Overcoming gaps: regional collaborative to optimize capacity management and predict length of stay of patients admitted with COVID-19.

OBJECTIVE: Ensuring an efficient response to COVID-19 requires a degree of inter-system coordination and capacity management coupled with an accurate assessment of hospital utilization including length of stay (LOS). We aimed to establish optimal practices in inter-system data sharing and LOS modeling to support patient care and regional hospital operations. MATERIALS AND METHODS: We completed a retrospective observational study of patients admitted with COVID-19 followed by 12-week prospective validation, involving 36 hospitals covering the upper Midwest. We developed a method for sharing de-identified patient data across systems for analysis. From this, we compared 3 approaches, generalized linear model (GLM) and random forest (RF), and aggregated system level averages to identify features associated with LOS. We compared model performance by area under the ROC curve (AUROC). RESULTS: A total of 2068 patients were included and used for model derivation and 597 patients for validation. LOS overall had a median of 5.0 days and mean of 8.2 days. Consistent predictors of LOS included age, critical illness, oxygen requirement, weight loss, and nursing home admission. In the validation cohort, the RF model (AUROC 0.890) and GLM model (AUROC 0.864) achieved good to excellent prediction of LOS, but only marginally better than system averages in practice. CONCLUSION: Regional sharing of patient data allowed for effective prediction of LOS across systems; however, this only provided marginal improvement over hospital averages at the aggregate level. A federated approach of sharing aggregated system capacity and average LOS will likely allow for effective capacity management at the regional level.

Timely Data for Targeted Quality Improvement Interventions: Use of a Visual Analytics Dashboard for Bronchiolitis.

BACKGROUND: Standard methods for obtaining data may delay quality improvement (QI) interventions including for bronchiolitis, a common cause of childhood hospitalization. OBJECTIVE: To describe the use of a dashboard in the context of a multifaceted QI intervention aimed at reducing the use of chest radiographs, bronchodilators, antibiotics, steroids, and viral testing in patients with bronchiolitis. METHODS: This QI initiative took place at Children's Minnesota, a large, not-for-profit children's health care organization. A multidisciplinary bronchiolitis workgroup developed a local clinical guideline and order-set. Delays in obtaining baseline data prompted a pediatric hospitalist and information technology specialist to modify a vendor's dashboard to display data related to bronchiolitis guideline metrics. Patients 2 months to 2 years old with a bronchiolitis emergency department (ED)/inpatient encounter in the period October 1, 2014 to April 30, 2018 were included. The primary outcome was a functioning dashboard; a process measure was the percentage of ED clinician logins. Outcome measures included the percent use of guideline metrics (e.g., bronchodilators) displayed on statistical process control charts (ED vs. inpatient). Balancing measures included length of stay, charge ratios, and hospital revisits. RESULTS: A workgroup (formed October 2015) implemented a bronchiolitis order-set and guideline (February 2016) followed by a bronchiolitis dashboard (August 2016) consolidating disparate data sources loaded within 2 to 4 days of discharge. In total, 35% of ED clinicians logged in. Leaders used the dashboard to target and track interventions such as a bronchodilator order alert. There were improvements in most outcome metrics; however, timing did not suggest direct dashboard impact. ED balancing measures were lower after implementation. CONCLUSION: We described use of a dashboard to support a multifaceted QI initiative for bronchiolitis. Leaders used the dashboard for targeted interventions but the dashboard did not directly impact the observed improvements. Future studies should assess reasons for low individual dashboard use.

Characterizing Patient-Generated Clinical Data and Associated Implications for Electronic Health Records.

Patient-facing technologies are increasingly utilized for direct patient data entry for potential incorporation into the electronic health record. We analyzed patient-entered data during implementation of a patient-facing data entry technology using an online patient portal and clinic-based tablet computers at a University-based tertiary medical center clinic, including entries for past medical history, past surgical history, and social history. Entries were assessed for granularity, clinical accuracy, and the addition of novel information into the record. We found that over half of patient-generated diagnoses were duplicates of lesser or equal granularity compared to previous provider-entered diagnoses. Approximately one fifth of patient-generated diagnoses were found to meet the criteria for new, meaningful additions to the medical record. Our findings demonstrate that while patient-generated data provides important additional information, it may also present challenges including generating inaccurate or less granular information.

Fluoxetine and aripiprazole treatment following prenatal immune activation exert longstanding effects on rat locomotor response.

AIM: Studies characterizing treatment interventions in a naturalistic setting suggest that antidepressant and antipsychotic medications may be equally effective in improving clinical outcome in individuals at high risk for first-episode psychosis. Of interest, both beneficial as well as potentially adverse effects have been observed following fluoxetine treatment in a mouse prenatal immune activation model of relevance to psychosis prevention. We sought to extend those findings by examining the effects of fluoxetine, as well as the antipsychotic medication aripiprazole, in a rat prenatal immune activation model. METHODS: Pregnant Sprague-Dawley rats were injected with poly I:C or saline on gestational day 14. Offspring of poly I:C and saline-treated dams received fluoxetine (10.0 mg/kg/d), aripiprazole (0.66 mg/kg/d), or vehicle from postnatal days 35 to 70. Locomotor responses to novelty, saline injection, and amphetamine (1 and 5 mg/kg) were determined at three months, i.e., 21 days following drug discontinuation. RESULTS: Both fluoxetine and aripiprazole had beneficial effects on behavioral response to amphetamine (1 mg/kg) at 3 months, ameliorating the impact of prenatal immune activation on offspring of poly I:C-treated dams. Significantly, both drugs also exerted effects in offspring of control (saline-treated) dams on locomotor response to injection. CONCLUSIONS: Fluoxetine and aripiprazole pretreatment of poly I:C offspring from postnatal days 35 to 70 stabilized response to amphetamine exposure persisting through 3 months of age, similar to earlier findings in mice that fluoxetine treatment following prenatal immune activation prevented altered locomotor response to amphetamine. The current data also confirm earlier findings of potential adverse behavioral effects in offspring of control dams following treatment with fluoxetine and antipsychotic medications, highlighting the potential for both therapeutic as well as safety concerns with exposure to preventive pharmacological treatments over the course of adolescent development. Further study is needed to determine clinical and epidemiological consequences of these pre-clinical findings.

Parietal lobe volume deficits in adolescents with schizophrenia and adolescents with cannabis use disorders.

OBJECTIVE: In early-onset schizophrenia (EOS), the earliest structural brain volumetric abnormalities appear in the parietal cortices. Early exposure to cannabis may represent an environmental risk factor for developing schizophrenia. This study characterized cerebral cortical gray matter structure in adolescents in regions of interest (ROIs) that have been implicated in EOS and cannabis use disorders (CUD). METHOD: T1-weighted magnetic resonance images were acquired from adolescents with EOS (n = 35), CUD (n = 16), EOS + CUD (n = 13), and healthy controls (HC) (n = 51). Using FreeSurfer, brain volume was examined within frontal, temporal, parietal and subcortical ROIs by a 2 (EOS versus no EOS) × 2 (CUD versus no CUD) design using multivariate analysis of covariance. In ROIs in which volumetric differences were identified, additional analyses of cortical thickness and surface area were conducted. RESULTS: A significant EOS-by-CUD interaction was observed. In the left superior parietal region, both "pure" EOS and "pure" CUD had smaller gray matter volumes that were associated with lower surface area compared with HC. A similar alteration was observed in the comorbid group compared with HC, but there was no additive volumetric deficit found in the comorbid group compared with the separate groups. In the left thalamus, the comorbid group had smaller gray matter volumes compared with the CUD and HC groups. CONCLUSIONS: These preliminary data indicate that the presence of a CUD may moderate the relationship between EOS and cerebral cortical gray matter structure in the left superior parietal lobe. Future research will follow this cohort over adolescence to further examine the impact of cannabis use on neurodevelopment.

Cochlear implant complications: utility of federal database in systematic analysis.

OBJECTIVES: To explore the suitability of the Manufacturer User Facility and Distributor Experience (MAUDE) database (which is maintained by the Food and Drug Administration and has a mandatory reporting requirement) for systemic analysis of cochlear implant complications and treatments and, in so doing, analyze trends in cochlear implant complications for 2 periods, 2002 and pre-1998. DATA SOURCES: All events from 2002 and from before 1998 were considered. Events and action taken were categorized and tabulated. DATA SYNTHESIS: Because there was no null hypothesis, statistical analysis (chi2) was only used in comparing the 2 time frames. CONCLUSIONS: Structural limitations of the database, in addition to disparate reporting quality, made systematic analysis difficult. It was noted that spontaneous device failure accounted for the greatest single number of adverse events for both 2002 and the pre-1998 period (267/654 [41%] and 74/129 [57%], respectively), confirming earlier studies. A statistically significant decrease in spontaneous device failure and a significant increase in infections from the pre-1998 period to 2002 was observed. Flap problems ranging from extrusion to infection that required explantation were less frequently reimplanted than other problems requiring explantation, such as device failure or trauma. We considered new directions, including close collaboration with the new MedSun reporting system and conclude that while a valuable resource for narrative data, the current structure of the MAUDE database is only modestly useful for analyzing trends in complications and cannot answer several crucial questions, including device type comparisons. We suggest changing the current report format to include patient age, duration of implant, presence of anatomical abnormalities, and details on spontaneous device failures.