This report is based on research conducted by Abt Associates in partnership with the MacColl Center for Health Care Innovation and Bailit Health Purchasing, Cambridge, MA, under contract to the Agency for Healthcare Research and Quality (AHRQ), Rockville, MD (Contract Nos. 290-2010-00004-I/ 290-32009-T). The findings and conclusions in this document are those of the authors, who are responsible for its contents; the findings and conclusions do not necessarily represent the views of AHRQ. Therefore, no statement in this report should be construed as an official position of AHRQ or of the U.S. Department of Health and Human Services.
Suggested citation: New Models of Primary Care Workforce and Financing Case Example #1: Stanford Coordinated Care. (Prepared by Abt Associates, in partnership with the MacColl Center for Health Care Innovation and Bailit Health Purchasing, Cambridge, MA, under Contract No.290-2010-00004-I/ 290-32009-T.) AHRQ Publication No. 16(17)-0046-1-EF. Rockville, MD: Agency for Healthcare Research and Quality; October 2016.
Table of Contents
Why Stanford Coordinated Care?
Overview of the Practice
Patient Population Description and Practice Panel Size
Model of Care
Care Coordination and Transitions
Team Building and Training
Access, Comprehensiveness, and Quality
Implications for Primary Care Staffing Models
The Agency for Healthcare Research and Quality (AHRQ) contracted with Abt Associates and its partners, the MacColl Center for Health Care Innovation and Bailit Health Purchasing, to conduct research on innovative ways to configure primary care workforce teams that can deliver fully comprehensive, high-quality care to the U.S. population. The purpose of the research is to offer models of primary care teams and the associated costs for consideration and discussion by policymakers and providers in the field of health services delivery. To explore existing innovative workforce configurations, AHRQ and the Abt project team conducted a literature review, explored extant data sources, convened a Technical Expert Workgroup of national experts in primary care workforce, and conducted site visits to primary care practices with innovative workforce configurations.
The case example report that follows provides an in-depth look at the workforce configuration of Stanford Coordinated Care Primary Care Practice (SCC), a primary care practice delivering high-quality, comprehensive care (located at 211 Quarry Rd., Palo Alto, CA 94304). Team members visited the practice on February 17, 2015. The data presented below were collected on or prior to the visit and reflect calendar year 2014.
Why Stanford Coordinated Care?
SCC provides primary care to Stanford University employees and their dependents who are medically complex, high risk, and high utilizers of health services. Understanding how comprehensive primary care is provided to this specific patient population was of interest because nationwide, the number and percentage of adults living with multiple chronic conditions who require complex care is growing1. SCC was also of interest because human-centered design techniques were incorporated into planning the physical space and practice processes. Patient input was also solicited to help design the services and workflow of SCC, and the practice maintains an active Patient Advisory Committee. SCC aims to fulfill most of its patients' needs and thereby minimize referrals by employing an interprofessional, team-based approach. Care is provided by phone, email (via patient portal), and home visits. As needed, SCC staff members accompany patients on specialist visits, or visit patients during hospitalization.
Overview of the Practice
Drs. Alan Glaseroff and Ann Lindsay were invited by Stanford Health Care to create SCC beginning in 2012. Stanford University is self-insured for its employees and their dependents and established an Accountable Care Organization (ACO) option for employees and dependents in January 2014. The SCC practice is a service of the health plan and offers care to high-risk, high-cost plan members in an attempt to reduce utilization (e.g., emergency room visits and hospitalizations), increase quality of care, and improve the health and satisfaction of its patients. All levels of employees and their dependents are eligible for the SCC practice, from nonexempt to senior professional staff members and faculty.
The design of the physical office space includes a large team room with 12 computer/work stations shared by all team members. The physical therapist (PT) and social worker have individual offices for treating patients. The care coordinators and physicians see patients in exam rooms that are shared by the full team.
Patient Population Description and Practice Panel Size
Eligibility criteria to be seen in the SCC practice include: three or more chronic conditions, five or more medications, and certain responses to questions about ability to manage conditions. Patients are recruited to the practice by an outreach coordinator who holds monthly “meet-and-greet“ open house sessions. Potential patients are surveyed about ongoing or persistent conditions as they may not understand or relate to the word “chronic.” They may be asked, “Do you find your needs are not met in a 15–20 minute visit?” The Stanford Hospital conducts employee screening that incorporates similar questions and refers eligible employees to SCC. Patients may hear about SCC through word of mouth and self-refer for eligibility screening.
SCC patients have risk factors that are about 2.5 times higher than average primary care patients in the Stanford system, and their costs are frequently three to four times the average. Excluded from the practice are:
- People with active acquired immunodeficiency syndrome (AIDS), who are referred to a human immunodeficiency virus (HIV)-oriented primary care clinic.
- Fragile elderly, who are referred to senior care.
- Individuals whose only health problems are severe mental health disorders.
The Stanford Coordinated Care practice, with 1.3 physician full-time equivalents (FTEs) spread over three physicians, currently has 320 patients and will reach maximum capacity at 400. SCC assigns patients to a physician and a care coordinator. MA care coordinators have an assigned patient panel. In contrast to many primary care practices, care coordinator panels contain patients from all physician panels. Dr. Glaseroff said this arrangement gives MAs an opportunity to learn from all three providers and understand their practice patterns. Physician panel sizes are approximately 300 patients per 1.0 FTE. Care coordinator panels are approximately 100 patients per 1.0 FTE. Patients whose first language is not English are assigned to care coordinators who speak their native language, when possible. Panel sizes were determined by experience, adding to panels until practice staff members were unable to take on more patients.
Model of Care
SCC used human-centered design principles to design their care delivery model. During the initial implementation year for the clinic, an Innovation Design Engineering Organization-trained consultant interviewed Stanford employees who chose or did not choose to use SCC. People who chose SCC described the following changes in the way they perceived care:
|Before Care at SCC||After Care at SCC|
|Felt alone||Became an empowered patient|
|Forced to be at the center||Felt supported and confident|
|Felt “studied”||Felt listened to|
|Was given facts||Received hands-on action|
|Passed between providers||Created personal relationships|
|Felt stalled in life||Had a sense of thriving|
|Felt resource intensive||Felt streamlined|
Trusting relationships are the foundation of the practice's care model. Dr. Glaseroff noted that, while patients may at times lack access to medical care, more often they do not trust the medical care they have access to, and as a result don't get care they need. SCC clinicians refer to their care as “artisanal,” meaning they use strong personal relationships to meet patient-specific needs rather than passing patients along an assembly line of services. In considering patient care, SCC clinicians consider both medical and social complexity, which they define as:
- What the patient is experiencing (e.g., medically, socially, behaviorally, including lack of trust)
- What the patient brings (e.g., issues of trauma such as early childhood adverse events)
Care is delivered primarily using an intensivist primary care team model, focusing on behavior change and self-management support, with the three in-house physicians providing primary care. A small number of patients who have primary care physicians outside the practice receive support from SCC's licensed clinical social worker (LCSW) and PT. These are patients who are satisfied with their current primary care provider but want the additional support of the LCSW and PT.
Care at SCC is delivered in two phases:
- Phase 1: Focus on patient goals and help patient achieve them. If patients don't achieve goals, SCC looks at the sources of dysfunction.
“We bombard them with love and affection. We have time to do the right thing.“
– SCC physician
- Phase 2: Dig deeper using the Adverse Childhood Experiences (ACE) Survey. The LCSW may use eye movement desensitization reprocessing and mindfulness to put childhood events into adult consciousness.
“The first thing we try to do with people is fix the health care—then we dig deeper and address their underlying psychosocial problems when simply building relationships is not enough.”
Most patients seen at SCC have difficulty managing their multiple chronic conditions, and at least 40% have mental health disorders. Therefore, in Phase 1, the practice assesses gaps in four domains of care:
- Medical neighborhood
- Social support
- Medical status and health trajectory
- Self-management and mental health
In addition to the care needs assessment, the practice administers Hibbard's Patient Activation Model (PAM) measurement tool at intake. Interviewees at SCC said that, often, their patients with low PAM scores do not have the information, confidence, or ability to maintain healthy behaviors. The SCC team may schedule weekly visits until patients begin to feel health improvements. The team intensively supports patients in registering and using its patient portal so patients have access to the team and to information about their conditions and treatments. A team member sits with the patient to help them register for the portal and practice using it.
“We go from 'What bothers you most?' to 'Where do you want to be in a year?' We use attention to patients' self-identified goals to build trust in relationships, and then we use action planning to determine the steps. We concentrate on diligent, immediate followup.”
SCC reassesses patients with the PAM at 6 months to measure changes in confidence. If the patient does not experience improvement in the PAM after about 6 months, the team begins Phase 2—a more intensive assessment phase—using the ACE survey administered by the social worker. Dr. Glaseroff cited a Center for Medicare & Medicaid Innovation demonstration project in six States using ACE concepts that showed that 58% of patients improved at least three points on the PAM, which is the threshold for reducing cost.
“Just as PTSD [Posttraumatic Stress Disorder] can be a driver of cost in the VA [U.S. Department of Veterans Affairs] system, adverse childhood events can be a hidden driver of costs in traditional practice. Early childhood experiences such as those uncovered using the Adverse Childhood Experiences Survey may be an underlying cause of health disparities.”
– Dr. Glaseroff
Exhibit 1. Stanford coordinated care team FTEs and roles
|1.3||MD||Three physicians combine for 1.3 FTEs|
|1.0||APRN||Care management, transitions|
|1.0||LCSW||Behavioral health counseling|
|.6||PT||Specialized in use of integrative physical therapy|
|1.0||Clinic Manager||On maternity leave at the time of the site visit|
|.2||Clinical Pharmacist/Certified Diabetes Educator||Performs both condition management and patient education roles|
|4.0||Care Coordinator||MA, navigator, health coach, coordinator|
|1.0||Community Project Coordinator||Community linkages, peer programs|
|Business Operations Support Workforce|
|1.0||General administrative (administrators, office managers, financial or human resources)|
|X||Information technology staff e.g., database/population manager, data analysts, EHR support for data retrieval and analysis||SCC has access to a biostatistician who was not engaged in the past year|
|Front Office Support Workforce|
|N/A||Medical receptionist||Position covered by rotating care coordinator staff|
|Contributed Staff Workforce|
|N/A||Labor contribution from a related party, e.g., a pharmacist who works for a parent organization who occasionally consults with the practice|
|N/A||Labor contribution from an unrelated party, e.g., a grant-funded or insurer-employed nurse case manager who works at the practice two days per week|
MD=medical doctor; FTE=full-time equivalent; APRN=advanced practice registered nurse; LCSW=licensed clinical social worker; PT=physical therapist; MA=medical assistant; EHR=electronic health record; N/A=not applicable; X=No specific FTE provided/role may be paid for and managed centrally
At SCC, in many primary care teams, a physician and medical assistant (MA) are the core team, or “teamlet.” SCC's core team (Exhibit 1) resembles what other primary care settings often call an extended care team: a nurse, social worker, PT, and community project coordinator are often involved in patient care and are considered part of the core team. The team is configured to provide commonly needed services in house to reduce the need for referrals outside the care team. The team offers independent visits with the APRN, LCSW, PT, and community project coordinator. Sometimes, providers or the APRN nurse specialist attend specialty visits with patients to make sure the specialist understands the patient's goals and to ask questions such as “How does this test give us more information?” or ”How do your recommendations fit with the patient's goals?”
Care Coordination and Transitions
Care coordinators are MAs with expanded roles. They cover for each other, and view each other's email, customer relationship management notes, and information in patient portals. They follow up on each other's prescription refills and other tasks.
Since all telephone calls come directly to the clinic, not to a call center, care coordinators take turns covering the reception desk, where they can also perform computer work (e.g., panel management). The front desk is not busy because of innovative processes the clinic developed. For example, each care coordinator greets patients when they sign in and takes them to the exam room. If a care coordinator is covering the front desk when an assigned patient comes for an urgent visit, another care coordinator takes over the desk. Patients are encouraged to use the patient portal, so about one-third of care is by email and care coordinators spend most of their time on email.
“We want to put medical assistant tasks, coaching, care coordination and health navigator tasks all in one person. We want the fewest hand-offs possible, with care provided by one person (in addition to the physician) who has a deep relationship with the patient.”
– SCC physician
Due in part to SCC, the broader Stanford University health system recently created four MA (care coordinator) personnel categories with increasing responsibility. The goal is to provide a career advancement plan that allows MAs across the Stanford system to increase their responsibilities and salaries.
The initial 2-hour intake visit for new patients includes:
- A 30-minute orientation to the clinic and care model by the care coordinator.
- A 60-minute clinical visit with a physician.
- A 30-minute action-oriented followup visit with a care coordinator.
Standard followup visits include:
- A 15-minute check-in with a care coordinator.
- A 30-minute clinical visit with a physician.
- A 15-minute “close-the-loop” followup visit with a care coordinator.
Care coordinators and physicians huddle either in person or electronically through functions of the electronic health record (EHR). Care coordinators follow patients through the entire visit: they greet them, room them, alert the physician to needs or changes, and scribe during the physician visit. Care coordinators administer all refills, check for needed tests, and perform medication reconciliation. Service protocols reside on a SharePoint site and are a standard part of each patient visit. Before each visit, care coordinators print patient Health Portraits (Appendix 1), so named by the Patient Advisory Committee. These documents contain information about physical health, including areas that are not well controlled. The practice team calls sharing the Health Portrait with the patient the “Visine moment”—a chance to get the red out, meaning red-highlighted areas in the Health Portrait that show the main health concerns for the patient. Care coordinators said they do not push behavior change, but simply present the profile information without preaching. Patients often request more information when reviewing the Health Portrait, which prompts a conversation. The Health Portrait review is a standard part of each patient visit.
At weekly team meetings, the entire SCC staff reviews all new patients who had an initial intake visit. Care coordinators present the patient, describing their life, housing, family, hobbies, and work, before the physician presents an overview of clinical issues. The team discusses possible next steps, and staff members offer suggestions for how they might help. Just-in-time appointments or other independent visits are arranged with the APRN, LCSW, or PT. As the team gets to know the patient, staff roles evolve with more independent visits with the PT, LCSW, or community resource staff person.
Exhibit 2 outlines tasks performed by SCC staff with detailed descriptions of the roles of each staff member related to each task.
Exhibit 2. Stanford coordinated care team tasks and roles
|Pre-visit planning, chart scrubbing||CC reviews care gaps and preventive services that are due and shares information with the physician for review ahead of the visit. MD is not involved in monitoring preventive services. CC monitors chronic care needs using six protocols for chronic illnesses. Care team holds informal morning huddles or electronic, synchronous, pre-visit huddles; MD looks for other indicated services and writes “huddle” in the subject line. CC checks patient action plans and whether milestones have been achieved.|
|Greeting the patient||MA at the front desk prints the Health Portrait in advance, alerts the care team when a patient arrives, and ensures the person is signed up for the patient portal. If needed and time permits, MA demonstrates how to use the patient portal.|
|Ordering lab tests and collecting vital signs||CC greets and rooms the patient, takes vital signs, and conducts blood or urine tests. CC meets with the MD outside the exam room to review vital signs and tests. MD conducts the exam with the CC scribing, enabling the CC to learn how the physician approaches the patient's health, the questions they ask, and what they recommend. The MD reviews patient problems while a computer screen is visible to the patient so if the patient asks about a treatment or condition, the CC and MD can search for an answer. This models how patients can get answers themselves. CCs may also perform blood draws and injections with APRN supervision.|
|Delivery of routine preventive services||CC is responsible for delivering routine preventive services following existing protocols.|
|Medication reconciliation or management||Reconciliation is by CCs and co-signed by clinicians. The clinical pharmacist uses specific treat-to-target protocols for illnesses and reviews medications for new patients for contraindications and to ensure that the medication regime is appropriate and cost-effective.|
|Reviewing and reconciling problem list||This is an MD role.|
|Patient navigation||This is a CC role.|
|Self-management goal setting and action planning||The Health Portrait review by the CC sets up the self-management conversation, leading to goal setting. The MD is in the room while the action plan is reviewed, but then leaves while the CC completes the plan and makes sure the patient understands it. The MD might consult if problems or questions arise. MD checks the plan for correctness and closes out the note. CCs follow up with patient action plans between clinic visits.|
|Patient telephone /email followup||This is a CC role.|
|Injections and venipuncture||This is performed by the CC after review by the RN.|
|Triaging phone calls and emails||Most requests are by email. CCs cover about 50% (e.g., insurance, referrals, routine refills), putting the patient on hold and asking the APRN, an MD, or other team member when necessary. Acute requests are given to a licensed professional on the team. Patients can also call the CC extension directly. CC can take patient history, but cannot make recommendations. CCs learn to take history, including questions to ask, through visit scribing.|
|Care and transition management of high-risk patients||This is an RN role.|
|Referral management||Referrals are by the MD except for routine preventive services, which are started by CCs and reviewed and sent by an MD.|
|Medication titration||This is an MD and pharmacist role.|
|Independent visits by non-providers (RN, MA, health coach)||These can occur with anyone on the team.|
CC, care coordinator; MA, medical assistant; MD, medical doctor; APRN, advance practice registered nurse
The team manages all specialist care for patients. Patients in the ACO are referred to Stanford specialists. Other patients are referred to either Stanford specialists or other community physicians. When a patient is empaneled to an SCC physician, a clinician sends a note to any specialists the patient is currently seeing and copies them on all visit notes.
All Stanford Health Care member practices use the Epic EHR, which allows staff messaging. When an SCC patient is admitted to the Stanford hospital, notification is sent to SCC. The SCC team may see patients during their hospital stay and/or in their homes after hospitalization to help with discharge care. The SCC nurse conducts home visits using the Care Transitions Program® developed by Eric Coleman, MD, MPH. The social worker or community project coordinator refers the patient to community resources such as the Chronic Disease Self-Management Program (CDSMP) or a Better Choices, Better Health® Workshop at the Stanford University campus that is taught by the community project coordinator.
The SCC intensivist model is now well developed. When asked what type of additional staff they would add at this point, the providers reported they would add a consulting psychiatrist “on quick dial” for help with diagnosis more than drug management.
Team Building and Training
Interviewees emphasized that all team members understand the clearly defined care goals, and the division of labor with defined roles. Team members are trained to use quality improvement (QI) methodology and collect QI data. Training is ongoing and mostly on-the-job, a strategy that aims to build trust between team members. Co-location and co-charting in the team room provides constant training and QI exposure, ensuring that all staff work at the top of license.
New care coordinators shadow an experienced care coordinator for a month, and receive training in the Epic EHR and other systems once their competencies are assessed. Dr. Glaseroff said that patients view care coordinators as part of the care team. Because they are in the exam room scribing for the physician, hearing the physician-patient conversation, and co-charting with the physician, care coordinators said they become familiar with the flow of care processes and know what to ask patients when requesting their histories and on telephone calls. Care coordinators meet and debrief often, providing opportunities to learn from experience. Friday team meetings include training on a topic determined by the staff.
Before SCC opened, a 2-day motivational interviewing (MI) training was held with the whole team. MI is a behavior-change method that encourages patient-provider communication; it is beyond most health professionals' regular training. MI training incorporates team-building exercises that promote better communication and break down the walls between team members. For care coordinators in particular, MI may enable them to elicit information the patient may not share with the doctor.
“Training here [is] very open. We receive training on medical conditions like diabetes, depression, domestic violence, and skills like phlebotomy, EKG [electrocardiogram] training, point-of-care testing, how to read the results, normal ranges, etc. We are also trained on protocols for vaccinations, [and] preventive screening. Case conferences and team meetings are another way of learning.”
-SCC care coordinator
Access, Comprehensiveness, and Quality
SCC is a “high-touch” practice. Average communication frequency with each patient is weekly, with most touches occurring by email. The practice maintains a “meaningful-touches tracker” to document the number and types of interactions each team has with its panel. Each provider team follows up with patients as needed. Communication data provided by the SCC team are below.
Daily Average of Touches Received by the SCC Team
- 11 incoming phone calls
- 6 voicemails
- 18 MyHealth messages
- 8 faxes
- 1 walk-in patient
Daily Average of Touches Provided by the SCC Team
- Made 22 outgoing phone calls
- Left 7 voicemails
- Sent 26 MyHealth messages
- Sent 5 faxes
SCC engages the community by working with specialists and offering health education opportunities. Specialty care referrals are made to Stanford Health Care or Palo Alto Medical Foundation specialists. Care for SCC patients referred to specialists is still managed by SCC.
The community project coordinator in the practice, trained in Stanford's CDSMP, currently leads a CDSMP workshop on campus for any interested participants.
SCC nurses engage with the community through home visits where they use the Care Transitions Program®.
Electronic Health Record
The practice uses the Patient Care Coordination note field in Epic to document patient goals, action plans, and followup arrangements. These are updated at every visit and placed into the progress note field. Ongoing data for each patient problem is entered into the overview section so other providers can see pertinent patient history without exhaustive chart review. Finished notes go into the patient instruction field and are stored in the patient record, uploaded to the patient portal, or mailed to patients not registered for the portal.
SCC receives reports on performance measures from the Stanford Health Care system with comparisons to other primary care practices to use as benchmarks. Examples of measures are the number of telephone calls answered within a specific time, percentage of patients scheduled on first call, and percentage of patients within specific clinical goals (e.g., hemoglobin A1c, low-density lipoprotein control, influenza vaccinations, screenings), and percent of patients on certain medications. Patient Risk Dashboards show the number of patients who are in “red zones" meaning they need followup for pain, smoking, body mass index, alcohol use, and other conditions. Stanford Health Plan subscribes to the Press Ganey Patient Satisfaction Surveys and scores are compiled for each practice for comparison.
Epic and a Stanford central data warehouse provide the Healthcare Effectiveness Data and Information Set (HEDIS) care gaps tool used by the practice. The central data warehouse sends SCC a weekly ”red report” on all patients; that includes patients whose illness is not in control and the tests and assessments they need. This information also appears in the SCC dashboard. SCC physicians and MAs quickly review all patients in need of services or clinically at risk.
The team aspires to achieve the Triple Aim of improved patient experience, quality of care, and lower cost, with the added dimension of staff satisfaction. For the two last reported quarters of 2014, patient experience ratings were in the 99th percentile, HEDIS measures were greater than the 90th percentile, and staff satisfaction was in the 99th percentile.
Additionally, the practice collects and uses QI and utilization data. These data demonstrated a 59% reduction in emergency department (ED) visits and a 29% reduction in hospital admissions between January and December 2014.
The practice focuses on attending to the biopsychosocial needs of their patients. The SCC Patient Advisory Committee, a group that was formed before the clinic began and was integral to its planning, meets every 2 months to respond to questions and issues. SCC has a suggestion box and has used a poker-chip method of measuring patient satisfaction suggested by the Patient Advisory Committee. In this method, when patients leave the office after a visit, they receive a poker chip to drop into one of five containers marked “poor” to “excellent.” The purpose of this method is to ascertain patient satisfaction in real time rather than waiting for patient survey reports that arrive well after events. SCC ended the poker-chip experiment after a few weeks because the vast majority of ratings were excellent, which did not help them find ways to improve.
“We interviewed patients to determine their needs, matched our services to them. Our patients are holding us to our promise.”
Implications for Primary Care Staffing Models
SCC tailored its assessment and categorization of patients to a high-risk, complex patient panel and aligned the composition of its care team to meet patients' needs almost entirely within the primary care practice. The practice was designed with human-centered techniques and maintains a Patient Advisory Committee to meet patient needs. The MA/care coordinators each have their own panel and one of their duties is to scribe during physician visits. This facilitates in-depth knowledge of patients and comprehensive service coordination and optimizes efficiency by offloading tasks from physicians.
SCC plans to spread its model in collaboration with the larger Stanford Health Plan. The practice will not be replicated wholesale, but modified for less complex patient panels. Spread to standard primary care practices might mean that care coordinators in a central and enhanced MA role would manage a panel of as many as 500 patients rather than the 100 at SCC. In teamlet models, which pair MAs and physicians in caring for a group of patients, patient panels are typically larger, for example 1500, although this size would be too large to develop SCC-style relationships. Stanford Health Plan's current projection for standard primary care practice anticipates a core team of four FTE providers, 20 care coordinators, and two RNs. This team would care for a panel of 10,000 patients. The next steps are currently in development.
We thank the entire staff of Stanford Coordinated Care, especially Ann Lindsay, MD, and Alan Glaseroff, MD, for hosting a site visit for the AHRQ New Models of Primary Care Workforce and Financing project.
1. Wu SY, Green A. Projection of Chronic Illness Prevalence and Cost Inflation (Santa Monica, Calif.: RAND, October 2000)., as cited in Bodenheimer T, Chen E, Bennett HD. Confronting the growing burden of chronic disease: can the U.S. health care workforce do the job? Health Aff (Millwood). 2009 Jan-Feb;28(1):64-74.