Practice-Based Care Management: Many Paths to Chronic Disease Care
AHRQ's 2012 Annual Conference Slide Presentation
Select to access the PowerPoint® presentation (720 KB).
Slide 1
Practice-Based Care Management: Many Paths to Chronic Disease Care
Jodi Summers Holtrop, PhD, MCHES
Michigan State University Department of Family Medicine
Slide 2
Care Management Delivery Models
Feature | Health Plan or Disease Management Company | Provider/Practice/PO |
---|---|---|
History | 1985: case mgmt 1994: disease mgmt | New → existing |
Staff Location | Centralized, non-clinical site | Integrated within practice |
Staff | Nurses | Nurses, or other professionals |
Disease focus | Disease-based specific condition such as diabetes, CHF | Variable |
Population | Plan members | All patients |
Targeting | Claims-based algorithms, customer requests | Clinician personal knowledge, clinical data, EMR, health risk assessment, risk score |
Delivery mode | Phone | In person, phone, mail, Email, tele-monitoring |
Slide 3
Provider-Delivered Care Management Pilot (PDCM Pilot)
- Collaborative project between Blue Cross Blue Shield of Michigan (BCBSM) and 5 state Physician Organizations (POs).
- Focus on Chronic Disease Care Management.
- $2M over 2 year Pilot.
Goals:
|
Slide 4
Provider-Delivered Care Management Grant (PDCM AHRQ Study)
- Assess the PDCM pilot.
- 3 year $1.8 grant from AHRQ.
Goals:
|
Slide 5
PDCM AHRQ Study: Convergent Mixed Methods Model
Patient Outcomes:
|
Images: Two arrows point from this box to two other boxes:
CER: Compare PDCM to HPDCM. |
Does CM participation:
- Improve clinical outcomes?
- Reduce utilization/cost?
Within PDCM:
|
Does a higher quality CM program and more integration lead to better outcomes? Which outcomes?
Are practice environment and normalization of CM mediators?
Slide 6
PDCM Pilot Monthly data flow process
BCBSM
Create member list:
- Member ID.
- Name.
- Date of birth.
- Gender.
- Risk score.
- High cost flags.
- Chronic disease flags.
- 12-month IP, ED counts.
- BCBSM CM/DM case status.
- New member flag.
- Lost eligibility flag (drop).
- Previous month PO responses.
PO
- Confirm member is patient of pilot PCP.
- Confirm patient has chronic condition.
- Confirm ability to provide CM if needed.
- Decide to accept / not accept.
Member list return file:
- Member ID.
- Acceptance status.
- Accept date.
- Reason for no accept.
- Chronic disease(s).
- Acceptance end date.
- Acceptance end reason.
Activity files:
- Encounters.
- Goals.
Image: Two arrows point in a clockwise circle under the above text to indicate that it is an ongoing process.
Slide 7
Compare PDCM to Health Plan Delivered Care Management (HPDCM)
Targeted = Number of patients having CM outreach attempt / Number of patients accepted
- HPDCM: 100%.
- PDCM: 62-82%.
Engaged = Number of patients having CM encounter / Number of patients having CM outreach attempt
- HPDCM: 17-18%.
- PDCM: 35-100%.
Point: PDCM engagement rates appear to be higher than HPDCM.
Note—still completing PDCM individual PO CM only rates; HPDCM rates are for the overall, not matched, populations.
Slide 8
Care Management Features
- Program Quality:
- Patient access.
- CMgr qualifications and personal attributes.
- Staff CM training.
- Resources for staff and patients.
- Program duration, dosage, features and length of visit.
- Integration in Practice:
- Practice staff participation in CM.
- CM team use.
- Location of CMgr visits with patients.
- CMgr visit documentation.
- Incentives and barriers for CM use.
Slide 9
Analysis of PO and Practice Care Management Features, Integration and Context
Image: A chart depicts the following process:
Practice visit—2 team members:
- Interviews.
- Observations.
→
- Audio files sent for transcription. →
- All transcripts for each practice assigned and coded by one of three coders. →
- Quotations pulled by quotes. →
- Immersion crystalization analysis. →
- Emergent themes ACROSS practices/PO's.
- Immersion crystalization analysis. →
- Coders review summary documents, add comments/quotes. →
- Revised summary documents v.1. →
- Reconciliation meeting for all practices within PO model—team and coders. →
- Revised summary documents v.2. →
- Reconciliation meeting across PO's—team and coders. →
- Revised summary documents v.3—Key features WITHIN practices/PO.
- Reconciliation meeting across PO's—team and coders. →
- Revised summary documents v.2. →
- Reconciliation meeting for all practices within PO model—team and coders. →
- Revised summary documents v.1. →
- Quotations pulled by quotes. →
- All transcripts for each practice assigned and coded by one of three coders. →
- Complete summary documents (both):
- Quality.
- Integration.
- NPT/implementation.
Lead RA:
- Practice summary (internal).
- Practice report and task diagram.
- Resources.
- Observation guide.
→
- 2 team members reconcile scores and notes. →
- Practice report sent to practice for member checking; modifications made. →
- Completed summary documents:
- Quality.
- Integration.
- NPT.
- Practice summary.
→
- Coders review summary documents, add comments/quotes. →
- Revised summary documents v.1. →
- Reconciliation meeting for all practices within PO model—team and coders. →
- Revised summary documents v.2. →
- Reconciliation meeting across PO's—team and coders. →
- Revised summary documents v.3—Key features WITHIN practices/PO.
- Reconciliation meeting across PO's—team and coders. →
- Revised summary documents v.2. →
- Reconciliation meeting for all practices within PO model—team and coders. →
- Revised summary documents v.1. →
- Completed summary documents:
- Practice report sent to practice for member checking; modifications made. →
Slide 10
Care Management Models: Practice Integration or Centralization at PO
PO | A | D | C | D | C | B | E | E |
---|---|---|---|---|---|---|---|---|
Model Type | Cent | Cent | Cent | Practice-based | Practice-based | Practice-based | Practice-based | Practice-based |
Practice Ownership | Ind | Ind | Ind | Ind | Ind | Owned | Owned | Owned |
C Mgr location | PO | PO | PO | Practice | Practice | Practice | Practice | Practice |
Patient Visits | Phone | Phone | Phone | Practice + phone | Practice + phone | Practice + phone | Practice + phone | Practice + phone |
Patient access | All | BCBS | BCBS | Diabetes | BCBS | All | All | All: BCBS focus |
Reported C Mgr team use | Lower | Lower | Low-med | Med; High for diabetes | Med-High | Higher | Higher | Higher |
Slide 11
Care Management Models: Within PDCM
Point: Within PDCM there is variation in practice integration.
- In 5 PO's, 8 "models" of delivery—2 distinct types.
- Integration: centralized → practice-based.
- One PO very centralized; more like health plan.
- Three PO's two different models happening.
- Two PO's highly integrated.
- Within practice-based models:
- Full-time C Mgr as practice team member only doing CM.
- Full-time nurse doing CM 1 day per week as part of job (included in job description of all practice nurses).
- Travel model with panel manager and C Mgr (panel manager in practice; C Mgr 1-3 days per week at practice).
Slide 12
How people talk about Integration
Point: Integration feels different than centralized to the participants.
Interviewer: If you would do anything to improve the CM program, what would it be?
- Centralized models:
- C Mgr: Increase referrals of patients to the CM, like pulling teeth to get referrals.
- Provider: C Mgr is a resource we can refer to.
- Integrated models:
- C Mgr: Hire more C Mgr's. We have so many patients with needs and can't keep up with the existing number of C Mgr's.
- Provider: Hire more C Mgrs. The CM is part of our team and how we practice care.
Slide 13
Implementation vs. Program Quality and Integration
PO | A | D | C | C | D | B | E | E |
---|---|---|---|---|---|---|---|---|
Quality score | 77 | 80 | 78 | 76 | 79 | 89 | 92 | 94 |
| 7 | 4 | 6 | 8 | 5 | 3 | 2 | 1 |
Integration score | 63 | 86 | 49 | 75 | 81 | 91 | 87 | 84 |
| 7 | 3 | 8 | 6 | 5 | 1 | 2 | 4 |
Normalization Process Model (NPM) score | 81 | 90 | 79 | 85 | 94 | 91 | 90 | 92 |
| 6 | 4 | 7 | 5 | 1 | 3 | 4 | 2 |
Slide 14
Program Quality
Point: Most PDCM programs were of high quality with a few exceptions.
Examples:
- MA delivered programs without much training.
- Not delivering motivational interviewing, rather telling patient what to do in one or a few brief calls.
- When restricted to only BCBSM patients on "lists" tended to work differently (out of patient flow, lacking provider selection of patient) rather than as team concept with this being just part of your care.
- Co-pays were reported to result in patient refusal of CM as compared to programs charging no patient co-pay or billing.
- Issues were the patient not knowing what the co-pay will be or wanting to pay it once they knew. Patients did not want to answer the phone for CM calls fearing they would get a bill.
Slide 15
Program Quality and Integration
Point: Lack of either quality or integration limits program use and likely effectiveness.
- Even if highly integrated, can it overcome problems of poor quality?
PROBABLY NOT.- Poorly trained or not properly credentialed C Mgr lacks cultural credibility and trust from team.
- Lack of time and support to do CM means it doesn’t happen.
- Even if your program is of high quality, can it overcome barriers of lack of integration?
PROBABLY NOT.- Hard to reach patients only by phone out of scope of normal care.
- C Mgr not known as practice team member.
- Lack of day to day huddles and ad hoc communication about patients (providers often don’t read EMR notes).
- Providers and practice members forget to refer; an extra step.
Slide 16
Summary
- Patients appear to participate in PDCM at higher rates than health plan care management.
- Typical community, non-research directed primary care practices can implement CM.
- PDCM program quality is generally high.
- PDCM integration in practice is highly variable.
- More CM integration appears to equal more team use of CM and we think CM participation.
Slide 17
Image: A photograph shows a flock of baby ducks at a concrete curb; one baby duck is making a valiant effort to climb the curb.
Note: THIS IS WHERE WE ARE—not much in the way of results, but getting an idea.
Slide 18
To be continued...
- Results are PRELIMINARY.
- Ongoing work includes:
- Verifying target population and specific practice and PO CM engagement rates.
- Gathering data on clinical values, claims.
- Analyzing emergent themes across practices.
- Further analysis:
- "Recipe" for successful care management—qualitative comparative analysis.
Slide 19
Research Team
- Michigan State University:
- Jodi Holtrop, Zhehui Luo, Qiaoling Chen, Laurie Fitzpatrick.
- University of Michigan:
- Gretchen Piatt, Lee Green (now U Alberta), Georges Potworowski (now SUNY Albany), Mike Fetters, Jean Malouin, Trudy Adler, Amy Kowalk.
- Altarum Institute:
- Rachelle May-Gentile, Anya Day, Brad Hinks, Lauren Wendel, Kristen Werner.
- Blue Cross Blue Shield of Michigan:
- Margaret Mason, Lisa Rajt, Min Tao, Ann Emeott, Guipeng Liu, Hsiu-Ching Chang, Darline El Reda.
- Physician Organizations:
- Ruth Clark, Mary Ellen Benzik, Cecilia Sauter, Jen Bailey, Cathy Heiman, Cara Seguin.
Slide 20
THANK YOU!