The Banner Psychiatric Center: A Model for Providing Psychiatric Crisis Care to the Community while Easing Behavioral Health Holds in Emergency Departments

Abstract Background: It is difficult to accurately predict whether a patient on the verge of a potential psychiatric crisis will need to be hospitalized. Machine learning may be helpful to improve the accuracy of psychiatric hospitalization prediction models. In this paper we evaluate and compare the accuracy of ten machine learning algorithms including the commonly used generalized linear model (GLM/logistic regression) to predict psychiatric hospitalization in the first 12 months after a psychiatric crisis care contact, and explore the most important predictor variables of hospitalization. Methods: Data from 2,084 patients with at least one reported psychiatric crisis care contact included in the longitudinal Amsterdam Study of Acute Psychiatry were used. The accuracy and area under the receiver operating characteristic curve (AUC) of the machine learning algorithms were compared. We also estimated the relative importance of each predictor variable. The best and least performing algorithms were compared with GLM/logistic regression using net reclassification improvement analysis. Target variable for the prediction models was whether or not the patient was hospitalized in the 12 months following inclusion in the study. The 39 predictor variables were related to patients’ socio-demographics, clinical characteristics and previous mental health care contacts. Results: We found Gradient Boosting to perform the best (AUC=0.774) and K-Nearest Neighbors performing the least (AUC=0.702). The performance of GLM/logistic regression (AUC=0.76) was above average among the tested algorithms. Gradient Boosting outperformed GLM/logistic regression and K-Nearest Neighbors, and GLM outperformed K-Nearest Neighbors in a Net Reclassification Improvement analysis, although the differences between Gradient Boosting and GLM/logistic regression were small. Nine of the top-10 most important predictor variables were related to previous mental health care use. Conclusions: Gradient Boosting led to the highest predictive accuracy and AUC while GLM/logistic regression performed average among the tested algorithms. Although statistically significant, the magnitude of the differences between the machine learning algorithms was modest. Future studies may consider to combine multiple algorithms in an ensemble model for optimal performance and to mitigate the risk of choosing suboptimal performing algorithms.

Download Full-text

Emergency Departments, Behavioral Health, And More

Health Affairs ◽

10.1377/hlthaff.2017.1188 ◽

2017 ◽

Vol 36 (10) ◽

pp. 1695-1695

Author(s):

Alan R. Weil

Keyword(s):

Behavioral Health ◽

Emergency Departments

Download Full-text

Care Coordination in Emergency Departments for Children and Adolescents With Behavioral Health Conditions

Pediatric Emergency Care ◽

10.1097/pec.0000000000001545 ◽

2018 ◽

Vol Publish Ahead of Print ◽

Cited By ~ 5

Author(s):

Sean Lynch ◽

Whitney Witt ◽

Mir M. Ali ◽

Judith L. Teich ◽

Ryan Mutter ◽

...

Keyword(s):

Children And Adolescents ◽

Behavioral Health ◽

Care Coordination ◽

Emergency Departments ◽

Health Conditions

Download Full-text

53.2 CRISIS CARE FOR YOUTH PRESENTING WITH ACUTE BEHAVIORAL HEALTH NEEDS

Journal of the American Academy of Child & Adolescent Psychiatry ◽

10.1016/j.jaac.2020.07.337 ◽

2020 ◽

Vol 59 (10) ◽

pp. S81

Author(s):

Eileen Twohy ◽

Molly Adrian ◽

Kalina Babeva ◽

Kyrill Gurtovenko ◽

Sophie King ◽

...

Keyword(s):

Behavioral Health ◽

Health Needs ◽

Crisis Care

Download Full-text

Prevalence of impairing behavioral health problems in ED patients and association with ED utilization.

Proceedings of IMPRS ◽

10.18060/23544 ◽

2019 ◽

Vol 2 (1) ◽

Author(s):

Setarah Mohammad Nader ◽

Paul Musey Jr., MD, MS, FACEP

Keyword(s):

Mental Health ◽

Suicidal Ideation ◽

Behavioral Health ◽

Emergency Departments ◽

Mental Health Problems ◽

Severe Depression ◽

Health Problems ◽

Screening Tools ◽

Convenience Sample ◽

Severe Anxiety

Background and Hypothesis: It has been observed that patients with poor mental health are relatively frequent users of the Emergency Departments (ED). The objective of this study is to evaluate the prevalence of numerous behavioral health domains (depression, anxiety, PTSD, substance abuse, and suicidality) in patients presenting to the Emergency Department and the association of each of these domains with ED utilization. Experimental Design or Project Methods: This prospective study seeks to enroll a convenience sample of 1000 Englishspeaking adults presenting to IU Health Methodist and Eskenazi Emergency Departments without psychiatric chief-complaints. Patients were assessed for behavioral health problems using the CAT-MHTM, PHQ-8 and GAD-7 screening tools, which were administered via tablet device. Additionally, data on disposition medical history, discharge diagnoses, and ED utilization in the 12 months before and after enrollment from electronic medical records and data from the Indiana Network for Patient Care (INPC) will be reviewed. Results: Over the course of five weeks, 375 patients have been enrolled. Of those 59.4% were female with an overall mean age of 46.1 (SD ± 16.4); 52.9% were white and 39.8% black/African American. Among enrollees 42.2% screened positive for depression, 29.7% for anxiety, and 1.3% for suicidal ideation. Patients who screened positive for depression were predominately females (76.1% vs 23.9%), those who screened positive for anxiety were also predominately females (71.6% vs. 28.4%). However, 3 out of the 5 (60%) patients that screened positive for suicidal ideation were males. The preliminary analysis of GAD-7 showed of those enrolled 215 (57.5%) had no anxiety, 157 (42%) had mild-severe anxiety. PHQ-8 scores showed 194 (51.9%) had no depression, 178 (47.5%) had mild-severe depression. Similarly, CAT-MH results showed 216 (57.8%) had no depression, 158 (42.2%) had mild-severe depression, while 263 (70.3%) had no anxiety and 111 (29.7%) had mild-severe anxiety. Full data analysis including comparative analysis of the CAT-MH with PHQ-8 and GAD-7 scores will take place after 1000 patients have been enrolled and data has been received from the INPC. Conclusion and Potential Impact: In our sample, almost half of patients that visit the ED have screened positive for mental health problems. We believe that early identification and appropriate referral may reduce inappropriate ED utilization.

Download Full-text

“Caring for a Crisis”: Care and Control in Community Mental Health

Frontiers in Psychiatry ◽

10.3389/fpsyt.2021.798599 ◽

2022 ◽

Vol 12 ◽

Author(s):

Christien Muusse ◽

Hans Kroon ◽

Cornelis Lambert Mulder ◽

Jeannette Pols

Keyword(s):

Mental Health ◽

Community Mental Health ◽

Open Door Policy ◽

Mental Health Team ◽

Psychiatric Crisis ◽

Crisis Care ◽

Specific Care ◽

And Control ◽

Psychiatric Beds ◽

The Relationship

In the debate on coercion in psychiatry, care and control are often juxtaposed. In this article we argue that this dichotomy is not useful to describe the more complex ways service users, care professionals and the specific care setting interrelate in a community mental health team (CMHT). Using the ethnographic approach of empirical ethics, we contrast the ways in which control and care go together in situations of a psychiatric crisis in two CMHT's: one in Trieste (Italy) and one in Utrecht (the Netherlands). The Dutch and Italian CMHT's are interesting to compare, because they differ with regard to the way community care is organized, the amount of coercive measures, the number of psychiatric beds, and the fact that Trieste applies an open door policy in all care settings. Contrasting the two teams can teach us how in situations of psychiatric crisis control and care interrelate in different choreographies. We use the term choreography as a metaphor to encapsulate the idea of a crisis situation as a set of coordinated actions from different actors in time and space. This provides two choreographies of handling a crisis in different ways. We argue that applying a strict boundary between care and control hinders the use of the relationship between caregiver and patient in care.

Download Full-text

Simulation-based medical education can be used to improve the mental health competency of emergency physicians

Australasian Psychiatry ◽

10.1177/1039856220901480 ◽

2020 ◽

Vol 28 (3) ◽

pp. 354-358

Author(s):

Andrew Coggins ◽

Dale Marchant ◽

Jane Bartels ◽

Brett Cliff ◽

Sandra Warburton ◽

...

Keyword(s):

Emergency Department ◽

Medical Education ◽

Emergency Departments ◽

Legal Issues ◽

Course Evaluations ◽

Emergency Physicians ◽

Multiple Choice Questions ◽

Emergency Department Nurses ◽

Psychiatric Crisis ◽

Simulation Based

Objective: We explored the feasibility of developing, running and evaluating a simulation-based medical education (SBME) workshop to improve the knowledge, skills and attitudes of emergency department (ED) doctors when called on to assess patients in psychiatric crisis. Method: We designed a four-hour workshop incorporating SBME and a blend of pre-reading, short didactic elements and multiple-choice questions (MCQs). Emergency department nurses (operating as SBME faculty) used prepared scripts to portray patients presenting in psychiatric crisis. They were interviewed in front of, and by, ED doctors. We collected structured course evaluations, Debriefing Assessment for Simulation in Healthcare (DASH) scores, and pre- and post-course MCQs. Results: The pilot workshop was delivered to 12 ED registrars using only existing resources of the Psychiatry and Emergency Departments. Participants highly valued both ‘level of appropriateness’ (Likert rating μ = 4.8/5.0) and ‘overall usefulness’ ( μ = 4.7/5.0) of the programme. They reported an improved understanding of the mental state and of relevant legal issues and rated the debriefings highly (participant DASH rating: n = 193; score μ = 6.3/7.0). Median MCQ scores improved non-significantly pre- and post-course (7.5/12 vs 10/12, p = 0.261). Conclusion: An SBME workshop with these aims could be delivered and evaluated using the existing resources of the Psychiatry and Emergency Departments.

Download Full-text