scholarly journals Building a Clinically Relevant Risk Model: Predicting Risk of a Potentially Preventable Acute Care Visit for Patients Starting Antineoplastic Treatment

2020 ◽  
pp. 275-289 ◽  
Author(s):  
Bobby Daly ◽  
Dmitriy Gorenshteyn ◽  
Kevin J. Nicholas ◽  
Alice Zervoudakis ◽  
Stefania Sokolowski ◽  
...  
Keyword(s):  
High Risk ◽  
Acute Care ◽  
Web Application ◽  
Risk Model ◽  
Care Delivery ◽  
Statistical Significance ◽  
Patient Specific ◽  
Cancer Center ◽  
Management Interventions ◽  
Care Visit

PURPOSE To create a risk prediction model that identifies patients at high risk for a potentially preventable acute care visit (PPACV). PATIENTS AND METHODS We developed a risk model that used electronic medical record data from initial visit to first antineoplastic administration for new patients at Memorial Sloan Kettering Cancer Center from January 2014 to September 2018. The final time-weighted least absolute shrinkage and selection operator model was chosen on the basis of clinical and statistical significance. The model was refined to predict risk on the basis of 270 clinically relevant data features spanning sociodemographics, malignancy and treatment characteristics, laboratory results, medical and social history, medications, and prior acute care encounters. The binary dependent variable was occurrence of a PPACV within the first 6 months of treatment. There were 8,067 observations for new-start antineoplastic therapy in our training set, 1,211 in the validation set, and 1,294 in the testing set. RESULTS A total of 3,727 patients experienced a PPACV within 6 months of treatment start. Specific features that determined risk were surfaced in a web application, riskExplorer, to enable clinician review of patient-specific risk. The positive predictive value of a PPACV among patients in the top quartile of model risk was 42%. This quartile accounted for 35% of patients with PPACVs and 51% of potentially preventable inpatient bed days. The model C-statistic was 0.65. CONCLUSION Our clinically relevant model identified the patients responsible for 35% of PPACVs and more than half of the inpatient beds used by the cohort. Additional research is needed to determine whether targeting these high-risk patients with symptom management interventions could improve care delivery by reducing PPACVs.

Perspectives on cancer care delivery from oncology patients who experienced unplanned acute care events.

2019 ◽  
Vol 37 (27_suppl) ◽  
pp. 201-201
Author(s):  
Tara L. Kaufmann ◽  
Katharine A. Rendle ◽  
Erin Aakhus ◽  
Andrea Bilger ◽  
Peter Edward Gabriel ◽  
...  
Keyword(s):  
Cancer Treatment ◽  
Acute Care ◽  
Care Delivery ◽  
System Level ◽  
Self Management ◽  
Structured Interviews ◽  
Patient Factors ◽  
Management Interventions ◽  
High Level ◽  
Oncology Providers

201 Background: Unplanned acute care is debilitating and burdensome for patients with advanced cancer and their caregivers. There is a pressing need to understand how available evidence-based practices (EBPs) to reduce acute care 1) align with the needs and priorities of patients and 2) are best implemented within large health systems. We are conducting a mixed methods study to assess patient- provider- and system-level factors that shape the decision to seek acute care during active cancer treatment in order to select and adapt EBPs for implementation. Here we present data from patients’ perspectives. Methods: Purposive sampling approach to identify solid tumor cancer patients on active treatment with unplanned acute care events at a large health system from Aug 2018-Jan 2019. We conducted semi-structured interviews to elicit patients’ perspectives on factors that shape their decision to seek acute care and to inform intervention strategies. Results: Forty-nine patients participated in this study. We identify several patient factors that intersect with the decision to seek care: self-management behaviors, guilt, negative ED perception, safety concerns, and trust. Patients attempt self-management prior to contacting their oncology team, which introduces variability in the duration and severity of reported symptoms. Delay is related to patients’ guilt for burdening their oncology team and to provider accessibility. Patients describe a high symptom threshold to seek care that is often coupled with a negative perception of the ED, but do prefer in-person evaluation for new and distressing symptoms for safety. They express a high level of trust in the oncology team and relative distrust of non-oncology providers. Conclusions: Our data suggest a conceptual model for patient factors that drive unplanned acute care. Patients identify three areas for improvement: 1) enhanced peer support and education to manage uncertainty about cancer treatment; 2) accessible portals for patient-clinician communication; and 3) home or clinic-based after-hours oncology symptom management. Interventions to target these needs should address patients’ emotional concerns and be well integrated within the oncology team.

A retrospective real-world major bleeding (MB) comparison of direct oral anticoagulants (DOAC) and low molecular weight heparin (LMWH) in genitourinary cancer-associated venous thromboembolism (GU-CAVTE) with reported randomized clinical trials (RTC).

2021 ◽  
Vol 39 (6_suppl) ◽  
pp. 410-410
Author(s):  
Ulyana Dashkevych ◽  
Eric Brucks ◽  
Kathylynn Saboda ◽  
Juan Chipollini ◽  
Hani M. Babiker ◽  
...  
Keyword(s):  
High Risk ◽  
Real World ◽  
Randomized Clinical Trials ◽  
Meta Analysis ◽  
Statistical Significance ◽  
Oral Anticoagulants ◽  
Direct Oral Anticoagulants ◽  
Retrospective Chart Review ◽  
Cancer Center ◽  
Recurrent Vte

410 Background: MB is a serious complication in patients with CAVTE receiving treatment with DOAC or LMWH. The most recent meta-analysis of the four major RCT showed that MB events rate were similar among the DOAC and LMWH group, however, it was noted that MB occurred at GU site 4.9 times more in DOAC than LMWH patients. While GUCA (e.g. bladder and testicular) are considered to be high-risk based on the Khorana Score, they were underrepresented among the RCT ( < 12%). We present a Real-World retrospective cohort study analyzing the MB rates in patients presenting with GU-CAVTE treated either by a DOAC or LMWH compared to those of the RTC. Methods: We performed a retrospective chart review of patients with a diagnosed GUCA and VTE who presented to The University of Arizona Cancer Center (UACC) and were subsequently placed on anticoagulant therapy with either a DOAC or LMWH from 11/2013-4/2020. MB outcome was defined as documented Hgb drop of ≥2 g/dL, ≥2 units of PRBC, MB in a critical site, or contributing to death. MB was extracted and compared from the SELECT D, ADAM VTE, and Caravaggio for DOAC and Hokusai for the LMWH control arm with the GUCA subgroup. Recurrent VTE was collected. In situations where there was insufficient data to categorize individuals, those individuals were excluded from the analysis. The proportion of MB reported in each study were compared using a binomial test. Results: Our review included 56 patients with similar baseline characteristics to the RCT, who were prescribed enoxaparin (n = 13), apixaban (n = 27) and rivaroxaban (n = 16). Our UACC data was compared to the RCT reported MB outcomes with rivaroxaban (12% vs 8%, [p = 0.63]), apixaban (11% vs 6%, [p = 0.40]), and LMWH (both 0 vs 1% [p = 0.67]). No statistical difference among DOAC selection [p = 0.90]. Our UACC rate of MB in patients with GUCA for both DOAC combined versus LWMH were 11.6% (5/43) and 0% [p = 0.1910], compared to the RCT GU subgroup was 5.7% (6/104) [p = 0.02] and 0.6% (1/175) [p = 1.0], respectively. Furthermore, our data found no statistical significance difference among the recurrent VTE rate among DOAC, LMWH, UACC Retrospective or RCT events. Conclusions: In agreement with the four major RCT, our study demonstrated that patients with high-risk GUCA and underlying VTE treated with a DOAC had a non-significant higher incidence of MB compared to those treated with LMWH. Further, our Real-World experience showed that GUCA DOAC had a significantly higher MB event rate compared to the RCT subgroup population. We acknowledge there are inherent biases in all retrospective studies and RCT. These data support the idea that DOAC should be further studied and used with caution in patients with a high risk of bleeding. We recommend LMWH being the safest anticoagulation modality for High-Risk Bleeding GU malignancy.

A framework for building a clinically relevant risk model.

2019 ◽  
Vol 37 (15_suppl) ◽  
pp. 6554-6554
Author(s):  
Robert Michael Daly ◽  
Dmitriy Gorenshteyn ◽  
Lior Gazit ◽  
Stefania Sokolowski ◽  
Kevin Nicholas ◽  
...  
Keyword(s):  
Positive Predictive Value ◽  
Predictive Value ◽  
Web Application ◽  
Risk Model ◽  
Statistical Significance ◽  
Quality Care ◽  
Improve Patient Care ◽  
Individual Risk ◽  
Multivariable Logistic Regression Model ◽  
Electronic Health Record Data

6554 Background: Acute care accounts for half of cancer expenditures and is a measure of poor quality care. Identifying patients at high risk for emergency department (ED) visits enables institutions to target resources to those most likely to benefit. Risk stratification models developed to date have not been meaningfully employed in oncology, and there is a need for clinically relevant models to improve patient care. Methods: We established and applied a predictive framework for clinical use with attention to modeling technique, clinician feedback, and application metrics. The model employs electronic health record data from initial visit to first antineoplastic administration for patients at our institution from January 2014 to June 2017. The binary dependent variable is occurrence of an ED visit within the first 6 months of treatment. The final regularized multivariable logistic regression model was chosen based on clinical and statistical significance. In order to accommodate for the needs to the program, parameter selection and model calibration were optimized to suit the positive predictive value of the top 25% of observations as ranked by model-determined risk. Results: There are 5,752 antineoplastic administration starts in our training set, and 1,457 in our test set. The positive predictive value of this model for the top 25% riskiest new start antineoplastic patients is 0.53. From over 1,400 data features, the model was refined to include 400 clinically relevant ones spanning demographics, pathology, clinician notes, labs, medications, and psychosocial information. At the patient level, specific features determining risk are surfaced in a web application, RiskExplorer, to enable clinician review of individual patient risk. This physician facing application provides the individual risk score for the patient as well as their quartile of risk when compared to the population of new start antineoplastic patients. For the top quartile of patients, the risk for an ED visit within the first 6 months of treatment is greater than or equal to 49%. Conclusions: We have constructed a framework to build a clinically relevant risk model. We are now piloting it to identify those likely to benefit from a home-based, digital symptom management intervention.

A personalized prediction model for hospital readmission risk for cancer patients.

2020 ◽  
Vol 38 (15_suppl) ◽  
pp. 7057-7057
Author(s):  
Jacob Tyler Shreve ◽  
Sarah Lee ◽  
Christina Felix ◽  
Rachel Benish Shirley ◽  
Cameron Beau Hilton ◽  
...  
Keyword(s):  
Cancer Patients ◽  
Hospital Readmission ◽  
Web Application ◽  
Risk Model ◽  
Learning Algorithm ◽  
Hospital Readmissions ◽  
Patient Specific ◽  
Unplanned Readmission ◽  
Discharge Disposition ◽  
Readmission Risk

7057 Background: Cancer patients (pts) are at high risk of unplanned hospital readmissions. Predicting which cancer patients are at higher risk of readmission would improve post-discharge follow-up/navigation, decrease cost, and improve pt outcomes. Methods: We conducted a retrospective cohort study of non-surgical cancer pts hospitalized at our center between 12/2014 to 7/2018. A machine learning algorithm was trained on 348 medical, sociodemographic and cancer-specific variables with a total of 1,801,944 data points. The cohort was randomly divided into training (80%) and validation (20%) subsets. Prediction performance was measured by area under the receiver operator characteristic curve (AUC). Results: A total of 5,178 hospitalizations were included, of which 45.1% were women, and 27.6% experienced an unplanned readmission within 30 days. The most frequently represented cancers were hematologic malignancies (30.5%), followed by GI (18.1%), lung (13.7%), and GU (10.9%). Significant variables that impacted the algorithm decision are ranked from the most to the least important, including: days from last admission; planned index chemotherapy admission; number of vascular access lines, drains, and airways in use; length of stay; cancer diagnosis; total ED visits in past 6 months; age; discharge lab values (sodium, albumin, alkaline phosphatase, bilirubin, platelets); number of prior admissions; and discharge disposition. The AUC for the validation subset was 0.80. To ease the translation of this model into the clinic, we developed a web application whereby users can supply the aforementioned variables to the model and receive a personalized prediction that highlights those variables most affecting a subject’s readmission risk status: www.Cancer-Readmission.com. Conclusions: A cancer-specific readmission risk model with high AUC for 30-days unplanned readmission has been developed. The model is embedded in a freely available web application that provides personalized, patient-specific predictions. Programs that integrate this model can identify cancer patients with a greater risk for unplanned hospital readmission, thus providing a personalized approach to prevent future unplanned readmissions.

scholarly journals Machine-Learning Modeling to Predict Hospital Readmission Following Discharge to Post-Acute Care Settings

2020 ◽  
Vol 4 (Supplement_1) ◽  
pp. 34-35
Author(s):  
Elizabeth Howard ◽  
John N Morris ◽  
Erez Schachter
Keyword(s):  
High Risk ◽  
Acute Care ◽  
Hospital Readmission ◽  
Readmission Rate ◽  
Care Delivery ◽  
Functional Limitations ◽  
Skilled Nursing Facility ◽  
Risk Groups ◽  
Total Sample ◽  
Post Acute Care

Abstract Increased attention to post-acute care (PAC) settings and available services to meet patients’ needs following acute hospital discharge is needed as these settings are being utilized increasingly in models of care delivery. The primary purpose was to generate a model to identify the most predictive factors relevant to hospital readmission within 90 days following discharge to one of three types of PAC sites: home with home care services (HC), skilled nursing facility (SNF), in-patient rehabilitation facility (IRF). Specific aims were to (1) examine number and characteristics of older adults discharged to the 3 PAC sites; (2) compare 90 day hospital readmission rate across sites and acuity level; and (3) examine assessment items across population and subgroups to identify variables most predictive of hospital readmission. 2015 assessment data from 3,592,995 Medicare beneficiaries were analyzed representing 1,536,908 from SNFs, 306,878 from IRFs, and 1,749,209 receiving HC services. Total sample 90-day readmission was 25.8 % . Patients discharged to IRF had lowest readmission rate (23.34%), and those receiving HC services had highest readmission rate (29.34%). Creation of risk subgroups however, revealed alternative outcomes. Among all patients in the low, intermediate and high risk groups, the lowest readmission rates occurred among SNF patients. Factor analysis of assessment variables indicated bladder and bowel incontinence and functional limitations were the most distinguishing factors between the very low and very high risk subgroups.

From Theory to Practice: Implementation of Strategies to Reduce Acute Care Visits in Patients With Cancer

2020 ◽  
pp. 85-94
Author(s):  
Bobby Daly ◽  
Laura C. Michaelis ◽  
John D. Sprandio ◽  
Jonathan T. Kapke ◽  
Ravi Kishore Narra ◽  
...  
Keyword(s):  
Acute Care ◽  
Care Delivery ◽  
Management Program ◽  
Lessons Learned ◽  
Cancer Center ◽  
Future Research ◽  
Community Practice ◽  
Patients With Cancer ◽  
Department Of Health ◽  
Risk Patients

Patients with cancer frequently seek acute care as a result of complications of their disease and adverse effects of treatment. This acute care comes at high cost to the health care system and often results in suboptimal outcomes for patients and their caregivers. The Department of Health and Human Services has identified this as a gap in our care of patients with cancer and has called for quality-improvement efforts to reduce this acute care. We highlight the efforts of three centers—a community practice, an academic practice, and a cancer center—to reduce acute care for their patients. We describe the foundational principles, the practice innovation and implementation strategy, the initial results, and the lessons learned from these interventions. Each of the described interventions sought to integrate evidence-based best practices for reducing unplanned acute care. The first, a telephone triage system, led to 82% of calls being managed at home and only 2% being directed to an emergency department (ED) or hospital. The second, a 24-hour continuity clinic, led to a 26% reduction in ED utilization for patients with cancer. The third, a digital symptom monitoring and management program for high-risk patients on active treatment, led to a 17% reduction in ED presentations. There is a need for innovative care delivery models to improve the management of symptoms for patients with cancer. Future research is needed to determine the elements of these models with the greatest impact and how successful models can be scaled to other institutions.

Advance electronic screening of patients with cancer leveraging a new EHR to increase efficiency of care delivery, driving clinical care pathways in a perioperative anesthesia assessment center (PAAC), at a comprehensive cancer center.

2018 ◽  
Vol 36 (30_suppl) ◽  
pp. 308-308
Author(s):  
Katy E. French ◽  
Shannon Popovich ◽  
B. Bryce Speer ◽  
Iris Recinos ◽  
Tayab Andrabi
Keyword(s):  
Cancer Patients ◽  
Patient Experience ◽  
Care Delivery ◽  
Day Surgery ◽  
Patient Specific ◽  
Cancer Center ◽  
Care Pathways ◽  
Electronic Screening ◽  
Cancellation Rate ◽  
Patient Care Delivery

308 Background: Most cancer institutions are using electronic health records (EHRs). With MACRA and MIPS, this number will continue to grow. EHR’s can be frustrating as some are seen as more cumbersome than useful to patient care delivery. In 2017, the MD Anderson cancer center (MDACC) PAAC assessed over 20,000 patients. We wanted to demonstrate positive utility of our EHR. We hypothesized that advanced electronic screening of our cancer patients, leveraging a questionnaire embedded in the EHR, could increase efficiency by driving patient specific perioperative care pathways leading to positive patient experience while not negatively affecting our low same day surgery cancellation rate of 2.07%. Methods: In March of 2016, MDACC launched a new EHR and the PAAC commenced an online medical and anesthesia screening questionnaire to be completed via a secure online portal by our cancer patients in advance of the surgery date. The questionnaire was developed and vetted by specialists from many areas including anesthesia, cardiology, internal medicine, and surgery. The answers to these questions were validated by a healthcare provider, and the patient was directed to one of two care pathways. Complex patients were scheduled for an in person clinic assessment prior to surgery and all other patients were scheduled for a phone call assessment. We used reports created within our EHR to track baseline and 2017 data for numbers of patients directed to the 2 different pathways and same day surgery cancellation data throughout this time frame. Results: Using the online patient entered questionnaire, we were able to improve clinic efficiency by implementing two patient care delivery pathways. Using our new EHR we saw an increase of 18.4% in our phone assessment pathway for patients. Patient experience improved as we were able to offer patients more options in the delivery of care saving them time. Conclusions: Use of our new EHR and IT analytics helped drive patient-specific care pathways, improve efficiency and the patient experience. MDACC same day surgery cancellation rate remained 2.07%.

scholarly journals New risk model is able to identify patients with a low risk of progression in systemic sclerosis

RMD Open ◽  
2021 ◽  
Vol 7 (2) ◽  
pp. e001524
Author(s):  
Nina Marijn van Leeuwen ◽  
Marc Maurits ◽  
Sophie Liem ◽  
Jacopo Ciaffi ◽  
Nina Ajmone Marsan ◽  
...  
Keyword(s):  
Machine Learning ◽  
Systemic Sclerosis ◽  
High Risk ◽  
Disease Progression ◽  
Risk Model ◽  
Immunosuppressive Drugs ◽  
Low Risk ◽  
Expert Assessment ◽  
Risk Of Progression

ObjectivesTo develop a prediction model to guide annual assessment of systemic sclerosis (SSc) patients tailored in accordance to disease activity.MethodsA machine learning approach was used to develop a model that can identify patients without disease progression. SSc patients included in the prospective Leiden SSc cohort and fulfilling the ACR/EULAR 2013 criteria were included. Disease progression was defined as progression in ≥1 organ system, and/or start of immunosuppression or death. Using elastic-net-regularisation, and including 90 independent clinical variables (100% complete), we trained the model on 75% and validated it on 25% of the patients, optimising on negative predictive value (NPV) to minimise the likelihood of missing progression. Probability cutoffs were identified for low and high risk for disease progression by expert assessment.ResultsOf the 492 SSc patients (follow-up range: 2–10 years), disease progression during follow-up was observed in 52% (median time 4.9 years). Performance of the model in the test set showed an AUC-ROC of 0.66. Probability score cutoffs were defined: low risk for disease progression (<0.197, NPV:1.0; 29% of patients), intermediate risk (0.197–0.223, NPV:0.82; 27%) and high risk (>0.223, NPV:0.78; 44%). The relevant variables for the model were: previous use of cyclophosphamide or corticosteroids, start with immunosuppressive drugs, previous gastrointestinal progression, previous cardiovascular event, pulmonary arterial hypertension, modified Rodnan Skin Score, creatine kinase and diffusing capacity for carbon monoxide.ConclusionOur machine-learning-assisted model for progression enabled us to classify 29% of SSc patients as ‘low risk’. In this group, annual assessment programmes could be less extensive than indicated by international guidelines.

scholarly journals Pathogens Associated With Repeat Versus Single Central-Line–Associated Bloodstream Infections, Acute-Care Hospitals, NHSN

2020 ◽  
Vol 41 (S1) ◽  
pp. s343-s344
Author(s):  
Margaret A. Dudeck ◽  
Katherine Allen-Bridson ◽  
Jonathan R. Edwards
Keyword(s):  
Acute Care ◽  
National Level ◽  
Bloodstream Infections ◽  
Acute Care Hospitals ◽  
Central Line ◽  
Patient Specific ◽  
Healthcare Associated Infections ◽  
Healthcare Associated ◽  
Repeat Infection ◽  
Single Versus

Background: The NHSN is the nation’s largest surveillance system for healthcare-associated infections. Since 2011, acute-care hospitals (ACHs) have been required to report intensive care unit (ICU) central-line–associated bloodstream infections (CLABSIs) to the NHSN pursuant to CMS requirements. In 2015, this requirement included general medical, surgical, and medical-surgical wards. Also in 2015, the NHSN implemented a repeat infection timeframe (RIT) that required repeat CLABSIs, in the same patient and admission, to be excluded if onset was within 14 days. This analysis is the first at the national level to describe repeat CLABSIs. Methods: Index CLABSIs reported in ACH ICUs and select wards during 2015–2108 were included, in addition to repeat CLABSIs occurring at any location during the same period. CLABSIs were stratified into 2 groups: single and repeat CLABSIs. The repeat CLABSI group included the index CLABSI and subsequent CLABSI(s) reported for the same patient. Up to 5 CLABSIs were included for a single patient. Pathogen analyses were limited to the first pathogen reported for each CLABSI, which is considered to be the most important cause of the event. Likelihood ratio χ2 tests were used to determine differences in proportions. Results: Of the 70,214 CLABSIs reported, 5,983 (8.5%) were repeat CLABSIs. Of 3,264 nonindex CLABSIs, 425 (13%) were identified in non-ICU or non-select ward locations. Staphylococcus aureus was the most common pathogen in both the single and repeat CLABSI groups (14.2% and 12%, respectively) (Fig. 1). Compared to all other pathogens, CLABSIs reported with Candida spp were less likely in a repeat CLABSI event than in a single CLABSI event (P < .0001). Insertion-related organisms were more likely to be associated with single CLABSIs than repeat CLABSIs (P < .0001) (Fig. 2). Alternatively, Enterococcus spp or Klebsiella pneumoniae and K. oxytoca were more likely to be associated with repeat CLABSIs than single CLABSIs (P < .0001). Conclusions: This analysis highlights differences in the aggregate pathogen distributions comparing single versus repeat CLABSIs. Assessing the pathogens associated with repeat CLABSIs may offer another way to assess the success of CLABSI prevention efforts (eg, clean insertion practices). Pathogens such as Enterococcus spp and Klebsiella spp demonstrate a greater association with repeat CLABSIs. Thus, instituting prevention efforts focused on these organisms may warrant greater attention and could impact the likelihood of repeat CLABSIs. Additional analysis of patient-specific pathogens identified in the repeat CLABSI group may yield further clarification.Funding: NoneDisclosures: None

Sign in / Sign up

Export Citation Format

Share Document