Can Primary Care Teams Improve Patient Access and

Can Primary Care Teams Improve Patient Access and Reduce Physician Burnout? Tom Bodenheimer MD Center for Excellence in Primary Care Department of Family and Community Medicine University of California, San Francisco TBodenheimer@fcm. ucsf. edu

Learning objectives By the end of this session participants will be able to: 1. Understand the concept of a primary care demandcapacity gap 2. Identify evidence supporting registered nurses and pharmacists providing care independent of physicians 3. Incorporate examples of high performing teams with the potential to improve patient access while reducing physician burnout

What we’ll talk about Primary care’s dilemmas Learn from the bright spots Bright-spot primary care teams Teaching practices

Dilemmas facing primary care Many patients lack prompt access to primary care because the number of primary care clinicians is insufficient to meet the population’s demand for care Yet physicians face high levels of burnout and cannot additional capacity Primary care teams -- that empower team members to share responsibility for patient care – could add capacity without increasing physician stress However, patients desire a trusting relationship with a personal physician; will patients accept care provided by teams? In summary, primary care is challenged to achieve 3 goals • Prompt access • Physician well-being • Continuous trusting relationships with patients

3 key questions Can teams add capacity? Can teams reduce burnout? Will patients accept teams? First, a few basic facts

Projected generalist physician supply vs. demand for US adults Shortage of 17, 000 by 2025 Demand: adult pop’n growth/aging Supply: family docs, general internal medicine docs Colwill et al. , Health Affairs, 2008: w 232 Petterson et al, Ann Fam Med 2015; 13: 107

NP/PAs to the rescue? New graduates each year % going into primary care • Nurse practitioners: about 14, 000 • Physician assistants: about 6, 000 • NPs: 50% • PAs: 32% Adding new GIM, Fam. Med, NPs, and PAs entering primary care each year, the primary care clinician to population ratio will still fall from 2005 to 2020. Colwill et al, Health Affairs Web Exclusive, April 29, 2008; Bodenheimer et al, Health Affairs 2009; 28: 64; Petterson et al, Ann Fam Med 2015; 13: 107.

Physician burnout Survey of 422 general internists and family physicians • 27%: definitely burning out • 30%: likely to leave the practice within 2 years Physician burnout is associated with poor patient experience and reduced patient adherence to treatment plans Linzer et al. Annals of Internal Medicine 2009; 151: 28 -36; Dyrbye, JAMA 2011; 305: 2009; Murray et al, JGIM 2001: 16, 452; Landon et al, Med Care 2006; 44: 234; Bodenheimer, Sinsky, Ann Fam Med 2014; 12: 573.

Panel sizes too large to manage Average primary care panel in US is about 2000 per FTE physician PCP with panel of 2000 average patients will spend 6 hours per day doing recommended preventive care [Yarnall et al. Am J Public Health 2003; 93: 635] PCP with panel of 2000 average patients will spend 8. 5 hours per day doing recommended chronic care [Ostbye et al. Annals of Fam Med 2005; 3: 209] If no team Large panel sizes without a team are a major contributor to poor access (demand > capacity) and clinician burnout

What do patients want from physicians? Detsky AS, JAMA 2011; 306: 2500; Safran DG, Ann Intern Med 2003; 138: 248 Competen ce • I want my physician to have the knowledg e needed to help me Empathy • I want my physicia n to care about me Familiarit y • I want to know my physician ; I want my physician to know me Continuit y • I want to see my personal physicia n when I need help Can we build teams that 1) Allow non-physicians to provide these four patient needs while 2) Adding capacity and 3) Reducing clinician burnout?

What we’ll talk about Primary care’s dilemmas Learn from the bright spots Bright-spot primary care teams Teaching practices

Learning from the bright spots In 1990 an organizer with Save the Children went to Vietnam to help reduce child malnutrition. The government told him: You have 6 months to make a difference. An impossible task. He went to rural villages and asked mothers to weigh every child in the village. He interviewed poor mothers with well-nourished kids and learned what they did. Then he organized the high-performing mothers to spread their practices to families with malnourished children. Chip and Dan Heath, How to Change Things When Change Is Hard

Learn from the bright spots The organizer found the bright spots and arranged for the bright spots to illuminate the dark corners In primary care there are bright spots and dark corners Bright spots exist at 3 levels: High-performing practices High-performing individuals High-performing teams

Learning from 23 bright-spot practices Martin’s Point. Evergreen Woods Group Health Olympia Fairview Rosemont Clinic Multnomah County Health Dept Harvard Vanguard Medford Theda. Care Allina BWH, MGH Amb Practi of the Future Mayo Red Cedar Medical Associates Clinic Ole Clinica Family Health Services Sebastopol Community Health Univ of Utah. Redstone La Clinica de la Raza Cleveland Clinic. Stonebridge North Shore Physicians Group Mercy Clinics Quincy, Office of the Future Newport News Family Practice West Los Angeles. VA South Central Foundation Bodenheimer et al, Ann Fam Med 2014: 12: 166 Sinsky et al, Ann Fam Med 2013: 11: 272

From these 23 bright-spot practices, we observed several common features The 10 Building Blocks of High-Performing Primary Care Bodenheimer et al, Ann Fam Med 2014: 12: 166 Teams

Bright spot primary care residency programs • We have visited 18 highlyregarded primary care residency clinics • They have different challenges than non-teaching clinics • We will discuss later in this presentation

What we’ll talk about Primary care’s dilemmas Learn from the bright spots Bright-spot primary care teams Teaching practices

Teams are difficult. Why bother? Who has worked with a primary care team? Did any team members independently care for patients without clinician involvement? Did the team improve any quality measures over what a lone clinician could achieve? If we want to succeed as a team, we need to put aside our own selfish, individual interests and start doing things my way Did the team reduce the work of any clinician? Did the team add capacity to see more patients without causing more work for a clinician? If the answers are all No, the team is not worth having

The 9 elements of high-performing teams Stable team structure Co-location Culture shift: share the care Defined roles with training and skills checks Standing orders Defined workflows Staffing ratios adequate to allow new roles Ground rules Communication: team meetings, huddles, minuteto-minute interactions

Stable team structure: teamlets Patient panel Clinician + MA teamlet RN, behavioral health professional, social worker, pharmacist, complex care manager 1 team, 3 teamlets

Primary care teaching practice model Patient Panel 1200 MA/LVN Resident Panel 300 Patient Panel 1200 NP/PA Continuity clinician Resident Panel 300 MA/LVN Resident Panel 300 NP/PA Continuity clinician 4 Residents Patient Panel 1200 MA/LVN NP/PA Continuity clinician MA/LVN 4 Residents 3 teamlets: each teamlet consisting of full-time NP/PA or faculty physician and 2 support personnel responsible for 1200 patients Each resident co-manages 300 of the 1200 patients with his/her teamlet

What is meant by stable teams/teamlets? In our visits to bright-spot practices, many had implemented the teamlet model The same people always work together Patients empaneled to a teamlet are always cared for by that team/teamlet Scheduling concerns often lead practices to have patients seen by different teams and team members to work in different teams on different days In bright-spot practices, team stability trumps scheduling convenience

Teamlets and patients’ needs (competence, empathy, familiarity, continuity) Patients prefer small practices over large practices [Rubin et al, JAMA 1993; 270: 835]. Teamlets break large practices into smaller entities that patients feel comfortable with Teams always work together; patients always seen by their team Patients get to know and trust their teamlet partners If services beyond what teamlet can provide, RNs, behaviorists, pharmacists, health coaches support several teamlets Small practices: the additional team members are in medical neighborhood: hospital, IPA, or health plan

Co-location Entire team works in one space (pod) Each teamlet sits together (clinician and MA) Minute to minute interaction is constant and easy Increases team collaboration and coordination [Mc. Naughton et al. BMC Health Serv Res 2013; 13: 486] Doctors first resist, then love co-location Can be done without changing architecture

Culture shift: share the care Share the care is a culture shift • From “I” -- clinician makes all decisions and non-clinician staff helps the clinician • To “We” -- the entire team shares responsibility for the health of their patient panel Sharing the care is not delegating tasks to non-clinician team members; it is re-allocating responsibilities Will all clinicians agree to share the care? Will all RNs, LPNs, MAs want to assume new responsibilities? • Of course not Start with bright spots -- enthusiastic clinicians, residents, and team members. More on sharing the care soon

Share the care: who does it now?

The 9 elements of high-performing teams ✔Stable team structure ✔Colocation ✔Culture shift: share the care Defined roles with training and skills checks Standing orders Defined workflows Staffing ratios adequate to allow new roles Ground rules Communication: team meetings, huddles, minuteto-minute interactions

You can’t share the care without standing orders RNs do diabetes refills without involving clinicians Appointme nt last 6 months Hb. A 1 c = 7. 5 or below Normal creatinine and potassium in last 6 months Yes How to refill Yes 3 month supply + 1 refill Yes or No No 1 month supply + order labs, give appt, no refill Yes No Yes 1 month supply + give appt, no refill No Yes 3 month supply + give appt, no refill No No Yes or No 1 month supply + give appt, no refill

Staffing ratios adequate to allow new roles You can’t share the care without adequate staff VA: adequately staffed care teams are associated with lower clinician burnout [Helfrich et al, Primary care receives a mere 5% of the US health care dollar [Mostashari Average staffing in US: 2. 68 staff/physician. To implement PCMH requires 4. 25 staff/physician [Patel et Insurers and provider organizations should pay 10 -12% of health care dollar to primary care JGIM 2013; 29 suppl 2: S 659], Group Health Cooperative in Seattle received 13% in 2010 [Koller et al. Health Affairs 2010; 29: 941] al Am J Manag Care 2013; 1`9: 509] A metric that should be collected: % of the health care dollar received by primary care et al, JAMA 2014; 311: 1855]

3 key questions Can teams add capacity? Can teams reduce burnout? Will patients accept teams?

Adding capacity with large panel sizes Dramatic shift from clinician-only care to sharing the care 2 share-the-care innovations to reduce burnout at same panel size • Widespread implementation of team documentation (scribes) • Re-allocating responsibility for subpanels of patients to RNs, pharmacists, physical therapists Barriers • F-f-s payment does not reimburse nonclinicians • Investment of time training and mentoring teams • Scope of practice laws • Physician acceptance • Staffing ratios (people to create teams) Paralysis with barriers. Make plan first, barrier worries later

Scribes to reduce documentation time: University of Utah “Care by Design” MA/scribe receives additional training MA/scribe takes history using EMR templates MD calls out lab, imaging, rx; MA enters them as pended orders that MD quickly OKs MDs see more patients (capacity up) and go home earlier (burnout down) Blash et al, UCSF Center for the Health Professions, April 2011 MD quickly reviews history, does physical exam, MA enters findings into EMR Revenue up, patient satisfaction up, provider satisfaction up, quality measures up

Scribes to reduce documentation time: UCLA internal medicine Reuben et al, JAMA Int Med 2014; 174: 1190 Study of 2 scribes (“physician partners”), one LVN, one college educated 75 minutes of physician time saved in each 4 -hour clinic session 79% of patients satisfied Patients more likely to report that physician spent enough time with them

Re-allocating responsibility to non-clinician professional team members Assume panel of 2000, creating 6000 visits/year • • • 1000 visits by patients with diabetes 1000 visits by patients with hypertension 1000 visits for uncomplicated low-back, knee, shoulder pain Assume RNs, pharmacists, PTs can independently care for 2/3 of these visits (no clinician needed) • Total non-clinician visits = 2000 Each clinician provides 4000 rather than 6000 visits/year Burnout drops because clinicians have fewer visits per day

Some evidence for re-allocating responsibilities RNs: RCT of patients with diabetes and elevated BP. Patients with RN management (including initiating meds and titrating doses) 3 times more likely to reach BP goal (p =. 003) than physician management [Denver et al, Diabetes Care 2003; 26: 2256] Pharmacists: RCT of pharmacist management of hypertension (including medications) compared with usual care. At 18 months, 72% BP control for pharmacist care vs. 57% in usual care group (p=. 003) [Margolis et al, JAMA 2013; 310: 46]

Some evidence for re-allocating responsibilities Patients with uncomplicated musculoskeletal injuries who directly access physical therapists without seeing a physician have better functional outcomes, greater satisfaction, and lower health care costs. [Ojha et al, Physical Therapy 2014; 94: 14; Overman et al, Phys Ther 1988; 68: 199]. Primary care behaviorists working as depression care managers in primary care improve depression outcomes compared with physician-only care and can reduce physician visits [Unutzer and Park, Primary Care 2012; 39: 415]

Some evidence for re-allocating responsibilities RNs, NPs, or PAs as complex care managers for patients with multidiagnosis, highutilizing patients can improve care, cut costs, and reduce clinician time spent • Bodenheimer T, Berry-Millett R. Care Management for Patients with Complex Healthcare Needs, Robert Wood Johnson Foundation, 2009. • Bodenheimer T. Strategies to Reduce Costs and Improve Care for High-Utilizing Medicaid Patients: Reflections on Pioneering Programs. Center for Health Care Strategies, 2013. • Hong C et al. Caring for High-Need, High-Cost Patients: What Makes for a Successful Care Management Program? Commonwealth Fund, August 2014.

Is there a business case for teams? Fee-forservice Payments on top of fee-forservice • Saving physician time allows physicians to see 1 – 2 extra patients per day (scribes) • RN visits: co-visits with physicians spending a few minutes and billing • for P 4 Pthe visit • RN doing Medicare annual wellness visits • RN providing Medicare chronic care management • PCMH additional payments Non fee-for-service (capitation or shared savings) decouples payment from visits

3 key questions Can teams add capacity? Can teams reduce burnout? Will patients accept teams?

Teams and burnout Some evidence that teams can save clinician time and reduce burnout: The less time clinicians do tasks below their level of training, the lower the burnout [Helfrich et al, JGIM 2013; 29 suppl 2: S 659] When nursing personnel take histories, stay in the visit to document care, arrange lab, imaging, pharmacy, patient education - physicians see more patients, more satisfied, increase revenues, go home earlier [Anderson and Halley, Fam Pract Manage 2008; 15(7): 35] Scribes performing EMR documentation saved general internists 75 minutes in a 4 -hour clinic session; better patient/clinician satisfaction [Reuben et al, JAMA Int Med 2014; 174: 1190]

Clinician Satisfaction with Teams Teamlet (work with same MA) Not satisfied 15% Neutral 15% Satisfied 70% Team (work with group of MAs) Not satisfied 35% Satisfied 37% No team (work with different MAs) Satisfied 11% Neutral 28% Not satisfied 61% UCSF Center for Excellence in Primary Care, 2012 survey of 16 primary care clinics

Team structure and clinician burnout 6. 00 Exhaustion Mean Score 5. 00 4. 00 Low Team Culture 3. 00 High Team Culture 2. 00 1. 00 0. 00 No Team Structure Teamlet Among clinicians reporting high team culture, stable teamlets were associated with lower exhaustion on the Maslach Burnout Inventory (p =. 002). Willard-Grace, J Am Bd Fam Med 2014; 27: 229 -238.

Clinician confidence that medical assistants will do a good job taking responsibility for panel 60% management 50% 48% 42% 40% 30% 29% 27% 20% 20% 13% 11% 10% 0% Immunizations Teamlet (n=26 -27) Cancer screenings Team (n=88 -91) Diabetes care No team (n=15) UCSF Center for Excellence in Primary Care, 2012 survey of 16 primary care clinics

3 key questions Can teams add capacity? Can teams reduce burnout? Will patients accept teams?

What do patients want from physicians? Detsky AS, JAMA 2011; 306: 2500; Safran DG, Ann Intern Med 2003; 138: 248 Competen ce • I want my physician to have the knowledg e needed to help me Empathy • I want my physicia n to care about me Familiarit y • I want to know my physician ; I want my physician to know me Continuit y • I want to see my personal physicia n when I need help Can patients get these things from non-physician team members?

Will patients accept care provided by non-clinician team members? Patients accept teams if • 1) The teams are visible and patients are always cared for by their own team • 2) The teams are small like a small practice (teamlets) • 3) Patients know their team members and the team members know their patients [Rodriguez et al, Medical Care 2007; 45: 19] California phone household survey Patients can trust nurses, physical therapists, and pharmacists, and just as patients can trust clinicians • • Familiarity andare continuity 81% surveyed willing to receive care from a team • Even if it means seeing their physician less • Blue Shield of California Foundation, Connectedness and Continuity, June 2012 • If they show empathy • If patients view them as competent • Dinc and Gastmans, Nursing Ethics 2013; 20: 501; Hall et al, Physical Therapy 2010; 90: 1099; Al. Ghurair et al, Patient Preference and Adherence 2012; 6: 663

What we talked about Primary care’s dilemmas Learn from the bright spots Bright-spot primary care teams Teaching practices

Unique Considerations of Teaching Clinics Faculty physicians and residents spend only 1 – 2 halfdays in teaching clinic Leads to challenges with: • Continuity • Access • Team based care

Traditional versus teamlet model Traditional model Over the week: Teamlet model

Resident room Co-location MA room Co-located workroom Attendings, RN, SW

Creating stable teams in residency teaching practices To succeed requires a new residency teaching program paradigm Not “hospital first, clinic second” Clinic First • Ambulatory learning is a top priority • Creating an operationally excellent clinic is paramount • The clinic is the curriculum

Creating stable teams in residency teaching practices Prioritize stable clinic teams so that residents and patients have the best possible continuity and feel comfortable in the clinic Consistent resident schedules to prioritize continuity and allow stable teams Separate inpatient and outpatient resident responsibilities Develop small core of clinic faculty that cares about the clinic and leads the teams Gupta R et al The road to excellence for primary care resident teaching clinics. Acad Med 2016, in press.

Take-home messages With stable teams that share the care • We can meet patient demand without more clinician burnout • We can care for our patients with competence and empathy, thereby deserving patients’ trust It is more difficult in resident teaching clinics but it can be done Bright spot practices show that we can do it

Great Primary Care Is a Beautiful Thing TBodenheimer@fcm. ucsf. edu