Logistic Regression Under Sparse Data Conditions

The impact of sparse data conditions was examined among one or more predictor variables in logistic regression and assessed the effectiveness of the Firth (1993) procedure in reducing potential parameter estimation bias. Results indicated sparseness in binary predictors introduces bias that is substantial with small sample sizes, and the Firth procedure can effectively correct this bias.

Download Full-text

Viscoelastometric Testing to Assess Hemostasis of COVID-19: A Systematic Review

Journal of Clinical Medicine ◽

10.3390/jcm10081740 ◽

2021 ◽

Vol 10 (8) ◽

pp. 1740

Author(s):

Marion Bareille ◽

Michaël Hardy ◽

Jonathan Douxfils ◽

Stéphanie Roullet ◽

Dominique Lasne ◽

...

Keyword(s):

Systematic Review ◽

Mechanical Strength ◽

Small Sample ◽

The Common ◽

Small Sample Sizes ◽

Definition Of ◽

Reported Data ◽

Prophylactic Anticoagulation ◽

The Impact ◽

Heparin Neutralization

Infection by SARS-CoV-2 is associated with a high risk of thrombosis. The laboratory documentation of hypercoagulability and impaired fibrinolysis remains a challenge. Our aim was to assess the potential usefulness of viscoelastometric testing (VET) to predict thrombotic events in COVID-19 patients according to the literature. We also (i) analyzed the impact of anticoagulation and the methods used to neutralize heparin, (ii) analyzed whether maximal clot mechanical strength brings more information than Clauss fibrinogen, and (iii) critically scrutinized the diagnosis of hypofibrinolysis. We performed a systematic search in PubMed and Scopus databases until December 31st, 2020. VET methods and parameters, and patients’ features and outcomes were extracted. VET was performed for 1063 patients (893 intensive care unit (ICU) and 170 non-ICU, 44 studies). There was extensive heterogeneity concerning study design, VET device used (ROTEM, TEG, Quantra and ClotPro) and reagents (with non-systematic use of heparin neutralization), timing of assay, and definition of hypercoagulable state. Notably, only 4 out of 25 studies using ROTEM reported data with heparinase (HEPTEM). The common findings were increased clot mechanical strength mainly due to excessive fibrinogen component and impaired to absent fibrinolysis, more conspicuous in the presence of an added plasminogen activator. Only 4 studies out of the 16 that addressed the point found an association of VETs with thrombotic events. So-called functional fibrinogen assessed by VETs showed a variable correlation with Clauss fibrinogen. Abnormal VET pattern, often evidenced despite standard prophylactic anticoagulation, tended to normalize after increased dosing. VET studies reported heterogeneity, and small sample sizes do not support an association between the poorly defined prothrombotic phenotype of COVID-19 and thrombotic events.

Download Full-text

Cannabis and breastfeeding

Paediatrics & Child Health ◽

10.1093/pch/pxaa037 ◽

2020 ◽

Vol 25 (Supplement_1) ◽

pp. S26-S28 ◽

Cited By ~ 1

Author(s):

Lisa Graves

Keyword(s):

Health Care Providers ◽

Small Sample ◽

Adverse Outcomes ◽

Care Providers ◽

High Quality ◽

High Quality Evidence ◽

Small Sample Sizes ◽

The Impact ◽

Quality Evidence

Abstract Cannabis is one of the most commonly used substances in Canada with 15% of Canadians reporting use in 2019. There is emerging evidence that cannabis is linked to an impact on the developing brain in utero and adverse outcomes in infants, children, and adolescents. The impact of cannabis during breastfeeding has been limited by studies with small sample sizes, follow-up limited to 1 year and the challenge of separating prenatal exposure from that during breastfeeding. In the absence of high-quality evidence, health care providers need to continue to engage women in conversation about the potential concerns related to breastfeeding and cannabis use.

Download Full-text

Real-world data evaluating immunotherapy markers and checkpoint inhibitor response among GI cancers in a large community-based oncology network.

Journal of Clinical Oncology ◽

10.1200/jco.2021.39.3_suppl.113 ◽

2021 ◽

Vol 39 (3_suppl) ◽

pp. 113-113

Author(s):

Carissa Jones ◽

Rebecca Lachs ◽

Emma Sturgill ◽

Amanda Misch ◽

Caressa Lietman ◽

...

Keyword(s):

Checkpoint Inhibitor ◽

Small Sample ◽

Tumor Type ◽

Real World Data ◽

Loss To Follow Up ◽

Kaplan Meier ◽

Gi Cancers ◽

Small Sample Sizes ◽

Line Of Therapy ◽

The Impact

113 Background: Checkpoint inhibitor (CPI) therapies have shown prolonged survival in patients (pts) with microsatellite instability (MSI). Tumor mutation burden (TMB) has also been associated with benefit from CPIs, with pembrolizumab recently approved for solid tumors with a TMB of ≥10 muts/mb. However, the validity of a tissue-agnostic approach has been debated given the high degree of TMB variation across tumor types. We sought to evaluate the impact of TMB and MSI on CPI response in pts with advanced gastrointestinal (GI) cancers who received NGS profiling. Methods: Patients within the Sarah Cannon network with GI cancer and comprehensive next generation sequencing (NGS) data were identified through Genospace, Sarah Cannon’s clinico-genomic analytics platform. Microsatellite (MS) status [high (MSI-H) vs. stable (MSS)] was defined by NGS. TMB-high (TMB-H) was defined as ≥10 muts/mb. Kaplan-Meier estimates were used to examine time to treatment failure (TTF), defined as the time from therapy start to start of next therapy, death, or loss to follow-up. Results: We identified 5,788 pts with GI cancers who received NGS, of which 48% (N=3,603) had colorectal cancer (CRC). MS status was evaluated on 4,219 (66%) NGS reports, TMB on 2,922 (46%) reports, and both on 2,863 (45%) reports. MSI-H and TMB-H co-occurred on 127 reports (4%), MSS/TMB-H occurred on 312 reports (11%), and MSS/TMB-Low (TMB-L) occurred on 2,424 reports (85%). No reports were MSI-H/TMB-L. Overall, 580 pts (14%) received CPI therapy (N=52 MSI-H/TMB-H, N=41 MSS/TMB-H, N=215 MSS/TMB-L). Independent of line of therapy or tumor type, TTF was significantly shorter for CPI vs. non-CPI therapies [median TTF (mTTF)=151 vs. 238 days, respectively, p=3.4x10-9]. As expected, MSI-H was also associated with longer CPI TTF compared to MSS (mTTF=727 vs. 124 days, p=3.1x10-11). Due to small sample sizes, further analyses were focused on pts with CRC (Table). TMB-H was independently associated with longer CPI TTF compared to TMB-L (Table, p=9.2x10-10). Similarly, the co-occurrence of MSI-H/TMB-H was also associated with longer CPI TTF (Table). However, there was no difference in CPI TTF for MSS/TMB-H and MSS/TMB-L (Table, p=0.45). All pts who received non-CPI therapies, regardless of MS/TMB group, had longer TTF than MSS/TMB-H and MSS/TMB-L pts who received CPI (Table). mTTF [days (N)] for pts with CRC by biomarker. Conclusions: Despite recent tissue-agnostic approvals, TMB does not appear to be a good biomarker of CPI response in pts with CRC. Rather, time to CPI failure is associated with the co-occurrence of TMB-H with MSI-H. Continued research is needed to identify better biomarkers of response to immunotherapy in GI cancers. [Table: see text]

Download Full-text

The impact of the 30 most cited articles on hip arthroscopy: what is the subject matter?

Journal of Hip Preservation Surgery ◽

10.1093/jhps/hnz067 ◽

2020 ◽

Vol 7 (1) ◽

pp. 14-21

Author(s):

Alexander von Glinski ◽

Emre Yilmaz ◽

Ryan Goodmanson ◽

Clifford Pierre ◽

Sven Frieler ◽

...

Keyword(s):

Surgical Treatment ◽

Hip Arthroscopy ◽

Randomized Clinical Trials ◽

Technical Note ◽

Small Sample ◽

Level Of Evidence ◽

Thomson Reuters ◽

Treatment Indications ◽

Small Sample Sizes ◽

The Impact

Abstract The purpose of this study was to identify the 30 most cited articles on hip arthroscopy and discuss their influence on recent surgical treatment. Due to advancements in hip arthroscopy, there is a widening spectrum of diagnostic and treatment indications. The purpose of this study was to identify the 30 most cited articles on hip arthroscopy and discuss their influence on contemporary surgical treatment. The Thomson Reuters Web of Science was used to identify the 30 most cited studies on hip arthroscopy between 1900 and 2018. These 30 articles generated 6152 citations with an average of 205.07 citations per item. Number of citations ranged from 146 to 461. Twenty-five out of the 30 papers were clinical cohort studies with a level of evidence between III and IV, encompassing 4348 patients. Four studies were reviewed (one including a technical note) and one a case report. We were able to identify the 30 most cited articles in the field of hip arthroscopy. Most articles were reported in high-impact journals, but reported small sample sizes in a retrospective setting. Prospective multi-arm cohort trials or randomized clinical trials represent opportunities for future studies.

Download Full-text

The scale of it all: postcanine tooth size, the taxon-level effect, and the universality of Gould's scaling law

Paleobiology ◽

10.1666/08089.1 ◽

2010 ◽

Vol 36 (2) ◽

pp. 188-203 ◽

Cited By ~ 19

Author(s):

Lynn E. Copes ◽

Gary T. Schwartz

Keyword(s):

Body Mass ◽

Scaling Law ◽

Small Sample ◽

Tooth Size ◽

Mathematical Methods ◽

Scaling Relationship ◽

Small Sample Sizes ◽

Relationship Of ◽

The Impact ◽

Level Effect

In a seminal paper in 1975, Gould proposed that postcanine occlusal area (PCOA) should scale metabolically (0.75) with body mass across mammals. By regressing PCOA against skull length in a small sample of large-bodied herbivorous mammals, Gould provided some marginal support for this hypothesis, which he then extrapolated as a universal scaling law for Mammalia. Since then, many studies have sought to confirm this scaling relationship within a single order and have found equivocal support for Gould's assertion. In part, this may be related to the use of proxies for both PCOA and body mass, small sample sizes, or the influence of a “taxon-level effect,” rendering Gould's scaling “universal” problematic.Our goal was to test the universality of Gould's prediction and the impact of the taxon-level effect on regressions of tooth size on body mass in a large extant mammalian sample (683 species spanning 14 orders). We tested for the presence of two types of taxon-level effect that may influence the acceptance or rejection of hypothesized scaling coefficients. The hypotheses of both metabolic and isometric scaling can be rejected in Mammalia, but not in all sub-groups therein. The level of data aggregation also influences the interpretation of the scaling relationship. Because the scaling relationship of tooth size to body mass is highly dependent on both the taxonomic level of analysis and the mathematical methods used to organize the data, paleontologists attempting to retrodict body mass from fossilized dental remains must be aware of the effect that sample composition may have on their results.

Download Full-text

Integrating multi-omics data with deep learning for predicting cancer prognosis

10.1101/807214 ◽

2019 ◽

Cited By ~ 3

Author(s):

Hua Chai ◽

Xiang Zhou ◽

Zifeng Cui ◽

Jiahua Rao ◽

Zheng Hu ◽

...

Keyword(s):

Deep Learning ◽

Proportional Hazards ◽

Proportional Hazards Model ◽

Small Sample ◽

The Cancer Genome Atlas ◽

Cancer Prognosis ◽

Omics Data ◽

Hazards Model ◽

Small Sample Sizes ◽

The Impact

AbstractMotivationAccurately predicting cancer prognosis is necessary to choose precise strategies of treatment for patients. One of effective approaches in the prediction is the integration of multi-omics data, which reduces the impact of noise within single omics data. However, integrating multi-omics data brings large number of redundant variables and relative small sample sizes. In this study, we employed Autoencoder networks to extract important features that were then input to the proportional hazards model to predict the cancer prognosis.ResultsThe method was applied to 12 common cancers from the Cancer Genome Atlas. The results show that the multi-omics averagely improves 4.1% C-index for prognosis prediction over single mRNA data, and our method outperforms previous approaches by at least 7.4%. A comparison of the contribution of single omics data show that mRNA contributes the most, followed by the DNA methylation, miRNA, and the copy number variation. In the case study for differential gene expression analysis, we identified 161 differentially expressed genes in the cervical cancer, among which 77 genes (65.8%) have been proven to be associated with cancer. In addition, we performed the cross-cancer test where the model trained on one cancer was used to predict the prognosis of another cancer, and found 23 pairs of cancers have a C-index larger than 0.5, with the largest value of 0.68. Thus, this study has provided a deep learning framework to effectively integrate multiple omics data to predict cancer prognosis.

Download Full-text

Covid-19 Patient Mortality Risk Classification Using Bayesian Binary Logistic Regression

Jurnal Matematika Statistika dan Komputasi ◽

10.20956/j.v18i1.14268 ◽

2021 ◽

Vol 18 (1) ◽

pp. 150-160

Author(s):

Muhammad Qolbi Shobri ◽

Ferra Yanuar ◽

Dodi Devianto

Keyword(s):

Logistic Regression ◽

Parameter Estimation ◽

Mortality Rate ◽

Bayesian Method ◽

Binary Logistic Regression ◽

Predictor Variables ◽

New Disease ◽

Death Risk ◽

Risk Of Death ◽

The World

At the end of 2019 the world was shocked by a new disease caused by SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus 2). The disease is called Covid-19 (Coronavirus Disease). The mortality rate due to disease is increasing every day. In Indonesia as of April 2021, confirmed Covid-19 patients who died reached 42,530 patients, seeing the high mortality rate of Covid-19 patients so it needs to be studied further so that the risk of death of these Covid-19 patients can be minimized. This research utilizing binary logistic regression with Bayesian method parameter estimation. In this study, the predictor variables used were in the form of categories that each category in the predictor variables was assumed to have the same risk of death risk of Covid-19 patients. The results of this study indicate that the number of comorbids has a significant effect on the risk of death of Covid-19 patients, the more the number of comorbids suffered by the patient, the higher the risk of death of the patient. The accuracy of this method in classifying data is 84.68%.

Download Full-text

A Review of Telemedicine Interventions for Weight Loss

Current Cardiovascular Risk Reports ◽

10.1007/s12170-021-00680-w ◽

2021 ◽

Vol 15 (9) ◽

Author(s):

Kelsey Ufholz ◽

Daksh Bhargava

Keyword(s):

Weight Loss ◽

Digital Literacy ◽

Small Sample ◽

Text Messages ◽

Treatment Access ◽

Self Monitoring ◽

Portion Control ◽

Phone Calls ◽

Small Sample Sizes ◽

The Impact

Abstract Purpose of Review Telemedicine has become popular as an alternative for in-person weight loss treatment during the COVID-19 pandemic. This review focuses on weight loss interventions utilizing real-time telemedicine. Recent Findings Telemedicine interventions are usually run as a weekly counseling and educational session or as a complement to a primarily Web-based intervention. A wide variety of healthcare professionals may provide the intervention. Common content includes portion control, increased physical activity, and relapse prevention. Self-monitoring is associated with intervention success. Modalities considered include online chats, text messages, phone calls, and videoconferences. Videoconferencing may be especially useful in capturing the interpersonal connection associated with in-person care but is understudied compared to other modalities. While many interventions show improvements in weight and weight-related outcomes, small sample sizes limit generalizability. Technology access and digital literacy are both necessary. Summary Telemedicine interventions can successfully help patients with obesity lose weight. Telemedicine interventions provide a safe, remote alternative and may expand treatment access to hard-to-reach populations. Further research is needed on telemedicine weight loss treatments for seniors, men, and ethnic minorities, as well as on the impact of long-term interventions.

Download Full-text

Neuroplasticity and Motor Rehabilitation in Multiple Sclerosis: A Systematic Review on MRI Markers of Functional and Structural Changes

Frontiers in Neuroscience ◽

10.3389/fnins.2021.707675 ◽

2021 ◽

Vol 15 ◽

Author(s):

Eleonora Tavazzi ◽

Marta Cazzoli ◽

Alice Pirastru ◽

Valeria Blasi ◽

Marco Rovaris ◽

...

Keyword(s):

Multiple Sclerosis ◽

Structural Changes ◽

Diffusion Tensor ◽

Structural Connectivity ◽

Structural Mri ◽

Resting State Fmri ◽

Small Sample ◽

Motor Rehabilitation ◽

Small Sample Sizes ◽

The Impact

Background: Motor rehabilitation is routinely used in clinical practice as an effective method to reduce progressive disability gain in multiple sclerosis (MS), but rehabilitation approaches are typically unstandardized, and only few studies have investigated the impact of rehabilitation on brain neuroplasticity.Objective: To summarize and critically analyze studies applying MRI markers of functional connectivity and structural changes to assess the effect of motor rehabilitation on brain neuroplasticity in MS.Methods: Literature search was performed using PubMed and EMBASE, selecting studies having as a subject motor rehabilitation and advanced MRI techniques investigating neuroplasticity in adult patients affected by MS.Results: Seventeen out of 798 papers were selected, of which 5 applied structural MRI (4 diffusion tensor imaging, 1 volumetric measurements), 7 applied functional fMRI (5 task-related fMRI, 2 resting-state fMRI) whereas the remaining 5 applied both structural and functional imaging.Discussion: The considerable data heterogeneity and the small sample sizes characterizing the studies limit interpretation and generalization of the results. Overall, motor rehabilitation promotes clinical improvement, paralleled by positive adaptive brain changes, whose features and extent depend upon different variables, including the type of rehabilitation approach. MRI markers of functional and structural connectivity should be implemented in studies testing the efficacy of motor rehabilitation. They allow for a better understanding of neuroplastic mechanisms underlying rehabilitation-mediated clinical achievements, facilitating the identification of rehabilitation strategies tailored to patients' needs and abilities.

Download Full-text