Search

Background Emergency Department (ED) overcrowding is an emerging risk to patient safety. This study aims to assess and compare the predictive ability of machine learning (ML) models for predicting frequent ED users. Method Korean Health Panel data from 2008 to 2015 were used for this study. Individuals with four or more visits per year were considered frequent ED users. Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM) as well as two ensemble models, namely Bagging and Voting, were trained and tested to examine their predictive performance. Results The ML classification algorithms identified frequent ED users with high precision (90%–98%) and sensitivity (87%–91%), whereas LR showed fair precision (65%) and sensitivity (67%). The ML algorithms showed a high area under the curve (AUC) values from 89% for SVM to 96% for Random Forest, while LR showed the lowest AUC (65%). The classification error varied among algorithms; LR had the highest classification error (24.07%) while RF had the least (3.8%). Conclusions Results show that ML classification algorithms are robust techniques to predict frequent ED users, and the variables in administrative health panels are reliable indicators for this purpose.

Background Conventional sunitinib dosing in metastatic renal cell carcinoma (mRCC) administers 50 mg daily on a 4 weeks on/2 weeks off (4/2) schedule. Not all patients tolerate this regimen and many undergo modifications to schedule, dose, or both. Material and Methods All patients with mRCC treated with sunitinib by the Saskatchewan Cancer Agency between January 1, 2006, and January 1, 2013, were included. Regimens were categorized as standard intermittent dosing (SID), modified intermittent schedule (MIS), modified intermittent dosing (MID), combination of modified schedule and dosing (MSD), or continuous dosing (CD). The primary objective was to compare overall survival (OS) between regimens. Secondary outcomes included progression-free survival (PFS), discontinuation due to adverse effects (AE), and medication cost. Results Among 161 patients, 18.0%, 51.6%, and 30.4% had favorable, intermediate, and poor Heng risk prognoses, respectively. A total of 140 (87.0%) received sunitinib as first-line therapy. MID was associated with longer OS compared with SID (estimated median 28.4 vs. 11.2 months). PFS was longer for MID, MSD, and CD compared with SID (estimated median 12.0, 9.0, and 8.0 months vs. 3.0 months, respectively). Adjustment for potential confounders did not negate these associations. SID also had higher average monthly drug costs than MIS, MID, and MSD. Overall discontinuation rate due to AE was high (24%). Conclusion An adjusted-dose sunitinib regimen is associated with improved OS and PFS over SID, with lower costs. The development of toxicities requiring dose reductions serves as a predictive biomarker for better outcomes.

Postoperative vomiting (POV) in children is frequent. Dextrose-containing intravenous fluids in the perioperative period have shown improvement of POV in adults. Similar studies have not been done in children.The primary purpose was to study the efficacy of intraoperative intravenous dextrose for antiemetic prophylaxis in children undergoing ambulatory surgery.A non-inferiority randomized clinical trial of healthy children (three to nine years old) undergoing ambulatory dental surgery was conducted. The control group received dexamethasone (0.15 mg·kgData from 290 patients were analyzed. Demographics and intraoperative anesthetic management were similar between groups. Vomiting in the PACU occurred in 7.6% and 3.5% of the dextrose and ondansetron groups, respectively, with a risk difference of 4.2% (95% confidence interval [CI], -1.0 to 9.5). Given that the upper limit of the 95% CI exceeded our non-inferiority margin, non-inferiority of dextrose compared with ondansetron was not shown.These results do not support the use of intravenous dextrose as a satisfactory alternative to ondansetron to prevent POV in ambulatory pediatric dental surgery patients.www.clinicaltrials.gov (NCT01912807); registered 18 July 2013.RéSUMé: CONTEXTE: Les vomissements postopératoires (VPO) sont fréquents chez l’enfant. Il a été démontré qu’en période périopératoire, les solutés intraveineux contenant du dextrose entraînaient une diminution des VPO chez l’adulte, mais des études similaires n’ont pas été réalisées auprès de populations pédiatriques.L’objectif principal était d’évaluer l’efficacité du dextrose intraveineux peropératoire en tant que prophylaxie antiémétique chez les enfants subissant une chirurgie ambulatoire. MéTHODE: Une étude clinique randomisée de non-infériorité a été réalisée auprès d’enfants en bonne santé (de trois à neuf ans) devant subir une chirurgie dentaire en ambulatoire. Le groupe témoin a reçu de la dexaméthasone (0,15 mg·kgCes résultats n’appuient pas l’utilisation de dextrose intraveineux en tant qu’alternative à l’ondansétron afin de prévenir les VPO chez les patients pédiatriques de chirurgie dentaire ambulatoire. ENREGISTREMENT DE L’éTUDE: www.gov (NCT01912807); enregistrée le 18 juillet 2013.

A pilot study of biomarker-driven targeted therapy in patients with platinum-resistant recurrent ovarian cancer has been started in Korea. Archival tumor samples were tested for HRD and PD-L1 status. Treatment arms will be allocated according to the test results. For HRD+ patients, we tested the synergistic effects of olaparib and other agents; treatment arms will randomly be allocated. (Arm 1: olaparib and cediranib; Arm 2: olaparib and durvalumab). For HRD- patients, we tested the role of biomarker-driven immunotherapy according to PD-L1 expression (Arm 3: durvalumab and chemotherapy in patients with high PD-L1 expression; Arm 4: durvalumab, tremelimumab, and chemotherapy in patients with low PD-L1 expression). Sixty-eight patients will be included from three Korean institutions within 1 year. The primary endpoint is the response rate according to RECIST 1.1 (6 months after treatment initiation). This trial has been registered with clinicaltrials.gov, and the registration number is NCT03699449.

Background: Despite the need for community services to support the intensive levels of care provided by dementia family caregivers, findings consistently indicate low rates of service utilization. This study aims to explore the extent of service use and examine the factors that may influence the use of services in a sample of family caregivers of Persons With Dementia (PWD).Methods: Two hundred and seventy-two family caregivers completed a questionnaire assessing caregiver burden, knowledge and use of services, dementia severity, frequency of problem behaviors and socio-demographic characteristics.Results: The rate of service use was modest with 39.9% using any of the available services. Reasons for non-use included lack of time (26.2%), perceived lack of need (18.5%), and presence of domestic help (21%). Separate regression models were evaluated to identify factors associated with caregiver service use and PWD service-use. Caregiver employment status (OR = 0.53, CI = 0.30–0.94) and knowledge of available services (OR = 4.22, CI = 2.23–7.98) contributed significantly to the model distinguishing caregiver service users from non-users. Knowledge of available services (OR = 2.14, CI = 1.23–3.71) was the only significant predictor of service use targeted at the PWD.Conclusions: Given that knowledge significantly increased the odds of service use, strategies should be designed to raise awareness about the availability of dementia support services in the community and provide a deeper understanding about the value of such services. Future research should employ larger sample sizes and explore factors not measured in this study that might potentially better explain use of services.

This cross-sectional study used the stress and coping paradigm to examine the factors associated with negative and positive adjustment outcomes among Asian family caregivers of persons with dementia (PWD) in Singapore. One hundred seven family caregivers completed measures assessing patient illness characteristics, general coping styles of caregivers, specific dementia management strategies, religion and spirituality, and caregiver adjustment outcomes of burden and gain. Multiple regressions revealed that behavioral problems in the PWD, dementia severity, and the use of behavioral disengagement and criticism as coping strategies were significant predictors of burden accounting for 48% of the explained variance [F(4,99)=23.12, P < 0.001]. The only significant predictor of gain was the use of encouragement as a specific dementia management strategy, explaining 18% of variance [F(3,102) = 7.39, P < 0.001]. Religion and spirituality predicted gain indirectly through the use of encouragement. Coping strategies had an independent effect on caregiver outcomes above and beyond PWD illness characteristics and caregiver characteristics. Hence, caregiver interventions should target coping strategies to improve outcomes of caregiving for dementia. Findings also support the need to examine religion and spirituality in future studies of caregiver adjustment outcomes and to explore the factors not measured in this study that might explain gain.

To describe and validate a new scale, Gain in Alzheimer care Instrument (GAIN), and to measure gains in dementia caregiving.Items in GAIN were derived through a qualitative study of family carers of persons with dementia (PWD). Content validity of GAIN was established by a team of dementia care experts (geriatrician, nurse, and social worker) and through focus group discussion with 15 family carers. Consecutive carers of PWD were administered a questionnaire containing GAIN and other measures.Ambulatory dementia clinic of a tertiary hospital and the local Alzheimer's Association.Family carers of PWD.The psychometric properties of GAIN were examined, in particular, exploratory factor analysis and construct validity through correlation with extant measures; Positive Aspects of Caregiving (PAC), Dementia Management Strategies Scale (DMSS), and Zarit Burden Interview (ZBI).Two hundred thirty-eight English-speaking carers completed the survey. Mean age of carers was 50.1 year (standard deviation [SD] 10.1) and females (68.1%) and child carers (81.3%) comprised the majority. Mean GAIN score was 30.5 (SD 6.5). Internal consistency of GAIN by Cronbach's alpha was 0.89 and test-retest reliability (2 weeks) by Intraclass Correlation Coefficient was 0.70. GAIN correlated strongly with PAC (r = 0.68, df = 236, p0.0001), moderately with DMSS (encouragement) (r = 0.35, df = 234, p0.0001) and DMSS (active management) (r = 0.42, df = 235, p0.0001), and modestly with DMSS (criticism) (r = -0.14, df = 236, p = 0.03) and ZBI (r = -0.15, df = 236, p = 0.02). Principal component analysis revealed a single component with an Eigen value1 at 5.28, accounting for 52.8% of the variance.GAIN is a psychometrically reliable and valid instrument for measuring gains in dementia caregiving. It adds an important dimension to carer assessment in dementia.

Objective To examine diagnostic agreement between Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V) Neurocognitive Disorders (NCDs) criteria and DSM, Fourth Edition (DSM-IV) criteria for dementia and International Working Group (IWG) criteria for mild cognitive impairment (MCI) and DSM-V's impact on diagnostic classifications of NCDs. The authors further examined clinical factors for discrepancy in diagnostic classifications between the different operational definitions. Methods Using a cross-sectional study in tertiary memory clinic, the authors studied consecutive new patients aged 55 years or older who presented with cognitive symptoms. Dementia severity was scored based on the Clinical Dementia Rating scale (CDR). All patients completed neuropsychological evaluation. Agreement in diagnostic classifications between DSM-IV/IWG and DSM-V was examined using the kappa test and AC1 statistic, with multinomial logistic regression for factors contributing to MCI reclassification as major NCDs as opposed to diagnostically concordant MCI and dementia groups. Results Of 234 patients studied, 166 patients achieved concordant diagnostic classifications, with overall kappa of 0.41. Eighty-six patients (36.7%) were diagnosed with MCI and 131 (56.0%) with DSM-IV–defined dementia. With DSM-V, 40 patients (17.1%) were classified as mild NCDs and 183 (78.2%) as major NCDs, representing a 39.7% increase in frequency of dementia diagnoses. CDR sum-of-boxes score contributed independently to differentiation of MCI patients reclassified as mild versus major NCDs (OR: 0.01; 95% CI: 0–0.09). CDR sum-of-boxes score (OR: 5.18; 95% CI: 2.04–13.15), performance in amnestic (OR: 0.14; 95% CI: 0.06–0.34) and language (Boston naming: OR: 0.52; 95% CI: 0.29–0.94) tests, were independent determinants of diagnostically concordant dementia diagnosis. Conclusion The authors observed moderate agreement between the different operational definitions and a 40% increase in dementia diagnoses with operationalization of the DSM-V criteria.

Objectives: Chemotherapy increases the risk of thromboembolism in patients with cancer. Although thrombocytopenia is a known side effect of chemotherapy, reactive thrombocytosis related to chemotherapy is uncommonly reported. The present study aimed to determine the incidence of gemcitabine-related thrombocytosis and the associated risk of thromboembolism. Methods: Medical records of 250 consecutive patients with a malignant disease who received gemcitabine-based therapy were reviewed. A multivariate analysis was done to determine factors associated with thromboembolism. Results: A total of 220 eligible patients with a median age of 63 years (range 26–83) were identified. Of these 220 patients, 95% had advanced malignancy and 59% had received prior chemotherapy. A total of 69% of patients received a platinum combination. In all, 46% patients experienced thrombocytosis following chemotherapy, with a median platelet count of 632 × 109/l (range 457–1,385). Twenty-three of the 220 patients experienced a vascular event within 6 weeks of treatment. Eleven patients with thrombocytosis experienced a vascular event compared with 10 patients without thrombocytosis (not significant). On multivariate analysis, leukocytosis (odds ratio 5.8, 95% confidence interval 2.1–15.8) and comorbid illnesses (odds ratio 4.1, 95% confidence interval 1.4–12.6) were correlated with thromboembolism. Conclusions: Although gemcitabine-based therapy has been associated with an increased incidence of thrombocytosis, it does not increase the risk of thromboembolism in cancer patients. Leukocytosis and comorbid illnesses do increase the risk of thromboembolism.

Introduction: This study aims to validate the Zarit Burden Interview as an instrument to measure the level of burden experienced by caregivers of patients with dementia (PWD) in Singapore. Materials and Methods: Adult family caregivers of PWD were recruited from the ambulatory dementia clinic of a tertiary hospital and the Alzheimer’s Disease Association. All subjects completed a battery of questionnaires which consisted of demographic questions and the following instruments: the Zarit Burden Interview (ZBI), Burden Assessment Scale (BAS), General Health Questionnaire (GHQ-28), Dementia Management Strategies Scale (DMSS), and the Revised Memory and Behaviour Problems Checklist (RMBPC). A subgroup of subjects also completed the ZBI for the second time 2 weeks after the fi rst survey. Results: A total of 238 subjects completed the survey. As hypothesised, the Zarit burden score was strongly correlated with BAS, GHQ-28, DMSS, and RMBPC scores (Pearson’s correlation coefficient: 0.53 to 0.73); caregivers who undertook the major role in caregiving, had spent >1 year in caregiving, or experienced financial problems had higher Zarit burden scores than those who were not main carers, with ≤1 year of caregiving, or reported no/minimal financial problems, respectively. The Cronbach’s alpha value for the ZBI items was 0.93; the intra-class correlation coefficient for the test-retest reliability of the Zarit burden score was 0.89 (n = 149). Conclusion: The results in this study demonstrated that the Zarit Burden Interview is a valid and reliable instrument for measuring the burden of caregivers of PWD in Singapore. Key words: Caregiver, Dementia, Scale, Validation