Opening Plenary Session Welcoming Remarks: Neda Ratanawongsa, MD, - - PDF document

Opening Plenary Session

Welcoming Remarks: Neda Ratanawongsa, MD, MPH and Geraldine E. Menard, MD, SGIM 37th Annual Meeting Chair and Co-Chair Moderators: Michael Steinman, MD, Julie Rosenbaum, MD, Chair and Co-Chair, Scientific Abstracts, SGIM 37th Annual Meeting

SCIENTIFIC ABSTRACT:

EREFERRALS AND ECONSULTS: DOWNSTREAM IMPACT ON ACCESS, UTILIZATION, AND COST IN A FEE- FOR-SERVICE SETTING. Nathaniel Gleason1; Jennifer J. Monacelli6; Chanda Ho2; Sara Ackerman3; Priya A. Prasad4; Michael Wang5; Don Collado1; Delthia McKinney5; Ralph Gonzales1. 1UC San Francisco, San Francisco, CA; 2UC San Francisco, San Francisco, CA; 3UC San Francisco, San Francisco, CA; 4UC San Francisco, San Francisco, CA; 5UCSF Medical Center, San Francisco, CA; 6Michigan State University, Grand Rapids, MI. (Tracking ID #1939845) BACKGROUND: Demand for specialty care services exceeds supply at many academic referral centers. Delays in access to specialty care are common, and are anathema to patient-centered care. Electronic referrals (eReferrals) and non-face-to-face consultations (eConsults) are mechanisms associated with significant improvements in the specialty care referral process in integrated delivery systems with aligned incentives. Translating these innovations to academic medical centers (AMCs) has been challenging; reimbursement remains largely fee-for-service, and Medicare and most commercial payers do not pay for eConsults. With support from the Medicaid waiver program (DSRIP), UCSF Medical Center implemented a new referral platform. The system is designed to improve information exchange at the point-of-referral and to provide an eConsult option, allowing PCPs to request timely input from specialists for data-oriented referral problems that the PCP does not perceive to require in-person evaluation of the patient by the

• specialist. In this study, we analyzed the downstream impact of this program on PCP referral rate, specialty care access, and costs.

METHODS: eReferral/eConsult Platform: In May 2012, UCSF introduced a structured referral platform within the electronic health record (EHR) (Epic Systems) with over 90 problem-specific, structured templates providing clinical decision support at the point-of-

• referral. An eConsult option was added in September 2012. The expected eConsult response time is 3 business days. Upon review, if

the specialist determines the case is not appropriate for eConsult, the specialist converts the eConsult to a standard new-patient visit. Specialists receive a payment corresponding to 0.5 wRVU per completed eConsult. PCPs receive 0.5 wRVU credit toward productivity targets. in recognition of the fact that the PCP maintains management responsibility for the referral problem. Study populations: The program included 8 primary care sites and 12 medicine subspecialty practices. The study population includes all eligible PCPs (n=178) and all referrals to participating practices for patients assigned to an eligible PCP during baseline (9/1/11 - 8/31/12) and study periods (9/1/12 - 4/30/13) (n=13,738 referrals representing 11,597 unique patients). (To account for seasonal effects on ED visits and hospitalizations, we used baseline and intervention periods of Oct-April 2012 vs Oct-April 2013 for the utilization and cost analysis). We obtained data on referral rates from the EHR scheduling database. Data on utilization and costs were

• btained from the University Health System Consortium. Analysis: We calculated the total referral rate per 100 primary care visits per

month, and the proportion of these sent as eConsults. To compare the trends in referral rate, we modeled the effect of time using a linear spline analysis with two knots to determine whether there were differences in the trajectory following the introduction of the structured referral system and eConsults. To measure impact on access to specialty care, the proportion of UCSF primary care patients who received specialty care input (office visit or eConsult) within 14 business days in the baseline and eConsult periods were compared using chi-square tests. Mean monthly ambulatory, emergency department, hospital utilization and professional fee costs

• ccurring within 120 days following each referral or eConsult were log transformed to minimize effects of outlier data. Average costs

were compared between baseline (n=6 months) and eConsult (n=6 months) periods with unpaired t-tests. RESULTS: Of eligible PCPs, 69% placed at least one eConsult. Based on the spline analysis, the referral rate declined from 12.42 per 100 primary care visits during the baseline period to 10.66 per 100 primary care visits in August 2012 (11% decrease; p=0.011). eConsults were added as an option in September 2012, and the referral rate remained stable though the remainder of the study. eConsults were used for 8.2% of total referrals. The proportion of patients who received specialty care input (office visit or eConsult) within 14 business days improved from 39% to 49% (p=0.001) across all participating specialties. Specialties with greatest improvements were pulmonary (from 20% to 50% (p=0.004)), GI (from 19% to 40% (p=001)) and nephrology (from 50% to 74% (p=-.05)). The proportion of referrals with an ED visit during the 120-day period following referral or eConsult decreased from 9.8 to 8.6%, and there was a significant reduction in total ED pro-fee costs (p= 0.016). Mean ambulatory and hospital pro-fees did not change between periods. CONCLUSIONS: This combined eReferral (enhanced referral platform) and eConsult program showed robust adoption and had significant impact on referral rate and specialty care access time. The significant impact on ED costs may represent a downstream benefit of improved access to care.

CLINICAL PRACTICE INNOVATION:

BRIDGING THE CHASM-ADVANCED ILLNESS MANAGEMENT: HIGHER QUALITY, LOWER COST Sandy Balwan; Ramiro Jervis; Joseph Conigliaro; Kristofer L. Smith. North Shore LIJ Health System, Manhasset, NY. (Tracking ID #1936650) STATEMENT OF PROBLEM OR QUESTION: Can delivering in-home care, to elderly patients with multiple chronic illnesses and functional impairment, focused on prompt response to changes in clinical condition and high quality transitions, result in lower total cost of care? OBJECTIVES OF PROGRAM/INTERVENTION: 1) Leverage information technology including real time notification to Improve care coordination between hospital, physician group, and home/community services. 2) Create an on-demand program to respond to patient's changes in clinical status through same day urgent visits, 24/7 telephonic triage and off-hours community paramedicine. 3) Improve patient specific metrics (hospital length of stay, number of hospital encounters, readmissions, patient satisfaction, percent of death at home). DESCRIPTION OF PROGRAM/INTERVENTION, INCLUDING ORGANIZATIONAL CONTEXT: The North Shore LIJ Health System Advanced Illness Management (AIM) House Calls Program is a multidisciplinary team of physicians, nurse practitioners and social workers with central support personnel including a nurse intake manager, practice manager, biller, data analyst, project manager, administrative assistant, medical coordinators, medical director, and director of quality and case

• management. Program eligible patients require two or more chronic conditions, a recent hospitalization or unmet symptom need, and

meet Medicare definition of home boundedness. Patients must also be willing to allow AIM providers to serve as their primary care

• provider. Patient Characteristics: Mean age was 85 years, 72.2% female and 27.6% male. Most needed assistance with 5-6 ADL's

(65%). The most common diagnosis were diabetes, protein-calorie malnutrition, decubitus ulcer of the skin, congestive heart failure, cardiac arrhythmias, COPD, renal failure, stroke, vascular disease, and Parkinson's and Huntington's disease. MEASURES OF SUCCESS: Quantitative measures to evaluate the program include hospital admission rate, 30 day readmission rate, hospital length of stay (LOS), percent of in-home post discharge medicine reconciliations performed within 48 hours, percent of patients with advanced care planning and percent of patients who die at home. Qualitative metrics included American Academy of Home Physicians surveys of patient satisfaction, likelihood to recommend the program, and whether the program met patient needs. FINDINGS TO DATE: Among the enrollees in the program, the percentage of patients who had a post-admission contact within 48 hours rose from 45% in quarter 3 of 2012, to 85% in quarter 3 of 2013. The acuity of patients entering the program, increased over time with the average HCC diagnosis risk score increasing by 46% by quarter 3, 2013. Mean hospital LOS decreased from Q4 in 2012 (8.61 days) to Q3 in 2013 (7.28 days). Compared to the year prior to enrollment, patient experienced a 37% decrease in hospital admissions. Hospital 30 day readmission rate was 17% for patients discharged to rehabilitation facilities and 12% for patients discharged to home. Post discharge in-home medication reconciliation was performed within 48 hours of discharge for 94% of patients in quarter 3, 2013 compared to 63% in quarter 2, 2012. Of the total patients enrolled in the program, 95% had an advanced care plan in place. In 2013 65% of patients were able to die at home. Patient satisfaction: When asked about likelihood to recommend the program, 84% responded "definitely yes" in September 2013 compared to 77% in October 2012. Seventy one percent responded "definitely yes" when asked if they were seen by the team within 36 hours for urgent medical problems in 2013, compared with 63% in 2012. When asked about receiving answers to medical questions the same day, 65% responded "definitely yes" in 2013 compared with 45% in 2012. Sixty five percent responded that the program reduced their trips to the ED in 2013 compared with 56% in 2012. Eighty percent of patients felt that the program improved their quality of life in 2013 compared with 71% in 2012. KEY LESSONS FOR DISSEMINATION: Implementation of a high quality advanced illness management program can decrease admission rates, hospital length of stay and increase patient satisfaction despite an increase in overall patient acuity. AIM programs can provide person-centric planning to balance curative and comfort care as disease progresses. Such programs must include an interdisciplinary team, and provide coordination of care between hospitals, physicians, home, and community services. However, the needs of the community can quickly

• utpace program resources. Our wait list is currently over one hundred with a 3-4 month wait for non-acute referrals. Programs must

also ensure buy in from senior leadership, finance and quality departments, and must leverage informatics to achieve high reliability. The employment of community paramedicine to assist with patient evaluation and triage during off hours is also critical to program success.

INNOVATION IN MEDICAL EDUCATION:

CARING FOR THE HOMELESS AND UNDERSERVED: AN ONLINE, SYSTEMS-BASED, INTERPROFESSIONAL CURRICULUM Davoren Chick1,3; April Bigelow2; F. Jacob Seagull3; Heather Rye5; Pamela Davis1,4; Brent C. Williams1. 1University of Michigan, Ann Arbor, MI; 2University of Michigan, Ann Arbor, MI; 3University of Michigan, Ann Arbor, MI; 4University of Michigan, Ann Arbor, MI; 5University of Michigan Health System, Ann Arbor, MI. (Tracking ID #1938597) NEEDS AND OBJECTIVES: Many medical residents lack formal training regarding social determinants of health, public healthcare systems, or special interprofessional care needs of the medically underserved. Internal medicine curricular milestones include an expectation that residents reflect awareness of socioeconomic barriers impacting patient care, but a national curriculum has not been

• formalized. Locally, University of Michigan internal medicine residents scored an average of 44% on a baseline needs assessment

quiz in these content areas. Residents who rotated through underserved clinical venues with a traditional informal curriculum showed no significant improvement in knowledge. To address identified needs, we developed a sharable online curriculum regarding public healthcare systems and bio-psychosocial care for the underserved. We stressed interdisciplinary education and innovative adult learning methods by forming an interprofessional curriculum design team and incorporating a novel, game-based learning tool. SETTING AND PARTICIPANTS: With support from a Graduate Medical Education Innovation grant from the University of Michigan Medical School, we formed a curriculum design team comprising the fields of nursing, social work, internal medicine, pediatrics, medical education, communications, and graphic design. The final curriculum was posted to a freely accessible website. Current participants include over 250 registered learners from numerous medical and nursing institutions across the United States. DESCRIPTION: Through iterative refinement, the curriculum design team identified two relevant curricular domains supporting care of at-risk populations. Each of the two domains was divided into knowledge content modules: Domain 1. Public Healthcare Systems includes modules on epidemiology of the homeless and medically uninsured, public health insurance systems, and public healthcare delivery systems; Domain 2. Bio-Psychosocial Model of Care includes modules on biomedical needs of the homeless, social determinants of health, the bio-psychosocial model of health, and interprofessional team care. We produced case-based learning modules for each of these content areas and packaged the curriculum in a professionally designed web site: "Caring with Compassion". The website includes an individualized learner dashboard, case-based didactic modules, extension resources, a trifold case formulation pocket guide, and a milestone-based assessment tool. The core knowledge curriculum is supplemented with a case- based online game that integrates core content into an enjoyable, low-risk learning experience. EVALUATION: Nationally, hundreds of residents have enrolled in this curriculum. Additionally, the curriculum is in use by nurse practitioner and medical students, and is now a core resource for University of Michigan Medical School and School of Nursing courses regarding care of underserved populations. Faculty across multiple institutions report assigning the website for independent learner review and using the resources during small group discussions. Qualitative feedback has been highly positive, with residency program directors stating that the curriculum addresses significant gaps in the traditional graduate medical curriculum. Preliminary quantitative outcome data demonstrates highly significant improvement in multiple choice examination scores following exposure to the modules and game for all assessed content areas [n, pre-test, post-test, T test p value]: epidemiology of populations in need (42, 49%, 81%, p<0.0001); health care delivery systems knowledge (19, 53%, 67%, p<0.002); homelessness and disease (20, 59%, 82%, p<0.0001); public health insurance (30, 52%, 80%, p<0.0001); team care (14, 45%, 64%, p<0.001); social determinants (29, 51%, 56%, p=0.022). DISCUSSION / REFLECTION / LESSONS LEARNED: This freely shared online curriculum addresses identified national and local gaps in socioeconomic and sociobehavioral knowledge for medical learners. Curricular components can be used individually or as a comprehensive curriculum. Medical educators and learners have expressed appreciation both for the cohesiveness of the compiled curriculum and for the flexibility of implementation options provided by the online modular format. Further plans include expansion

• f faculty development tools, expansion of the learning game through addition of a second game case scenario, and continual revisions

based on curricular feedback. Challenges moving forward include securing time and resources to update content, maintain the website, and measure and monitor learner outcomes. High quality, freely accessible, interprofessional online curricula are highly appreciated resources for academic medical and nursing institutions. Shared online curricula are an important and valued direction for the future of a national medical curriculum and for academic recognition of medical education scholarship. ONLINE RESOURCE URL (OPTIONAL): http://caringwithcompassion.org

CLINICAL VIGNETTE:

WOMAN, INTERRUPTED: AN UNUSUAL CAUSE OF PSYCHOSIS Sara Attalla; Deborah Hemel. Montefiore Medical Center, Bronx, NY. (Tracking ID #1939021) LEARNING OBJECTIVE 1: Recognize hallmark neuropsychiatric symptoms of anti-NMDA receptor encephalitis LEARNING OBJECTIVE 2: Initiate treatment as early as possible to optimize chances for recovery CASE: A 31 year-old female with no past medical or psychiatric history presented with acute behavior

• changes. According to her family, the patient was in her fully functional state of health until one week prior

when they noted changes in mood; she cried frequently, appeared anxious, and had difficulty sleeping, all of which were uncharacteristic. On the day of admission she was disoriented, paranoid and aggressive, prompting her family members to call EMS. Vital signs were within normal limits. CT head was unremarkable. Lumbar puncture (LP) revealed 25 white blood cells with 96% lymphocytes. On initial exam she was alert and oriented but had poor recollection of recent events. Shortly afterwards the patient began screaming uncontrollably and attempted to leave the hospital. She became combative, and appeared to be responding to internal stimuli. She required several rounds of Haldol and Ativan for sedation. Over the next several days, she experienced numerous similar episodes requiring remarkable physical and chemical restraint. An extensive infectious work up was negative, and a brain MRI showed no gross abnormalities. EEG demonstrated cerebral dysfunction in the L temporal and R parasaggital regions. On hospital day 7, a CT scan of the patient's abdomen and pelvis was performed as there was growing concern for a paraneoplastic process. Imaging revealed a large mass encompassing both ovaries. She was taken to the OR on hospital day 9 for bilateral salpingo-oopherectomy. Pathologic examination revealed a low-grade immature ovarian teratoma. She was subsequently started on high dose steroids and plasmapheresis. She experienced minimal neurologic improvement and so IVIG was initiated. Shortly after her last treatment, she began to show recovery in communication and comprehension. Her mental status improved, and she was discharged home on hospital day 44. CSF later returned positive for anti-NMDA receptor antibodies, confirming the diagnosis. Upon outpatient follow-up ten weeks after presentation, neurologic function was noted to be nearing her baseline, although she had no recollection of her hospitalization

• r the events leading up to it.

DISCUSSION: Paraneoplasic limbic encephalitis is a challenging illness to diagnose and treat. A form of this that has been gaining increasing attention is anti-N-methyl-D-aspartate (NMDA) receptor encephalitis, which was identified in 2007. Manifestations of this illness include dramatic psychotic symptoms, altered consciousness, seizure-like activity, catatonia, dyskinesias, and autonomic disturbances. It is believed to be an autoimmune mediated disorder that results in excitotoxicity of the NMDA receptors found mainly in the limbic

• system. Anti-NMDA receptor encephalitis is prevalent in young women and is often associated with benign

tumors, the most notable being ovarian teratomas. Common workup may reveal lymphocytic pleocytosis in CSF and abnormal frontotemporal activity on EEG. Definitive diagnosis is established by the presence of antibodies in CSF. First line treatment includes removal of tumor, if there is one, and combinations of plasmapheresis, steroids and IVIG. Studies have shown that prognosis is improved when the disease is identified and treated

• promptly. However, there is typically a significant delay in diagnosing anti-NMDA receptor encephalitis as

patients are often misdiagnosed with a primary psychiatric illness. It is therefore important to think of this illness once common etiologies of encephalitis have been excluded.