Improving Diabetes Outcomes and Overcoming Communication Barriers for Vulnerable Communities Through Health IT Self Management Support
Improving Diabetes Outcomes and Overcoming Communication Barriers for Vulnerable Communities Through Health Information Technology (Health IT) Self Management Support
Neda Ratanawongsa, MD, MPH
September 20, 2011
- Projects funded by:
- AHRQ 5R21HS014864.
- AHRQ 5R18HS017261.
- McKesson Foundation.
- Centers for Disease Control and Prevention (CDC) Division of Diabetes Translation grant.
- California Diabetes Program.
- The funders had no role in design, data collection, analysis, or presentation.
Defining our communities
- Practice-based researchers in CA safety net:
- Persons with diabetes in San Francisco Community Health Network (CHNSF).
- Primary care clinicians.
- Health plan / insurers overseeing care for persons with diabetes.
- Representative of larger CA communities.
- 10-year evolution: disparities → interventions → practice-/population-based implementation.
- Disparities in diabetes health due to communication barriers in traditional health care.
- Diabetes self-management health IT intervention:
- Practice-based research → population-based implementation.
- Impact on quality of life and self-management.
- Learning opportunities and next steps.
Limited Health Literacy (LHL)
- Over half of public hospital pts.
- Average reading level for Medicaid patient: grade 5.
- Impact on health outcomes:
- Poorer knowledge of chronic conditions.
- Worse self-care.
- Higher utilization of services.
- Worse health outcomes:
- Poor glycemic control (AOR 2.03; p = 0.02).
Institute of Medicine (IOM) 2004.
Scott 2002, Williams 1998, Baker 2003, IOM 2004; Schillinger 2002.
... And Poor Communication with Clinicians
Image: Bar graph presents the following data:
|MD Uses Words Not Understood*||MD Gives You Test Results w/o Explanation*||Pt Confused About Medical Care*||MD Understands Problems Doing Rx**|
Limited English Proficiency (LEP) & Lack of Language-Concordant Care
- 6.2 million (19%) in California 2000 census.
- LEP Asian immigrants using interpreters report unasked questions about care (30% vs. 21%, P <.001).
- LEP Latinos with language discordant MD had increased odds of poor control (AOR 1.98).
- Less likely to report receiving self-mgmt advice.
Wilson 2005; Fernandez 2010; Green AR 2005; Lopez-Quintero 2009
Health IT for Self-Management Support
- Self-management support improves behaviors, satisfaction, and outcomes.
- Desired by patients with LHL and LEP.
- Automated telephone self-management (ATSM):
- 97% of adults in CA have phone.
- Relatively inexpensive and efficient.
- Control jargon, volume, pace, and language.
- Effective in diverse, low income patients.
Improving Diabetes Efforts Across Language and Literacy (IDEALL)
- Developed with users.
- Preferred language.
- Weekly surveillance.
- Touch tone response.
- Tailored education.
- Language-concordant care managers respond to out-of-range triggers.
Image: A graphic shows relationships between Patient, primary care provider (PCP), nurse practitioner (NP) Care manager, and ATSM.
- Notify clinics.
IDEALL Development Process
- Identify priority population/condition and objectives.
- Harness registry and network to identify population.
- Develop queries to solicit questions and concerns.
- Write and revise health education (cooperative process).
- Pilot questions and health education responses with pts.
- Translate and adapt toward cultural appropriateness.
- Record and code.
- Design callback algorithm (scenarios) and trigger reports.
- Train clinical staff.
Health IT Can Promote Patient-Centered Diabetes Care (IDEALL)
- Randomized trial: ATSM, group visits, & usual care.
- 339 patients with poorly controlled DM:
- 43% Spanish- and 11% Cantonese-speaking.
- 94% completed ≥1 call → 84% ≥1 action plan.
- High PCP satisfaction:
- Perceived activated pts & higher quality of care.
- Overcoming barriers to LEP & med mgmt.
IDEALL Program Outcomes
- Interpersonal communication with providers.
- Self-management behaviors (diet, exercise).
- Functional status & days confined to bed.
- Detected adverse / potentially adverse events.
- $65,167 for set-up and ongoing costs.
- $32,333 for ongoing costs only.
Schillinger 2009; Sarkar 2008; Handley 2008.
" I became more aware of what I put in my system and that I need to do something greater than what I've been doing to lose more weight... (ATSM narratives) talked about a woman who lost weight... I liked that... I could walk in those shoes."
"I had already made a moral promise that this week I would give 100%, that I would exercise and get sweaty, and I did it."
"It elevated my self-esteem so that I could 'get fired up' and really respond because it was up to me to gain control of my diabetes. In other words, one needs to do their part.
Potential for Medicaid Partnership
- Improve coordination of care.
- Support patients and clinicians.
- Promote personal control over services.
- Harness IT to reduce disparities.
- Survey of CA Medicaid managed care plans:
- Few had chronic care mgmt targeting LEP/LHL.
- 68% planning to expand programs for diabetes.
- Barriers: cost of broad implementation and IT.
SMART Steps: Partnering to Put Research Into Practice
- San Francisco Health Plan (SFHP): nonprofit govt-sponsored Medicaid managed-care plan:
- Linguistically diverse vulnerable population.
- SFHP recruitment for members from 4 clinics.
- SFHP implementation.
- Evaluation by UCSF.
Quasi-Experimental Study Design
- SFHP did not want control group (no intervention).
- Lack of staff to 'scale up' quickly.
- Wait list with 6-mo crossover, recruiting in waves.
Image: Graphic shows the following process:
Intervention Wave 1
Wait-List Wave 1 →
Intervention Wave 2
Wait-List Wave 2 →
Intervention Wave 3
Wait-List Wave 3 →
Intervention Wave 4
Wait-List Wave 4 →
Intervention: ATSM + health coach
- 27 weeks of ATSM calls.
- SFHP health coach for follow-up calls:
- Tailored training & scripts.
Image: A sample table shows tailored training and scripts.
- Engagement in ATSM:
- % completing calls.
- Differences by language.
- Compare intervention (combined) vs. waitlist in change from baseline to 6-month:
- Summary of Diabetes Self-Care Activities.
- Quality of life (SF-12).
Toobert 2000, Ware 1996.
Participants With 6-Month F/U (n=249)
|Characteristic||Intervention (n=125)||Wait-List (n=124)|
|Age in years, mean (SD)||56.6 (7.9)||54.9 (8.6)|
Black / African-American
Asian / Pacific Islander
White / Caucasian
|Born Outside the U.S.||86%||85%|
|8th grade education or less||39%||47%|
|Limited health literacy||47%||40%|
|Income ≤$20,000 / Yr||61%||60%|
|Hgb A1c >8.0%||30%||24%|
Completed Calls by Language For Those Exposed to All Weeks
Image: A line graph shows the following data:
|All Languages (n=273)||57.4||55.7||54.9||53.8||55.7||54.9||54.2||57.9||56.7||57.1||61.9||52.4||52.7||52.7||54.6||53.1||54.6||53.1||54.9||56.0||54.2||54.9||56.8||55.7||52.4||52.0||51.6|
Engagement by Language Among Patients Exposed to 27 Calls (n=273)
|Completed ≥1 Call, %||85%||90%||81%||80%||0.07|
|Number of completed calls, median (IQR)||19|
Change in Quality of Life at 6 Mos (n=249)
|Standardized Effect Size*||p-value|
* Controlling for baseline value.
Change in Self-Care at 6 Mos (n=249)
|Overall Self-Care||0.2 (0.1, 0.4)||0.29||<0.01|
|Glucose monitoring||0.7 (0.2, 1.3)||0.30||<0.01|
|0.6 (0.2, 0.9)||0.32||<0.01|
|0.0 (-0.2, 0.2)||0.02||0.82|
- Partnering with LHL / LEP patients:
- Bicultural and bilingual content.
- Unmet need for language-concordant support.
- Practice-based research:
- Innovate and create from within.
- Invest in the safety net providers.
- Partnership with Medicaid managed care plan:
- Population-based implementation.
- Long-term relationships.
- Phone-based population recruitment.
- Health coaches:
- Tailoring, training, and turnover.
- Bicultural as well as bilingual staff.
- Assessing fidelity: data collection & feedback.
- Quasi-experimental designs beyond randomized controlled trial (RCT) .
Handley 2011 (in press).
- Scope: develop new content for health promotion self-mgmt across health conditions.
- Platform: mHealth beyond telephone outreach.
- Linkages to patient-centered medical home, including electronic health records.
- Fidelity to mission.
- Reach and sustainability:
- Within our health system.
- Medicaid and other insurers.
- Dean Schillinger.
- Margaret Handley.
- Judy Quan.
- Urmimala Sarkar.
- Catalina Soria, Claudia Barrera, and Phiona Tan.
- Kelly Pfeifer, Allison Lum, Dawn Surratt, Karen Fong, Jennifer Beach, Juanita Gonzalez, Nicole Lam, Stanley Tan, and San Francisco Health Plan.
- California Diabetes Program.