free resources and information about how to get started in data collection on the social determinants of health

3. Association of Asian Pacific Community Health Organizations (AAPCHO)
Non-profit, national association established in 1987 with a mission to improve health status and access of Asian Americans, Native Hawaiians and other Pacific Islanders.
Represents 35 community health organizations, mostly health centers serving primarily underserved Asian Americans, Native Hawaiians and other Pacific Islanders (AA&NHPI).
BPHC funded National Cooperative Agreement holder providing T/TA to health centers serving AA&NHPI.
The Association of Asian Pacific Community Health Organizations (AAPCHO) is a national association of 35 community health centers and organizations serving primarily Asian Americans, Native Hawaiians, and other Pacific Islanders. We are dedicated to promoting advocacy, collaboration, and leadership that improves the health status and access of these medically underserved communities. Together, our CHC members serve over 500,000 patients nationally and are at the forefront in providing community responsive, financially affordable and culturally and linguistically appropriate primary health care services.

4. Storytelling with our data SELECTED SDOH DATA

5. Average AA&NHPI Patients at AAPCHO Health Centers, 2015
29 AAPCHO members, 1375 national FQHCs
Data Source: UDS 2015

6. AAPCHO* HEALTH CENTER PATIENTS BY RACE, 2015 (% KNOWN)
AAPCHO member HCs served over 530,000 patients
AA&NHPI patients accounted for almost 70% of those served
NATIVE HAWAIIANS
11.8%
PACIFIC ISLANDERS
3.7%
Guamanian/Chamorro
0.1%
Micronesian
0.6%
Other Pacific Islanders
0.8%
Samoan
2.0%
Tongan
0.2%
ASIAN GROUPS
65.2%
Cambodian
0.6%
Chinese
50.7%
Filipino
4.6%
Japanese
0.7%
Korean
1.2%
Laotian
0.2%
Mien
Mixed AA&NHOPI
Other Asian
3.7%
Thai
0.1%
Vietnamese
2.6%
*AAPCHO has 29 AA&NHOPI-serving FQHCs

7. AAPCHO HEALTH CENTER PATIENTS BY LIMITED ENGLISH PROFICIENCY (LEP), 2000-2015
More than half of AAPCHO member HC patients were best served in a language other than English
For some AAPCHO members, LEP patients represented up to 100% of their patient populations
English
39.3%
Cantonese
22.5%
Mandarin
22.3%
Unknown
4.8%
Chinese
3.7%
Vietnamese
2.2%
Other
2.3%
Korean
0.9%
Cambodian
0.4%
Spanish
0.5%
Tagalog
0.2%
Samoan
Burmese
0.1%
Hakka
Japanese
Karen
Mien
Laotian
Chuukese
<.10%
Indonesian
Malay
Thai
Tongan
*AAPCHO has 29 AA&NHOPI-serving HCs

8. AAPCHO members had a higher proportion of geriatric (age 65+) patients from 2007-2015
VL, please update.

9. AAPCHO and national HCs’ proportions of uninsured patients have been on a declining trend while Medicaid has simultaneously risen.

10. SELECTED CLINICAL DATA

11. SELECTED CLINICAL DIFFERENCES
AAPCHO( %)
National ( %)
% Hepatitis B
2.6%
0.2%
% TB
0.09%
0.04%
% Abnormal breast findings
0.83%
0.6%
% Abnormal cervical findings
0.94%
0.8%

12. SELECTED QUALITY OF CARE MEASURES
AAPCHO ( %)
National (%)
Cervical cancer screening
58.3%
56.0%
Hypertensive patients with patients with controlled BP
67.6%
68.7%
Diabetic patients with HbA1c > 9
26.1%
29.8%
Notice AAPCHO has lower percentage with controlled BP

13. SELECTED ENCOUNTER & CO$T DATA

14. AAPCHO had a higher average number of enabling service (ES) encounters compared to the national average

15. In 2015, average AAPCHO Enabling Services FTEs were higher than the national average overall and particularly for case management, patient education, and interpretation
VL update this slide with 2015 data

16. AAPCHO total costs for enabling services per HC were higher than national average
VL update this slide with 2015 data
Statistically significant

17. Average AAPCHO medical costs per medical encounter were similar to the national average
AAPCHO centers spend significantly more per pt and per encounter so we REALLY need the data on SDH and ES to justify why we are doing so! Look what the UDS data says. How do we defend without data?
No significant difference, t-test for medical costs per medical encounter.
Average costs per patient therefore much higher and statistically significant.
The graph is the average cost/MEDICAL encounter, which doesn’t include ES etc. It’s really just the medical staff and Lab. For some reason AAPCHO CHCs still have a higher cost for medical encounters. Also, the results of t-tests don’t really matter here, because we are not comparing another sample to the national HCs, we are really just comparing AAPCHO members (which weren’t randomly selected) to the national HCs. Higher average for AAPCHO members is higher average no matter statistically significant or not. T-test is telling us what’s the chance to have randomly selected a sample like the AAPCHO members from the national FQHCs (significant result means not likely, so the sample couldn’t have been selected from the national FQHCs, so the sample must have some different characteristic).
AAPCHO Members ($)
National HCs ($)
Costs per patient
1,030.45
826.84
Costs per encounter
215.45
207.21

18. CONTRIBUTION OF NON-CLINICAL SERVICES
AAPCHO centers have better health outcomes compared to national averages for pap test, hypertension, and diabetes. This may be due to the higher number of enabling services and resulting quality outcomes provided by AAPCHO centers compared to national health centers.
This is what the UDS is showing, we need data to show it. We can’t make this conclusion with just UDS data, need pt level data.

19. ENABLING SERVICES (ES)
non-clinical services that are provided to health center patients that promote, support and assist in the delivery of health care and facilitate access to quality patient care NACHC/MGMA 2000

20. AAPCHO’S ENABLING SERVICES ACCOUNTABILITY PROJECT (ESAP)
PARTICIPATING CENTERS:
Charles B. Wang Community Health Center NY, NY
International Community Health Services Seattle, WA
Kalihi-Palama Health Center Honolulu, HI
Waianae Coast Comprehensive Health Center Waianae, HI
This project developed out of the recognition by AAPCHO health centers that enabling services play an important role in the health of their patients, many of whom are immigrant, and face cultural and linguistic barriers to care.
We believe that if these services are valued and if we can show this with our data, then adequate and sufficient funding should be explicitly directed to pay for them, to assure they are provided where and when needed in sufficient volume and with appropriate staffing levels.
Further, the lack of data on enabling services impedes the ability to quantify the need and demand for services, or to demonstrate their effectiveness.
So in 2004, we embarked on a pilot project to develop a standardized data collection model for enabling services.
AAPCHO, in collaboration with 4 CHCs pilot sites, NACHC and other experts, implemented findings of the MGMA/NACHC Health Center Enabling Services Report and developed a practical system for CHCs.  To address differences in staffing and IT capacity among multiple CHCs, we applied a simpler enabling services tracking system.  

21. LACK OF COMPREHENSIVE ES DATA ON UDS
UDS TABLE 5- STAFFING AND UTILIZATION
Personnel by Major Service Category
FTEs
(a)
Clinic Visits
(b)
Patients
(c)
Case Managers
Patient/Community Education Specialists
Outreach Workers
Transportation Staff
Eligibility Assistance Workers
Interpretation Staff
Community Health Workers
Other Enabling Services
So you’re thinking, but aren’t we already required to submit ES data to UDS?
HCs are required to submit some information on their ES to the BPHC as part of their annual Data System reports but the UDS poorly tracks the scope of ES actually provided and needed by the health center. The current UDS fails to demonstrate the critical impact of these services and the need to adequately finance them, requiring them to report partial information about ES staffing for case managers, patient/community education specialists, outreach workers, transportation staff, eligibility assistance workers, interpretation staff and other enabling services. UDS also provides a limited accounting of patient users and patient visits for ES but only for those seen by case managers and education specialists and also does not indicate which patients are actually receiving these services.
Case managers and health educators do not currently document patient visits or the time spent with each patient.

22. ESAP GOALS
Develop standard data collection protocol and database for ES at health centers nationally
Describe utilization of ES at health centers and the patients who use them
Evaluate the impact of enabling services on health access, outcomes and utilization of primary care
Disseminate findings to health centers and policy makers to guide effective resource allocation
Facilitate research and expansion opportunities to other health centers and networks
CMS example- CMS had proposed a daily encounter rate based on national average cost-per-visit for the Medicare program. This rate is calculated by aggregating all of the data from all FQHCs and dividing their costs by their total visits. In other words, CMS will develop a single payment rate based on the average of ALL health centers, adjusting for geography and an upward adjuster for patients who are new or doing their initial primary care screening. This will adversely affect those health centers that provide a high number of ES since the proposed rate do not take them into account. Having ES data will show all the wrap-around services that we provide.

23. DOMAINS OF STANDARDIZED ES
CATEGORY
CODE
Minutes
CASE MANAGEMENT ASSESSMENT
CM001
CASE MANAGEMENT TREATMENT AND FACILITATION
CM002
CASE MANAGEMENT REFERRAL
CM003
FINANCIAL COUNSELING/ELIGIBILITY ASSISTANCE
FC001
HEALTH EDUCATION/SUPPORTIVE COUNSELING
HE001
INTERPRETATION
IN001
OUTREACH
OR001
TRANSPORTATION
TR001
OTHER
OT001
Enabling Services Accountability Project (ESAP)
The ONLY standardized data system to track and document
non-clinical enabling services that help patients access care.

24. SAMPLE ENABLING SERVICES EMR TEMPLATE

25. .

26. DATA EXAMPLE CHC 5,150 27,712 33 69% 91% Medicaid, 52%; Self Pay, 19%
Health Center
# of Patients
# of ES Encounters
Average Age
Gender (%Women)
Non-English Speaking Patients
Most Common Insurance
Most Common ES used
Most Common ES Provider type
CHC
5,150
27,712 33
69%
91%
Medicaid, 52%;
Self Pay, 19%
CM-Assessment followed by CM-Treatment
Social Worker, Social Work Assistant
If 91% of your patients are LEP and the most commonly used ES was CM-assessment, what does this say about your staff? Shouldn’t you get to brag about the fact that you have staff who’s not only providing an ES but is providing it in another language?!

27. ESAP OUTCOMES: DIABETES
One of the health outcomes research questions we looked at were: Do users of Enabling Services have better outcomes on diabetes and children’s immunizations when compared with non-users?
We first looked at the impact of ES on HbA1C. This study included patients aged between 18 and 75 as of 2007 with a diagnosis of type 1 or type 2 diabetes. Similar to the last study, patients were categories into two groups: one enabling services user group with all patients that had at least one enabling service in 2007; and one enabling services non-user group with all patients who didn’t have any enabling services in As seen in the figure, ES users had lower average HbA1c level than non-users, and patients using enabling services were more likely to have controlled diabetes with HbA1c levels below 7% than non-users (53% vs. 42%).
Data source: all ESAP sites ES and clinical data
T Test: P<0.05

28. ESAP OUTCOMES: APPROPRIATE CHILD IMMUNIZATION
In this health outcomes example, we assessed the impact of enabling services on appropriate child immunization. With 2007 data collected using the standardized pilot protocol and definitions, this study looked at all nine categories of enabling services and two performance measures including diabetes and children immunizations. Further, it exposed key demographic differences between enabling services users and non-users at FQHCs. The goals of this study is to provide a better understanding of the relationship between enabling services utilization and process/health outcomes by AA & NHOPIs, and to provide useful information to help policy makers effectively address health center needs as they strive to improve access and quality care to medically underserved AA & NHOPIs and other safety net patients. 
Data source: all ESAP sites ES and clinical data

29. DEMONSTRATING HEALTH CENTER VALUE
RESPONSE
Standardized data on interventions
NEED
Standardized data on
patient risk
We then pieced together some more research and were seeing that some patients did not experience worse outcomes when they had more SDH barriers, so knowing which patients had which interventions as well as the volume and time clinic staff spend on those interventions with patients (through our ES data collection) gives much needed context to understand patient outcome data.
Together both sets of data help demonstrate the value of health centers in effectively meeting the needs of complex patients and benefiting the overall health system. Simply stating that health center patients are sicker and experience more social determinants of health barriers doesn’t prove the extra work (and the extra funds needed) that CHCs are doing to address.
BOTH are necessary to demonstrate health center value

30. Why Is it Important to collect standardized data on the social determinants of health?

31. Why collect data on social determinants of Health (SDH)
Why collect data on social determinants of Health (SDH)? SDH drive outcomes before patients receive care
How well do we know our patients?
Are services addressing SDH incentivized and sustainable?
Are community partnerships adequate and integrated?

32. Standardized data to see the big picture
why is addressing social determinants of health (SDH) important to community health centers?
Provide Better Care
Reduced Costs
Community Connection
Standardized data to see the big picture
Take answers from audience
Understanding SDOH can help clinics provide better care
By identifying and understanding the people you serve—including their challenges and risks—you can tailor your services more appropriately and provide better care
Reduced costs
By treating root causes and better integrating services, health centers can reduce the volume of unnecessary visits
Connection to the community
By recognizing issues facing the community and addressing these issues through partnership and advocacy, health centers cans further solidify their role in the community
Clinics can leverage their existing relationships and standing in communities—patients are more likely to trust their motivation for collecting data
The data itself is the crux of this work; without a large enough sample size, we can’t draw meaningful conclusions
Standardized protocol in terms of what data is being collected is critically important; Some clinics are already collecting data on SDH, but without a standard data set, it’s impossible to aggregate responses on the population level and see the big picture

33. Assessment Tool to Identify Needs in EHR Protocol to Respond to Needs
What is prApare?
Protocol for Responding to & Assessing Patients’ Assets, Risks & Experiences:
A national standardized patient risk assessment protocol designed to engage patients in assessing & addressing social determinants of health (SDH).
PRAPARE = SDH screening tool + implementation/action process +
Assessment Tool to Identify Needs in EHR
Protocol to Respond to Needs

34. PRAPARE was designed to accelerate systemic change
for insured and uninsured patients

35. Dissemination Develop PRAPARE tool
Timeline of Phase 1
Year 1
2014
Develop PRAPARE tool
Year 2
2015
Pilot PRAPARE implementation in EHR and explore data utility
Year 3
2016
PRAPARE Implementation & Action Toolkit
Dissemination

36. Developing PRAPARE Criteria
Experience of Existing Protocols
Stakeholder Feedback
Identified 15 Core Social Determinants of Health
Sensitivity
Burden of Data Collection
Action-ability
Literature Review
Aligned with National Initiatives:
* Healthy People 2020
* ICD-10
* Meaningful Use Stage 3
* NQF on Risk Adjustment
Criteria

38. What makes PRAPARE Unique and feasible
17 core questions
9 of which already collected by health centers through federal reporting (UDS)
All align with national initiatives
Design
Vetted and stakeholder engaged development process
In the EHR to facilitate assessment & interventions (free templates)
Conversation starter and patient-centered
Common core yet flexible:
Able to make more granular and/or add questions
Focus on standardizing the need, not question
Can be used in combination with other tools/data

39. PRAPARE aligns with other national initiatives
PRAPARE Domain
UDS
ICD-10
IOM
Meaningful Use (2 and 3)
HP2020
RWJF County Health
Race/Ethnicity X
Farmworker Status
Veteran Status
Seeking comments
English Proficiency
Income
Insurance Status
Housing
Education
Employment
Material Security
Social Integration
Stress
Transportation
Also includes neighborhood and optional questions (incarceration history, refugee status, safety, domestic violence)

40. Money & Resources Tab:
If information was previously collected on any of the measures, the answer responses will populate on the left hand side in the gray boxes with the date of last documentation
You can select previous if the answer choices have not changes for the last time the information was collected
Again, you will see Add XXX diagnoses jump buttons populate depending on the patient’s answer choices

43. What we’ve learned from implementation

44. prapare pilot testing implementation teams and Electronic Health Records
Other EHRs in Development or Interested:
Greenway
Allscripts
Athena
Cerner

45. Sample Workflows: a Variety of Approaches!

46. Pilot results Easy to administer
Possible to implement using various workflows and staffing models
Builds patient-provider relationship
Identifies new needs
Leads to positive changes at the patient, health center, and community/population levels
Facilitates collaboration with community partners

47. PILOT Data Results High stress
SDH risks vary by community
Most common risks*:
High stress
Having less than a high school education
Uninsured
Unemployed
Preference for language other than English
But, patients are very socially integrated,
Half of patients see people they care about 5+ times a week
* Excludes low income

48. patients experience multiple SDH risk factors (typically 4-7, excluding low income)
Percent of Patients with Number* of SDH “Tallies”
N = 2,694 patients for all teams
This health center pilot population had highest burden of chronic illness.
* Excludes low income
3 CHCs
1 CHC
2 CHCs
1 CHC
7 CHCs

49. Positive Correlation Between SDH Factors and Hypertension: All Teams

50. HOW PRAPARE DATA HAS BEEN USED TO IMPROVE CARE DELIVERY AND HEALTH OUTCOMES
Better Understand INDIVIDUAL Patient’s Socioeconomic Situation
Build services in-house for same-day use as clinic visit (children’s book corner, food banks, clothing closets, wellness center, transportation shuttle, etc)
Ensure prescriptions and treatment plan match patient’s socioeconomic situation
Guide work of local foundations (ex: New York housing)
Build partnerships with local community based organizations to offer bi-directional referrals and discounts on services (ex: Iowa transportation)
Better Understand Needs of Patient POPULATION
Streamline care management plans for better resource allocation (ex: Hawaii)
First and foremost: about good clinical care
Transform Care Experience for Individual Patients
Inform Care Team Members:
Knows patient risks/needs to inform counseling and referrals
Understand context of patient health concerns
Empower patients through:
Staff sensitivity training
Patient-centered phrasing of questions
Brings awareness to SDH issues affecting health
Improve patient population management
Identify disparities
Inform allocation of resources and services
Mobilize community resources
Encourage local partnerships and bi-directional referrals
Opportunity to meaningfully share data
Guide local policy reform
Inform both Medicaid and Medicare ACO discussions (ex: Iowa, New York)
Drive STATE and NATIONAL Care Transformation
Inform payment reform and APM discussions with state agencies (e.g., Medicaid) on caring for complex patients (ex: Oregon, Hawaii)
Create risk score to inform risk adjustment (ex: Hawaii)

51. PRAPARE Implementation and Action Toolkit www.nachc.org/prapare
Chapter 1: Understand the PRAPARE Project
Chapter 2: Engage Key Stakeholders
Chapter 3: Strategize the Implementation Process
Chapter 4: Technical Implementation with EHR Templates
Chapter 5: Develop Workflow Models
Chapter 6: Develop a Data Strategy
Chapter 7: Understand and Evaluate Your Data
Chapter 8: Build Capacity to Respond to SDH Data
Chapter 9: Respond to SDH Data with Interventions
Chapter 10: Track Enabling Services

52. PRAPARe Readiness Assessment
Culture of Organization
Leadership and Management
Workflow Process Improvement
Technology
Paper form:
Online form:
ADD link. Best to do as organization but report back

53. Phase 2: PRAPARE Learning and Action Network (PLAN)
Build health center capacity to deploy a common, standardized patient-centered SDH assessment protocol
Create a PRAPARE community of users with shared learning
Accelerate spread of PRAPARE
Document impact
With support from the Kresge Foundation:
Audience
Health Centers
PCAs/HCCNs
Breadth
Online PRAPARE Engagement Platform (promote shared learning, track use, provide resources and access to experts)
PRAPARE Readiness resources for supporting members
Depth
PRAPARE Engagement Grants
PRAPARE Train the Trainer Academy

54. PRAPARE IS A NATIONAL MOVEMENT
Health centers and/or PCAs in every state including AK have expressed interest in PRAPARE
Health centers in over 30 states including AK downloaded PRAPARE EHR Templates
7000+ hits on PRAPARE website
Health and hospital systems are interested in PRAPARE
What pilot sites are doing in NC pilot will benefit others with resources specific to special populations and circumstances unique to CA, etc
National movement! Think of what we could do in terms of research, in terms of advocacy, in terms of informing policy, payment, and care transformation and accelerating community change.
Nearly 200 members are on our PRAPARE listserv
Nearly 100 EHR template downloads from PRAPARE website

55. RESOURCES AVAILABLE NOW
Visit
PRAPARE Tool
PRAPARE Implementation and Action Toolkit
Electronic Health Record PRAPARE Templates
Readiness Assessment
Webinars
PRAPARE Overview
EHR and Workflow-specific
Frequently Asked Questions
Contact: Michelle Jester at
Visit
AAPCHO’s Enabling Services Accountability Project
protocol for data collection of non- clinical enabling services
Enabling Services Data Collection Implementation Guide
White Papers, Best Practices, Studies
Contact Tuyen Tran at

56. HOW QUESTIONS ARE ASKED JUST AS CRITICAL
RESOURCES ON EMPATHIC INQUIRY
Waianae Demo:
OPCA Demo:
Play video

57. For more information on PRAPARE, visit www.nachc.org/prapare
This project was made possible with funding from: