Behavioral Health Integration and the Psychiatric NP Role Manuel A. Castro, MD, Medical Director of BHI Kathleen Peniston, PMHNP-BC
Objectives Understand the growing need to integrate behavioral health into Primary Care Provider s Office. Understand the CHS Virtual Model for BH Integration Understand how standardization of screening tools and treatment algorithms are critical to improving patient care within ambulatory care settings. Identify tools and measurements for evaluating the effectiveness of an integrated behavioral health program. Understand and describe the role of the Psychiatric NP as part of the BH Integration Team.
Behavioral Health Integration at CHS https://www.youtube.com/watch?v=6- wmz0wjw2i
Our Story Clinically Integrated Why? The Case for Virtual Care Virtual Care Models in Hospital Acute Care Behavioral Health Primary Care Integration Outcomes Lessons Learned and Next Steps
Achieving the Triple Aim Carolinas HealthCare System will develop a transformative, clinically integrated and sustainable system of high quality, patient and family centered care to serve the behavioral health needs of patients, their families and the community. 5
Carolinas HealthCare System 39 hospitals and 900+ care locations in North Carolina, South Carolina and Georgia More than 7,800 licensed beds More than 11 million patient encounters in 2013 3,000+ system-employed physicians, 14,000+ nurses and more than 60,000 employees $1.5 billion in community benefit in 2013 More than $8 billion in annual revenue The region s only Level I trauma center One of five academic medical centers in North Carolina One of the largest HIT and EMR systems in the country Cerner s largest contract
The State of Behavioral Health One in five adults suffers from a diagnosable mental disorder. The average annual Medicaid spend per person is only $4,000, but that jumps to $38,000 annually with one mental health and one substance use diagnosis. Untreated mental health and substance abuse disorders cost the US $250-$500 billion per year. $193 billion per year in lost workplace earnings due to untreated mental illness. Even beyond the United States, mental illness is the #1 cause of disability life years worldwide, vastly outnumbering those caused by cardiovascular disease and cancer. With proposer diagnosis and effective treatment, the recovery rate for patients with mental illness is 60-80%. But in today s environment, the effective recovery rate is only 5-10% due to such limited resources and infrastructure.
All Healthcare Spending $2.7 Trillion Behavioral Health $100 150 Billion
1971 Spending on behavioral health is shrinking relative to overall healthcare spending 1991 2001 2011 11.1 8.4 5.9 4.6 Share of national health spending
Psychiatry Workforce US: 40,000 Psychiatrist 9,000 Psychiatric NPs Most are located in Urban Areas Half of all the counties in the US don t have a single practicing Mental Health professional. 29-1066 Psychiatrists." U.S. Bureau of Labor Statistics. U.S. Bureau of Labor Statistics, n.d. Web. 30 Nov. 2014. AANP National NP Database 2016 10
Psychiatry Workforce
Psychiatric Facilities - Nationwide 12
NC Mental Health Workforce Shortage 95 counties have no addiction psychiatrist, 87 without any MD with addiction specialty. 70 counties have no Child and Adolescent Psychiatrist. 60 counties across North Carolina nave no practicing Psychiatric NPs. 28 counties across North Carolina have no practicing Psychiatrist. 18 counties across North Carolina have only one. www.ihealthbeat.org NC Health Professions Data System, with data from the NC BON 13
Fragmented System Unique emr Referral Coordinat or PCP Patient Medic al ED Psyc h ED Psychiatri st Transpor t 13
Crisis = Call to Action
Why Primary Care? Stigma is lower Roughly 80% of all antidepressants are prescribed by non-psychiatrists 70% of visits are Psycho-Socially related 1 Million Active Patients Annals of General Psychiatry. 14 (13) 2015,WJM Jan 170, No.1, 1999 Psychosomatics 41:5 Sept 2000, Psychiatric Services Jan 2006
We couldn t get a psychiatrist, but perhaps you d like to talk about your skin. Dr. Perry here is a dermatologist. 2/3 of PCP s report poor access to mental health services for their patients
PERCENT Prevalence of Major Depressive Disorder in Chronic Disease 60% 50% 51% 40% 42% 30% 20% 10% 11% 17% 23% 27% 0% Alzheimer's Cardiovascular Cerebrovascular Diabetes Cancer Parkinson's National Center for Chronic Disease Prevention and Health Population. Published 2012
Cost of a Chronic Disease with a Mental Illness Absent Comorbity $18,870 $17,200 $10,030 $3,730 $11,650 $9,770 $10,980 $12,280 $3,840 $5,480 A STHMA CA NCER CHF COPD D IABETES
40,000 suicides annually 64% of these patients saw their Primary Care Provider within 30 days. 38% people made a healthcare visit the previous week. Estimated cost of each Suicide $1 million American Foundation for Suicide Prevention: http://www.visualwebcaster.com/apaintegratedcare
Behavioral Health Integration: IMPACT PCP Patient BHP/Care Manager New Roles Core Program Consulting Psychiatrist Other Behavioral Health Clinicians Additional Clinic Resources Substance Treatment, Vocational Rehabilitation, CMHC, Other Community Resources Outside Resources 22
BHP Deconstructing the BHP by Function BH Screener BH Diagnoser BH Treater BH Navigator BH Coach/Activator BH Communicator
IMPACT / Collaborative Care Model 2 year Randomized Control Trial: 1801 Adults with Depression 12 months-50% reduction of depressive symptoms 45% IMPACT model 19% usual care participants 4 years $3,300 in savings in health care spend per patient Repeated in 80 Randomized Trials
24-Month Health Care Costs IMPACT Enrollees vs. Control Patients
Going Virtual 26
Consumerism Patient feedback: I appreciate having this service available at my primary care. I would much rather get taken care of at one place (my PCP office) rather than have to drive especially to Charlotte. This is the coolest thing ever! Virtual Heart Success CMC Lincoln HF 30 Day All Cause Readmission Rate 2010-2013 40 20 0 19.39 19.77 19.61 19.56 1.01 1.05 0.90 17.680.50 9.82 19.24 18.83 2010 2011 2012 2013 ytd Readmissi on Observed
Virtual Team in ED 4,688 patients received a Telepsych consult in 2014 (75% increase) Currently offering Virtual Care services in 21 Emergency Departments 28
CHS Telepsychiatry Growth and Projected Volume for 2017 1400 1200 1000 800 600 400 200 1138 1152 1121 1150 1175 1204 1225 1078 948 960 934 958 980 1000 1021 940 877 886 898 793 783 636 626 641 668 710 747 766 783 802 816 731 738 718 661 558 538 534 499 345 350 354 396 419 422 272 264 295 258 275 256 139 151 166 196 242 247 240 242 261 0 2013 2014 2015 2016 2017
BHI Team Call Back Protocol Motivational Interviewing Symptom Management Mindfulness Chart Reviews Patient Education Online Therapy Health Coach BHP (LCSW, LPC) Patient Education Psych Assessments Diagnostic Clarification Crisis Intervention Tx Planning Psychoeducation Practice Onboarding Educational Services Chart Reviews Clinical support and oversight for team Evidenced Based Medication Algorithm Psychiatrist ACP Pharm D Patient and Provider Education Access to Medication Chart Review Evidenced Based Medication Algorithm
Why a Psychiatric Nurse Practitioner
Our Model We ve got your back The key to making teambased medical care work is helping the patient feel that his or her relationship with the primary-care provider is at its center. What we want to accomplish: Improve early detection Timely access to services Referrals to the right level of care Drive cost effective & clinically effective treatment Support the Primary Care Provider Support the Patient Suzanne Koven is a primary care doctor at Massachusetts General Hospital in Boston and writes the column "In Practice" at the Boston Globe.
Screening is the Driver
Treatment and Patient Engagement Evidenced Based Treatment Standardized tools in the PCP setting enhances screening diagnosis, and treatment planning Patient Engagement Recovery
Technology Utilized
Program Outcomes: Return on Investment Disease Severity Clinical Outcomes Healthcare Utilization Cost of Care Depression symptoms Weight/BMI Inpatient visits Overall Anxiety symptoms Suicide ideations HgB A1C Cholesterol (Total, triglycerides, LDL, HDL) Inpatient days ED visits Inpatient Ambulatory Ambulatory visits (Primary/specialty) ED Avoidable ED/IP visits
Evaluation Framework P R E P O S T Clinical Indicators (labs) were compared at baseline and 12 months after enrollment. If patient had multiple clinical test results the score closest to baseline (but prior to BHI enrollment) was considered a baseline measurement and score closest to the end of 1-year mark was considered as post measurement. Healthcare utilization and charges were also compared for 1-year pre and 1-year post- enrollment time windows
Depression Compares baseline and the latest reported PHQ-9 scores *p-value <.05 indicates statistically significant change within the same patients overtime (Pre-post analysis using paired t-test procedure) ** Patient who had at least 6 months follow-up, started out with elevated PHQ-9 score (>=10) and had at least 30 days between the first and last assessment 38
Anxiety Compares baseline and the latest reported GAD-7 scores *p-value <.05 indicates statistically significant change within the same patients overtime (Pre-post analysis using paired t-test procedure) ** Patient who had at least 6 months follow-up, started out with elevated GAD-7 score (>=10) and had at least 30 days between the first and last assessment 39
Suicidal Ideations* Compares baseline and the latest reported PHQ-9 (Q9) or CSSRS assessments * Based on Question 9 on the PHQ-9 Scale and CSSRS Assessments ** Patient who had at least 6 months follow-up and at least 30 days between the first and last assessment ***p-value <.05 indicates statistically significant change 40
Weight n=980 Mean (± Standard Deviation) Mean change p-value* Baseline 192.8 (±57.8) 6 months 194.8 (±57.7) 2.0 (±18.4) p=.0006 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 41
BMI Mean (± Standard Deviation) Mean change p-value* Baseline 30.9 (±8.7) 1-year 31.2 (±8.6) 0.32 (±2.8) p=.0006 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 42
HgB A1C Mean (± Standard Deviation) Mean change p-value* Baseline 7.8 (±2.3) 1-year 7.4 (±1.8) 0.4 (±1.8) p=.013 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 43
Lipids: Total Cholesterol Mean (± Standard Deviation) Mean change p-value* Baseline 190.4 (±53.4) 1-year 179 (±43.4) 11.5 (±41.9) p=.009 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 44
Lipids: Triglycerides Mean (± Standard Deviation) Mean change p-value* Baseline 166.9 (±106.9) 1-year 173.1 (±123.7) 6.2 (±88.7) p=.5 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 45
Lipids: LDL Mean (± Standard Deviation) Mean change p-value* Baseline 109.6 (± 44) 1-year 98 (± 37.9) 11.7 (±31.2) p=.0007 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 46
Lipids: HDL Mean (± Standard Deviation) Mean change p-value* Baseline 47.3 (± 13.9) 1-year 46.4 (± 14.2) 0.87 (±8.5) p=.32 *p-value <.05 indicates statistically significant change (statistical significance doesn t necessarily indicate clinical significance) NOTE: To allow for pre-post comparison patients had to have at least 2 measurements (explains decrease in sample size) 47
Healthcare Utilization: Inpatient Care 12% of patients had at least one IP admission 1-year prior or 1-year post-enrollment (averaging 0.2 visits per patient) Only 3% of patients had more than 1 IP stay Overall length of IP stay was 1285 days pre- and 1433 days post-intervention, respectively There was a $930,050 increase in overall billed charges that could have been caused by increase in total IP days in the year after intervention 48
Healthcare Utilization: Avoidable Inpatient Care There was 25% reduction in avoidable inpatient visits (from 53 visits pre- to 38 visits post-intervention). Inpatient visits were classified as avoidable using AHRQ Prevention Quality Indicator (PQI) methodology 49 The decrease in avoidable visits resulted in $130,833 savings in total billed charges (average of $8722 saved per patient)
Healthcare Utilization: ED Avg $480 per person Avg $690 per person Total number of ED only visits that did not result in admission decreased by ~ 8% (109 less visits) 36% of the patients had at least one ED visit 1-year before the program start date averaging 1(± 2.6) visits per person 34% had at least one ED visit 1-year after the program start date averaging 0.9 (± 2.3) visits per person About 10% of the patients had 3 or more visit a year Although number of overall ED visits decreased the complexity of these visit must have increased resulting in of average billed charges of $690 per visit 1-year after intervention compared to $480 1-year before intervention 50
Healthcare Utilization: Ambulatory Overall billed charges associated with ambulatory care have increased by $470,077 averaging $343 increase per patient Increase in total charges concurred with the rise in total patient visits. Average ambulatory charge per person Mean (± Standard Devition) 1-year PRE $2534 (±3708) 1-year POST $2877 (±4193) Mean change $343 (± 4619) p-value* p=.006 *p-value <.05 indicates statistically significant change (statistical significance doesn t always indicate clinical significance) 51
System Investment - 2016 Incremental Salary Expense 1,268,791 salary and benefits BHI by the Numbers 7,049 Unique Patients 64,299 Patient Encounters 2,392 Patients currently under active mgmt. 21 Primary Care Practices 4 Pediatric Clinics 70 Care Management Clinics Teammate Interventions 2016 Health Coach [VALUE] Total Patient Interventions 64,299 MD/ACP [VALUE] BHP [VALUE] Existing Resources Utilized 1 Manager 1 Program Coordinator 1 On Boarding Specialist 12 BHPs + 2 PRN 13 Health Coaches Per Patient Expense = $183.56 PharmD/BHI Support 0.2%
Next Steps / Lessons Learned Next Steps: Expansion into Pediatrics Standardize screenings across Pediatric and OB/GYN clinics. Increase patients enrolled by expanding the program to incorporate additional CHS Pediatric Clinics Capture Parent depression screenings to provide services to the entire family. Incorporate other tools, such as texting to engage patients in our Patient Outreach Program. Virtual Therapy, Transition Clinics and Virtual BH Outpatient Sites. Lessons Learned: Leadership support. One EMR is very helpful. Standardize screenings Who When Screener in the EMR It Takes a Village!!!
Contact Information Kathleen.Peniston@carolinashealthcare.org Manuel.Castro@carolinashealthcare.org 2017 NPSS ASHEVILLE, NC