LACE Index Research Articles

Predicting Outcomes after Hospitalisation for COPD Exacerbation Using Machine Learning

BackgroundEarly readmission and death are critical adverse outcomes following hospitalisation due to exacerbation of chronic obstructive pulmonary disease (ECOPD). This study aimed to develop and validate machine learning models to enhance the prediction of these outcomes after ECOPD hospitalisation.MethodsUtilising a nationwide database, data from the index ECOPD hospitalisation and the preceding year were collected. Prediction models for 30-day readmission and death were developed using logistic lasso regression, random forest, extreme gradient boosting (XGBoost), and neural network, with the LACE index serving as a reference. Model performance was assessed with receiver operating characteristic (ROC) curves and calibration plots from the validation dataset. Key predictors were identified using SHapley Additive exPlanations.ResultsThe study included 101 011 hospitalisations in the development dataset and 17 565 in the validation dataset. The rates of 30-day readmission and death were 29.1% and 4.3%, respectively. AmoXGBoost outperformed other models, achieving an area under the ROC curve of 0.721 (95% CI, 0.713–0.729) for readmission and 0.809 (95% CI, 0.794–0.824) for death, both exceeding the corresponding values for the LACE index (0.651 and 0.641). All machine learning models demonstrated good calibration. The number of hospitalisations in the previous year and the lowest haemoglobin level during the index hospitalisation were the top predictors of readmission and death, respectively.ConclusionsApplying machine learning techniques to large-scale data effectively improves the prediction of early readmission and death following ECOPD hospitalisation. Identifying critical prognostic factors could enhance targeted post-discharge care for this high-risk patient group.

Timely data on the dynamic drug supply is crucial in addressing the ongoing drug poisoning crisis. The team is exploring ways to share information with emergency responders, healthcare providers, government officials, and pwSUD.

ObjectiveTransitioning from hospitals to primary care brings significant challenges, including increased readmission rates, mortality, and costs due to information gaps. Alberta, Canada, addressed this by integrating 1000+ health systems and implementing Connect Care (CC), a Clinical Information System (CIS), to enhance patient safety and care coordination. In 2020, the Primary Health Care Integration Network (PHCIN) released the Home to Hospital to Home (H2H2H) Transitions Guideline and metrics to improve patient outcomes and system integration. ApproachThis study examines provincial data on H2H2H transitions measures from acute care hospitals using CC between April 1, 2022, and October 31, 2023, to assess transitions and support improvement efforts. CC data assesses transition components, including confirming primary care providers at hospital discharge, utilizing the LACE Readmission Risk Index, and ensuring timely discharge summary (DS) signoffs. It links with administrative data to evaluate post-discharge outcomes like primary care physician follow-up, unplanned readmissions, and Emergency Department (ED) visits. Results47 CC-implemented sites show nearly 80% discharges identified a primary care provider; less than 5% incorporated the LACE index in DS. About 90-93% of DS were signed within 24-72 hours. Approximately 62% of moderate-risk and 55% of high-risk discharges received timely follow-up. Readmission and ED visit rates within 7-30 days varied from 2-8% and 5-11% respectively. ConclusionIn conclusion, adopting CC and H2H2H transition metrics facilitates integrated care measurement, highlighting the need for risk index inclusion and high-risk discharge follow-up for improved patient transitions. Sustained initiatives are vital for optimal outcomes and system integration in Alberta.

External validation and comparative analysis of the HOSPITAL score and LACE index for predicting readmissions among patients hospitalised with community-acquired pneumonia in Australia.

ObjectiveCommunity-acquired pneumonia (CAP) is a leading cause of emergency hospitalisations globally and is associated with high readmission rates. Specific score systems developed for all medical conditions such as the HOSPITAL score and the LACE index can also usefully predict CAP readmissions. However, there is limited evidence regarding their performance in the Australian healthcare settings.MethodsThis multicentre retrospective study analysed adult CAP discharges from two metropolitan hospitals in South Australia between 1 January 2018 and 31 December 2023. Data for determining the HOSPITAL score and the LACE index were derived from electronic medical records. Demographic characteristics of patients readmitted within 30 days were compared with those who were not readmitted. The scores were evaluated for overall performance, discriminatory power and calibration, with discriminatory power assessed using the concordance statistic (C-statistic).ResultsOver 6years, 7245 CAP discharges were recorded, with 1329 (18.3%) readmissions within 30days. The mean (s.d.) age of the cohort was 74.4 (17.8) years. Readmitted patients were more likely to have multiple morbidities and frailty than those not readmitted (P<0.05). They also had a higher mean number of emergency department presentations and hospital admissions in the previous year and a longer initial hospital stay (P<0.05). Overall, the mean (s.d.) HOSPITAL score and LACE index were 3.4 (2.1) and 9.3 (3.6), respectively. Among readmissions, 28.4% occurred in patients with a HOSPITAL score >4 (intermediate and high-risk group), while 25.8% occurred in patients in the high-risk LACE category (LACE index>10). The C-statistic for the HOSPITAL score and LACE index was 0.62 (95% CI 0.61-0.64) and 0.63 (95% CI 0.61-0.65), respectively, with no significant difference in the area under the receiver operating characteristic curves (P>0.05).ConclusionsThe predictive abilities of the HOSPITAL score and the LACE index for CAP readmissions are modest and comparable in an Australian setting.

1492P The role of LACE index in predicting 30-day unplanned readmission in patients receiving immunotherapy

1492P The role of LACE index in predicting 30-day unplanned readmission in patients receiving immunotherapy

Assessing racial bias in healthcare predictive models: Practical lessons from an empirical evaluation of 30-day hospital readmission models

ObjectiveDespite increased availability of methodologies to identify algorithmic bias, the operationalization of bias evaluation for healthcare predictive models is still limited. Therefore, this study proposes a process for bias evaluation through an empirical assessment of common hospital readmission models. The process includes selecting bias measures, interpretation, determining disparity impact and potential mitigations. MethodsThis retrospective analysis evaluated racial bias of four common models predicting 30-day unplanned readmission (i.e., LACE Index, HOSPITAL Score, and the CMS readmission measure applied as is and retrained). The models were assessed using 2.4 million adult inpatient discharges in Maryland from 2016 to 2019. Fairness metrics that are model-agnostic, easy to compute, and interpretable were implemented and apprised to select the most appropriate bias measures. The impact of changing model’s risk thresholds on these measures was further assessed to guide the selection of optimal thresholds to control and mitigate bias. ResultsFour bias measures were selected for the predictive task: zero-one-loss difference, false negative rate (FNR) parity, false positive rate (FPR) parity, and generalized entropy index. Based on these measures, the HOSPITAL score and the retrained CMS measure demonstrated the lowest racial bias. White patients showed a higher FNR while Black patients resulted in a higher FPR and zero-one-loss. As the models’ risk threshold changed, trade-offs between models’ fairness and overall performance were observed, and the assessment showed all models’ default thresholds were reasonable for balancing accuracy and bias. ConclusionsThis study proposes an Applied Framework to Assess Fairness of Predictive Models (AFAFPM) and demonstrates the process using 30-day hospital readmission model as the example. It suggests the feasibility of applying algorithmic bias assessment to determine optimized risk thresholds so that predictive models can be used more equitably and accurately. It is evident that a combination of qualitative and quantitative methods and a multidisciplinary team are necessary to identify, understand and respond to algorithm bias in real-world healthcare settings. Users should also apply multiple bias measures to ensure a more comprehensive, tailored, and balanced view. The results of bias measures, however, must be interpreted with caution and consider the larger operational, clinical, and policy context.

Characteristics of Patients Receiving Complex Case Management in an Acute Care Hospital.

Improving transitions in care is a major focus of health care planning. In the research team's prior intervention study, the length of stay (LOS) was reduced when patients at high risk for readmission were identified early in their acute care stay and received complex management. This study will describe the characteristics of patients receiving complex case management in an urban acute care hospital. Acute care hospital. This was a retrospective chart review of patients in a previous quality assurance study. A random selection of patients who previously underwent high-risk screening using the LACE (Length of stay; Acuity of the admission; Comorbidity of the patient; Emergency department use) index and received complex case management (the intervention group) were reviewed. The charts of a random selection of patients from the previous comparison group were also reviewed. Patient characteristics were collected and compared using descriptive statistics. In the intervention group, more patients had their family physicians (FPs) documented (93.1% [81/87] vs. 89.2% [66/74]). More patients in the intervention group (89.7% [77/87] vs. 85.1% [63/74]) lived at home prior to admission. More patients in the intervention group had a family caregiver involved (44.8% [39/87] vs. 41.9% [31/74]). At discharge, more patients in the intervention group (87.1% [74/85]) were discharged home compared with the comparison group (78.4% [58/74]). (1) Having an identified FP, living at home, and having family caregiver(s) characterized those with lower LOS and discharged home. (2) Case management, risk screening, and discharge planning improve patient outcomes. (3) This study identified the importance of having a FP and engaged family caregivers in improving care outcomes.

#1604 LACE index as predictors of 30 day rehospitalisation among patients on maintenance hemodialysis, cohort study

Abstract Background and Aims Over 35% of patients on maintenance dialysis are readmitted to the hospital within 30 days of hospital discharge. Validated risk assessment tools such as the LACE index (Length of stay, Acuity of admission, Co-morbidities, and Emergency department visits) have been developed to identify patients at high risk of hospital readmission in non-dialysis patients but have not been tested in dialysis patients. Method Data was collected from electronic medical records of patients on maintenance hemodialysis in Diaverum outpatient clinics hospitalized during 2023. Data included Age, gender, duration of hospital stay, Charlson comorbidity index (CCS), urgent or planned, and cause of Hospitalization. Multivariable logistic regression models with 30-day readmission for the same reason as the dependent outcome were used to identify readmission risk factors and whether the LACE score validly predicts 30-day rehospitalization among our cohort. Results Among (4829) patients received dialysis in Diaverum Saudi Arabia during 2023. (1487) episodes of hospitalization were reported due to various reasons rate of hospitalization was 0.54 Episodes per patient-years at risk, and 372(25%) of hospitalized patients were re-hospitalized for the same reason within 30 days. The proportion of patients with 30 days of rehospitalization for the same reason was higher in patients with post-initial discharge LACE index ≥10 in comparison to patients with LACE index&amp;lt;10, 348(32.6%) &amp; 24(5.7) respectively P-VALUE&amp;lt;0.001. also significantly higher in patients ≥65years old than those younger 163(27.7%) &amp;209(23.3%) respectively P-VALUE&amp;lt;0.05, those in high risk group (defined as those with higher comorbidity index ≥6 and/or lower Individual Patient Performance Score&amp;lt;75) than those who are not in high risk group 215(35%) &amp; 157(17.9%) respectively P-VALUE&amp;lt;0.001. Multivariate logistic regression including covariates significantly associated with higher 30-day rehospitalization to avoid confounder effect with model including ( age, gender, dialysis duration, body mass index, comorbidity index, diabetes, vascular access, cause of hospitalization and inclusion in high-risk group ) showed significantly higher rehospitalization in patients with post-initial discharge LACE index ≥10 in comparison to patients with LACE index&amp;lt;10 ODDS RATIO (95%CI) 7.95(5.1-12.5) P-VALUE&amp;lt;0.001. independent of other covariates in the model Conclusion LACE index at discharge is an independent predictor of (30) days re-hospitalization in patients on maintenance Hemodialysis. Further analysis needed for a specific model for HD patients taking into consideration the unique characteristics of this cohort of patients.

#100 Hospital readmissions for fluid overload among individuals with diabetes mellitus

Abstract Background and Aims Hospital readmissions due to recurrent fluid overload in diabetes can be avoided with evidence-based interventions. To identify at-risk individuals, we evaluated the incidence and risk factors associated with fluid overload-related hospital readmissions and developed risk prediction models for readmissions for fluid overload in people with diabetes. Method Single-centre retrospective cohort study of 1531 adults with diabetes and hospitalized in Singapore General Hospital for fluid overload, congestive heart failure, pulmonary edema and generalized edema between 2015 and 2017. Patients were included if they had diabetes mellitus identified by the SingHealth Diabetes Registry according to diagnosis codes, laboratory results and/or the use of glucose-lowering medication and had follow up visits, but excluded if their estimated glomerular filtration rate (eGFR) &amp;lt;15 ml/min/1.73 m2. Baseline data on demographics, comorbidities, biochemistry and medication use during admission were retrieved from electronic medical records up till 30 December 2018. The outcomes were fluid overload- related readmissions within 30 days and 90 days of discharge of the index hospitalization. Multivariable Cox proportional-hazards regression analysis was performed with inclusion of covariates with p≤0.10 on univariate analysis. Multivariable regression models developed were compared with the LACE index with net reclassification index (NRI). Results Among 1531 patients, median age was 73 (IQR 64, 81) and readmissions for fluid overload within 30 days and 90 days occurred in 131 (8.6%) and 264 (17.2%) patients, respectively. Individuals with 30- or 90-day readmissions were more likely to have prior hospitalization for fluid overload and more frequent ED visits within 6 months before the index hospitalization, cardiovascular disease, hypertension, and required high-dose intravenous furosemide during the index hospitalization, or current smoker. Clinical and clinico-psychosocial models for 30- and 90-day readmissions for fluid overload were developed (Table 1). The clinical and clinico-psychosocial models showed better discrimination and calibration than LACE index for predicting 30- and 90-day readmissions for fluid overload. The NRIs for clinical models were i. 30-day readmissions: +22.1%, ii. 90-day readmissions: +27.9% respectively, while the NRIs for clinico-psychosocial models were i. 30-day readmissions: +22.4%, ii. 90-day readmissions: +28.9% respectively (Fig. 1). Conclusion Patients with diabetes who are at-risk of readmission for fluid overload can be identified for targeted interventions to reduce preventable hospitalizations.

#87 Hospital readmissions among individuals with severe kidney disease hospitalized for fluid overload

Abstract Background and Aims Fluid overload is a common presentation and a leading cause for hospitalizations in severe kidney disease, but interventions can help reduce readmissions. To identify at-risk individuals, we evaluated the incidence and risk factors associated with fluid overload and all-cause hospital readmission among individuals with severe kidney disease. Method Single-center retrospective cohort study of patients hospitalized in Singapore General Hospital for fluid overload between 2015 and 2017. Fluid overload was identified from discharge codes for fluid overload, heart failure, pulmonary edema, and generalized edema. Patients were included if their estimated glomerular filtration rate (eGFR) was between 11 and 30 ml/min/1.73 m2, but excluded if they had urgent indications for dialysis or were lost to follow-up (Fig. 1). Baseline data on demographics, comorbidities, biochemistry and medication use during admission were retrieved from electronic medical records up till 30th December 2018. The outcomes were fluid overload-related and all-cause readmissions within 30 days of discharge of the index hospitalization. Multivariable logistic regression analysis (stepwise method) was performed with inclusion of covariates with P ≤ .10 on univariate analysis. Results Among 783 patients, the median age was 75 (IQR 67, 83) years and the admission eGFR was 20.9 (IQR 16.1, 25.4) ml/min/1.73 m2 at index hospitalization. 30-day readmission for fluid overload and all-causes occurred in 10.6% and 26.8% of the cohort, respectively (Table 1). Individuals with fluid overload readmissions were more likely to have atherosclerotic cardiovascular disease (ASCVD), atrial fibrillation (AF), prior hospitalization for fluid overload within 6 months, lower serum sodium and bicarbonate, but higher serum potassium at admission (Table 1). They were less likely to have hypertension, prior nephrology consult within 3 months and renin-angiotensin system (RAS) blocker prescription at discharge. In multivariable analysis, 30-day readmission for fluid overload was independently associated with ASCVD (adjusted odds ratio (aOR) 1.81, 95% CI 1.08-3.03, P = .02), AF (aOR 1.93, 95% CI 1.13-3.29, p = 0.02), higher serum potassium (aOR 1.61, 95% CI 1.14-2.26, P = .007), high-dose intravenous furosemide (aOR 1.66, 95% CI 1.02-2.67, P = .04), and inversely associated with prior nephrology consult (aOR, 0.51, 95% CI 0.29-0.89, P = .02) and RAS blocker prescription at discharge (aOR 0.61, 95% CI 0.38-0.99, P = .04). Individuals with all-cause readmissions were more likely to have AF, more frequent emergency department visits within 6 months, lower eGFR and serum sodium, but higher serum potassium on admission, longer length of stay and higher LACE index (Table 1). They were less likely to have hypertension, RAS blocker and statin prescription at discharge. 30-day readmission for all-causes was independently associated with higher number of emergency department visits (aOR 1.21, 95% CI 1.04-1.40, P = .02), higher LACE index (aOR 1.08, 95% CI 0.001-1.17, P = .048), and inversely associated with mineralocorticoid receptor blocker (aOR 0.55, 95% CI 0.31-0.97, P = .04) and statin prescription at discharge (aOR 0.51, 95% CI 0.33-0.78, P = .002) in multivariable analysis. Conclusion Risk stratification of patients with severe kidney disease who are at-risk of fluid overload and all-cause readmissions may allow for targeted interventions to reduce hospital readmissions.

Predicting 1year readmission for heart failure: A comparative study of machine learning and the LACE index.

There is a lack of tools for accurately identifying the risk of readmission for heart failure in elderly patients with arrhythmia. The aim of this study was to establish and compare the performance of the LACE [length of stay ('L'), acute (emergent) admission ('A'), Charlson comorbidity index ('C') and visits to the emergency department during the previous 6months ('E')] index and machine learning in predicting 1 year readmission for heart failure in elderly patients with arrhythmia. Elderly patients with arrhythmia who were hospitalized at Sichuan Provincial People's Hospital between 1 June 2018 and 31 May 2020 were enrolled. The LACE index was calculated for each patient, and the area under the receiver operating characteristic curve (AUROC) was calculated. Six machine learning algorithms, combined with three variable selection methods and clinically relevant features available at the time of hospital discharge, were used to develop machine learning models. AUROC and area under the precision-recall curve (AUPRC) were used to assess discrimination. Shapley additive explanations (SHAP) analysis was used to explain the contributions of the features. A total of 523 patients were enrolled, and 108 patients experienced 1year hospital readmission for heart failure. The AUROC of the LACE index was 0.5886. The complete machine learning model had the best predictive performance, with an AUROC of 0.7571 and an AUPRC of 0.4096. The most important predictors for 1year readmission were educational level, total triiodothyronine (TT3), aspartate aminotransferase/alanine aminotransferase (AST/ALT), number of medications (NOM) and triglyceride (TG) level. Compared with the LACE index, the machine learning model can accurately identify the risk of 1year readmission for heart failure in elderly patients with arrhythmia.

Using Machine Learning Models to Identify Factors Associated With 30-Day Readmissions After Posterior Cervical Fusions: A Longitudinal Cohort Study.

Readmission rates after posterior cervical fusion (PCF) significantly impact patients and healthcare, with complication rates at 15%-25% and up to 12% 90-day readmission rates. In this study, we aim to test whether machine learning (ML) models that capture interfactorial interactions outperform traditional logistic regression (LR) in identifying readmission-associated factors. The Optum Clinformatics Data Mart database was used to identify patients who underwent PCF between 2004-2017. To determine factors associated with 30-day readmissions, 5 ML models were generated and evaluated, including a multivariate LR (MLR) model. Then, the best-performing model, Gradient Boosting Machine (GBM), was compared to the LACE (Length patient stay in the hospital, Acuity of admission of patient in the hospital, Comorbidity, and Emergency visit) index regarding potential cost savings from algorithm implementation. This study included 4,130 patients, 874 of which were readmitted within 30 days. When analyzed and scaled, we found that patient discharge status, comorbidities, and number of procedure codes were factors that influenced MLR, while patient discharge status, billed admission charge, and length of stay influenced the GBM model. The GBM model significantly outperformed MLR in predicting unplanned readmissions (mean area under the receiver operating characteristic curve, 0.846 vs. 0.829; p < 0.001), while also projecting an average cost savings of 50% more than the LACE index. Five models (GBM, XGBoost [extreme gradient boosting], RF [random forest], LASSO [least absolute shrinkage and selection operator], and MLR) were evaluated, among which, the GBM model exhibited superior predictive performance, robustness, and accuracy. Factors associated with readmissions impact LR and GBM models differently, suggesting that these models can be used complementarily. When analyzing PCF procedures, the GBM model resulted in greater predictive performance and was associated with higher theoretical cost savings for readmissions associated with PCF complications.

Utility of the LACE index to assess risk of mortality and readmission in patients with spinal infections

Retrospective cohort study. To assess the utility of the LACE index for predicting death and readmission in patients with spinal infections (SI). SIs are severe conditions, and their incidence has increased in recent years. The LACE (Length of stay, Acuity of admission, Comorbidities, Emergency department visits) index quantifies the risk of mortality or unplanned readmission. It has not yet been validated for SIs. LACE indices were calculated for all adult patients who underwent surgery for spinal infection between 2012 and 2021. Data were collected from a single academic teaching hospital. Outcome measures included the LACE index, mortality, and readmission rate within 30 and 90 days. In total, 164 patients were analyzed. Mean age was 64.6 (± 15.1) years, 73 (45%) were female. Ten (6.1%) patients died within 30 days and 16 (9.8%) died within 90 days after discharge. Mean LACE indices were 13.4 (± 3.6) and 13.8 (± 3.0) for the deceased patients, compared to 11.0 (± 2.8) and 10.8 (± 2.8) for surviving patients (p = 0.01, p < 0.001), respectively. Thirty-seven (22.6%) patients were readmitted ≤ 30 days and 48 (29.3%) were readmitted ≤ 90 days. Readmitted patients had a significantly higher mean LACE index compared to non-readmitted patients (12.9 ± 2.1 vs. 10.6 ± 2.9, < 0.001 and 12.8 ± 2.3 vs. 10.4 ± 2.8, p < 0.001, respectively). ROC analysis for either death or readmission within 30 days estimated a cut-off LACE index of 12.0 points (area under the curve [AUC] 95% CI, 0.757 [0.681–0.833]) with a sensitivity of 70% and specificity of 69%. Patients with SI had high LACE indices that were associated with high mortality and readmission rates. The LACE index can be applied to this patient population to predict the risk of early death or unplanned readmission.

Abstract P155: Identifying Inpatient and Outpatient Factors That Reduce Heart Failure Exacerbation Readmissions

Acute heart failure 30-day readmissions are associated with increased patient morbidity and mortality. Our aim is to identify systemic factors that can reduce heart failure readmissions, including the implementation of a heart failure care coordination team. We assessed the hypothesis that individualized post-discharge education and summary from the heart failure care coordination team will reduce 30-day heart failure readmissions.This retrospective study included 213 patients, 18 years and older admitted with NYHA Class II-IV and ACC/AHA Stage B-D heart failure exacerbations from July 2021- March 2022, at a community hospital. Collected variables included demographics, medical comorbidities, labs, echocardiographic findings, inpatient orders, LACE index score, discharge follow-up, receipt of an after-visit summary, length of stay, and 30-day readmission. Descriptive data was summarized as counts and percentages for categorical data, and mean or median for continuous data. Between-group parametric data means were compared by calculating the t-test for independent variables, or medians using the Mann-Whitney U test for data that were nonparametric. Categorical variables were compared by calculating the chi-square test for association or Fisher’s exact test, as appropriate. The primary study objective was tested using Logistic regression modeling. Throughout this study, a p-value ≤0.05 (two-tail) was considered statistically significant. The medical records of all available reduced treatment plan patients were compared to preserved treatment plan patients in approximately equal numbers. Pair-wise t-tests required the greatest number of subjects. Using the conventional power level of 80%, a two-tailed alpha of 0.05, and assuming a medium effect size, it was estimated that this comparison will require at least 114 subjects. In this study, 189 of 213 patients were reached by the heart failure care coordination team. 49/189 were readmitted in 30 days and 140/189 were not. For the readmission group, the percentage of patients who were CKD Stage 3 or 4 (39.4%, 28 of 71) was higher than those who were not (21.1%, 30 of 142) with a p-value of 0.005. The percentage of patients who had discharge BNP checked (39.1%, 18 of 46) was higher than those who were not (24.1%, 40 of 166) with a p-value of 0.043. A higher LACE index score was associated with readmission as compared to a lower score, with a p-value of &lt; 0.001. In conclusion, presence of CKD stage 3 or 4, discharge BNP, and a higher LACE-index score were associated with 30-day heart failure readmission. Counterintuitively, the heart failure care coordination team did not have an impact on this endpoint, contrary to our hypothesis. Future studies with a larger sample size and focus on the outpatient care team efforts may be helpful in elucidating the team's potential in preventing heart failure readmission.

Causes for 30-day readmissions and accuracy of the LACE index in regional Victoria, Australia.

Clinicians and funders continue to search for ways to reduce costs without sacrificing quality of care. Ongoing research should focus on innovative care models that identify patients at high-risk for hospitalisation and thereby reduce healthcare costs. This study examined readmission rates, comorbidity profiles and the performance of the LACEi (Length of stay, Acuity of admission, Charlson Comorbidity Index, ED admissions in the previous 6 months index) to predict the risk of 30-day readmissions in a regional population. Furthermore, we tested a novel clinician-orientated classification for the causes of 30-day readmissions. Using a nested case-control design, data were extracted from administrative health records using 30-day readmission status as the outcome. We defined cases as discharges within 30 days before readmission and controls without a discharge within 30 days before admission between 1 July 2020 and 30 June 2022. The study was conducted at South West Healthcare in Victoria, Australia. All adult medical patients were discharged alive from the facility. We excluded planned readmissions, surgical and obstetric admissions, dialysis, transfers to alternative facilities and discharges against medical advice. Thirty-day readmission rate, comorbidity profile for all admissions, LACEi for all admissions, the performance of the LACEi in our setting and the causes leading to readmission using a clinician-orientated classification tool. Comorbidity burden, male sex and age > 65 years were associated with increased readmission risk but not length of stay. The LACEi demonstrated modest predictive ability to identify high-risk patients for readmissions (area under the receiver operating characteristic curve = 0.59). Additional variables were needed to increase accuracy. The novel classification identified 42% of readmissions as potentially avoidable. Our study identified comorbidity burden, male sex and age ≥ 65 years as critical indicators for readmission risk. Although the LACEi showed moderate predictive ability, additional variables were needed for increased accuracy. Over 40% of readmissions were potentially avoidable, and nearly two thirds occurred within 14 days of discharge from the hospital.

Chronic care management: An opportunity to drive quality outcomes in oncology care.

444 Background: CMS approved Chronic (CCM) and Principal Care Management (PCM) to improve patient care. These services are typically non-face-to-face; providers can bill ≥ twenty minutes /month. In oncology, this may help with comorbidity and symptom management to reduce the risk and expense of avoidable hospitalizations and ER visits and pay providers for time spent. In 2021, a system of community cancer centers initiated a program with a partner to manage chronic conditions more effectively at home. Approximately three hundred people were enrolled. A review of the data found that patients did not meet the medical criteria for the program, and care was not billable. An internal program was created to improve the enrollment process and financial sustainability. A clinical needs assessment was completed with our oncologists, outlining revised criteria for patient enrollment. Qualified patients were enrolled by their attending oncologist during the initial face-to-face visit and assigned a dedicated RN. Methods: Three hundred unenrolled, actively treating patients were randomly selected from January 2023 to May 2023 by a group of data analysts not associated with the program to be compared to enrolled patients who have been enrolled on average six months. Patients were compared by age, gender, cancer type, and race. The LACE Index for Readmission, a validated tool using the length of stay, acuity of admission, comorbidities, and ER visits to predict the risk of readmission, was done retrospectively by RNs for all patients. Enrolled patients received a comprehensive clinical needs assessment by RNs, and a provider-approved care plan was created. RNs provided patients with education on lifestyle modification and assisted in setting achievable goals. The care plan was monitored with weekly calls, providing the opportunity to respond to patients’ needs proactively. Symptom and medication management were provided, and psychological and nutritional needs were addressed. This coordination of services was documented in the electronic health record, and superbills were created and sent to billing and coding monthly. Results: See table. Conclusions: The purpose of this study was to compare hospital admissions between the groups. Our findings suggest that CCM and PCM may reduce unplanned admissions, with physician engagement pivotal in driving appropriate enrollment. The current program is labor-intensive and further investigation is required to determine the most cost-effective way of scaling the program with the possible use of artificial intelligence, RPM, or other technology.[Table: see text]

Evaluation of the Retrospective LACE Index in Predicting the Risk of Readmission in Patients with Hereditary Angioedema in an Emergency Department.

This study aimed to calculate the LACE index in patients who admitted to the emergency department (ED) with hereditary angioedema (HA) diagnosed and to predict recurrent admissions of patients. In this single-center study, patients aged 18 or higher who were admitted to the ED diagnosed with HA were included over a 12-year period. 35 patients diagnosed with code E88.0 were evaluated according to electronic file records. The number of admissions to the ED in the last 6 months was 2. The LACE index was 4, and risk was 71.4%. The patients admitted to the hospital in the last 30 days had a higher rate of admission to the hospital in the last 6 months (p < 0.001). The LACE index at admission predicted 30 days admission with (AUC = 0.75, 95% CI (0.56-0.91)) acceptable discrimination. The LACE index and the number of admissions in the last 6 months included in the evaluation can be considered predictive in recurrent ED admissions of HA patients. However, the distribution of LACE-risk groups is no priority. Therefore, the low-, medium-, or high-risk level of LACE index values should be not taken into consideration in readmission of such patients.

1603P The ability of the LACE index to predict 30-day readmissions in oncology patients

1603P The ability of the LACE index to predict 30-day readmissions in oncology patients

WED-341 - The combination of LACE index, MELD on discharge and a previous history of decompensation defines risk groups for early readmission in patients with liver cirrhosis

WED-341 - The combination of LACE index, MELD on discharge and a previous history of decompensation defines risk groups for early readmission in patients with liver cirrhosis

The LACE index and risk factors of 14-day versus 30-day readmissions in children.

Risk factors for readmissions in children differ from those for adults, yet little is known about whether the LACE index (Length of stay, Acute admission, Charlson comorbidity index, Emergency department visits in the previous 6 months) developed for adults retains its validity when applied to the prediction of readmissions in children or within shorter intervals of time after discharge. In this retrospective cohort analysis of 4 256 patients aged≤18 hospitalized at one academic medical center in Taiwan, 2019, we first evaluated the performance of a LACE index model and the three other multivariate logistic regression models in their predictions of hospital readmissions in children using the same time intervals of 30 days. We then used multinomial logistic regression to analyze the characteristics and risk factors for readmissions that occur in the first 14 days with those that occur between 15 to 30 days after discharge. The optimal cutoff of LACE index score for children in the current study was 7, which is less than the optimal score of 10 in the original derivation for adults. The predictive model with the least discriminatory power was based on LACE score alone, whereas our model that included only two variables (length of stay and the number of admissions in the past one year) was found to outperform LACE index. Multinomial logistic regression results revealed that children who had supplemental private health insurance had lower readmission rates than those insured by National Health Insurance program only, suggesting a disparity by insurance status. Some risk factors of readmission within 14 days such as acute admissions and supplemental private health insurance were not found to be relevant for longer-term predictions, while age, which was a valid predictor of readmission within 15 to 30 days, did not affect the prediction of shorter-term readmissions. Prior healthcare utilizations and a higher illness burden were found to be greater contributors to readmissions beyond 14 days. Our simpler model outperformed the more complex LACE tool in identifying children at risk of readmission. Shorter-term readmissions can be attributed to different causes than 30-day readmissions, suggesting a need for different screening tools, interventions, and care support. There remains some socioeconomic disparity associated with readmission rates in the current healthcare system.

Identification of Factors Associated With 30-Day Readmissions After Posterior Lumbar Fusion Using Machine Learning and Traditional Models: A National Longitudinal Database Study.

Retrospective cohort study. To identify factors associated with readmissions after PLF using machine learning and logistic regression (LR) models. Readmissions following posterior lumbar fusion (PLF) place significant health and financial burden on the patient and overall healthcare system. The Optum Clinformatics® Data Mart database was used to identify patients who underwent posterior lumbar laminectomy, fusion, and instrumentation between 2004 and 2017. Four machine learning models and a multivariable LR model were used to assess factors most closely associated with 30-day readmission. These models were also evaluated in terms of ability to predict unplanned 30-day readmissions. The top performing model (Gradient Boosting Machine; GBM) was then compared to the validated LACE index in terms of potential cost savings associated with implementation of the model. A total of 18,981 patients were included, of which 3,080 (16.2%) were readmitted within 30 days of initial admission. Discharge status, prior admission, and geographic division were most influential for the LR model, while discharge status, length of stay, and prior admissions had greatest relevance for the GBM model. GBM outperformed LR in predicting unplanned 30-day readmission (mean AUC 0.865 vs. 0.850, P<0.0001). Use of GBM also achieved a projected 80% decrease in readmission-associated costs relative to those achieved by the LACE index model. Factors associated with readmission vary in terms of predictive influence based on standard logistic regression and machine learning models used, highlighting the complementary roles these models have in identifying relevant factors for prediction of 30-day readmissions. For posterior lumbar fusion procedures, Gradient Boosting Machine yielded greatest predictive ability and associated cost savings for readmission. 3.