Difference In Median Length Of Stay Research Articles

Clinical significance of an elevated on-admission beta-hydroxybutyrate in acutely ill adult patients without diabetes.

To determine the relationship between point-of-care β-hydroxybutyrate (BHB) concentration and outcomes in adult patients without diabetes admitted through ED. This was a prospective study from 10 March to 2 July 2021. Admitted patients without diabetes had capillary BHB sampled in ED. Outcomes of length-of-stay (LOS), composite mortality/ICU admission rates and clinical severity scores (Quick Sepsis Organ Failure Assessment score/National Early Warning Score [qSOFA/NEWS]) were measured. BHB was assessed as a continuous variable and between those with BHB above and equal to 1.0 mmol/L and those below 1.0 mmol/L. A total of 311 patients were included from 2377 admissions. Median length-of-stay was 4.1 days (IQR 2.1-9.8), 18 (5.8%) died and 37 (11.8%) were admitted to ICU. Median BHB was 0.2 mmol/L (IQR 0.1-0.4). Twenty-five patients had BHB ≥1.0 mmol/L and five were >3.0 mmol/L. There was no significant difference in median LOS for patients with BHB ≥1.0 mmol/L compared to non-ketotic patients, 5.3 days (IQR 2.2-7.5) versus 4.1 days, respectively (IQR 2.0-9.8) (P = 0.69). BHB did not correlate with LOS (Spearman ρ = 0.116, 95% confidence interval: 0.006-0.223). qSOFA and NEWS also did not differ between these cohorts. For those 25 patients with BHB ≥1.0 mmol/L, an infective/inflammatory diagnosis was present in 11 (44%), at least 2 days of fasting in 10 (40%) and ethanol intake >40 g within 48 h in 4 (16%). Routine BHB measurement in patients without diabetes does not add to clinical bedside assessment and use should be limited to when required to confirm a clinical impression.

330. PRE-EMPTIVE ENDOLUMINAL VACUUM THERAPY WITH A SURGICAL DRAIN FOR REDUCTION OF ANASTOMOTIC LEAK RATE FOLLOWING MINIMALLY INVASIVE ESOPHAGECTOMY

Abstract Background Endoluminal vacuum therapy (EVT) is an effective treatment for anastomotic leak (AL) following esophagectomy. A novel application of EVT is to utilise its effect to improve microcirculation and granulation at the anastomosis pre-emptively to reduce the risk of AL. We describe a unique method of positioning a fenestrated surgical drain with applied vacuum suction at the level of the anastomosis following minimally invasive esophagectomy (MIE), with the aim of reducing AL rate and postoperative morbidity. Methods A retrospective cohort comparison was performed on consecutive MIEs from 2019 to 2023. All operations were performed by experienced esophagogastric surgeons at a tertiary centre in the UK. Included patients were all adults undergoing MIE for esophageal cancer. Patients who underwent pre-emptive EVT (pEVT) with surgical drain following MIE were compared with a control group who underwent MIE alone. All patients followed the same enhanced recovery protocol with pEVT removed two to five days post-operatively. The primary outcome assessed was AL, secondary outcomes included: length of stay (LOS), readmission to Intensive care, complications (Clavien-Dindo) and 90-day mortality with minimum 90-days follow-up. Results 72 patients were included, 25 patients underwent MIE with pEVT compared with 47 in the control. The AL rate in the pEVT group was 8% compared with 21.3% in the control (RRR 62%, p = 0.13). All ALs following pEVT were managed successfully with EVT alone. There was a reduced risk of aspiration pneumonia (RR:0.37, p = 0.22) and readmission to ICU (RR:0.24, p = 0.049) following pEVT. There was 0%, 90-day mortality in both groups and no significant difference in median LOS (7 vs 9 days, p = 0.72 (pEVT vs Control)). However, a higher risk of stricturing following pEVT was noted (RR:3.76, p = 0.04). Conclusion Our data suggests pEVT with a surgical drain may be a safe and effective method of reducing AL rate following MIE. Additionally, it confers the benefit of being an effective treatment for AL should this complication arise. There was significant increase in benign stricturing requiring dilatation following pEVT, however, this may be attributable to other operative factors. Our results support the need for further prospective studies such as a randomised controlled trial.

1667 VARIATION IN ACUTE GERIATRIC MEDICINE ACTIVITY AND OUTCOMES IN SCOTLAND BEFORE, DURING AND AFTER THE COVID PANDEMIC LOCKDOWNS

Abstract Introduction The Scottish Care of Older People (SCoOP) collaborative reports outcomes of acute geriatric medicine admissions across Scottish hospitals. The covid pandemic caused major and variable restructuring of acute services across the country. Their impact on activity and outcomes is unknown. Methods We collated all SMR01/SMR01E hospital episodes from Public Health Scotland from 1st April 2017 to 31st March 2022 where over 50% of the total episode was spent under acute geriatric medicine (code AB) and the diagnosis was not stroke. Activity and outcomes in 19 major hospitals were compared across financial years 2017-19 (before-), 2020/21 (during-) and 2021/22 (after lockdowns). Results Admissions fell by 15% to 36954 in 2020/21 from an average 42566 before recovering to 41971 in 2021/22. Some hospitals were busier than ever in 2020/21 while others saw large reductions in activity. Age, sex and social deprivation profiles differed between hospitals (p<0.001) but remained similar within each hospital at all timepoints. Mortality at 30 days post-admission was 8.5% higher in 2020/21 (17.9% v 16.5% in other years, p<0.001), with 2-fold differences across hospitals. Mean median length of stay (LOS) across hospitals was 11.7 days, compared to 13.6 days in 2017-20, p<0.001. There were up to 17-fold differences in median LOS between hospitals (2-34 days), p<0.001. The impact of the pandemic on LOS within each hospital was variable. Readmission rates at 7 days post-discharge were broadly similar across all years but two-fold differences between hospitals were seen each year (2020/21 4.8%-9.8%, mean 6.8%, p<0.001). Conclusion The impact of the pandemic on the activity and outcomes of individual hospitals was widely disparate, probably reflecting variation in how each hospital service responded to the pandemic. Overall activity and LOS were lower and mortality was higher in 2020/21.

Preoperative CT-Based Skeletal Muscle Mass Depletion and Outcomes after Total Laryngectomy.

To assess the role of preoperative CT-based skeletal muscle mass depletion on postoperative clinical outcomes and survival in patients who underwent total laryngectomy for cancer. Patients operated on between January 2011 and March 2020 were retrospectively included. Skeletal muscle area and intra- and inter-muscular fat accumulation were measured at the third lumbar vertebral level on preoperative CT scans. Skeletal muscle mass depletion was defined based on pre-established cut-off values. Their association with postoperative morbidity, length of stay (LOS), costs, and survival was assessed. A total of 84 patients were included, of which 37 (44%) had preoperative skeletal muscle mass depletion. The rate of postoperative fistula (23% vs. 35%, p = 0.348), cutaneous cervical dehiscence (17% vs. 11%, p = 0.629), superficial incisional surgical site infections (SSI) (12% vs. 10%, p = 1.000), and unplanned reoperation (38% vs. 37%, p = 1.000) were comparable between the two patient groups. No difference in median LOS was observed (41 vs. 33 days, p = 0.295), nor in treatment costs (119,976 vs. 109,402 CHF, p = 0.585). The median overall survival was comparable between the two groups (3.43 vs. 4.95 years, p = 0.09). Skeletal muscle mass depletion alone had no significant impact on postoperative clinical outcomes or survival.

Does an Adductor Canal Block Influence Patient-Reported Outcomes at One Year Following Total Knee Arthroplasty?

Introduction Total knee arthroplasty (TKA) for osteoarthritis is performed to improve knee function and quality of life. Adductor canal block (ACB) (with posterior capsule local anaesthetic) and periarticular local anaesthetic infiltration (LA) alonearecommon methods used for post-operative pain control following TKA. The primary aim of this study was to investigate the influence of ACB compared to LA alone on knee function at one year in patients undergoing primary TKA. The secondary aims were to investigate the influence on health-related quality of life (HRQoL), patient satisfaction at one year, and length of stay (LOS) in hospital following TKA. Methods During a three-year period, 1396 patients who underwent TKA at the study centre completed pre-operative and one-year post-operative questionnaires. Data collected included patient demographics, co-morbidities, LOS, Oxford Knee Score (OKS), EuroQol 5-Dimension (EQ-5D) and satisfaction scores. Patients who had a spinal anaesthetic (n=1096) with either ACB (with posterior capsule local anaesthetic) (n=224) and LA infiltration alone (n=872) were compared. Results There was a significant improvement in pre-operative to one-year post-operative knee function as measured by OKS overall (15.1, p<0.001), however there was no significant difference between ACB and LA groups (p=0.37). There was no significant difference in change (p=0.43) or one-year post-operative (p=0.70) HRQoL, as measured by EQ-5D. There was also no significant difference in one-year patient satisfaction between groups (p=0.57). There was no significant difference in median LOS between groups (four days with ACB, four days with LA), although patients younger than 55 yearsundergoing ACB had a statistically significant median of one day reduced LOS compared with the LA group (p=0.01). Conclusions ACB when compared with LA alone was not associated with a difference in change in knee function, HRQoL, or patient satisfaction at one year following TKA. There was also no difference between groups in length of stay post-operatively. However, ACB was associated with a shorter length of stay in patients younger than 55, and this may be a group that would benefit from this intervention.

Institutional outcomes of blunt liver and splenic injury in the Arizona-Texas-Oklahoma-Memphis-Arkansas Consortium era.

The Arizona-Texas-Oklahoma-Memphis-Arkansas Consortium practice management guideline was created to standardize management of blunt liver or spleen injury across pediatric trauma centers. We describe our outcomes since guideline adoption at our institution and hypothesize that blunt liver or spleen injury may be managed more expeditiously than currently reported without compromising safety. A retrospective cohort study was conducted on patients younger than 18 years presenting with blunt liver and/or splenic injuries from March 2016 to March 2021 at one participating center. A total of 199 patients were included. There were no clinically relevant differences for age, body mass index, or sex among the cohort. Isolated splenic injuries (n = 91 [46%]) and motor vehicle collisions (n = 82 [41%]) were the most common injury and mechanism, respectively. The overall median length of stay (LOS) was 1.2 days (interquartile range, 0.45-3.3 days). Intensive care unit utilization was 23% (n = 46). There was no statistically significant difference in median LOS among patients with isolated solid organ injuries, regardless of injury grade. There were no readmissions associated with non-operative management. The Arizona-Texas-Oklahoma-Memphis-Arkansas Consortium guideline fosters high rates of nonoperative management with low intensive care unit utilization and LOS while demonstrating safety in implementation, irrespective of injury grade. Therapeutic/Care Management; Level IV.

Race and Remdesivir: Examination of Clinical Outcomes in a Racially and Ethnically Diverse Cohort in New York City

ObjectiveTo compare clinical characteristics and examine in-hospital length of stay (LOS) differences for COVID-19 patients who received remdesivir, by race or ethnicity.DesignRetrospective descriptive analysis comparing cumulative LOS as a proxy of recovery time.SettingA large academic medical center serving a minoritized community in Northern Manhattan, New York City.ParticipantsInpatients (N=1024) who received remdesivir from March 30, 2020–April 20, 2021.MethodsWe conducted descriptive analyses among patients who received remdesivir. Patients were described by proxies of social determinants of health (SDOH): race and ethnicity, residence, insurance coverage, and clinical characteristics. We calculated median hospital LOS as the cumulative incidence of hospitalized patients who were discharged alive, and tested differences between groups by using the Gray test. Patients who died or were discharged to hospice were censored at 29 days.Main Outcome MeasuresThe primary outcome was hospital LOS. The secondary outcome was in-hospital mortality.ResultsMedian LOS was 11.9 days (95% CI, 10.8-13.2) overall, with Black patients having the shortest (10.0 days, 95% CI, 8.0-13.2) and Asian patients having the longest (16.2 days, 95% CI, 8.3-27.2) LOS. A total of 214 patients (21%) died or were discharged to hospice, ranging from 16.5% to 23.7% of patients who identified as Black and Other (multiracial, biracial, declined), respectively.ConclusionsCOVID-19 has disproportionately burdened communities of color. We observed no difference in median LOS between racial or ethnic groups, which supports the notion that the heterogeneous effect of remdesivir in the literature may be explained in part by underrecruitment or participation of Black, Hispanic, and Asian patients in clinical trials.

Abstract 12261: Access to Care among Hispanic Patients Hospitalized for Lower Extremity Peripheral Artery Disease

Introduction: Compared to other racial/ethnic groups, Hispanics are less likely to access regular primary care, which can delay detection and management of PAD. It is unknown whether in admissions for PAD, a trend towards more burden is observed for Hispanic vs. non-Hispanic White (NHW) patients. We therefore examined emergency department (ED) and elective admissions, length of stay (LOS), and costs accrued by ethnicity in patients hospitalized for primary PAD. Methods: The 2011-2017 National Inpatient Sample data were used to identify admissions in Hispanics and NHWs and whether the patient sought care via the ED and if the admission was elective. Yearly and overall percentages for each were determined, and temporal trends evaluated with the Cochran-Mantel-Haenszel test, stratified by ethnicity. Differences in median LOS and cost by ethnicity were evaluated with the Wilcoxon rank sum test. Results: From 2011-2017, there were 1,018,220 PAD hospitalizations (39% female, 14% Hispanic) and the percentage of patients who accessed inpatient care via the ED increased in Hispanics and NHWs, but was consistently higher in Hispanics (Fig. 1, all P trend &lt;0.0001). Across all years, 58.0% (82,047 of 141,551) of admissions in Hispanics were via the ED compared to 36.7% (321,932 of 876,669) in NHWs ( d =0.48) and elective admissions were higher in NHWs (396,294; 45.2%) than in Hispanics (38,661; 27.3%, d =0.43). Median LOS (4.5 vs. 3.7 days) and cost ($63,813 vs. $52,368) were higher in Hispanics than NHWs ( p &lt;.0001). Conclusions: Hispanic patients are more likely than NHWs to seek emergent, non-elective inpatient PAD care and to be admitted for longer and accrue greater costs. They are also increasingly more likely to access this care via the ED which may provide fewer barriers than other settings, especially for uninsured patients or those without legal immigration status. More accessible outpatient care may potentially reduce ED burden and improve PAD management for this population.

Neutrophil count, lymphocyte count and neutrophil-to-lymphocyte ratio in relation to response to adjunctive dexamethasone treatment in community-acquired pneumonia

BackgroundIt is hypothesised that community-acquired pneumonia (CAP) patients with more severe disease or inflammation might benefit more from adjunctive corticosteroid treatment. Neutrophil count, lymphocyte count and neutrophil-lymphocyte ratio (NLR) have been associated with inflammation and disease severity in CAP. We investigated the interaction between these parameters and adjunctive dexamethasone effects on clinical outcomes in CAP. MethodsWe conducted a post hoc analysis of the randomised placebo-controlled Santeon-CAP trial (n = 401), which showed a positive effect of adjunctive oral dexamethasone on length of stay (LOS) in CAP patients. White blood cell (WBC) count, neutrophil count, NLR (highest tertile vs. lowest two tertiles) and lymphocyte count (lowest tertile vs. highest two tertiles) were examined as potential effect modifiers of treatment with dexamethasone on LOS (primary outcome) and ICU-admission, 30-day mortality and hospital readmission. ResultsWBC differential counts were available for 354 patients. The effect of dexamethasone on LOS was more pronounced in high WBC count, high neutrophil count or high NLR subgroups (difference in median LOS of 2 days versus zero days in the reference subgroups, p for interaction < 0.05). There was no effect modification for the other outcomes. Patients with low WBC and low neutrophil counts did not benefit from dexamethasone, while hospital readmission rate was higher in those treated with dexamethasone (6% vs. 11%). ConclusionsWBC count and/or neutrophil might be easily available biomarkers to guide selection of CAP patients who are more likely to benefit from adjunctive dexamethasone treatment. Future prospective trials are needed to confirm this predictive potential.

Effect of a geriatric management unit on the outcomes of hospitalized older adults with cancer in Mexico.

12039 Background: Geriatric assessments and interventions improve the outcomes of hospitalized older adults with cancer, but their implementation in developing countries is limited. We studied the effect of a specialized geriatrician-led inpatient geriatric management unit compared with a conventional internal medicine ward on the outcomes of hospitalized Mexican older adults with cancer. Methods: This retrospective study included patients aged ≥65 with solid tumors who had a cancer-related hospitalization at a public academic center in Mexico City between March 2015 and October 2018. Patients hospitalized in the geriatric management unit (cases) were paired in a 1:2 fashion with those in internal medicine wards (controls). Pairing was done by age (+/- 5 years), tumor type, and admission date (+/- 3 months). We studied the effect of being hospitalized in the geriatric management unit on length of stay (LOS), incidence of delirium, hospital-acquired complications, and in-hospital mortality. Multivariate logistic regression models for each outcome were created using variables which were significant on univariate analysis. Results: 300 patients (100 cases, 200 controls, median age 75) were included. The most common tumors were gastrointestinal (GI) (53%) and genitourinary (25%). Both groups were comparable regarding baseline comorbidities (Charlson index 8.5 vs. 7.7, p = 0.99) and illness severity at admission (NEWS2 score 2.6 vs. 2.3, p = 0.82). No difference in median LOS was found between cases and controls (9.1 vs. 9.5 days, p = 0.34). Diagnosis of a GI tumor (OR 3.4, 95% CI 1.3-5.5), hospital-acquired complications (OR 4.9, 95% CI 2.5-7.3), and delirium (OR 5.5, 95% CI 2.3-8.7) were associated with longer LOS. 14% of patients in both groups had delirium. Hospitalization in the geriatric management unit reduced the risk of delirium (OR 0.35, 95% CI 0.1-0.9), while a higher Charlson index (OR 1.2, 95% CI 1.0-1.4), NEWS2 score (OR 1.2, 95% CI 1.1-1.4), and hospital-acquired complications (OR 7.3, 95% CI 2.9-18.5) increased it. 34% of patients developed hospital-acquired complications. Diagnosis of a GI tumor (OR 1.9, 95% CI 1.1-3.3) and higher NEWS2 score (OR 1.2, 95% CI 1.1-1.4) increased the risk of hospital-acquired complications. No differences in in-hospital mortality were seen between cases and controls (12% vs. 10%, p = 0.59). A higher NEWS2 score at admission (OR 1.4, 95% CI 1.2-1.7) and delirium (OR 10.7, 95% CI 3.2-36.3) increased the risk of death. Conclusions: Among older Mexican adults hospitalized for a cancer-related diagnosis, receiving care in a geriatric management unit led to a significant decrease in the risk of delirium. No improvements were seen in LOS, complications, or in-hospital mortality, which were associated with tumor and patient-related characteristics. Geriatric co-management can lead to improved geriatric outcomes in developing countries with limited resources.

Transitions of Care Coordination Intervention Identifies Barriers to Discharge in Hospitalized Stroke Patients.

Background: Prolonged hospital lengths of stay increase costs, delay rehabilitation, and expose acute ischemic stroke patients to hospital-acquired infections. We designed and implemented a nurse-driven transitions of care coordinator (TOCC) program to facilitate the transition of care from the acute care hospital setting to rehabilitation centers and home.Methods: This was a single-blinded, prospective, randomized pilot study of 40 participants to evaluate the feasibility of implementing a TOCC program led by a stroke nurse navigator in hospitalized acute ischemic stroke patients. The intervention consisted of a stroke nurse navigator completing eight specific tasks, including meeting with stroke patients and their families, facilitating communication between team members at multi-disciplinary rounds, assisting with referrals to rehabilitation facilities, providing stroke education, and arranging stroke clinic follow-up appointments, which were confirmed to be completed by independent study personnel. The primary outcome was to assess the feasibility of the program. The secondary outcomes included comparing hospital length of stay (LOS) and patient satisfaction between the TOCC and usual care groups. We also explored the association between patient-level variables and LOS.Results: The TOCC program was feasible with all pre-specified components completed in 84.2% (95% CI: 60.4–96.6%) and was not significantly different from the assumed completion rate of 75% (p = 0.438). There was no significant difference in median LOS between the two groups [TOCC 5.95 days (4.02, 9.57) vs. usual care 4.01 days (2.00, 10.45), false discovery rate (FDR)-adjusted p = 0.138]. There was a trend toward higher patient median satisfaction in the TOCC group [TOCC 35.00 (33.00, 35.00) vs. usual care 30 (26.00, 35.00), FDR-adjusted p = 0.1] as assessed by a questionnaire at 30 days after discharge. The TOCC study allowed us to identify patient variables (gender, insurance, stroke severity, and discharge disposition) that were significantly associated with longer hospital LOS.Conclusion: A TOCC program is feasible and can serve as a guide for future allocation of resources to facilitate transitions of care and avoid prolonged hospital stays.

Abstract 14555: Trends, and Impact of Atrial Fibrillation in Patients With Cardiac Amyloidosis

Introduction: The prevalence of atrial fibrillation (AF) in patients with cardiac amyloidosis (CA) can be as high as 40%, depending on the amyloid protein involved. There is a lack of data, however, concerning the temporal trends and prognostic implications of AF in these patients. Hypothesis: AF is associated with worse outcomes in patients with CA Methods: We identified 21,661 patients with the primary diagnosis of CA from the 2006 to 2014 National Inpatient Sample, of which 5,279 (25%) had AF. We used trend analysis to assess the temporal trends in the prevalence of AF from 2006 to 2014. We compared overall mortality, stroke, length of stay (LOS), and total cost in patients with those without AF. Results: The prevalence of AF increased from 1.3% in 2006 to 17.7% in 2014 (p&lt;0.0001). Patients with AF were older and had more comorbidities. After propensity matching, overall mortality (8.7% vs 7.5%; p= 0.027) and rates of stroke (21.0% vs 18.0%; p&lt;0.0001) were significantly higher in patients with CA and AF, compared to those with CA alone. There was no difference in median LOS (5 (3,9) vs 5(3,10); p= 0.065) and median total hospital cost ($36163 vs $34706; p= 0.114). The strongest predictors for overall mortality were hemodialysis, stroke, cardiac arrest, and congestive heart failure. Conclusions: Over a decade period, the prevalence of AF increased among CA patients and was associated with a higher risk of stroke and overall mortality.

Safety and Efficacy of an Enhanced Recovery Protocol After Hepatic Resection.

Enhanced recovery protocols (ERPs) have shown to improve outcomes in multiple specialties and were recently applied to hepatic resections. The objective of this study was to determine the safety and efficacy of ERP in hepatic resection. Between 2013-2017, 208 patients underwent hepatectomy. The ERP included early ambulation, early oral intake, and multimodal analgesia. Primary study end points were hospital length of stay (LOS) and overall morbidity; secondary end points were return of bowel function (ROBF), 30-day readmission, and 90-day mortality. Major hepatectomies were selected for separate analysis. Overall, pre-ERP (N = 99) and ERP (N = 109) were similar in demographics. ERP patients had earlier oral intake and ROBF with similar overall morbidity. Although median LOS was 5days, 43% of ERP patients had LOS ≤4days vs. 27% in the pre-ERP cohort (P = .02). 30-day readmission was similar (12%), and 90-day mortality was 2.8% vs. 3.0% (pre-ERP vs. ERP, P = .90). In major hepatectomies, pre-ERP (N = 41) and ERP (N = 33) demographics and operative characteristics were similar. ERP patients had earlier oral intake and ROBF with similar morbidity and mortality. There was no significant difference in median LOS; however, 36% of the major hepatectomy ERP patients had LOS ≤4days vs. 17% of pre-ERP patients, P = .06. In conclusion, ERP can be safely implemented in hepatectomy, with earlier oral intake and ROBF, shorter LOS in some patients, and no increase in morbidity or mortality.

Association of patterns of multimorbidity with length of stay: A multinational observational study.

The aim of this study was to identify the combinations of chronic comorbidities associated with length of stay (LOS) among multimorbid medical inpatients.Multinational retrospective cohort of 126,828 medical inpatients with multimorbidity, defined as ≥2 chronic diseases (data collection: 2010-2011). We categorized the chronic diseases into comorbidities using the Clinical Classification Software. We described the 20 combinations of comorbidities with the strongest association with prolonged LOS, defined as longer than or equal to country-specific LOS, and reported the difference in median LOS for those combinations. We also assessed the association between the number of diseases or body systems involved and prolonged LOS.The strongest association with prolonged LOS (odds ratio [OR] 7.25, 95% confidence interval [CI] 6.64-7.91, P < 0.001) and the highest difference in median LOS (13 days, 95% CI 12.8-13.2, P < 0.001) were found for the combination of diseases of white blood cells and hematological malignancy. Other comorbidities found in the 20 top combinations had ORs between 2.37 and 3.65 (all with P < 0.001) and a difference in median LOS of 2 to 5 days (all with P < 0.001), and included mostly neurological disorders and chronic ulcer of skin. Prolonged LOS was associated with the number of chronic diseases and particularly with the number of body systems involved (≥7 body systems: OR 21.50, 95% CI 19.94-23.18, P < 0.001).LOS was strongly associated with specific combinations of comorbidities and particularly with the number of body systems involved. Describing patterns of multimorbidity associated with LOS may help hospitals anticipate resource utilization and judiciously allocate services to shorten LOS.

166 Single Centre Experience with Heart Function Clinic and Heart Failure Admissions

Heart failure admissions contribute significantly to health care expenditure. Multidisciplinary heart failure clinic (HFC) may reduce readmission rate and mortality. We conducted a retrospective audit of heart failure admissions under the Cardiology Service in 2018. 178 of 198 admitted patients were analysed. Baseline characteristics: male 67.4%, average age 72.6y. Co-morbidities: ischaemic heart disease (47.4%), atrial fibrillation (50.9%), diabetes (37.1%), hypertension (64.6%), lung pathology (27.4%) and chronic kidney disease (43.4%). Heart failure with preserved ejection fraction (HFpEF) accounted for 30.9% and heart failure with reduced ejection fraction (HFrEF) accounted for 60.5% of patients. HFpEF patients were older than HFrEF patients (76.3 vs 70.5y, p=0.01). HFpEF were less likely to attend HFC (38.9% of all HFpEF patients), compared with 50.9% of all HFrEF patients attending HFC. Length of stay (LOS) for index admission was similar in both groups (4.1 vs 4.3 days in HFpEF and HFrEF respectively). There was no difference in median total admissions over 12 months between the groups. Patients attending HFC had numerically increased median LOS during index admission (6.0 vs 4.9 days, p=0.07). There was a trend towards increased severity of HFrEF patients attending HFC. The burden of HFpEF on inpatient services are similar to those of HFrEF, with no difference in median LOS or admission rate. Despite this, they are under-represented in HFC. Increased LOS in patients attending HFC likely represents referral bias of unwell patients with severe HFrEF.

889. Impact of an Infectious Disease Telehealth (IDt) Service on S. aureus Bacteremia (SAB) Outcomes in 15 Small Community Hospitals

BackgroundInfectious diseases (ID) consultation improves SAB readmission rates, compliance with care bundles and mortality. Small community hospitals (SCHs) (which comprise 70% of US hospitals) often lack access to on-site ID physicians. IDt is one way to overcome this barrier, but it is unknown if IDt provides similar clinical benefits to traditional ID consultation. Our study aims to evaluate the impact of IDt on patient outcomes at 15 SCHs (bed range: 16–146) within the Intermountain Healthcare system in Utah.MethodsBaseline demographics, Charlson Comorbidity Index (CCI), hospital length of stay (LOS), and mortality (in-hospital, 30- and 90-day) were collected using an electronic health record database and health department vital records on all patients with a positive S. aureus blood culture from January 1, 2009 through December 31, 2018. Data from January 2014 through Sep 2016 were excluded to avoid potential influence of a concurrent antimicrobial stewardship study. Starting in October 2016 an IDt program (staffed by an ID physician and pharmacist) provided consultation for SCH providers and patients using electronic consultation and encrypted two-way audiovisual communication.Statistical analyses were performed using Fisher’s exact test or χ 2 test for categorical variables and Mann–Whitney U test for nonparametric continuous data.ResultsIn total, 625 patients with SAB were identified: 127 (20%) received IDt and 498 (80%) did not (non-IDt). The two groups (IDt vs. non-IDt) were similar in median age (66 vs. 62 years; P = 0.76), percent male (62% vs. 58%; P = 0.35), and median baseline CCI (4 vs. 4; P = 0.54). There were no statistically significant differences in median LOS (5 vs. 5 days; P = 0.93) or in-hospital mortality (2% in both groups). The IDt group had a lower 30-day (9% vs. 15%; P = 0.049) and 90-day mortality (13% vs. 21%; P = 0.034).ConclusionIDt consultation was associated with a decrease in 30- and 90-day mortality for SCH SAB cases. Early transfer of critically ill patients might have affected LOS and in-hospital mortality. Post-discharge care factors might also contribute to 30- and 90-day mortality. While more work is needed to identify other factors associated with the effect of IDt on SAB, these data support the use of IDt to increase access to care and improve SAB outcomes in SCHs.Disclosures All Authors: No reported Disclosures.

Multimorbidity and healthcare resource utilization in Switzerland: a multicentre cohort study

BackgroundMultimorbidity is associated with higher healthcare resource utilization, but we lack data on the association of specific combinations of comorbidities with healthcare resource utilization. We aimed to identify the combinations of comorbidities associated with high healthcare resource utilization among multimorbid medical inpatients.MethodsWe performed a multicentre retrospective cohort study including 33,871 multimorbid (≥2 chronic diseases) medical inpatients discharged from three Swiss hospitals in 2010–2011. Healthcare resource utilization was measured as 30-day potentially avoidable readmission (PAR), prolonged length of stay (LOS) and difference in median LOS. We identified the combinations of chronic comorbidities associated with the highest healthcare resource utilization and quantified this association using regression techniques.ResultsThree-fourths of the combinations with the strongest association with PAR included chronic kidney disease. Acute and unspecified renal failure combined with solid malignancy was most strongly associated with PAR (OR 2.64, 95%CI 1.79;3.90). Miscellaneous mental health disorders combined with mood disorders was the most strongly associated with LOS (difference in median LOS: 17 days) and prolonged LOS (OR 10.77, 95%CI 8.38;13.84). The number of chronic diseases was strongly associated with prolonged LOS (OR 9.07, 95%CI 8.04;10.24 for ≥10 chronic diseases), and to a lesser extent with PAR (OR 2.16, 95%CI 1.75;2.65 for ≥10 chronic diseases).ConclusionsMultimorbidity appears to have a higher impact on LOS than on PAR. Combinations of comorbidities most strongly associated with healthcare utilization included kidney disorders for PAR, and mental health disorders for LOS.

66 Impact of High-Risk HEART Score on Readmission Rates, Observation Unit Utilization, and Length of Stay

The HEART Score, which assesses patient history, electrocardiogram (ECG), age, risk factors, and troponin level, is a clinical decision tool used in the emergency department (ED) evaluation of chest pain to calculate the six-week risk for potential major adverse cardiac events (MACE). Low-risk (HEART score ≤ 3) and high-risk (HEART score >3) groups have been shown to carry a 1.7% and 12-17% rate of a MACE within 6-weeks, respectively. The objective of this study was to evaluate differences in 60-day readmission rate, patient placement, and length of stay between low- and high-risk patient groups, according to HEART Score. In this single-institution, retrospective cohort study, ED patients seen during a five-month period were dichotomized into low-risk (HEART score ≤ 3) and high-risk (HEART score >3) groups. Each HEART score component (history, ECG, age, risk factors, and troponin) was given a sub-score between 0-2 based on clinical or diagnostic severity. Two physicians independently reviewed individual patient history and ECG, with adjudication performed by a third physician reviewer. The primary endpoint was 60-day readmission rate. Secondary endpoints included patient placement (discharge, observation, inpatient) and length of stay (LOS). A total of 558 ED patients (n=321 low risk; n=237 high risk) were included. Baseline demographics were well matched between groups. High-risk patients had a greater 60-day readmission rate compared to low-risk patients (18.7% vs. 25.9%, respectively; p=0.046). The high-risk cohort was more likely to be admitted to hospital observation units (40.9% vs. 24.6%; p<0.01) and inpatient settings (23.31% vs. 5.61%; p<0.01). In contrast, low-risk patients had greater discharge rates compared to those considered high-risk (69.5% vs. 35%; p<0.01). The difference in median LOS between low- and high-risk groups admitted to inpatient settings was not statistically significant (43.0 hours vs. 41.4 hours; p=0.97). However, median ED LOS was greater for low-risk patients (4.3 hours vs. 3.88 hours; p=0.01), and median observation unit LOS was greater for high-risk patients (23.6 hours vs. 19.3 hours; p=0.02). In calculating composite HEART score, high-risk patients were more likely to score greatest (2/2) on age and risk factor components (50.2% and 64.9%, respectively), compared to the ECG, history and troponin factors (13.78%, 21,1%, and 1.30%, respectively). High-risk HEART score was associated with greater 60-day readmission rates, and observation unit and inpatient admissions. This finding may reflect the clinical realities of treating more complex patient populations. The age and risk factor components were the most significant contributors to high-risk HEART scores, and may be weighed more heavily by physicians during clinical assessments of emergent chest pain.

Outcomes in video-assisted thoracoscopic surgery lobectomies: challenging preconceived notions

BackgroundMost thoracic surgical procedures in the United States are being performed by general surgeons (GSs) without any advanced training. With the recent approval of computed tomography screening for lung malignancy in high-risk populations, the number of thoracic oncologic resections is expected to rise. Previous literature has demonstrated consistently worsened outcomes for patients undergoing thoracic surgical procedure when done by nonthoracic fellowship–trained surgeons. Using the American College of Surgeons National Surgical Quality Improvement Project database, we examined short-term outcomes in patients undergoing video-assisted thoracoscopic surgery (VATS) lobectomy for malignancy. Materials and methodsData were obtained from the American College of Surgeons National Surgical Quality Improvement Project from 2010-2015. We identified patients who had an International Classification of Disease 9 diagnosis of lung cancer (162) who underwent VATS lobectomy (current procedural terminology 32663). We included only adults (≥18y) and elective cases. We excluded patients who had preoperative diagnosis of sepsis, contaminated wound class, or those patients with missing American Society of Anesthesiologists classification, morbid obesity, functional status, length of stay (LOS), or sex, and race information. We identified two groups by specialty: GS versus cardiothoracic (CT) surgeon. We then performed univariate analysis. We then performed propensity score analysis using a 1:3 ratio of general surgery patients to CT patients. Outcomes of interest included 30-d postoperative mortality, 30-d postoperative morbidity, and LOS. ResultsA total of 4105 patients were identified, 607 performed by GSs, 3508 performed by CT surgeons. The mean age for patients who underwent lobectomies by GSs was 68.6 versus 67.8 in the CT surgeon group (P < 0.05). The majority were female (58.09% GS versus 57.74% CT surgeon). There was a statistically significant difference in race between groups; patients were more likely to be African American in the CT surgeon group. Operative time was lower in the GS group as opposed to the CT surgeon group 179 min versus 196 (P < 0.01). Univariate analysis (mortality <0.1 CT surgeon and GS) and 1:3 propensity score matched analysis (0.08 GS% versus 0.08% CT surgeon) failed to demonstrate a significant difference in mortality. There was a statistically significant difference in median LOS between groups (6.2 GS versus 5.1 CT surgeon). Univariate and propensity matched analyses of pneumonia, sepsis, wound infection, deep vein thrombosis, transfusion requirement, myocardial infarction stroke, postoperative renal insufficiency, failure to wean, pulmonary embolism, reintubation, and deep organ space infection all failed to demonstrate a statistically significant difference between our groups of interest. Urinary tract infection was noted to be higher in the GS group operating room 2.29 as compared to the CT surgeon group (P value 0.02). ConclusionsIn this large observational study, we found that VATS lobectomies performed by GS compared to the matched CT surgeon cohort had shorter operative time, and there was no difference in major postoperative morbidity or mortality. However, LOS was higher and there was increased risk of urinary tract infection in the GS compared to matched CT surgeon cohort.

PCV61 - Inpatient Admission Length Of Stay And Costs Among Vascular Procedure Patients Receiving Perclose Proglide Suture-Mediated Closure Compared To Patients Receiving Surgical Cutdown For Closure Of Large Bore Arterial Access

To compare length of stay (LOS) and costs among patients undergoing procedures requiring closure of large bore arterial access using the Perclose ProGlide® Suture-Mediated Closure System (ProGlide) versus manual surgical cutdown (Cutdown) in a real-world setting. This retrospective study analyzed patients undergoing procedures that require arterial closure (Transcatheter Aortic Valve Replacement/Repair (TAVR), Endovascular Abdominal Aortic Aneurysm Repair (AAA), Thoracic Endovascular Aortic Repair (TAA), or Balloon Aortic Valvuloplasty (BAV)) from 2013 -2017 in the IBM Explorys Research Database. Patients receiving ProGlide during the procedure were directly matched and compared to patients treated by the same providers that were closed via manual surgical cutdown. The LOS for the procedure admission and procedure complications were assessed. Costs per day of admission were estimated using claims data from TAVR, AAA, TAA, or BAV procedure admissions in the Truven Health MarketScan Research Database. A generalized linear model was used to quantify the cost associated with differences in LOS. A total of 839 ProGlide patients and 839 matched Cutdown patients met the inclusion criteria. Proglide patients experienced lower rates of complications, including infection. Average LOS for ProGlide patients was nearly four days shorter than Cutdown patients (mean: 9.0 vs. 5.1 days; median: 7.0 vs. 3.0; p<0.001). While controlling for index procedure, age, and gender, LOS was a significant predictor of costs, with each day associated with a 4% increase in total cost for the hospitalization (p<0.05). Using the difference in median LOS, predicted total costs for the procedure admission were $20,599 lower among ProGlide patients compared to Cutdown. The use of Perclose ProGlide was associated with significantly shorter length of stay for the procedure admission compared to surgical Cutdown. Future analyses should focus on directly quantifying the cost-savings and patient benefit from lower complication rates and shortened LOS.