scholarly journals Development and validation of a machine learning model predicting illness trajectory and hospital utilization of COVID-19 patients—a nationwide study

2021 ◽  
Author(s):  
Michael Roimi ◽  
Rom Gutman ◽  
Jonathan Somer ◽  
Asaf Ben Arie ◽  
Ido Calman ◽  
...  
Keyword(s):  
Critical Care ◽  
Clinical Status ◽  
Patient Characteristics ◽  
Hospitalized Patients ◽  
Hospital Utilization ◽  
Clinical State ◽  
Care Hospital ◽  
System Utilization ◽  
The Impact ◽  
Day By Day

Abstract Objective The spread of COVID-19 has led to severe strain on hospital capacity in many countries. We aim to develop a model helping planners assess expected COVID-19 hospital resource utilization based on individual patient characteristics. Materials and Methods We develop a model of patient clinical course based on an advanced multistate survival model. The model predicts the patient's disease course in terms of clinical states—critical, severe, or moderate. The model also predicts hospital utilization on the level of entire hospitals or healthcare systems. We cross-validated the model using a nationwide registry following the day-by-day clinical status of all hospitalized COVID-19 patients in Israel from March 1st to May 2nd, 2020 (n = 2,703). Results Per-day mean absolute errors for predicted total and critical-care hospital-bed utilization were 4.72 ± 1.07 and 1.68 ± 0.40 respectively, over cohorts of 330 hospitalized patients; AUCs for prediction of critical illness and in-hospital mortality were 0.88 ± 0.04 and 0.96 ± 0.04, respectively. We further present the impact of patient influx scenarios on day-by-day healthcare system utilization. We provide an accompanying R software package. Discussion The proposed model accurately predicts total and critical-care hospital utilization. The model enables evaluating impacts of patient influx scenarios on utilization, accounting for the state of currently hospitalized patients and characteristics of incoming patients. We show that accurate hospital-load predictions were possible using only a patient’s age, sex, and day-by-day clinical state (critical, severe or moderate). Conclusion The multistate model we develop is a powerful tool for predicting individual-level patient outcomes and hospital-level utilization.

scholarly journals Development and validation of a machine learning model for predicting illness trajectory and hospital resource utilization of COVID-19 hospitalized patients - a nationwide study

2020 ◽  
Author(s):  
Michael Roimi ◽  
Rom Gutman ◽  
Jonathan Somer ◽  
Asaf Ben Arie ◽  
Ido Calman ◽  
...  
Keyword(s):  
Machine Learning ◽  
Critical Care ◽  
Critical Illness ◽  
Length Of Stay ◽  
Hospital Mortality ◽  
Hospital Utilization ◽  
Care Hospital ◽  
Hospital Resource ◽  
Hospital Bed ◽  
The Impact

Background: The spread of COVID-19 has led to a severe strain on hospital capacity in many countries. There is a need for a model to help planners assess expected COVID-19 hospital resource utilization. Methods: Retrospective nationwide cohort study following the day-by-day clinical status of all hospitalized COVID-19 patients in Israel from March 1st to May 2nd, 2020. Patient clinical course was modelled with a machine learning approach based on a set of multistate Cox regression-based mod- els with adjustments for right censoring, recurrent events, competing events, left truncation, and time-dependent covariates. The model predicts the patient's entire disease course in terms of clinical states, from which we derive the patient's hospital length-of-stay, length-of-stay in critical state, the risk of in-hospital mortality, and total and critical care hospital-bed utilization. Accuracy assessed over eight cross-validation cohorts of size 330, using per-day Mean Absolute Error (MAE) of predicted hospital utilization averaged over 64 days; and area under the receiver operating characteristics (AUROC) for individual risk of critical illness and in-hospital mortality, assessed on the first day of hospitalization. We present predicted hospital utilization under hypothetical incoming patient scenarios. Findings: During the study period, 2,703 confirmed COVID-19 patients were hospitalized in Israel. The per-day MAEs for total and critical-care hospital- bed utilization, were 4.72 ± 1.07 and 1.68 ± 0.40 respectively; the AUROCs for prediction of the probabilities of critical illness and in-hospital mortality were 0.88 ± 0.04 and 0.96 ± 0.04, respectively. We further present the impact of several scenarios of patient influx on healthcare system utilization, and provide an R software package for predicting hospital-bed utilization. Interpretation: We developed a model that, given basic easily obtained data as input, accurately predicts total and critical care hospital utilization. The model enables evaluating the impact of various patient influx scenarios on hospital utilization and planning ahead of hospital resource allocation.

scholarly journals Evaluation of Clinical Outcomes after Introduction of a Dedicated Infectious Diseases- Critical Care Medicine Service in Critical Care Units

2021 ◽  
Author(s):  
Polina Trachuk ◽  
Vagish Hemmige ◽  
Ruth Eisenberg ◽  
Kelsie Cowman ◽  
Victor Chen ◽  
...  
Keyword(s):  
Critical Care ◽  
Infectious Diseases ◽  
Clinical Outcomes ◽  
Broad Spectrum ◽  
Critical Care Medicine ◽  
Care Hospital ◽  
Intervention Period ◽  
Broad Spectrum Antibiotics ◽  
Antibiotic Utilization ◽  
The Impact

Abstract Objective Infection is a leading cause of admission to intensive care units (ICU), with critically ill patients often receiving empiric broad-spectrum antibiotics. Nevertheless, a dedicated infectious diseases (ID) consultation and stewardship team is not routinely established. An ID-Critical Care Medicine (ID-CCM) pilot program was designed at a 400-bed tertiary care hospital in which an ID attending was assigned to participate in daily rounds with the ICU team, as well as provide ID consultation on select patients. We sought to evaluate the impact of this dedicated ID program on antibiotic utilization and clinical outcomes in patients admitted to the ICU. Method In this single site retrospective study, we analyzed antibiotic utilization and clinical outcomes in patients admitted to an ICU during post-intervention period from January 1, 2017 to December 31, 2017 and compared it to antibiotic utilization in the same ICUs during the pre-intervention period from January 1, 2015 to December 31, 2015. Results Our data showed a statistically significant reduction in usage of most frequently prescribed antibiotics including vancomycin, piperacillin-tazobactam and cefepime during the intervention period. When compared to pre-intervention period there was no difference in-hospital mortality, hospital length of stay and re-admission. Conclusion With this multidisciplinary intervention, we saw a decrease in the use of the most frequently prescribed broad-spectrum antibiotics without a negative impact on clinical outcomes. Our study shows that the implementation of an ID-CCM service is a feasible way to promote antibiotic stewardship in the ICU and can be used as a strategy to reduce unnecessary patient exposure to broad-spectrum agents.

scholarly journals Perception and Impact of Life Events in Medically Hospitalized Patients With Somatic Symptom and Related Disorders

2018 ◽  
Vol 8 (11) ◽  
pp. 699-705
Author(s):  
Katherine K. Bedard-Thomas ◽  
Simona Bujoreanu ◽  
Christine H. Choi ◽  
Patricia I. Ibeziako
Keyword(s):  
Life Events ◽  
Somatic Symptom ◽  
Functional Disability ◽  
De Novo ◽  
Patient Characteristics ◽  
Hospitalized Patients ◽  
Health Care Use ◽  
Life Event ◽  
Academic Events ◽  
The Impact

OBJECTIVES: We describe the prevalence and perceived impact of life events reported by medically hospitalized patients with somatic symptom and related disorders (SSRD) and highlight patient characteristics and outcomes associated with highly impactful life events. METHODS: Retrospective chart reviews were conducted of patients with SSRD at a tertiary pediatric hospital who were seen by the psychiatry consultation service and completed various instruments while medically admitted, including a de novo life events checklist. Descriptive statistics, correlations, χ2 tests, and internal consistency analyses were used. RESULTS: Charts of 70.2% of patients with SSRD who completed the life events checklist (N = 172; age range 8–25 years) were reviewed. Of those studied, 94% reported at least 1 life event in the last year, with academic events most prevalent, 81% reported life events across multiple domains, and 56% perceived the life event(s) as having a great impact on their lives. Patients who perceived more great impact life events were older, from households with lower median incomes, had higher self-reported somatization, greater functional disability, more comorbid psychiatric diagnoses, required more psychotropic medications, and had longer medical admissions. CONCLUSIONS: Findings reveal that although the majority of medically hospitalized patients with SSRD reported at least 1 relevant life event, it was the patients’ perception of the impact of the life event(s) that correlated with high levels of disability and health care use. An assessment of the perception of life events in patients with SSRD may help hospitalists and interdisciplinary providers identify high-risk patients for whom early psychiatry referrals can be made.

Development of an Ordinal Scale Treatment Endpoint for Adults Hospitalized With Influenza

2020 ◽  
Author(s):  
Nelson Lee ◽  
Stephanie W Smith ◽  
David S C Hui ◽  
Ming Ye ◽  
Nathan Zelyas ◽  
...  
Keyword(s):  
Respiratory Failure ◽  
Acute Care ◽  
Clinical Status ◽  
Neuraminidase Inhibitor ◽  
Hospitalized Patients ◽  
Ordinal Scale ◽  
Clinical Endpoints ◽  
Subacute Care ◽  
The Impact ◽  
Ordinal Outcome

Abstract Background An obstacle in influenza therapeutics development is the lack of clinical endpoints, especially in hospitalized patients. A single time-point binary outcome measure is limited by patients’ diverse clinical trajectories and low event rates. Methods A 6-point ordinal scale with ascending clinical status severity (scoring: discharged; subacute care; acute care without/with respiratory failure; intensive care unit [ICU]; death) was proposed to study outcomes of adults hospitalized with influenza. Individual patient data from 2 active surveillance cohorts’ datasets (2015/2016−2017/2018; Edmonton, Hong Kong) was used for evaluation. The impact of neuraminidase inhibitor (NAI) treatment on longitudinal ordinal outcome changes over 30 days was analyzed using mixed-effects ordinal logistic regression and group-based trajectory models. Results Patient (n = 1226) baseline characteristics included age (mean 68.0 years), virus-type (A 78.1%, B 21.9%), respiratory failure (57.2%), ICU admittance (14.4%), and NAI treatment within 5 days of illness (69.2%). Outcomes at 30 days included discharged (75.2%), subacute care (13.7%), acute care (4.5%), and death (6.6%). Two main clinical trajectories were identified, predictive by baseline scoring (mean ± SD, 4.3 ± 0.6 vs 3.5 ± 0.6, P < .001). Improved outcomes with NAI treatment within 5 days were indicated by significantly lower clinical status scores over time (unadjusted odds ratio [OR], 0.53; 95% confidence interval [CI], .41−.69; P < .001; adjusted OR, 0.62; 95% CI, .50−.77; P < .001, for baseline score, age, and within-patient correlations). In subanalysis, influenza vaccination was also associated with lower scores (adjusted OR, 0.67; 95% CI, .50−.90; P = .007). Analyses of binary endpoints showed insignificant results. Conclusions The ordinal outcome scale is a potentially useful clinical endpoint for influenza therapeutic trials, which could account for the diverse clinical trajectories of hospitalized patients, warranting further development.

scholarly journals Variation Among Hospitals in the Continuity of Care for Older Hospitalized Patients: a Cross-sectional Cohort Study

2021 ◽  
Author(s):  
James S. Goodwin ◽  
Shuang Li ◽  
Jie Zhou ◽  
Yong-Fang Kuo ◽  
Ann Nattinger
Keyword(s):  
Logistic Regression ◽  
Length Of Stay ◽  
Regression Model ◽  
Continuity Of Care ◽  
Logistic Regression Model ◽  
Large Degree ◽  
Patient Characteristics ◽  
Hospitalized Patients ◽  
Cross Sectional ◽  
The Impact

Abstract Background: Little is known about how continuity of care for hospitalized patients varies among hospitals. We describe the number of different general internal medicine physicians seeing hospitalized patients during a medical admission and how that varies by hospital. Methods: We conducted a retrospective study of a national 20% sample of Medicare inpatients from 01/01/16 to 12/31/18. In patients with routine medical admissions (length of stay of 3-6 days, no Intensive Care Unit stay, and seen by only one generalist per day), we assessed odds of receiving all generalist care from one generalist. We calculated rates for each hospital, adjusting for patient and hospital characteristics in a multi-level logistic regression model. Results: Among routine medical admissions with 3- to 6-day stays, only 43.1% received all their generalist care from the same physician. In those with a 3-day stay, 50.1% had one generalist providing care vs. 30.8% in those with a 6-day stay. In a two-level (admission and hospital) logistic regression model controlling for patient characteristics and length of stay, the odds of seeing just one generalist did not vary greatly by patient characteristics such as age, race/ethnicity, comorbidity or reason for admission. There were large variations in continuity of care among different hospitals and geographic areas. In the highest decile of hospitals, the adjusted mean percentage of patients receiving all generalist care from one physician was >84.1%, vs. <24.1% in the lowest decile. This large degree of variation persisted when hospitals were stratified by size, ownership, location or teaching status. Conclusions: Continuity of care provided by generalist physicians to medical inpatients varies widely among hospitals. The impact of this variation on quality of care is unknown.

scholarly journals The Impact of a Pediatric Antibiotic Standard Dosing Table on Dosing Errors

2013 ◽  
Vol 18 (3) ◽  
pp. 220-226 ◽  
Author(s):  
Mohammed A. Aseeri
Keyword(s):  
Pediatric Patients ◽  
Tertiary Care Hospital ◽  
Tertiary Care ◽  
Patient Characteristics ◽  
Care Hospital ◽  
Implementation Phase ◽  
Before And After ◽  
The Impact ◽  
Standardization Policy ◽  
Antibiotic Dosing

OBJECTIVE The goal of this study was to compare the rate of dosing errors for antibiotic orders in pediatric patients before and after the implementation of an antibiotic standard dosing table with precalculated dosage for different weight ranges at a tertiary care hospital. METHODS A retrospective study of 300 antibiotic prescriptions for pediatric patients in three different settings (ambulatory care, inpatient, and emergency department) at a tertiary care hospital assessed the appropriateness of antibiotic dosing. The need for an antibiotic dosing standardization policy was identified after finding that more than 30% of patients experienced a dose variation of ±10% of the recommended daily dose. An antibiotic dosing standardization policy was implemented with an antibiotic standard dosing table for different weight ranges, and a hospital wide-education program was conducted to increase awareness of this new practice and its benefits. Three months after implementation, a random sampling of 300 antibiotic prescriptions collected from the same settings as the pre-intervention period was evaluated for compliance with the new policy and its effect on the number of antibiotic dosing errors. RESULTS Six hundred prescriptions were included in this study (300 in the pre-implementation phase and 300 in the post-implementation phase). Patient characteristics were similar in both groups in terms of sex, age, and weight. Physician compliance with the antibiotic dosing standardization policy after its implementation was 62%. The dosing standardization policy reduced the rate of dosing errors from 34.3% to 5.06% (p=0.0001), and weight documentation on the antibiotic prescription improved from 65.8% to 85.7% (p=0.0001). CONCLUSIONS Implementation of an antibiotic dosing standardization policy significantly reduced the incidence of dosing errors in antibiotics prescribed for pediatric patients in our hospital.

Clinical and economic outcomes attributable to carbapenem-resistant Enterobacterales and delayed appropriate antibiotic therapy in hospitalized patients

2021 ◽  
pp. 1-11
Author(s):  
Kirati Kengkla ◽  
Yuttana Wongsalap ◽  
Natthaya Chaomuang ◽  
Pichaya Suthipinijtham ◽  
Peninnah Oberdorfer ◽  
...  
Keyword(s):  
Antibiotic Therapy ◽  
Tertiary Care ◽  
Hospital Costs ◽  
Mortality Rates ◽  
Hospitalized Patients ◽  
Economic Outcomes ◽  
Care Hospital ◽  
Carbapenem Resistant ◽  
Medication Costs ◽  
The Impact

Abstract Objective: To assess the impact of carbapenem resistance and delayed appropriate antibiotic therapy (DAAT) on clinical and economic outcomes among patients with Enterobacterales infection. Methods: This retrospective cohort study was conducted in a tertiary-care medical center in Thailand. Hospitalized patients with Enterobacterales infection were included. Infections were classified as carbapenem-resistant Enterobacterales (CRE) or carbapenem-susceptible Enterobacterales (CSE). Multivariate Cox proportional hazard modeling was used to examine the association between CRE with DAAT and 30-day mortality. Generalized linear models were used to examine length of stay (LOS) and in-hospital costs. Results: In total, 4,509 patients with Enterobacterales infection (age, mean 65.2 ±18.7 years; 43.3% male) were included; 627 patients (13.9%) had CRE infection. Among these CRE patients, 88.2% received DAAT. CRE was associated with additional medication costs of $177 (95% confidence interval [CI], 114–239; P < .001) and additional in-hospital costs of $725 (95% CI, 448–1,002; P < .001). Patients with CRE infections had significantly longer LOS and higher mortality rates than patients with CSE infections: attributable LOS, 7.3 days (95% CI, 5.4–9.1; P < .001) and adjusted hazard ratios (aHR), 1.55 (95% CI, 1.26–1.89; P < .001). CRE with DAAT were associated with significantly longer LOS, higher mortality rates, and in-hospital costs. Conclusion: CRE and DAAT are associated with worse clinical outcomes and higher in-hospital costs among hospitalized patients in a tertiary-care hospital in Thailand.

scholarly journals Efficacy of scopolamine transdermal patch in children with sialorrhea in a pediatric tertiary care hospital

2020 ◽  
Vol 20 (1) ◽  
Author(s):  
Majed Al Jeraisy ◽  
Maissa AlFuraih ◽  
Raghad AlSaif ◽  
Bushra AlKhalifah ◽  
Hazza AlOtaibi ◽  
...  
Keyword(s):  
Relative Risk ◽  
Neurological Disorders ◽  
Rating Scale ◽  
Clinical Status ◽  
Tertiary Care ◽  
Family Satisfaction ◽  
Phone Call ◽  
Care Hospital ◽  
Transdermal Patch ◽  
The Impact

Abstract Background Drooling is common in children with neurological disorders, but its management is very challenging, Scopolamine transdermal patch (STP) appears to be useful in controlling drooling, although it is not approved for this indication and there are limited clinical studies about its effectiveness. This study aimed (1) to assess the impact of STP use on the severity of drooling and on the frequency of emergency department (ED) and hospital readmission (RA) visits related to drooling, and (2) to determine the level of family satisfaction with STP when used in children with neurological disorders. Methods This is a retrospective cohort study of all pediatric patients aged 3–14 years, with non-progressive neurodevelopmental disability, who used STP for more than one year during the period between April 2015 and July 2018 (n = 44). Data on demographics, clinical status, comorbidities, STP dose and duration, other medications, ED and RA visits were collected. Follow-up phone-call interviews with parents/caregivers were performed using a parent-reported frequency and severity rating scale of sialorrhea. Absolute and relative risk reductions were calculated to assess the impact of STP on ED and RA visits. Significance was considered at p-value of ≤ 0.05. Results STP use showed significant reduction in severity of drooling (p < 0.001), wiping of the child’s mouth (p < 0.001), bibs or clothing changes (p < 0.001), choking and aspiration of saliva (p = 0.001). The Relative Risk Reduction of the drooling-related ED and RA visits were 86% and 67% respectively. Nearly two-thirds (60%) of caregivers were satisfied with using STP. Conclusions This is the first study of its kind done in Saudi Arabia demonstrating favorable impact of STP use by children on the consequences associated with drooling and with the frequency of ER and RA visits due to drooling. Development of a medication use protocol is recommended to standardize STP treatment in order to optimize its effectiveness. This study serves as baseline information for future prospective interventional studies.

scholarly journals MP12: Emergency department boarding: predictors and outcomes

CJEM ◽  
2018 ◽  
Vol 20 (S1) ◽  
pp. S44-S44
Author(s):  
L. Salehi ◽  
V. Jegatheeswaran ◽  
J. Herman ◽  
P. Phalpher ◽  
R. Valani ◽  
...  
Keyword(s):  
Critical Care ◽  
Patient Care ◽  
High Volume ◽  
Patient Characteristics ◽  
Patient Specific ◽  
Medicine Service ◽  
Comorbidity Burden ◽  
Ed Overcrowding ◽  
Ed Visits ◽  
The Impact

Introduction: Delays in transfer to an in-patient bed of admitted patients boarded in the ED has been identified as one of the chief drivers of ED overcrowding. Our study aims to replicate findings from a previous study in identifying patient characteristics associated with increased boarding time, and the impact of increased boarding time on in-patient length of stay (IPLOS). Methods: We conducted a retrospective single-centre observational study during the period between January 1, 2015 December 31, 2015 at a very high volume community hospital (~ 75,000 ED visits/year). All patients admitted from the ED to Medicine, Pediatrics, Surgery, and Critical Care were identified. The mean time to in-patient bed (TTB), as well as patient-specific and institutional factors that were associated with prolonged boarding times ( 12 hours) were identified. Mean IP LOS was calculated for those with prolonged boarding times and compared to those without prolonged boarding times. Results: There were 8,096 unique admissions during the study period. Patients admitted to the Medicine service exhibited significantly higher boarding times than those admitted to other services, with a mean boarding time of 17.4 hrs, as compared to 4.2 hrs, 5.7 hrs, and 4.0 hrs for those admitted to Surgery, Critical Care and Pediatrics respectively. Within Medicine patients, there was a statistically significant greater odds of prolonged boarding time for patients who were older, had a greater comorbidity burden, and required more specialized in-patient care (i.e. an isolation bed or telemetry bed). Medicine patients with prolonged boarding times also experienced 0.7 days longer IP LOS, even after correcting for age and comorbidity (mean adjusted IP LOS 10.6 days versus 11.3 days). Conclusion: Within our study period, older, sicker patients and those patients requiring more resource-intensive in-patient care have the longest ED boarding times. These prolonged ‘boarding’ times are associated with significantly increased IP LOS.

A Nationwide Study of the Impact of Dysphagia on Hospital Outcomes Among Patients With Dementia

2016 ◽  
Vol 32 (1) ◽  
pp. 5-11 ◽  
Author(s):  
Suchitra Paranji ◽  
Neethi Paranji ◽  
Scott Wright ◽  
Shalini Chandra
Keyword(s):  
Mechanical Ventilation ◽  
Patient Characteristics ◽  
Hospitalized Patients ◽  
Skilled Nursing ◽  
Percutaneous Endoscopic Gastrostomy Placement ◽  
Endoscopic Gastrostomy ◽  
Diagnosis Of Dementia ◽  
The Impact ◽  
New Evidence ◽  
Mean Charge

Objectives: To assess the impact of dysphagia on clinical and operational outcomes in hospitalized patients with dementia. Design: Retrospective cohort study. Setting: 2012 Nationwide Inpatient Sample. Participants: All patients discharged with a diagnosis of dementia (N = 234,006) from US hospitals in 2012. Measurements: Univariate and multivariate regression models, adjusting for stroke and patient characteristics, to assess the impact of dysphagia on the prevalence of comorbidities, including pneumonia, sepsis, and malnutrition; complications, including mechanical ventilation and death; and operational outcomes, including length of stay (LOS) and total charges for patients with dementia. Results: Patients having dementia with dysphagia (DWD) had significantly higher odds of having percutaneous endoscopic gastrostomy placement during the admission (odds ratio [OR]: 13.68, 95% confidence interval [CI]: 12.53-14.95, P < .001), aspiration pneumonia (OR: 6.27, 95% CI: 5.87-6.72, P < .001), pneumonia (OR: 2.84, 95% CI: 2.67-3.02, P < .001), malnutrition (OR: 2.5, 95% CI: 2.27-2.75, P < .001), mechanical ventilation (OR: 1.69, 95% CI: 1.51-1.9, P < .001), sepsis (OR: 1.52, 95% CI: 1.39-1.67, P < .001), and anorexia (OR: 1.29, 95% CI: 1.01-1.65, P = .04). Mean LOS was 2.16 days longer (95% CI: 1.98-2.35, P < .001), mean charge per case was US$10,703 higher (95% CI: US$9396-US$12,010, P < .001), and the odds of being discharged to a skilled nursing, rehabilitation, or long-term facility was 1.59 times higher (95% CI: 1.49-1.69, P < .001) in the DWD cohort compared to patients having dementia without dysphagia. Conclusion: Dysphagia is a significant predictor of worse clinical and operational outcomes including a 38% longer LOS and a 30% increase in charge per case among hospitalized patients with dementia. Although these findings may not be surprising, this new evidence might bring heightened awareness for the need to more thoughtfully support patients with dementia and dysphagia who are hospitalized.

Sign in / Sign up

Export Citation Format

Share Document