Antecedent treat-and-release diagnoses prior to sepsis hospitalization among adult emergency department patients: a look-back analysis employing insurance claims data using Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) methodology

Diagnosis ◽  
2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Najlla Nassery ◽  
Michael A. Horberg ◽  
Kevin B. Rubenstein ◽  
Julia M. Certa ◽  
Eric Watson ◽  
...  
Keyword(s):  
Claims Data ◽  
Back Analysis ◽  
Altered Mental Status ◽  
Diagnostic Error ◽  
Study Cohort ◽  
Electrolyte Disorders ◽  
Look Back ◽  
Disease Pair ◽  
Ed Visits ◽  
Pair Analysis

Abstract Objectives The aim of this study was to identify delays in early pre-sepsis diagnosis in emergency departments (ED) using the Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) approach. Methods SPADE methodology was employed using electronic health record and claims data from Kaiser Permanente Mid-Atlantic States (KPMAS). Study cohort included KPMAS members ≥18 years with ≥1 sepsis hospitalization 1/1/2013–12/31/2018. A look-back analysis identified treat-and-release ED visits in the month prior to sepsis hospitalizations. Top 20 diagnoses associated with these ED visits were identified; two diagnosis categories were distinguished as being linked to downstream sepsis hospitalizations. Observed-to-expected (O:E) and temporal analyses were performed to validate the symptom selection; results were contrasted to a comparison group. Demographics of patients that did and did not experience sepsis misdiagnosis were compared. Results There were 3,468 sepsis hospitalizations during the study period and 766 treat-and-release ED visits in the month prior to hospitalization. Patients discharged from the ED with fluid and electrolyte disorders (FED) and altered mental status (AMS) were most likely to have downstream sepsis hospitalizations (O:E ratios of 2.66 and 2.82, respectively). Temporal analyses revealed that these symptoms were overrepresented and temporally clustered close to the hospitalization date. Approximately 2% of sepsis hospitalizations were associated with prior FED or AMS ED visits. Conclusions Treat-and-release ED encounters for FED and AMS may represent harbingers for downstream sepsis hospitalizations. The SPADE approach can be used to develop performance measures that identify pre-sepsis.

Rate of sepsis hospitalizations after misdiagnosis in adult emergency department patients: a look-forward analysis with administrative claims data using Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) methodology in an integrated health system

Diagnosis ◽  
2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Michael A. Horberg ◽  
Najlla Nassery ◽  
Kevin B. Rubenstein ◽  
Julia M. Certa ◽  
Ejaz A. Shamim ◽  
...  
Keyword(s):  
Emergency Department ◽  
Claims Data ◽  
Diagnostic Error ◽  
Administrative Claims ◽  
Sepsis Diagnosis ◽  
Forward Analysis ◽  
Look Back ◽  
Disease Pair ◽  
Ed Visits ◽  
Pair Analysis

Abstract Objectives Delays in sepsis diagnosis can increase morbidity and mortality. Previously, we performed a Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) “look-back” analysis to identify symptoms at risk for delayed sepsis diagnosis. We found treat-and-release emergency department (ED) encounters for fluid and electrolyte disorders (FED) and altered mental status (AMS) were associated with downstream sepsis hospitalizations. In this “look-forward” analysis, we measure the potential misdiagnosis-related harm rate for sepsis among patients with these symptoms. Methods Retrospective cohort study using electronic health record and claims data from Kaiser Permanente Mid-Atlantic States (2013–2018). Patients ≥18 years with ≥1 treat-and-release ED encounter for FED or AMS were included. Observed greater than expected sepsis hospitalizations within 30 days of ED treat-and-release encounters were considered potential misdiagnosis-related harms. Temporal analyses were employed to differentiate case and comparison (superficial injury/contusion ED encounters) cohorts. Results There were 4,549 treat-and-release ED encounters for FED or AMS, 26 associated with a sepsis hospitalization in the next 30 days. The observed (0.57%) minus expected (0.13%) harm rate was 0.44% (absolute) and 4.5-fold increased over expected (relative). There was a spike in sepsis hospitalizations in the week following FED/AMS ED visits. There were fewer sepsis hospitalizations and no spike in admissions in the week following superficial injury/contusion ED visits. Potentially misdiagnosed patients were older and more medically complex. Conclusions Potential misdiagnosis-related harms from sepsis are infrequent but measurable using SPADE. This look-forward analysis validated our previous look-back study, demonstrating the SPADE approach can be used to study infectious disease syndromes.

scholarly journals Symptom-Disease Pair Analysis of Diagnostic Error (SPADE): a conceptual framework and methodological approach for unearthing misdiagnosis-related harms using big data

2018 ◽  
Vol 27 (7) ◽  
pp. 557-566 ◽  
Author(s):  
Ava L Liberman ◽  
David E Newman-Toker
Keyword(s):  
Public Health ◽  
Big Data ◽  
Conceptual Framework ◽  
Diagnostic Performance ◽  
Diagnostic Errors ◽  
Diagnostic Error ◽  
Administrative Claims ◽  
Similar Data ◽  
Disease Pair ◽  
Pair Analysis

BackgroundThe public health burden associated with diagnostic errors is likely enormous, with some estimates suggesting millions of individuals are harmed each year in the USA, and presumably many more worldwide. According to the US National Academy of Medicine, improving diagnosis in healthcare is now considered ‘a moral, professional, and public health imperative.’ Unfortunately, well-established, valid and readily available operational measures of diagnostic performance and misdiagnosis-related harms are lacking, hampering progress. Existing methods often rely on judging errors through labour-intensive human reviews of medical records that are constrained by poor clinical documentation, low reliability and hindsight bias.MethodsKey gaps in operational measurement might be filled via thoughtful statistical analysis of existing large clinical, billing, administrative claims or similar data sets. In this manuscript, we describe a method to quantify and monitor diagnostic errors using an approach we call ‘Symptom-Disease Pair Analysis of Diagnostic Error’ (SPADE).ResultsWe first offer a conceptual framework for establishing valid symptom-disease pairs illustrated using the well-known diagnostic error dyad of dizziness-stroke. We then describe analytical methods for both look-back (case–control) and look-forward (cohort) measures of diagnostic error and misdiagnosis-related harms using ‘big data’. After discussing the strengths and limitations of the SPADE approach by comparing it to other strategies for detecting diagnostic errors, we identify the sources of validity and reliability that undergird our approach.ConclusionSPADE-derived metrics could eventually be used for operational diagnostic performance dashboards and national benchmarking. This approach has the potential to transform diagnostic quality and safety across a broad range of clinical problems and settings.

scholarly journals JADE: A tool for medical researchers to explore adverse drug events using health claims data

2014 ◽  
Vol 05 (03) ◽  
pp. 621-629 ◽  
Author(s):  
S.K. Sauter ◽  
C. Rinner ◽  
L.M. Neuhofer ◽  
M. Wolzt ◽  
W. Grossmann ◽  
...  
Keyword(s):  
Adverse Drug Events ◽  
Claims Data ◽  
Age Groups ◽  
Rule Learning ◽  
Health Claims ◽  
Drug Event ◽  
Study Cohort ◽  
Defined Daily Doses ◽  
Health Claims Data ◽  
Medical Researchers

SummaryObjective: The objective of our project was to create a tool for physicians to explore health claims data with regard to adverse drug reactions. The Java Adverse Drug Event (JADE) tool should enable the analysis of prescribed drugs in connection with diagnoses from hospital stays.Methods: We calculated the number of days drugs were taken by using the defined daily doses and estimated possible interactions between dispensed drugs using the Austria Codex, a database including drug-drug interactions. The JADE tool was implemented using Java, R and a PostgreSQL database.Results: Beside an overview of the study cohort which includes selection of gender and age groups, selected statistical methods like association rule learning, logistic regression model and the number needed to harm have been implemented.Conclusion: The JADE tool can support physicians during their planning of clinical trials by showing the occurrences of adverse drug events with population based information.Citation: Edlinger D, Sauter SK, Rinner C, Neuhofer LM, Wolzt M, Grossmann W, Endel G, Gall W. JADE: A tool for medical researchers to explore adverse drug events using health claims data. Appl Clin Inf 2014; 5: 621–629http://dx.doi.org/10.4338/ACI-2014-04-RA-0036

Missed acute myocardial infarction in the emergency department–standardizing measurement of misdiagnosis-related harms using the SPADE method

Diagnosis ◽  
2020 ◽  
Vol 0 (0) ◽  
Author(s):  
Adam L. Sharp ◽  
Aileen Baecker ◽  
Najlla Nassery ◽  
Stacy Park ◽  
Ahmed Hassoon ◽  
...  
Keyword(s):  
Emergency Department ◽  
Adverse Event ◽  
Chest Pain ◽  
Administrative Data ◽  
Diagnostic Performance ◽  
Event Rate ◽  
Public Health Problem ◽  
Diagnostic Error ◽  
Adverse Event Rate ◽  
Disease Pair

AbstractObjectivesDiagnostic error is a serious public health problem. Measuring diagnostic performance remains elusive. We sought to measure misdiagnosis-related harms following missed acute myocardial infarctions (AMI) in the emergency department (ED) using the symptom-disease pair analysis of diagnostic error (SPADE) method.MethodsRetrospective administrative data analysis (2009–2017) from a single, integrated health system using International Classification of Diseases (ICD) coded discharge diagnoses. We looked back 30 days from AMI hospitalizations for antecedent ED treat-and-release visits to identify symptoms linked to probable missed AMI (observed > expected). We then looked forward from these ED discharge diagnoses to identify symptom-disease pair misdiagnosis-related harms (AMI hospitalizations within 30-days, representing diagnostic adverse events).ResultsA total of 44,473 AMI hospitalizations were associated with 2,874 treat-and-release ED visits in the prior 30 days. The top plausibly-related ED discharge diagnoses were “chest pain” and “dyspnea” with excess treat-and-release visit rates of 9.8% (95% CI 8.5–11.2%) and 3.4% (95% CI 2.7–4.2%), respectively. These represented 574 probable missed AMIs resulting in hospitalization (adverse event rate per AMI 1.3%, 95% CI 1.2–1.4%). Looking forward, 325,088 chest pain or dyspnea ED discharges were followed by 508 AMI hospitalizations (adverse event rate per symptom discharge 0.2%, 95% CI 0.1–0.2%).ConclusionsThe SPADE method precisely quantifies misdiagnosis-related harms from missed AMIs using administrative data. This approach could facilitate future assessment of diagnostic performance across health systems. These results correspond to ∼10,000 potentially-preventable harms annually in the US. However, relatively low error and adverse event rates may pose challenges to reducing harms for this ED symptom-disease pair.

Viral Encephalitis In Adults: A Narrative Review

2021 ◽  
Vol 16 ◽  
Author(s):  
Siciliano Valentina ◽  
Rosà Tommaso ◽  
Del Vecchio Pierluigi ◽  
D'Angelillo Anna ◽  
Brigida Mattia ◽  
...  
Keyword(s):  
Viral Encephalitis ◽  
Viral Infections ◽  
Direct Action ◽  
Ventilatory Support ◽  
Signs And Symptoms ◽  
Altered Mental Status ◽  
Poor Quality ◽  
Electrolyte Disorders ◽  
The Central Nervous System

: Viral infections of the central nervous system cause frequent hospitalization. The pathogenesis of viral encephalitis involves both the direct action of invading pathogens and the damage generated by the inflammatory reaction they trigger. The type of signs and symptoms presented by the patient depends on the severity and location of the ongoing inflammatory process. Most of the viral encephalitides are characterized by an acute development, fever, variable alterations in consciousness (confusion, lethargy, even coma), seizures (focal and generalized) and focal neurologic signs. The specific diagnosis of encephalitis is usually based on lumbar puncture. Cerebrospinal fluid examination should be performed in all patients unless absolutely contraindicated. Also, electroencephalogram and neuroimaging play a prominent role in diagnosis. Airway protection, ventilatory support, the management of raised intracranial pressure and correction of electrolyte disorders must be immediately considered in a patient with altered mental status. The only therapy strictly recommended is acyclovir in HSV encephalitis. The use of adjunctive glucocorticoids has poor-quality evidence in HSV, EBV, or VZV encephalitis. The role of antiviral therapy in other types of viral encephalitis is not well defined.

scholarly journals P085: Remember that patient you saw last week – Characteristics of patients experiencing unanticipated death following emergency department discharge

CJEM ◽  
2020 ◽  
Vol 22 (S1) ◽  
pp. S95-S95
Author(s):  
R. Hoang ◽  
K. Sampsel ◽  
A. Willmore ◽  
K. Yelle-Labre ◽  
V. Thiruganasambandamoorthy ◽  
...  
Keyword(s):  
Emergency Department ◽  
High Risk ◽  
Medical Errors ◽  
Medical Error ◽  
Tertiary Care ◽  
Vital Signs ◽  
Diagnostic Error ◽  
Sufficient Information ◽  
Risk Area ◽  
Ed Visits

Background: The emergency department (ED) is an at-risk area for medical error. We measured the frequency and characteristics of patients with unanticipated death within 7 days of ED discharge and whether medical error contributed. Aim Statement: This study aimed to calculate the frequency of patients experiencing death within 7 days after ED discharge and determine whether these deaths were related to their index ED visit, were unanticipated, and whether possible medical error occurred. Measures & Design: We performed a single-centre health records review of 200 consecutive cases from an eligible 458,634 ED visits from 2014-2017 in two urban, academic, tertiary care EDs. We included patients evaluated by an emergency physician who were discharged and died within 7 days. Three trained and blinded reviewers determined if deaths were related to the index visit, anticipated or unanticipated, or due to potential medical error. Reviewers performed content analysis to identify themes. Evaluation/Results: Of the 200 cases, 129 had sufficient information for analysis, translating to 44 deaths per 100,000 ED discharges. We found 13 cases per 100,000 ED discharges were related and unanticipated deaths and 18 of these were due to potential medical errors. Over half (52.7%) of 129 patients displayed abnormal vital signs at discharge. Patients experienced pneumonia (27.1%) as their most common cause of death. Patient characteristic themes were: difficult historian, multiple complaints, multiple comorbidities, acute progression of chronic disease, recurrent falls. Provider themes were: failure to consider infectious etiology, failure to admit high-risk elderly patient, missed diagnosis. System themes included multiple ED visits or recent admission, no repeat vital signs recorded. Discussion/Impact: Though the frequency of related and unanticipated deaths and those due to medical error was low, these results highlight opportunities to potentially enhance ED discharge decisions. These data add to the growing body of ED diagnostic error literature and emphasize the importance of identifying potentially high risk patients as well as being cognizant of the common medical errors leading to patient harm.

scholarly journals Mental Health Costs of Inflammatory Bowel Diseases

2020 ◽  
Vol 27 (1) ◽  
pp. 40-48 ◽  
Author(s):  
Eva Szigethy ◽  
Sean M Murphy ◽  
Orna G Ehrlich ◽  
Nicole M Engel-Nitz ◽  
Caren A Heller ◽  
...  
Keyword(s):  
Mental Health ◽  
Inflammatory Bowel Disease ◽  
Bowel Disease ◽  
Claims Data ◽  
The United States ◽  
Cost Of Care ◽  
Medical Utilization ◽  
Administrative Claims ◽  
Inflammatory Bowel ◽  
Ed Visits

Abstract Background Mental health diagnoses (MHDs) were identified as significant drivers of inflammatory bowel disease (IBD)-related costs in an analysis titled “Cost of Care Initiative” supported by the Crohn’s & Colitis Foundation. In this subanalysis, we sought to characterize and compare IBD patients with and without MHDs based on insurance claims data in terms of demographic traits, medical utilization, and annualized costs of care. Methods We analyzed the Optum Research Database of administrative claims from years 2007 to 2016 representing commercially insured and Medicare Advantage insured IBD patients in the United States. Inflammatory bowel disease patients with and without an MHD were compared in terms of demographics (age, gender, race), insurance type, IBD-related medical utilization (ambulatory visits, emergency department [ED] visits, and inpatient hospitalizations), and total IBD-related costs. Only patients with costs >$0 in each of the utilization categories were included in the cost estimates. Results Of the total IBD study cohort of 52,782 patients representing 179,314 person-years of data, 22,483 (42.6%) patients had at least 1 MHD coded in their claims data with a total of 46,510 person-years in which a patient had a coded MHD. The most commonly coded diagnostic categories were depressive disorders, anxiety disorders, adjustment disorders, substance use disorders, and bipolar and related disorders. Compared with patients without an MHD, a significantly greater percentage of IBD patients with MHDs were female (61.59% vs 48.63%), older than 75 years of age (9.59% vs 6.32%), white (73.80% vs 70.17%), and significantly less likely to be younger than 25 years of age (9.18% vs 11.39%) compared with those without mental illness (P < 0.001). Patients with MHDs had significantly more ED visits (14.34% vs 7.62%, P < 0.001) and inpatient stays (19.65% vs 8.63%, P < 0.001) compared with those without an MHD. Concomitantly, patients with MHDs had significantly higher ED costs ($970 vs $754, P < 0.001) and inpatient costs ($39,205 vs $29,550, P < 0.001) compared with IBD patients without MHDs. Patients with MHDs also had significantly higher total annual IBD-related surgical costs ($55,693 vs $40,486, P < 0.001) and nonsurgical costs (medical and pharmacy) ($17,220 vs $11,073, P < 0.001), and paid a larger portion of the total out-of-pocket cost for IBD services ($1017 vs $905, P < 0.001). Conclusion Patients whose claims data contained both IBD-related and MHD-related diagnoses generated significantly higher costs compared with IBD patients without an MHD diagnosis. Based on these data, we speculate that health care costs might be reduced and the course of patients IBD might be improved if the IBD-treating provider recognized this link and implemented effective behavioral health screening and intervention as soon as an MHD was suspected during management of IBD patients. Studies investigating best screening and intervention strategies for MHDs are needed.

Retrospective claims analysis of palliative care costs and survival in a U.S. metastatic renal cell (mRCC) population.

2012 ◽  
Vol 30 (5_suppl) ◽  
pp. 433-433
Author(s):  
Henry J. Henk ◽  
Connie Chen ◽  
Agnes Benedict ◽  
Jane Sullivan ◽  
April Teitelbaum
Keyword(s):  
Health Care ◽  
Health Care Costs ◽  
Health Plan ◽  
Index Date ◽  
Claims Data ◽  
Cohort Analysis ◽  
Hospital Costs ◽  
Best Supportive Care ◽  
Study Cohort ◽  
Care Costs

433 Background: Survival and costs outcomes for patients with mRCC receiving palliative or best supportive care (BSC) after stopping active therapy have been poorly characterized. This information is important to understand how resources are utilized at the end of life and to put current treatment costs into perspective. The objective of this retrospective database analysis was to examine survival and costs associated with BSC after receiving 1 or 2 lines of mRCC treatment. Methods: A retrospective cohort analysis using claims data from commercially insured or Medicare Advantage (MCR) enrollees of a large US health plan, with medical and pharmacy benefits. The study cohort consisted of patients with an index diagnosis for RCC [ICD-9-CM 189.0] from 1/1/07 to 6/30/10 initiating any of the following treatments from 30 days prior to index date through disenrollment: sunitinib, temsirolimus, sorafenib, bevacizumab, everolimus, pazopanib, cytokines. Patients were required to have a 6 mos. continuous enrollment ± index date (patients disenrolling due to death within the 6 mos. were retained). Lines of therapy (LOT) were identified based on prescription fill and administration dates, began following the last LOT and continued until disenrollment. Health care costs reported represent the health plan + patient paid amount. Results: The overall study cohort (n=274) was 73% male; mean (±SD) age 63.3 ± 11.1 yr. with the majority of patients commercially insured (80% vs 20% MCR). The majority started BSC following 1st LOT (68% vs 32%). Median survival from start of BSC was similar following 1st and 2nd LOT (126 and 118 days). The mean (median) duration of BSC after 1 LOT was 223 (114) days and 176 (109) days for 2 LOT. Total health care costs incurred during BSC averaged $50,187 ± 96,984 and $37,294 ± 51,101 and monthly costs were similar ($10,284 ± 17979) after 1 and 2 LOT, respectively. In both cases, inpatient hospital costs represented the largest proportion of these costs (47%) while outpatient costs represented 36%. Conclusions: Our study estimating BSC survival and costs in patients with mRCC based on US claims data found monthly cost of $10, 284. These estimates suggest that BSC costs are not insignificant.

Targeted therapy at the end of life and the impact on health care utilization.

2014 ◽  
Vol 32 (31_suppl) ◽  
pp. 116-116
Author(s):  
Heidi Yeung ◽  
Parsa Salehi ◽  
James Don Murphy
Keyword(s):  
Targeted Therapy ◽  
End Of Life ◽  
Odds Ratio ◽  
Adjusted Odds Ratio ◽  
Large Population ◽  
Metastatic Breast ◽  
Care Utilization ◽  
Study Cohort ◽  
Ed Visits ◽  
The Impact

116 Background: The use of targeted therapy has steadily increased over the past decade, though the impact of targeted agents on patterns of care at the end-of-life life remains unknown. The purpose of this study was to explore the influence of targeted therapy on end-of-life care in a large population-based database. Methods: We identified 14,398 patients from the SEER-Medicare linked database with metastatic breast, lung, or colorectal cancer diagnosed between 2000 and 2009 who received conventional cytotoxic chemotherapy or targeted therapy in the last 3 months of life. Multivariate logistic and linear regression models were used to determine the impact of targeted therapy on the following endpoints in the last 3 months of life: emergency department (ED) visits, hospitalizations, and hospice utilization. Analyses were adjusted for differences in patient age, gender, race, comorbidity, socioeconomic status, and geography. Results: Among the whole study cohort 83% of patients received chemotherapy alone, 12% received chemotherapy with targeted therapy, and 5% received targeted therapy alone. The delivery of any targeted therapy in the last 3 months increased across the study period, from 1.5% in 2000 to 28% in 2009. Compared to patients treated with chemotherapy alone, those treated with targeted therapy alone had lower rates of ED visits (adjusted odds ratio [aOR]=0.81, p=0.01), lower rates of hospitalization (adjusted odds ratio [aOR]=0.69, p<0.0001), no difference in overall hospice utilization rates, though had longer stays on hospice (5.1 days longer, p<0.0001). Compared to patients treated with chemotherapy alone, those treated with both chemotherapy and targeted therapy had no difference in ED visits or hospitalization admission rates, though had decreased rates of hospice utilization (aOR = 0.79, p<0.0001), and shorter stays on hospice (2.7 days longer, p<0.0001). Conclusions: This study found that targeted therapy was associated with varying patterns of healthcare utilization at the end-of-life, though these differences could be influenced by unknown confounding variables. Future research should focus on defining the specific impact of targeted therapy on quality of life at the end-of-life.

Using deep learning with registry linked claims data to predict hospitalization during chemotherapy: Feasibility study.

2019 ◽  
Vol 37 (27_suppl) ◽  
pp. 270-270
Author(s):  
Laura Elizabeth Panattoni ◽  
Li Li ◽  
Catherine R. Fedorenko ◽  
Emily Silgard ◽  
Scott White ◽  
...  
Keyword(s):  
Deep Learning ◽  
Claims Data ◽  
Predictive Accuracy ◽  
Machine Learning Algorithms ◽  
Receiver Operating Curve ◽  
Future Research ◽  
Primary Tumor Site ◽  
Model Complex ◽  
Ed Visits ◽  
Chemotherapy Initiation

270 Background: Approximately half of cancer patients undergoing outpatient chemotherapy experience unplanned Emergency Department (ED) visits and Inpatient (IP) stays. Current machine learning algorithms that identify high-risk patients are based on pre-treatment variables which can not detect changes in risk over time. Deep learning recurrent neural networks can model complex longitudinal patient histories. This study tests the feasibility of using an interpretable recurrent neural network to predict a patient’s daily likelihood of ED and unplanned IP stays in the six months following chemotherapy initiation. Methods: Medicare and commercial claims data were linked with cancer registry records for patients in Washington State from 2011 to 2017. The study included patients diagnosed with any primary tumor site, excluding leukemia, and treated with chemotherapy. We used the Reverse Time Attention model (RETAIN) with a 1:10 case-control match and included registry elements; diagnoses, procedures, medication, and utilization pre-and post-chemotherapy initiation. Patients were randomly divided into internal training, validation, and test sets (75%, 10%, 15%). Model accuracy was measured by the areas under the receiver operating curve (ROC) and precision-recall curve (PRC), and the Youden sensitivity and specificity. Results: Of the 15,400 eligible patients; 4,037 (26.2%) visited the ED a median of 1 time (6,080 total visits); 5,116 (33.2%) had a median of 1 IP stay (7,839 total stays). Both models had good predictive accuracy: The top 20 predictors for ED visits included 5 chemotherapy regimes, 12 procedures, and 2 tumor characteristics; IP stays included all chemotherapy regimes. Conclusions: The promising performance of RETAIN supports the possibility of building a tool capable of estimating daily hospitalization risk. However, future research, particularly with alternative data sources, may be required to predict hospitalization in a real time clinical setting. [Table: see text]

Sign in / Sign up

Export Citation Format

Share Document