Predicting In-Hospital Mortality at Admission to the Medical Ward: A Big-Data Machine Learning Model

BackgroundGeneral medical wards admit high-risk patients. Artificial intelligence algorithms can use big data for developing models to assess patients’ risk stratification. The aim of this study was to develop a mortality prediction machine learning model using data available at the time of admission to the medical ward.MethodsWe included consecutive patients (ages 18-100) admitted to medical wards at a single medical center (January 1, 2013-December 31, 2018). We constructed a machine learning model using patient characteristics, comorbidities, laboratory tests, and patients’ emergency department (ED) management. The model was trained on data from the years 2013 to 2017 and validated on data from the year 2018. The area under the curve (AUC) for mortality prediction was used as an outcome metric. Youden index was used to find an optimal sensitivity-specificity cutoff point.ResultsOf the 118,262 patients admitted to the medical ward, 6311 died (5.3%). The single variables with the highest AUCs were medications administered in the ED (AUC = 0.74), ED diagnosis (AUC = 0.74), and albumin (AUC = 0.73). The machine learning model yielded an AUC of 0.924 (95% confidence interval [CI]: 0.917-0.930). For Youden index, a sensitivity of 0.88 (95% CI: 0.86-0.89) and specificity of 0.83 (95% CI: 0.83–0.83) were observed. This corresponds to a false-positive rate of 1:5.9 and negative predictive value of 0.99.ConclusionA machine learning model outperforms single variables predictions of in-hospital mortality at the time of admission to the medical ward. Such a decision support tool has the potential to augment clinical decision-making regarding level of care needed for admitted patients.     

Elevated ANA titers in patients with severely abnormal gastrointestinal motility

We studied a group of six patients with clinical, radiological, and/or manometric features of severely abnormal gastrointestinal motility. Symptoms suggestive of esophageal, small bowel, or colonic involvement were present from 1 1/2 to 40 years. All patients had elevated antinuclear antibody (ANA) titers. None had clinical or radiographic features suggestive of progressive systemic sclerosis or other connective tissue diseases. Two patients had pathologic examinations of intestinal specimens, and these did not show changes suggestive of progressive systemic sclerosis. We conclude that patients with severe gastrointestinal motility disorders can have elevated ANA titers without features of progressive systemic sclerosis or other connective tissue diseases.This study was supported by the National Institutes of Health grant AM 25965 and grant RR 59 from the General Research Centers Program, Division of Research Resources, National Institutes of Health.     

A case report of an intracaval extrathoracic solitary fibrous tumour

Solitary fibrous tumours are infrequent neoplasms based in the pleura that are predominantly benign with malignant pathology and behaviour described in 10–36% of cases. Extrathoracic solitary fibrous tumours (ESFTs) have been considered separately to their intrathoracic counterparts and comprise a third of all solitary fibrous tumours. The extrathoracic location was identified as an adverse prognostic factor for local recurrence but not for metastatic disease. So far, there have not been any reports of solitary fibrous tumours demonstrating caval infiltration. We present a case of a benign ESFT infiltrating into the perirenal inferior vena cava. Together with extrauterine leiomyomas, ESFTs should also be considered as a differential diagnosis for the rare benign lesions invading the inferior vena cava.     

Promoting head CT exams in the emergency department triage using a machine learning model

Neuroradiology - In this study, we aimed to develop a novel prediction model to identify patients in need of a non-contrast head CT exam during emergency department (ED) triage. We collected data...     

Can laboratory evaluation differentiate between coronavirus disease-2019, influenza,and respiratory syncytial virus infections? A retrospective cohort study

AimTo identify clinical and laboratory parameters that can assist in the differential diagnosis of coronavirus disease 2019 (COVID-19), influenza, and respiratory syncytial virus (RSV) infections.MethodsIn this retrospective cohort study, we obtained basic demographics and laboratory data from all 685 hospitalized patients confirmed with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), influenza virus, or RSV from 2018 to 2020. A multiple logistic regression was employed to investigate the relationship between COVID-19 and laboratory parameters.ResultsSARS-CoV-2 patients were significantly younger than RSV (P = 0.001) and influenza virus (P = 0.022) patients. SARS-CoV-2 patients also displayed a significant male predominance over influenza virus patients (P = 0.047). They also had significantly lower white blood cell count (median 6.3 × 10⁶ cells/μ) compared with influenza virus (P < 0.001) and RSV (P = 0.001) patients. Differences were also observed in other laboratory values but were insignificant in a multivariate analysis.ConclusionsMale sex, younger age, and low white blood cell count can assist in the diagnosis of COVID-19 over other viral infections. However, the differences between the groups were not substantial enough and would probably not suffice to distinguish between the viral illnesses in the emergency department.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an RNA virus causing coronavirus disease 2019 (COVID-19). First identified in the Chinese province of Hubei in late 2019, COVID-19 was declared a global pandemic by the World Health Organization in March 2020 (1).As of July 2021, there were more than 180 million confirmed COVID-19 cases and more than four million patients who died due to the disease complications (2). Moreover, the disease caused a substantial economic and social burden (3), and affected health care quality (4-7).The diagnosis of COVID-19 is currently determined primarily by molecular methods and antigen tests (8,9). Radiographic diagnosis is possible as well (10,11). This practice often consumes valuable time and expensive equipment (12). There is a growing need to accelerate the diagnostic process by enabling point-of care diagnosis in various ambulatory settings, while keeping it accurate to ensure the necessary precautionary measures (13).The clinical presentation of SARS-CoV-2 infection resembles that of other respiratory viruses, with predominant symptoms of fever, cough, fatigue, and dyspnea (14-17). Hematological abnormalities, including leukopenia, lymphopenia, and thrombocytopenia, are common among COVID-19 patients, as well as elevated levels of C-reactive protein (CRP), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), and ferritin (14,15,18-21). Some of these inflammatory markers correlated with disease severity and mortality (22,23).The influenza season of 2021 in the Northern hemisphere was relatively weak in contrast with predictions. Low to zero rates of influenza were detected in several countries. This was attributed to social distancing, masks wearing, and a reduced number of air travelers (24). Despite a growing number of vaccinated individuals (25), the emergence of new SARS-CoV-2 variants suggest that COVID-19 is here to stay. Seasonal viruses such as influenza virus and respiratory syncytial virus (RSV) could rebound in the following winter, with the loosening of restrictions.Differentiating between COVID-19 and other respiratory viral illnesses on clinical grounds alone can be very challenging. These viral infections share similarities in the transmission route and symptoms (26-28). Several small studies attempted to delineate the differences in the clinical presentation of SARS-CoV-2 and influenza infections (29-31). In this study, we aimed to identify demographic and laboratory parameters that can assist in the early differentiation between SARS-CoV-2, influenza, and RSV infections in the emergency department.     

Twenty-five-year experience with ventricular septal defect in infants and children

Three hundred eighty-one children with isolated VSD were catheterized from 1960 through 1985, 228 of whom were first catheterized in the first year of life. Defect size was defined by the right-to-left ventricular systolic pressure ratio. Patients with small VSD had a good outcome, with no pulmonary vascular disease. Seventy-five percent of patients with moderate VSDs did not require surgery, and 58% had normal pulmonary artery pressures documented on a second catheterization. Of those patients with large VSDs who underwent serial catheterizations in the first year of life before surgery, 62% had partial closure, and pulmonary vascular resistance rose in 21%. None developed pulmonary vascular disease after surgery. Fewer large-restrictive than nonrestrictive VSDs required surgery in infancy (12% vs 51%; p less than 0.005), and more closed enough to never require surgery (62% vs 27%; p less than 0.005, chi square). Surgery is rarely necessary for patients with small and moderate-sized VSDs in the first year and is usually never necessary because of the expected decrease in size. Few patients with large-restrictive VSDs have an increase in pulmonary vascular resistance, many defects decrease in size, and most patients never need surgery. Patients who have a nonrestrictive VSD have an excellent prognosis with appropriate management.     

Risk factors for positive postpartum depression screen in women with private health insurance and access to care

Abstract

Objective: To determine risk factors for a positive postpartum depression screen among women with private health insurance and 24/7 access to care.

Study design: Retrospective cohort study of all patients delivered by a single MFM practice from April 2015 to September 2016. All patients had private health insurance and 24/7 access to care. All patients were scheduled to undergo the Edinburgh Postnatal Depression Scale (EPDS) at their 6-week postpartum visit and a positive screen was defined as a score of 10 or higher, or a score greater than zero on question 10 (thoughts of selfharm). Using logistic regression, risk factors for postpartum depression were compared between women with and without a positive screen.

Results: Of the 1237 patients delivered, 1113 (90%) were screened with the EPDS. 81 patients (7.3, 95%CI 5.9–9.0%) of those tested had a positive screen. On regression analysis, risk factors associated with a positive screen were nulliparity (aOR 1.8, 95%CI 1.1, 2.9), cesarean delivery (aOR 1.7, 95%CI 1.1, 2.8), non-White race (aOR 2.0, 95%CI 1.1, 3.5), and a history of depression or anxiety (aOR 4.6, 95%CI 2.6, 8.1). Among the 100 women with a history of depression or anxiety, selective serotonin reuptake inhibitor (SSRI) use in the postpartum period was not associated with a reduced risk of a positive screen (25.5% in those taking an SSRI versus 18.4% of those not taking an SSRI, p?=?.39).

Conclusions: Among women with private health insurance and access to care, the incidence of a positive screen for postpartum depression is approximately 7%. The use of an SSRI did not eliminate this risk. All women should be screened for postpartum depression.     

An Analysis of Teacher Investment in the Context of a Family–School Intervention for Children with ADHD

Intervention researchers have often failed to assess treatment integrity; when integrity is examined, the focus is typically on whether the steps of intervention have been applied and not on quality of implementation. In the few studies that have investigated intervention quality, the emphasis has been on how intervention is delivered and not how it is received or the degree of participant engagement. This study was designed to examine participant engagement, specifically teacher investment, in the context of family interventions for children with attention-deficit/hyperactivity disorder (ADHD) that were linked with the school. The interventions included Family–School Success (FSS), a program that engages families and schools in a problem solving partnership, and Coping with ADHD through Relationships and Education (CARE), a program designed to provide education and support to families. Participants included the families and teachers of children in grades 2 through 6; 45 sets of families and teachers participated in FSS and 48 were in CARE. The Teacher Investment Questionnaire (TIQ) was designed to assess teacher engagement in intervention. The TIQ is a rating scale for clinicians to be completed during and after treatment. The findings provided support for the reliability and validity of the TIQ. Level of teacher involvement was demonstrated to decline with advancing grade level for FSS, but not CARE. Parent ratings of the quality of the family–school relationship as well as the level of teacher support for homework, assessed at baseline, were shown to be significantly related to clinician ratings of teacher investment post-intervention. The findings highlight the importance of assessing participant engagement in intervention. Strategies to refine the assessment of teacher investment were discussed.