Population-split-based risk assessment model of venous thromboembolism in Chinese medical inpatients

Objective Inpatients with high risk of venous thromboembolism (VTE) usually face serious threats to their health and economic conditions. Many studies using machine learning (ML) models to predict VTE risk neglected an important statistical phenomenon, "fuzzy feature", and achieved inferior results. Considering the effect of "fuzzy feature", our study aims to develop a VTE risk assessment model suitable for Chinese medical inpatients. Materials and Methods Inpatients in the medical department of Peking Union Medical College Hospital (PUMCH) from January 2014 to June 2016 were collected. A new ML VTE risk assessment model was built through population splitting. First patients were classified into different groups based on values of VTE risk factors, then trustless groups were filtered out, and finally ML models were built on training data in unit of groups. Predictive performances of our method, five traditional ML models, and the Padua model were compared. Results The "fuzzy feature" was verified on the whole dataset. Compared with the Padua model, the proposed model showed higher sensitivities and specificities on training data, and higher specificities and similar sensitivities on test data. Standard deviations of predictive validity of five ML models were larger than the proposed model. Discussion The proposed model was the only one which showed advantages on both sensitivity and specificity over Padua model. Its robustness was better than traditional ML models. Conclusion This study built a population-split-based ML model of VTE for Chinese medical inpatients and it may help clinicians stratify VTE risk and guide prevention more efficiently.

Risk Assessing Medical Inpatients for Hospital-Acquired Venous Thrombosis: The Medical Inpatients Thrombosis and Hemostasis (MITH) Study

Introduction: Multiple regulatory agencies and professional societies recommend risk assessment of hospitalized medical patients for hospital-acquired (HA) venous thromboembolism (VTE) and provision of pharmacologic prophylaxis to those at risk. Extant risk assessment models (RAMs) include risk factors not knowable or difficult to assess at admission and often do not include risk factors reflecting illness acuity (such as laboratory studies and vital signs at admission). We developed a RAM for HA-VTE that reports absolute VTE risk, as opposed to arbitrary risk categories, using only objective risk factors measured within the first 24 hours of admission. Methods: The study setting was a combined academic and community 540-bed teaching hospital in northwest Vermont (The University of Vermont Medical Center). Using validated electronic health record (EHR) derived phenotypes (computable phenotypes), we captured all medical admissions between 2010-2019 and examined patient demographics, past medical history, and presenting vital and laboratory measures as potential risk factors for HA-VTE. As risk assessment should happen within 24 hours of admission, we only assessed risk factors knowable within this timeframe. Individuals with VTE at admission were excluded. Key outcome and risk factor definitions were validated using chart review. Bayesian logistic regression with a least absolute shrinkage and selection operator (LASSO) prior probability distribution was used to select risk factors for the model. Variables with a t-statistic ≥1.5 or ≤-1.5 were included in the final model. Full or prophylactic anticoagulation use was adjusted for in the final model. Model performance was assessed using bootstrap resampling to estimate area under the receiver operating characteristic (AUC) curve and calibration slope with 95% confidence interval (CI). Results: There were 62,468 medical admissions in the study period with 219 HA-VTE events. Chart review demonstrated the positive predictive value of our HA-VTE computable phenotype to be 84% and the negative predictive value 99%. Mean age was 65 years and 51% were male. Comorbid conditions were common in this hospitalized population, including active cancer (29%), congestive heart failure (25%), diabetes (27%), hypertension (59%), and prior myocardial infarction (13%). Seven risk factors met the criteria for inclusion in the final model: prior history of VTE (OR 2.7; 95% CI 1.8, 3.8), red cell distribution width ≥14.7% (OR 1.6; 95% CI 1.2, 2.2), creatinine ≥2.0 mg/dL or on dialysis (OR 2.0; 95% CI 1.4, 2.8), serum sodium <136 MEq/L (OR 1.5; 95% CI 1.1, 2.1), active cancer (OR 1.4; 95% CI 1.1, 2.0), malnutrition based on prior reported weight loss (OR 2.1; 95% CI 1.3, 3.3), and low hemoglobin (<13.6 g/dL in men, <12.1 g/dL in women; OR 1.5; 95% CI 1.0, 2.1). The unadjusted AUC of the RAM was 0.73 with an unadjusted calibration slope 1.09 (Figure 1). The optimism-adjusted AUC was 0.68 (95% CI 0.64, 0.71) and the optimism-adjusted calibration slope was 0.87 (95% CI: 0.72, 1.03). Discussion: We developed and internally validated a RAM for HA-VTE during medical hospitalization which incorporates simple, objective risk factors knowable within the first 24 hours of admission. Unlike most prior RAMs, this model also incorporates risk factors reflecting illness severity such as laboratory results. The RAM has good fit and calibration and will be moved forward to external validation. Future applications include incorporating the RAM into hospital admission workflows and assessing VTE prophylaxis rates and the incidence of HA-VTE and HA-bleeding. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Prevalence of adrenal insufficiency among patients with euvolemic hyponatremia

Background: The diagnosis of SIAD requires the exclusion of secondary adrenal insufficiency (AI) among patients with euvolemic hyponatremia (EuVHNa). Studies have suggested about 2.7% to 3.8% of unselected patients presenting to the emergency room with EuVHNa have undiagnosed AI and it is as high as 15% among patients admitted to specialized units for evaluation of hyponatremia. Objective: To study the prevalence of AI among inpatients with EuVHNa in a general medical ward setting Methods: This was a prospective, single centre observational study conducted among general medical inpatients with EuVHNa, defined as patients with a serum Na <135 mmol/L, clinical euvolemia and urine spot sodium >30mmol/L. Additionally patients with recent vomiting, current renal failure, diuretic use and those with uncontrolled hyperglycemia were excluded. Adrenal functions were assessed by a modified ACTH stimulation test called Acton Prolongatum™ stimulation test (APST). A cut off cortisol value of <18mg/dl after 60 minutes of ACTH injection was used to diagnose AI. Results: One hundred and forty-one (141) patients were included and underwent an APST. APST suggested 20/141 (14.2%) had undiagnosed AI. The most common cause of AI (9/20) was secondary AI because of the use of steroids including inhaled steroids and indigenous medicines contaminated with steroids. In 5 (3.5%) patients hypopituitarism was newly diagnosed. Despite primary adrenal insufficiency not commonly presenting as EuVHNa, 2/20 patients had primary adrenal insufficiency. Conclusions: AI is much commoner in our country among inpatients with EuVHNa primarily driven by exogenous steroid use and undiagnosed hypopituitarism.

Do not attempt cardiopulmonary resuscitation (DNACPR) decisions for older medical inpatients: a cohort study

ObjectivesA decision not to attempt cardiopulmonary resuscitation in the event of cardiorespiratory arrest requires a discussion between the doctor and the patient and/or their relatives. We aimed to determine how many older patients admitted to acute medical wards had a pre-existing 'do not attempt cardiopulmonary resuscitation' (DNACPR) decision, how many had one recorded on the ward and how many of those who died had a DNACPR decision in place.MethodsA prospective cohort study, using data from medical records, of 481 consecutive patients aged ≥65 years admitted to the six acute medical wards of the John Radcliffe Hospital, Oxford.Results105/481 (22%) had a DNACPR decision at ward admission, 30 of which had been made in the emergency unit. A further 45 decisions were recorded on the ward, mostly after discussion with relatives. Of the 37 patients who died, 36 had a DNACPR decision. For the 20 deceased patients whose DNACPR decision was recorded during their admission, the median time from documentation to death was 4 days with 7/20 (35%) recorded the day before death.ConclusionsOlder patients with multimorbidity need the opportunity to discuss the role of CPR earlier in their care and preferably before acute hospital admission.

Comparing accuracy of bedside ultrasound examination with physical examination for detection of pleural effusion

Background In detecting pleural effusion, bedside ultrasound (US) has been shown to be more accurate than auscultation. However, US has not been previously compared to the comprehensive physical examination. This study seeks to compare the accuracy of physical examination with bedside US in detecting pleural effusion. Methods This study included a convenience sample of 34 medical inpatients from Calgary, Canada and Spokane, USA, with chest imaging performed within 24 h of recruitment. Imaging results served as the reference standard for pleural effusion. All patients underwent a comprehensive lung physical examination and a bedside US examination by two researchers blinded to the imaging results. Results Physical examination was less accurate than US (sensitivity of 44.0% [95% confidence interval (CI) 30.0–58.8%], specificity 88.9% (95% CI 65.3–98.6%), positive likelihood (LR) 3.96 (95% CI 1.03–15.18), negative LR 0.63 (95% CI 0.47–0.85) for physical examination; sensitivity 98% (95% CI 89.4–100%), specificity 94.4% (95% CI 72.7–99.9%), positive LR 17.6 (95% CI 2.6–118.6), negative LR 0.02 (95% CI 0.00–0.15) for US). The percentage of examinations rated with a confidence level of 4 or higher (out of 5) was higher for US (85% of the seated US examination and 94% of the supine US examination, compared to 35% of the PE, P < 0.001), and took less time to perform (P < 0.0001). Conclusions US examination for pleural effusion was more accurate than the physical examination, conferred higher confidence, and required less time to complete.

219 Have New National Guidelines Changed Practice in Referrals with Suspected Ureteric Colic?

Aim To assess how many patients had a non-contrast CT as first-line investigation for suspected renal colic, and how this was affected by the release of the 2019 NICE guidelines. A secondary aim was to assess the proportion of patients who had CT KUBs that demonstrated a ureteric calculus. Method CT KUB scans performed over two separate 3 month periods were identified. These reflected periods of time before and after the new NICE guidelines. Electronic records were used to assess if the patients had an ultrasound performed as the initial diagnostic investigation, instead of a CT scan. The results of the scans were reviewed to identify if a ureteric or renal calculus had been positively identified. Results In the period before the new guidelines, 61 patients were scanned. 4 had an ultrasound to assess for a stone prior to a CT. All of these patients were medical inpatients. None of the ultrasounds diagnosed a stone. 22/61 patients had CT-proven stones (36%) In the period after the new guidelines, 79 patients were scanned. 12 had an ultrasound to assess for a stone prior to a CT. 8 were medical patients, 3 were surgical and 1 was gynaecological. 1 ultrasound diagnosed a stone. 28/79 patients had CT-proven stones (35.4%) Conclusions The release of new guidelines did not improved compliance with suggested imaging pathways. This clearly demonstrates an area for improvement. It is also worth noting that only 1/3 of referrals with suspected renal colic did actually have a stone, which has implications for specialty referral pathways.

Disease-Related Malnutrition and Sarcopenia Predict Worse Outcome in Medical Inpatients: A Cohort Study

(1) Background: Both sarcopenia and disease-related malnutrition (DRM) are unfortunately underdiagnosed and undertreated in our Western hospitals, which could lead to worse clinical outcomes. Our objectives included to determine the impact of low muscle mass (MM) and strength, and also DRM and sarcopenia, on clinical outcomes (length of stay, death, readmissions at three months, and quality of life). (2) Methodology: Prospective cohort study in medical inpatients. On admission, MM and hand grip strength (HGS) were assessed. The Global Leadership Initiative on Malnutrition (GLIM) criteria were used to diagnose DRM and EWGSOP2 for sarcopenia. Assessment was repeated after one week and at discharge. Quality of life (EuroQoL-5D), length of stay (LoS), readmissions and mortality are reported. (3) Results: Two hundred medical inpatients, median 76.0 years-old and 68% with high comorbidity. 27.5% met GLIM criteria and 33% sarcopenia on admission, increasing to 38.1% and 52.3% on discharge. Both DRM and sarcopenia were associated with worse QoL. 6.5% died and 32% readmission in 3 months. The odds ratio (OR) of mortality for DRM was 4.36 and for sarcopenia 8.16. Readmissions were significantly associated with sarcopenia (OR = 2.25) but not with DRM. A higher HGS, but not MM, was related to better QoL, less readmissions (OR = 0.947) and lower mortality (OR = 0.848) after adjusting for age, sex, and comorbidity. (4) Conclusions: In medical inpatients, mostly polymorbid, both DRM but specially sarcopenia are associated with poorer quality of life, more readmissions, and higher mortality. Low HGS proved to be a stronger predictor of worse outcomes than MM.