Search

Abstract Background Waste and fraud are important problems for health insurers to deal with. With the advent of big data, these insurers are looking more and more towards data mining and machine learning methods to help in detecting waste and fraud. However, labeled data is costly and difficult to acquire as it requires expert investigators and known care providers with atypical behavior. Methods In this work we show how recent advances in machine learning can be used to set up a workflow that can aid investigators in discovering practitioners or groups of practitioners with unusual resource use in order to more efficiently combat waste and fraud. We combine three different techniques, which have not been used in the context of healthcare insurance anomaly detection: categorical embeddings to deal with high-cardinality categorical variables, state-of-the-art unsupervised anomaly detection techniques to detect anomalies and Shapley additive explanations (SHAP) to explain the model output. Results The method has been evaluated on providers with a known anomalous profile and with the help of experts of the largest health insurance fund in Belgium. The quantitative experiments show that categorical embeddings offer a significant improvement compared to standard methods and that the state-of-the-art unsupervised anomaly detection techniques generally show an improvement over traditional methods. In a practical setting, the proposed workflow with SHAP was able to detect a previously unknown, anomalous trend among general practitioners. Conclusions The proposed workflow is able to detect known care providers with atypical behaviour and helps expert investigators in making informed decisions concerning possible fraud or overconsumption in the health insurance field.

IntroductionEquipoise, generally defined as uncertainty about the relative effects of the treatments being compared in a trial, is frequently referenced as an ethical standard for the conduct of randomized clinical trials. However, it seems to be defined in several different ways and may be used differently by different individuals. We explored how clinical researchers, chairs of research ethics boards, and philosophers of science define and reason with this term.MethodsWe completed semi-structured interviews about clinical trial ethics with 15 clinical researchers, 15 research ethics board chairs, and 15 philosophers of science/bioethicists. Each participant was asked a standardized set of 10 questions, 4 of which were specifically about equipoise. All interviews were conducted telephonically and transcribed. Responses were grouped and analysed via a modified grounded theory method.ResultsForty-three respondents defined equipoise in 7 logically distinct ways, and 2 respondents could not explicitly define it. The most common definition, offered by 14 respondents (31%), defined "equipoise" as a disagreement at the level of a community of physicians. There was significant variability in definitions offered between and within groups. When asked how they would "operationalize" equipoise - i.e. check or test for its presence - respondents provided 7 alternatives, the most common being in relation to a literature review (15/45, 33%). The vast majority of respondents (35/45, 78%) felt the concept was helpful, though many acknowledged that the lack of a clear definition or operationalization was problematic.ConclusionThere is significant variation in definitions of equipoise offered by respondents, suggesting that parties within groups and between groups may be referring to different concepts when they reference "equipoise". This non-uniformity may impact fairness and transparency and opens the door to potential ethical problems in the evaluation of clinical trials - for instance, a patient may understand equipoise very differently than the researchers enrolling her in a trial, which could cause her agreement to participate to be based upon false premises.

The Rift Valley fever virus (RVFV) causes haemorrhagic fever, encephalitis, and permanent blindness and has been listed by the WHO as a priority pathogen. To study RVFV pathogenesis and identify small-molecule antivirals, we established a novel In Vivo model using zebrafish larvae. Pericardial injection of RVFV resulted in ~4 log10 viral RNA copies/larva, which was inhibited by the antiviral 2′-fluoro-2′-deoxycytidine. The optical transparency of the larvae allowed detection of RVFVeGFP in the liver and sensory nervous system, including the optic tectum and retina, but not the brain or spinal cord. Thus, RVFV-induced blindness likely occurs due to direct damage to the eye and peripheral neurons, rather than the brain. Treatment with the JAK-inhibitor ruxolitinib, as well as knockout of stat1a but not stat1b, enhanced RVFV replication to ~6 log10 viral RNA copies/larva and ultra-bright livers, although without dissemination to sensory neurons or the eye, thereby confirming the critical role of stat1 in RVFV pathogenesis.

Background: Waste and fraud are important problems for health insurers to deal with. With the advent of big data, these insurers are looking more and more towards data mining and machine learning methods to help in detecting waste and fraud. However, labeled data is costly and difficult to acquire as it requires expert investigators and known care providers with atypical behavior. Methods: In this work we show how recent advances in machine learning can be used to set up a workflow that can aid investigators in discovering practitioners or groups of practitioners with unusual resource use in order to more efficiently combat waste and fraud. We combine three different techniques, which have not been used in the context of healthcare insurance anomaly detection: categorical embeddings to deal with high-cardinality categorical variables, state-of-the-art unsupervised anomaly detection techniques to detect anomalies and Shapley additive explanations (SHAP) to explain the model output. Results: The method has been evaluated on providers with a known anomalous profile and with the help of experts of the largest health insurance fund in Belgium. The quantitative experiments show that categorical embeddings offer a significant improvement compared to standard methods and that the state-of-the-art unsupervised anomaly detection techniques generally show an improvement over traditional methods. In a practical setting, the proposed workflow with SHAP was able to detect a previously unknown, anomalous trend among general practitioners. Conclusions: The proposed workflow is able to detect known care providers with atypical behaviour and helps expert investigators in making informed decisions concerning possible fraud or overconsumption in the health insurance field. [ABSTRACT FROM AUTHOR]

The Rift Valley fever virus (RVFV) causes haemorrhagic fever, encephalitis, and permanent blindness and has been listed by the WHO as a priority pathogen. To study RVFV pathogenesis and identify small-molecule antivirals, we established a novel In Vivo model using zebrafish larvae. Pericardial injection of RVFV resulted in ~4 log10 viral RNA copies/larva, which was inhibited by the antiviral 2′-fluoro-2′-deoxycytidine. The optical transparency of the larvae allowed detection of RVFVeGFP in the liver and sensory nervous system, including the optic tectum and retina, but not the brain or spinal cord. Thus, RVFV-induced blindness likely occurs due to direct damage to the eye and peripheral neurons, rather than the brain. Treatment with the JAK-inhibitor ruxolitinib, as well as knockout of stat1a but not stat1b, enhanced RVFV replication to ~6 log10 viral RNA copies/larva and ultra-bright livers, although without dissemination to sensory neurons or the eye, thereby confirming the critical role of stat1 in RVFV pathogenesis. [ABSTRACT FROM AUTHOR]

Chronic subjective tinnitus is a prevalent symptom, which has many similarities with chronic pain. Central sensitization is considered as a possible underlying mechanism of both symptoms. Central sensitization has already been investigated in chronic pain populations but not in patients with chronic subjective tinnitus. Therefore, the main objective of this cross-sectional study was to compare signs and symptoms, indicative for central sensitization, in tinnitus patients with and without chronic idiopathic neck pain, patients with chronic idiopathic neck pain only, and healthy controls. Also, differences in psychological and lifestyle factors, possibly influencing the association between central sensitization and tinnitus, were examined as well as correlations between signs and symptoms of central sensitization, and tinnitus, pain, psychological and lifestyle factors. Differences in signs and symptoms of central sensitization were examined using the self-report Central Sensitization Inventory and QST protocol (local and distant mechanical and heat hyperalgesia, conditioned pain modulation). Tinnitus, pain, psychological and lifestyle factors were evaluated using self-report questionnaires. Symptoms of central sensitization and local mechanical hyperalgesia were significantly more present in both tinnitus groups, compared to healthy controls, but were most extensive in the group with chronic tinnitus+chronic idiopathic neck pain. Distant mechanical hyperalgesia, indicative for central sensitization, was only observed in the group with both chronic tinnitus+chronic idiopathic neck pain. This group also displayed a significantly higher psychological burden and poorer sleep than patients with chronic tinnitus only and healthy controls. Signs and symptoms of central sensitization were also shown to be associated with tinnitus impact, pain-related disability, psychological burden and sleep disturbances. This study shows preliminary evidence for the presence of central sensitization in patients with chronic tinnitus+chronic idiopathic neck pain. This could be explained by the higher perceived tinnitus impact, psychological burden and sleep problems in this group. Trial registration: This study is registered as NCT05186259 (www.clinicaltrials.gov).

Tuberous sclerosis complex (TSC) is a multisystem genetic disorder caused by pathogenic variants in TSC1 and TSC2 genes. TSC patients present with seizures and brain abnormalities such as tubers and subependymal giant cells astrocytoma (SEGA). Despite common molecular and clinical features, the severity of the disease varies greatly, even intrafamilially. The second hit hypothesis suggests that an additional, inactivating mutation in the remaining functional allele causes a more severe phenotype and therefore explains the phenotypic variability. Recently, second hit mutations have been detected frequently in mTORopathies. To investigate the pathophysiological effects of second hit mutations, several mouse models have been developed. Here, we opted for a double mutant zebrafish model that carries a LOF mutation both in the tsc2 and the depdc5 gene. To the best of our knowledge, this is the first time a second-hit model has been studied in zebrafish. Significantly, the DEP domain-containing protein 5 (DEPDC5) gene has an important role in the regulation of mTORC1, and the combination of a germline TSC2 and somatic DEPDC5 mutation has been described in a TSC patient with intractable epilepsy. Our depdc5−/−x tsc2−/− double mutant zebrafish line displayed greatly increased levels of mammalian target of rapamycin (mTORC1) activity, augmented seizure susceptibility, and early lethality which could be rescued by rapamycin. Histological analysis of the brain revealed ventricular dilatation in the tsc2 and double homozygotes. RNA-sequencing showed a linear relation between the number of differentially expressed genes (DEGs) and the degree of mTORC1 hyperactivity. Enrichment analysis of their transcriptomes revealed that many genes associated with neurological developmental processes were downregulated and mitochondrial genes were upregulated. In particular, the transcriptome of human SEGA lesions overlapped strongly with the double homozygous zebrafish larvae. The data highlight the clinical relevance of the depdc5−/− x tsc2−/− double mutant zebrafish larvae that showed a more severe phenotype compared to the single mutants. Finally, analysis of gene-drug interactions identified interesting pharmacological targets for SEGA, underscoring the value of our small zebrafish vertebrate model for future drug discovery efforts.

IntroductionTo determine the effect of parental participation in hospital care on neonatal and parental outcomes in low- and middle-income countries (LMICs) and to identify the range of parental duties in the care of hospitalized neonates in LMICs.MethodsWe searched CINAHL, CENTRAL, LILACs, MEDLINE, EMBASE and Web of Science from inception to February 2022. Randomized and non-randomized studies from LMICs were eligible if parents performed one or more roles traditionally undertaken by healthcare staff. The primary outcome was hospital length-of-stay. Secondary outcomes included mortality, readmission, breastfeeding, growth, development and parental well-being. Data was extracted in duplicate by two independent reviewers using a piloted extraction form.ResultsEighteen studies (eight randomized and ten non-randomized) were included from seven middle-income countries. The types of parental participation included hygiene and infection prevention, feeding, monitoring and documentation, respiratory care, developmental care, medication administration and decision making. Meta-analyses showed that parental participation was not associated with hospital length-of-stay (MD −2.35, 95% CI −6.78–2.07). However, parental involvement was associated with decreased mortality (OR 0.46, 95% CI 0.22–0.95), increased breastfeeding (OR 2.97 95% CI 1.65–5.35) and decreased hospital readmission (OR 0.36, 95% CI 0.16–0.81). Narrative synthesis demonstrated additional benefits for growth, short-term neurodevelopment and parental well-being. Ten of the eighteen studies had a high risk of bias.ConclusionParental participation in neonatal hospital care is associated with improvement in several key neonatal outcomes in middle-income countries. The lack of data from low-income countries suggests that there remains barriers to parental participation in resource-poor settings.Systematic review registration[https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=187562], identifier [CRD42020187562].

Implication Statement The Community Health and Social Medicine (CHASM) Incubator is a social impact venture that gives medical and other health care students the opportunity to develop initiatives that sustainably promote health equity for, and in partnership with, community partners and historically marginalized communities. Students learn how to develop projects with project management curricula, are paired with community health mentors, and are given seed micro-financing. As the first community health incubator driven by medical students, CHASM provides a framework for students interested in implementing sustainable solutions to local health disparities which extends the service-learning opportunities offered in existing curricula.

Epilepsy is a common disorder of the brain characterized by spontaneous recurrent seizures, which develop gradually during a process called epileptogenesis. The mechanistic processes underlying the changes of brain tissue and networks toward increased seizure susceptibility are not fully understood. In rodents, injection of kainic acid (KA) ultimately leads to the development of spontaneous epileptic seizures, reflecting similar neuropathological characteristics as seen in patients with temporal lobe epilepsy (TLE). Although this model has significantly contributed to increased knowledge of epileptogenesis, it is technically demanding, costly to operate and hence not suitable for high-throughput screening of anti-epileptic drugs (AEDs). Zebrafish, a vertebrate with complementary advantages to rodents, is an established animal model for epilepsy research. Here, we generated a novel KA-induced epilepsy model in zebrafish larvae that we functionally and pharmacologically validated. KA was administered by pericardial injection at an early zebrafish larval stage. The epileptic phenotype induced was examined by quantification of seizure-like behavior using automated video recording, and of epileptiform brain activity measured via local field potential (LFP) recordings. We also assessed GFP-labeled GABAergic and RFP-labeled glutamatergic neurons in double transgenic KA-injected zebrafish larvae, and examined the GABA and glutamate levels in the larval heads by liquid chromatography with tandem mass spectrometry detection (LC-MS/MS). Finally, KA-injected larvae were exposed to five commonly used AEDs by immersion for pharmacological characterization of the model. Shortly after injection, KA induced a massive damage and inflammation in the zebrafish brain and seizure-like locomotor behavior. An abnormal reorganization of brain circuits was observed, a decrease in both GABAergic and glutamatergic neuronal population and their associated neurotransmitters. Importantly, these changes were accompanied by spontaneous and continuous epileptiform brain discharges starting after a short latency period, as seen in KA rodent models and reminiscent of human pathology. Three out of five AEDs tested rescued LFP abnormalities but did not affect the seizure-like behavior. Taken together, for the first time we describe a chemically-induced larval zebrafish epilepsy model offering unique insights into studying epileptogenic processes in vivo and suitable for high-throughput AED screening purposes and rapid genetic investigations.

Epilepsy has a peak incidence during the neonatal to early childhood period. These early onset epilepsies may be severe conditions frequently associated with comorbidities such as developmental deficits and intellectual disability and, in a significant percentage of patients, may be medication‐resistant. The use of adult rodent models in the exploration of mechanisms and treatments for early life epilepsies is challenging, as it ignores significant age‐specific developmental differences. More recently, models developed in immature animals, such as rodent pups, or in three‐dimensional organoids may more closely model aspects of the immature brain and could result in more translatable findings. Although models are not perfect, they may offer a more controlled screening platform in studies of mechanisms and treatments, which cannot be done in pediatric patient cohorts. On the other hand, more simplified models with higher throughput capacities are required to deal with the large number of epilepsy candidate genes and the need for new treatment options. Therefore, a combination of different modeling approaches will be beneficial in addressing the unmet needs of pediatric epilepsy patients. In this review, we summarize the discussions on this topic that occurred during the XVI Workshop on Neurobiology of Epilepsy, organized in 2022 by the Neurobiology Commission of the International League Against Epilepsy. We provide an overview of selected models of early onset epilepsies, discussing their advantages and disadvantages. Heterologous expression models provide initial functional insights, and zebrafish, rodent models, and brain organoids present increasingly complex platforms for modeling and validating epilepsy‐related phenomena. Together, these models offer valuable insights into early onset epilepsies and accelerate hypothesis generation and therapy discovery. [ABSTRACT FROM AUTHOR]