High-cost Patients Research Articles

Out-of-Pocket Payment for Ambulatory Blood Pressure Monitoring Among Commercially Insured in the United States.

Clinical guidelines increasingly recommended ambulatory blood pressure monitoring (ABPM) for hypertension diagnosis and management. Yet, ABPM is used infrequently in the United States, possibly because of low insurance coverage and high patient costs. We sought to analyze out-of-pocket payments (OPPs) for ABPM among privately insured patients. We conducted a retrospective analysis using IBM® MarketScan® commercial claims of beneficiaries aged ≥18 years receiving ABPM from January 2012 to December 2018. The date of first ABPM claim (Healthcare Common Procedure Coding System codes 93784, 93786, 93788, or 93790) was considered the index date. Patients with 12 months of continuous enrollment preindex and 30-day postindex were included. Per beneficiary OPP was calculated by aggregating all ABPM-related OPPs within the 30-day postindex window (ABPM episode). Of 22,317 beneficiaries receiving ABPM, 62% had $0 OPP and 38% had OPP >$0. Among the latter, median OPP per beneficiary for an ABPM episode was $23 (interquartile range [IQR], $14, $32), driven primarily by full ABPM claims (median, $22; IQR, $14, $24). Among individual components, scan analysis and report claims (median, $25; IQR, $13, $49) had the greatest OPP. The median OPP per ABPM episode did not change substantively from 2012 through 2018. Among commercially insured in the United States, nearly 4-in-10 have an OPP for ABPM. Though most OPPs are relatively modest, some patients incur substantial OPP. Our findings highlight the need for policymakers to ensure adequate ABPM coverage in the commercial insurance marketplace.

Hemorrhagic reversible cerebral vasoconstriction syndrome: A retrospective observational study.

Reversible cerebral vasoconstriction syndrome (RCVS) is characterized by recurrent thunderclap headaches associated with segmental vasoconstriction of cerebral arteries, which may result in intracranial hemorrhage (ICH). There is a lack of contemporary data available regarding the ICH burden in RCVS cohort. Our aim of the study is to assess the ICH burden, associated risk factors, and discharge outcome of ICH in patients with RCVS. All patients diagnosed with RCVS in the 2016 Nationwide Readmission Database were identified using ICD-10 code after excluding patients with the concurrent diagnosis of primary angiitis. ICH was defined as both intraparenchymal (IPH), subarachnoid hemorrhage (SAH), and subdural hematoma (SDH). Categorical and continuous variables were assessed by the Rao-Scott Chi-square test and the Wilcoxon signed-rank sum test respectively. We used a multivariable survey-weighted logistic model to determine the association between ICH and RCVS patient-level characteristics. A total of 799 patients were identified with RCVS. Total hospitalization of ICH was 43.4% [(95% CI 36.4-50.4%); (n = 346)] including SAH 35.9% [(95% CI 29.7-42.1%); (n = 287)], IPH 13.1% [(95% CI 9.5-16.7%); (n = 105)] and SDH 3.6% [(95% CI 1.5-5.6%); (n = 28)]. Patients with hemorrhagic RCVS (H-RCVS) had a mean age (years ± SE) of 47.4 ± 1.1 vs. 45.5 ± 1.2years in R-RCVS (p = 0.247); and were predominantly female (84.0% vs. 68.8%; p = 0.001); with longer inpatient stays (10.9 vs. 6.8days; p = 0.016); and a higher inpatient cost ($44,300 vs. $21,350; p < 0.001). On multivariable analyses, higher odds of ICH were female sex 2.57 (95% CI 1.45-4.55; p = 0.001), middle age-group (45-64years) 1.87 (CI: 1.11-3.15; p = 0.018) and older age group (> 64years) 3.72 (CI: 1.15-12.03; p = 0.029). About 67.0% of all H-RCVS patients were discharged home, with no observed inpatient mortality. Intracerebral hemorrhage is the most common vascular complication in hospitalized RCVS patients, resulting in longer hospitalizations with more invasive procedures and higher healthcare expenditure. However, overall outcomes are excellent regardless of types of ICH, with no inpatient mortality observed in patients with hemorrhagic RCVS. Female sex and middle to older age-group are associated with higher odds of ICH.

Promotion of collagen deposition during skin healing through Smad3/mTOR pathway by parathyroid hormone-loaded microneedle

Skin wounds are associated with huge economic and emotional burdens for millions of people annually and are a challenge for health workers worldwide. At present, for skin defects after traumatic accidents, especially large-area skin defects, newly developed strategies such as the use of emerging biomaterials and cell therapy could be considered as options besides classic skin grafts. However, the new strategies have to deal with problems such as immune rejection and high costs for patients. An insufficient understanding of the mechanisms of skin wound healing further hinders the development of innovative treatment approaches. In this study, we developed a parathyroid hormone (PTH)-loaded phase-transition microneedle (PTMN) patch to deliver PTH subcutaneously in an efficient manner and change microneedle patch daily to achieve intermittent and systematic drug administration. By evaluating wound closure, re-epithelialization, collagen deposition, and extracellular matrix (ECM) expression in a Sprague–Dawley rat model of traumatic skin wounds, we demonstrated that intermittent systemic administration of PTH using our PTMN patches accelerated skin wound healing. Further, we demonstrated that the use of the patch may accelerate skin wound healing depending on the activation of the transforming growth factor (TGF)-β/Smad3/mammalian target of rapamycin (mTOR) cascade pathway. Our results suggest that the PTH-loaded PTMN patch may be a novel therapeutic strategy for treating skin wounds.

The self-assessment INTERMED predicts healthcare and social costs of orthopaedic trauma patients with persistent impairments

Objective: To use the self-assessment INTERMED questionnaire to determine the relationship between biopsychosocial complexity and healthcare and social costs of patients after orthopaedic trauma. Design: Secondary prospective analysis based on the validation study cohort of the self-assessment INTERMED questionnaire. Setting: Inpatients orthopaedic rehabilitation with vocational aspects. Subjects: In total, 136 patients with chronic pain and impairments were included in this study: mean (SD) age, 42.6 (10.7) years; 116 men, with moderate pain intensity (51/100); suffering from upper (n = 55), lower-limb (n = 51) or spine (n = 30) pain after orthopaedic trauma; with minor or moderate injury severity (severe injury for 25). Main measures: Biopsychosocial complexity, assessed with the self-assessment INTERMED questionnaire, and other confounding variables collected prospectively during rehabilitation. Outcome measures (healthcare costs, loss of wage costs and time for fitness-to-work) were collected through insurance files after case settlements. Linear multiple regression models adjusted for age, gender, pain, trauma severity, education and employment contract were performed to measure the influence of biopsychosocial complexity on the three outcome variables. Results: High-cost patients were older (+3.6 years) and more anxious (9.0 vs 7.3 points at HADS-A), came later to rehabilitation (+105 days), and showed higher biopsychosocial complexity (+3.2 points). After adjustment, biopsychosocial complexity was significantly associated with healthcare (ß = 0.02; P = 0.003; expß = 1.02) and social costs (ß = 0.03; P = 0.006, expß = 1.03) and duration before fitness-to-work (ß = 0.04; P < 0.001, expß = 1.04). Conclusion: Biopsychosocial complexity assessed with the self-assessment INTERMED questionnaire is associated with higher healthcare and social costs.

High-need high-cost patients: A Concept Analysis.

High-need high-cost (HNHC) patients are variously defined in the literature as the small subset of the population who account for the majority of US health care costs. Lack of consensus on the defining attributes of HNHC patients has challenged the effectiveness of interventions aimed to improve disease management and reduce costs. Guided by the Walker and Avant method of concept analysis, a literature review of 2 databases (PubMed and CINAHL) was conducted. Three main subgroups of HNHC patients were identified: adults with multiple chronic conditions and functional disability, the frail elderly, and patients under 65 years old with a disability or behavioral health condition. HNHC patients are categorized by a feedback loop of acute-on-chronic health conditions, preventable health service utilization, and fragmented care. Antecedents that predispose becoming a HNHC patient include challenges accessing timely care, low socioeconomic status, unmet support, and social factors such as isolation and inadequate.

An evaluation of community pharmacies’ actions under value-based payment

ObjectiveTo determine actions taken by community pharmacies to be successful under a value-based pharmacy program (VBPP). MethodAn exploratory sequential mixed methods approach was used to evaluate pharmacies participating in the VBPP, with qualitative data collected and analyzed in the first phase, followed by quantitative measurement through a 30-item survey instrument in the second phase. ResultsThe qualitative data showed that participating pharmacies were more involved with adherence and cardiovascular and diabetes metrics than with other metrics. Depression metrics received the lowest overall involvement. For total cost of care, different approaches were used; 5 pharmacies used the dashboard to identify likely high-cost patients they could try to manage, and 4 pharmacies monitored adherence to avoid complications that could contribute to increased cost. For the survey response rate was 72.6% (n = 53). The mean perception of level of success was 53.06 ± 20.15 (mean ± SD). Activities with the highest priority were adherence (1.98 ± 0.97) and diabetes care (2.04 ± 0.83), and the activity with the lowest priority was depression care (3.60 ± 1.10). The most frequently mentioned challenge was time availability, and the most common improvement suggestion was better communication between the insurer and providers. ConclusionIn conclusion, this study found that community pharmacies were transforming their practices to be successful under a commercial value-based payment program. The pharmacies tended to build on care processes already established (e.g., medication adherence, patients with diabetes or cardiovascular conditions) and developed new processes to address emerging metrics and associated patient needs (e.g., collecting and documenting blood pressure and hemoglobin A1c levels). Future research is needed to identify best practices for patient care and pharmacy success under broad VBPPs such as the one studied here.

Identifying Patients with Persistent Preventable Utilization Offers an Opportunity to Reduce Unnecessary Spending.

Improving care for high-cost patients is increasingly important for improving the value of healthcare. Most prior research has focused on identifying patients with high costs, but the extent to which these costs are potentially preventable remains unclear. To identify patients with persistent preventable utilization and compare their characteristics with high-cost patients. Descriptive analysis using Medicare claims data from 2013 to 2014. Medicare fee-for-service and dual-eligible beneficiaries (N = 515,689) from the New York metropolitan area who were continuously enrolled in Medicare Parts A and B in 2013 and 2014. The primary analysis focuses on patients with persistent preventable utilization (at least one preventable emergency department visit, hospitalization, or 30-day readmission in both 2013 and 2014) and high-cost patients in 2014 (top 10% of total annual spending). We compared demographic, medical, behavioral, and social characteristics and total and preventable healthcare utilization between these two groups. Patients with persistent preventable utilization accounted for 4.8% of the overall patient population, 13.4% of overall costs, but 46.2% of preventable costs among all Medicare patients. Compared with high-cost patients, patients with persistent preventable utilization had lower median healthcare costs ($33,383 vs. $56,552), but their median potentially preventable costs were seven times higher ($7151 vs. $928). We also found that 1.9% of patients could be categorized in both the persistent preventable utilization group and the high-cost group. This subset of patients had the highest median Medicare costs and preventable costs and represented over 30% of total preventable spending and 9.4% of overall costs among all Medicare patients. Designing and targeting interventions for patients with persistent preventable utilization may offer an important opportunity to reduce unnecessary utilization and promote high-value care.

Evaluation of the Incremental Healthcare Economic Burden of Patients with Atrial Fibrillation Treated with Direct-Acting Oral Anticoagulants and Hospitalized for Major Bleeds in the USA.

Direct-acting oral anticoagulants (DOACs) are associated with risk of major bleeding. This study evaluated the incremental healthcare economic burden of patients with atrial fibrillation (AF) treated with DOACs and hospitalized with a major bleed (MB). Adult patients with AF treated with DOACs and hospitalized with MB or no MB hospitalizations during January1, 2015-April30, 2018 were extracted from MarketScan claims databases. The index date was defined as the first MB hospitalization for patients with MB and a random date during DOAC usage for patients without MB. Healthcare resource utilization and costs were evaluated for index hospitalizations of patients with MB and during the 6-month period prior to index dates and a variable follow-up period of 1-12months for both patients with and those without MB. Multivariable regression analyses were performed to evaluate the incremental burden of MB vs. non-MB status on all-cause hospital days and healthcare costs. Of the overall AF patient population using DOACs (N = 152,305), 7577 (5.0%) had a hospitalization for MB. Greater proportions of those who had an MB hospitalization were older and female compared to patients without MB (mean age 76.1 vs. 70.1years; 44.1% vs. 40.5% female, respectively). For index MB hospitalizations, mean length of stay (LOS) was 5.3days and cost was $32,938. In adjusted analyses, patients with MB had 3.6 more hospital days, $10,609 higher inpatient cost, $9613 higher outpatient medical cost, and $18,910 higher total healthcare costs for all causes per patient during follow-up (all p < 0.001). Including index MB hospitalization costs in the follow-up, all-cause total adjusted healthcare costs were almost two times higher for patients with vs. without MB ($96,590 vs. $49,091, p < 0.001). Among a large US nationally representative sample of patients with AF treated with DOACs, the cost of MB hospitalization was substantial. Furthermore, healthcare costs following MB events were nearly 40% higher compared to those of patients with AF without an MB.

Particulate-Aggregated Adhesives with Exudate-Sensitive Properties and Sustained Bacteria Disinfection to Facilitate Wound Healing.

Wound-associated infections create additional suffering and come at a high cost for patients and their families, which urgently require wound disinfection biomaterials with improved healing efficacy. Here, we report an adhesive with sustained bacteria disinfection ability, which is aggregated from hydrogen-bonded polymer particulates. The particulate-aggregated adhesive shows strong binding ability on different surfaces from rigid substrates to soft skins. Moreover, water-sensitive mechanical properties are shown in wound exudates, resulting from the dissociation of hydrogen bonds under the competition of water and thus the sustained release of particulates. Synergizing with the strong binding ability, exudate-sensitive behaviors, and sustained release of antibacterial particulates, the adhesive achieves sustained bactericidal activity and can facilitate the healing process in bacteria-infected skin wounds.

Economic outcomes of insurer-led care management for high-cost Medicaid patients.

To evaluate the impact of the Community-Based Care Management (CBCM) program on total costs of care and utilization among adult high-need, high-cost patients enrolled in a Medicaid managed care organization (MCO). CBCM was a Medicaid insurer-led care coordination and disease management program staffed by nurse care managers paired with community health workers. Retrospective cohort analysis. We obtained deidentified health plan claims data, enrollment information, and the MCO's monthly registry of the top 10% of costliest patients. The analysis included 896 patients enrolled in CBCM over the course of 2 years (January 2016 to December 2017) and a propensity score-matched cohort of high-cost patients (n = 2152) who received primary care at sites that did not participate in CBCM during the same time period. The primary outcomes were total costs of care and utilization in the 12-month period after enrollment. Secondary outcomes included utilization by care setting: outpatient, inpatient, emergency department, pharmacy, postacute care, and all other remaining sites. We used zero-inflated gamma and Poisson regression models to estimate average differences in postperiod costs and utilization between CBCM enrollees versus non-CBCM enrollees. We did not observe meaningful differences in total costs or visit frequency among CBCM enrollees relative to non-CBCM enrollees. Although our study found no association between the CBCM program and subsequent cost or utilization outcomes, understanding why these outcomes were not achieved will inform how future Medicaid programs are designed to achieve better patient outcomes and lower costs.

Development of a Value-based Algorithm for Inpatient Triage of Elderly Hip Fracture Patients.

The purpose of this study was to combine a validated middle-age and geriatric trauma risk assessment tool (STTGMA) with a novel cost-prediction tool to create an objective triage tool for elderly hip fractures that would guide value-based care initiatives. From October 2014 to January 2018, all patients aged ≥55 years who were admitted with a primary diagnosis of hip fracture to a single level 1 trauma center were enrolled. Upon evaluation in the emergency department, demographics, injury severity, and functional status were recorded to calculate the trauma triage score (STTGMARisk). A model to predict high-cost hip fracture patients was created using similar variables (STTGMACost). Three hundred sixty-one consecutive operative hip fracture patients were enrolled. Inpatient mortalities were skewed toward STTGMARisk3 with 21.4% of patients in this high-risk group ultimately expiring during their hospitalization. High-cost patients were correctly skewed to the STTGMACost2 and STTGMACost3 groups with 88.9% of all high-cost operatively treated hip fracture correctly triaged to these cohorts. Statistically significant variations were found in cost within each STTGMARisk group. A simple risk score calculated upon admission (STTGMARisk and STTGMACost) was able to be used as a triage tool not only to differentiate increased mortality risk but also to predict high-cost patients based on resource utilization in hip fracture patients. Prognostic, level II.

Embedded Pragmatic Trials in Dementia Care: Realizing the Vision of the NIA IMPACT Collaboratory.

Close to 6 million Americans have Alzheimer's disease (AD) or Alzheimer's disease and related dementia (AD/ADRD). These high-need, high-cost patients are vulnerable to receiving poor quality uncoordinated care, ultimately leading to adverse health outcomes, poor quality of life, and misuse of resources. Improving the care of persons living with dementia (PLWD) and their caregivers is an urgent public health challenge that must be informed by high-quality evidence. Although prior research has elucidated opportunities to improve AD/ADRD care, the adoption of promising interventions has been stymied by the lack of research evaluating their effectiveness when implemented under real-world conditions. Embedded pragmatic clinical trials (ePCTs) in healthcare systems have the potential to accelerate the translation of evidence-based interventions into clinical practice. Building from the foundation of the National Institutes of Healthcare Systems Collaboratory, in September 2019 the National Institute on Aging Imbedded Pragmatic AD/ADRD Clinical Trials (IMPACT) Collaboratory was launched. Its mission is to build the nation's capacity to conduct ePCTs within healthcare systems for PLWD and their caregivers by (1) developing and disseminating best practice research methods, (2) supporting the design and conduct of ePCTs including pilot studies, (3) building investigator capacity through training and knowledge generation, (4) catalyzing collaboration among stakeholders, and (5) ensuring the research includes culturally tailored interventions for people from diverse backgrounds. This report presents the rationale, structure, key activities, and markers of success for the overall NIA IMPACT Collaboratory. The articles that follow in this special Issue describe the specific work of its 10 core working groups and teams. J Am Geriatr Soc 68:S1-S7, 2020.

Access to Chagas disease treatment in the United States after the regulatory approval of benznidazole.

Approximately 300,000 persons in the United States (US) are infected with Trypanosoma cruzi, the protozoan that causes Chagas disease, but less than 1% are estimated to have received antiparasitic treatment. Benznidazole was approved by the US Food and Drug Administration (FDA) for treatment of T. cruzi infection in 2017 and commercialized in May 2018. This paper analyzes factors that affect access to benznidazole following commercialization and suggests directions for future actions to expand access. We applied an access framework to identify barriers, facilitators, and key actors that influence the ability of people with Chagas disease to receive appropriate treatment with benznidazole. Data were collected from the published literature, key informants, and commercial databases. We found that the mean number of persons who obtained benznidazole increased from just under 5 when distributed by the CDC to 13 per month after the commercial launch (from May 2018 to February 2019). Nine key barriers to access were identified: lack of multi-sector coordination, failure of health care providers to use a specific order form, lack of an emergency delivery system, high medical costs for uninsured patients, narrow indications for use of benznidazole, lack of treatment guidelines, limited number of qualified treaters, difficulties for patients to make medical appointments, and inadequate evaluation by providers to determine eligibility for treatment. Our analysis shows that access to benznidazole is still limited after FDA approval. We suggest six areas for strategic action for the pharmaceutical company that markets benznidazole and its allied private foundation to expand access to benznidazole in the US. In addition, we recommend expanding the existing researcher-clinician network by including government agencies, companies and others. This paper’s approach could be applied to access programs for benznidazole in other countries or for other health products that target neglected populations throughout the world.

Case-mix for assessing primary care value (CPCV).

Measuring primary care (PC) performance and designing payment systems that reward value rather than volume have been a great challenge due in large part to lack of reliable risk adjustment mechanisms pertinent to primary care. Using risk scores designed for total resource needs to assess PC performance or set PC payment rates is inadequate because high-cost patients may not have high needs in PC and vice versa. The greatest challenge in developing a risk algorithm for PC is that significant components of PC providers' workload are unobservable but needed in the modeling. In this study, we sought to overcome this challenge by analyzing 5,172,773 patients in the U.S. Veterans Affairs (VA) healthcare system to identify potential proxies of the unobservable PC workload. By combining the number of PC visits and prescription drug classes, we formed a proxy for the expected PC workload, which enabled us to develop a case-mix algorithm pertaining to primary care. The resultant algorithm with high explanatory power (R2 = 0.702) is based on a publicly available patient classification system to account for patient comorbidities and thus can be used by other health systems to compare PC performance, workload, staffing levels, and to set more equitable payment rates.

Hochkostenversicherte in Deutschland: Leistungs- und Kostenprofile

BackgroundStudies from different countries have shown that a small number of insured persons (high-cost patients) are responsible for a large portion of health care spending. At the same time, it is assumed that some of these costs could be saved by a better management of this group of people. The aim of this article is to analyze the performance and cost profiles of high-cost patients, to put them in an international comparison, and to derive a better management approach. MethodsRetrospective observation study based on statutory health insurance data from two statutory health insurances for the year 2013. Study population: top 5 %, as well as top 1 % of the most expensive insured persons. Identification of characteristics of high-cost patients and international comparison with the Netherlands, the USA, Canada, Spain, England and Japan. Results5 % of insured persons account for almost half of the total costs and the most expensive 1 % of 22 %. These high-cost patients in Germany are, on average, 20 years older than the general population. Almost every person of the high-cost population was prescribed at least one medication during the study period (99.2 %), and 85.8 % had at least one hospital stay. Hospital care accounts for the biggest part of total costs: 75 % together with drugs. The average per capita costs caused by one of the 5 % most expensive insured persons in the year under review are 20 times higher than that of the other 95 % of insured persons. High-cost patients are generally more multimorbid and have higher mortality rates. The most common diagnoses of these patients are hypertension, lipoprotein metabolism disorder and back pain. ConclusionSimilar to other developed countries, Germany faces the challenge to develop and implement adequate intervention approaches addressing the special requirements of high-cost insured persons. This paper provides a first basis. The analogies of high-cost patients in Germany and other countries illustrate the need for transnational research and intervention approaches on this specific issue. More in-depth work is needed to investigate the potentials of Predictive Modelling and integrated care approaches to the management of this group of insured persons.

Analysis of modifiable and nonmodifiable risk factors in patients undergoing pituitary surgery.

Pituitary adenomas (PAs) are among the most common intracranial tumors. Understanding the clinical effects of various modifiable risk factors (MRFs) and nonmodifiable risk factors (NMRFs) is important in guiding proper treatment, yet there is limited evidence outlining the influence of MRFs and NMRFs on outcomes of PA resection. The aim of this study was to analyze MRFs and NMRFs in patients undergoing resection for PAs. Using the 2016 and 2017 National Readmission Database, the authors identified a cohort of 9472 patients undergoing microscopic or endoscopic resection of a PA. Patients with nonoverlapping MRFs and NMRFs were analyzed for length of stay (LOS), hospital cost, readmission rates, and postoperative complications. From the original cohort, a subset of 373 frail patients (as defined by the Johns Hopkins Frailty Index) were identified and propensity matched to nonfrail patients. Statistical analysis included 1-way ANOVA, Tukey multiple comparisons of means, odds ratios, Wald testing, and unpaired Welch 2-sample t-tests to compare complications, outcomes, and costs between each cohort. Perioperative outcomes and hospital readmission rates were tracked, and predictive algorithms were developed to establish precise relationships between relevant risk factors and neurosurgical outcomes. Malnourished patients had significantly longer LOSs when compared to nonmalnourished patients (p < 0.001). There was a significant positive correlation between the number of MRFs and readmission at 90 days (p = 0.012) and 180 days (p = 0.020). Obese patients had higher rates of postoperative neurological injury at the 30-day follow-up (p = 0.048) compared to patients with normal BMI. Within this NMRF cohort, frail patients were found to have significantly increased hospital LOS (p < 0.001) and total inpatient costs compared to nonfrail patients (p < 0.001). Predictive analytics showed that frail patients had significantly higher readmission rates at both 90-day (p < 0.001) and 180-day follow-ups (p < 0.001). Lastly, rates of acute postsurgical infection were higher in frail patients compared to nonfrail patients (p < 0.001). These findings suggest that both MRFs and NMRFs negatively affect the perioperative outcomes following PA resection. Notable risk factors including malnutrition, obesity, elevated lipid panels, and frailty make patients more prone to prolonged LOS, higher inpatient costs, and readmission. Further prospective research with longitudinal data is required to precisely pinpoint the effects of various risk factors on the outcomes of pituitary surgery.

Person-centred and efficient care delivery for high-need, high-cost patients: primary care professionals\u2019 experiences

BackgroundHigh-need, high-cost (HNHC) patients, who typically have complex and long-term care demands, contribute considerably to the high work pressure of primary care professionals (PCPs). To improve patient as well as provider experiences, it is crucial to take into account the PCPs’ perspective in designing health care strategies for HNHC patients. Therefore, this study aimed to create insight into PCPs’ experienced barriers and possible solutions with regards to person-centred, efficient care delivery to HNHC patients.MethodsWe conducted a qualitative study using focus group interviews with PCPs at a Dutch primary care group. A semi-structured interview guide was developed for the interviews. Qualitative content analysis was employed deductively by means of a categorisation matrix. The matrix was based on the components retrieved from the SELFIE framework for integrated care for multi-morbidity.ResultsForty-two PCPs participated in five focus group interviews. Discussed barriers and solutions were related to the core of the SELFIE framework (i.e. the individual and environment), and particularly four of the six health system components in the framework: service delivery, leadership & governance, workforce, and technologies & medical products. Many discussed barriers revolved around the complex biopsychosocial needs of HNHC patients: PCPs reported a lack of time (service delivery), insufficiently skilled PCPs (workforce), and inefficient patient information retrieval and sharing (technologies & medical products) as barriers to adequately meet the biopsychosocial needs of HNHC patients.ConclusionsThis qualitative study suggests that primary care is currently insufficiently equipped to accommodate the complex biopsychosocial needs of HNHC patients. Therefore, it is firstly important to strengthen primary care internally, taking into account the experienced lack of time, the insufficient number of equipped PCPs and lack of inter-professional information retrieval and sharing. Secondly, PCPs should be supported in cooperating and communicating more efficiently with health services outside primary care to adequately deliver person-centred, efficient care. As a prerequisite, it is crucial to direct policy efforts at the design of a strong system of social and community services. In terms of future research, it is important to assess the feasibility and effects of re-designing primary care based on the provided recommendations.

4375 Developing team science for practical applications of artificial intelligence in health systems to improve value and outcomes: A case study in reducing avoidable emergency department use

OBJECTIVES/GOALS: Health care systems are complex, dynamic, and varied. Advances in artificial intelligence (AI) are enabling healthcare systems to use their own data to elicit patterns and design suitable interventions. To realize this potential, computer scientists and clinicians need an effective, practical, and replicable approach to collaboration METHODS/STUDY POPULATION: In this study, computer scientists partnered with clinicians to investigate predictors of avoidable emergency department use. The team sought an approach to computational medicine that could increase the relevance and impact of prediction to solve pressing problems in the health system. The team adopted an emergent architecture that engaged system leaders, computer scientists, data scientists, health services researchers, and practicing clinicians with deep ambulatory and inpatient knowledge to form the initial questions that shaped the prediction model; to understand nuances of coding and recording in source data and the implications for models; and to generate insights for promising points of intervention. The team recorded decisions and challenges as it progressed to analyze its function. RESULTS/ANTICIPATED RESULTS: Most avoidance models focus on a narrow time period around target events, or on high cost patients and events. This interdisciplinary team used their insights into the health system’s workflows and patient population to adopt a longitudinal approach to their prediction models. They used AI to build models of behavior in the system and consider prevention points across clinical units, time, and place. The holistic, systemwide focus enabled the team to generate insights that the system leaders and subsequently specific clinical units could apply to improve value and outcomes. A facilitated team process using learning system and cooperative network principles allowed a large and modular interdisciplinary team to build a transparent AI modeling process that yielded actionable insights into hypercomplex workflows. DISCUSSION/SIGNIFICANCE OF IMPACT: An architecture for involving diverse stakeholders in computational medicine projects can increase the relevance and impact of AI for solving care delivery problems in complex health systems. Translational science and computational medicine programs can foster this type of engagement and encourage a whole system perspective.

Transitional Care Models for High-Need, High-Cost Adults in the United States: A Scoping Review and Gap Analysis.

This scoping review explored research literature on the integration and coordination of services for high-need, high-cost (HNHC) patients in an attempt to answer the following questions: What models of transitional care are utilized to manage HNHC patients in the United States? and How effective are they in reducing low-value utilization and in improving continuity? U.S. urban, suburban, and rural health care sites within primary care, veterans' services, behavioral health, and palliative care. Utilizing the Joanna Briggs Institute and PRISMA guidelines for scoping reviews, a stepwise method was applied to search multiple databases for peer-reviewed published research on transitional care models serving HNHC adult patients in the United States from 2008 to 2018. All eligible studies were included regardless of quality rating. Exclusions were foreign models, studies published prior to 2008, review articles, care reports, and studies with participants younger than 18 years. The search returned 1,088 studies, of which 19 were included. Four studies were randomized controlled trials and other designs included case reports and observational, quasi-experimental, cohort, and descriptive studies. Studies focused on Medicaid, Medicare, dual-eligible patients, veterans, and the uninsured or underinsured. High-need, high-cost patients were identified on the basis of prior utilization patterns of inpatient and emergency department visits, high cost, multiple chronic medical diagnoses, or a combination of these factors. Tools used to identify these patients included the hierarchical condition category predictive model, the Elder Risk Assessment, and the 4-year prognostic index score. The majority of studies combined characteristics of multiple case management models with varying levels of impact. .

O9.7. TECHNOLOGY ASSISTED RELAPSE PREVENTION PROGRAM REDUCES HOSPITALIZATION IN SCHIZOPHRENIA PATIENTS AT HIGH RISK FOR RELAPSE

BackgroundPsychiatric hospitalization is a major driver of cost in the treatment of schizophrenia. During the first six months after a hospitalization, patients are at increased risk for return to the hospital. Targeting treatment during this period may be especially valuable. Mobilizing a range of treatment enhancements using both smart phones and computers to support patients by individualizing treatment to specific needs may be especially beneficial. The Improving Care Reducing Cost (ICRC) program sought to determine whether a novel, multicomponent, and technology-enhanced approach to relapse prevention in outpatients following a psychiatric hospitalization could reduce days spent in a hospital after discharge.MethodsThe ICRC study was a quasi-experimental clinical trial in outpatients with schizophrenia conducted between February 2013 and April 2015 at 10 sites in the US that provide both in and outpatient. Data were obtained from 89 participants who received usual relapse prevention services, followed by a second cohort of 349 participants who received technology-enhanced relapse prevention program. Both groups were followed for 6 months. Patients were between 18 and 60 years old; had a diagnosis of schizophrenia, schizoaffective disorder, or psychotic disorder not otherwise specified; and were currently hospitalized or had been hospitalized within the past 30 days at consent. The first cohort received usual care; the second cohort received the Health Technology Program (HTP) a technology-enhanced relapse prevention program. HTP included medication treatment guided by a computer decision support system for the prescriber, a smartphone application for patients that supported medication adherence and other coping strategies, a web-based patient and family psycho-educational intervention, and web-accessed cognitive behavioral therapy for paranoia and hallucinations. A mental health technology coach provided technical support, and developed a personalized, structured, relapse prevention plan with each participant that identified individual relapse precipitants and determined which HTP components should be employed to address them. All patients received computers and Android smartphones to insure access to the interventions. Days spent in a psychiatric hospital during 6 months after discharge was assessed. The Heinrichs Carpenter Quality of Life Scale was completed at baseline and six months.ResultsThe study included 438 patients. Control participants (N = 89; 37 females) were enrolled first and received usual care for relapse prevention, and followed by 349 participants (128 females) who received the HTP. Days of hospitalization were reduced by 4 days (Mean days: b = -4.25, 95% CI: -8.29; -0.21, P = 0.039) during follow-up in the intervention condition compared to control. Finally, using Heinrichs Carpenter Quality of Life total score at month 6 as an outcome, we found no significant effect of HTP (β = 0.02, t (345) = 0.43, P = 0.668).DiscussionRecently hospitalized patients with schizophrenia who received an integrated technology informed relapse prevention program (HTP) experienced fewer days in the hospital compared to those who received usual care in the six months following their discharge. Given the high patient burden and costs of even a single day spent in a psychiatric hospital, estimated at $1358 per day based on inflation adjusted results from a recent study, our findings imply total savings in psychiatric inpatient expenditures of $5772 during the first 6 months after discharge on average. However, reduction in hospitalization days did not result in a parallel improvement in functioning as assessed by the Quality of Life Scale.Although the control and experimental cohorts were comparable in many characteristics, the quasi–experimental design represented by sequential cohorts rather than a true concurrent randomized controlled trial represents a limitation. The results of the study suggest that technology enhanced treatments that are tailored to patient needs can be implemented in a range of clinical settings in the US to patients at high risk of hospitalization and that the intervention can reduce subsequent hospitalization days. Future research should address limitations in the current study design and will benefit from the development of technology applications that can be available on a single flexible platform.