Regression Modeling Strategies

Construction of a ceRNA network to reveal a vascular invasion associated prognostic model in hepatocellular carcinoma

The aim of this study is to explore the prognostic value of vascular invasion (VI) in hepatocellular carcinoma (HCC) by searching for competing endogenous RNAs (ceRNA) network and constructing a new prognostic model for HCC. The differentially expressed genes (DEGs) between HCC and normal tissues were identified from GEO and TCGA. StarBase and miRanda prediction tools were applied to construct a circRNA-miRNA-mRNA network. The DEGs between HCC with and without VI were also identified. Then, the hub genes were screened to build a prognostic risk score model through the method of least absolute shrinkage and selection operator. The prognostic ability of the model was assessed using the Kaplan-Meier method and Cox regression analysis. In result, there were 221 up-regulated and 47 down-regulated differentially expressed circRNAs (DEcircRNAs) in HCC compared with normal tissue. A circRNA-related ceRNA network was established, containing 11 DEcircRNAs, 12 DEmiRNAs, and 161 DEmRNAs. Meanwhile, another DEG analysis revealed 625 up-regulated and 123 down-regulated DEGs between HCC with and without VI, and then a protein-protein interaction (PPI) network was built based on 122 VI-related DEGs. From the intersection of DEGs within the PPI and ceRNA networks, we obtained seven hub genes to build a novel prognostic risk score model. HCC patients with high-risk scores had shorter survival time and presented more advanced T/N/M stages as well as VI occurrence. In conclusion a novel prognostic model based on seven VI-associated DEGs within a circRNA-related ceRNA network was constructed in this study, with great ability to predict the outcome of HCC patients.

Effect of mean heart rate on 30-day mortality in ischemic stroke with atrial fibrillation: Data from the MIMIC-IV database

Background

The relationship of mean heart rate (MHR) with 30-day mortality in ischemic stroke patients with atrial fibrillation in the intensive care unit (ICU) remains unknown. This study aimed to investigate the association between MHR within 24 h of admission to the ICU and 30-day mortality among patients with atrial fibrillation and ischemic stroke.

Methods

This retrospective cohort study used data on US adults from the Medical Information Mart for Intensive Care-IV (MIMIC-IV, version 1.0) database. Patients with ischemic stroke who had atrial fibrillation for and first time in ICU admission were identified from the MIMIC-IV database. We used multivariable Cox regression models, a restricted cubic spline model, and a two-piecewise Cox regression model to show the effect of the MHR within 24 h of ICU admission on 30-day mortality.

Results

A total of 1403 patients with ischemic stroke and atrial fibrillation (mean [SD] age, 75.9 [11.4] years; mean [SD] heart rate, 83.8[16.1] bpm; 743 [53.0%] females) were included. A total of 212 (15.1%) patients died within 30 days after ICU admission. When MHR was assessed in tertials according to the 25th and 50th percentiles, the risk of 30-day mortality was higher in participants in group 1 (< 72 bpm; adjusted hazard ratio, 1.23; 95% CI, 0.79–1.91) and group 3 (≥82 bpm; adjusted hazard ratio, 1.77; 95% CI, 1.23–2.57) compared with those in group 2 (72–82 bpm). Consistently in the threshold analysis, for every 1-bpm increase in MHR, there was a 2.4% increase in 30-day mortality (adjusted HR, 1.024; 95% CI, 1.01–1.039) in those with MHR above 80 bpm. Based on these results, there was a J-shaped association between MHR and 30-day mortality in ischemic stroke patients with atrial fibrillation admitted to the ICU, with an inflection point at 80 bpm of MHR.

Conclusion

In this retrospective cohort study, MHR within 24 h of admission was associated with 30-day mortality (nonlinear, J-shaped association) in patients with ischemic stroke and atrial fibrillation in the ICU, with an inflection point at about 80 bpm and a minimal risk observed at 72 to 81 bpm of MHR. This association was worthy of further investigation. If further confirmed, this association may provide a theoretical basis for formulating the target strategy of heart rate therapy for these patients.

ToPP: Tumor online prognostic analysis platform for prognostic feature selection and clinical patient subgroup selection

Patients with cancer with different molecular characterization and subtypes result in different response to anticancer therapeutics and survival. To identify features that are associated with prognosis is essential to precision medicine by providing clues for target identification, drug discovery. Here, we developed a tumor online prognostic analysis platform (ToPP) which integrated eight multi-omics features and clinical data from 68 cancer projects. It provides multiple approaches for customized prognostic studies, including 1) Prognostic analysis based on multi-omics features and clinical characteristics; 2) Automatic construction of prognostic model; 3) Pancancer prognostic analysis in multi-omics data; 4) Explore the impact of different levels of feature combinations on patient prognosis; 5) More sophisticated prognostic analysis according to regulatory network. ToPP provides a comprehensive source and easy-to-use interface for tumor prognosis research, with one-stop service of multi-omics, subtyping, and online prognostic modeling. The web server is freely available at http://www.biostatistics.online/topp/index.php.

•

ToPP platform integrated eight multi-omics and clinical data from 68 cancer projects

•

ToPP provides multi-omics combination and subgroup selection for prognostic analysis

•

ToPP provides automatic construction of prognostic model for public and custom data

•

Users can perform prognostic analysis based on regulatory network or pathways in ToPP

Prognostic models of diabetic microvascular complications: a systematic review and meta-analysis

BACKGROUND

Many prognostic models of diabetic microvascular complications have been developed, but their performances still varies. Therefore, we conducted a systematic review and meta-analysis to summarise the performances of the existing models.

METHODS

Prognostic models of diabetic microvascular complications were retrieved from PubMed and Scopus up to 31 December 2020. Studies were selected, if they developed or internally/externally validated models of any microvascular complication in type 2 diabetes (T2D).

RESULTS

In total, 71 studies were eligible, of which 32, 30 and 18 studies initially developed prognostic model for diabetic retinopathy (DR), chronic kidney disease (CKD) and end stage renal disease (ESRD) with the number of derived equations of 84, 96 and 51, respectively. Most models were derived-phases, some were internal and external validations. Common predictors were age, sex, HbA1c, diabetic duration, SBP and BMI. Traditional statistical models (i.e. Cox and logit regression) were mostly applied, otherwise machine learning. In cohorts, the discriminative performance in derived-logit was pooled with C statistics of 0.82 (0.73‑0.92) for DR and 0.78 (0.74‑0.83) for CKD. Pooled Cox regression yielded 0.75 (0.74‑0.77), 0.78 (0.74‑0.82) and 0.87 (0.84‑0.89) for DR, CKD and ESRD, respectively. External validation performances were sufficiently pooled with 0.81 (0.78‑0.83), 0.75 (0.67‑0.84) and 0.87 (0.85‑0.88) for DR, CKD and ESRD, respectively.

CONCLUSIONS

Several prognostic models were developed, but less were externally validated. A few studies derived the models by using appropriate methods and were satisfactory reported. More external validations and impact analyses are required before applying these models in clinical practice.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42018105287.

Impact of YouTube Advertising on Sales with Regression Analysis and Statistical Modeling: Usefulness of Online Media in Business

Computer technology plays a prominent role in almost every aspect of daily life including education, health care, online shopping, advertising, and even in homes. Computers help to make daily tasks much easier and convenient. Among social media, YouTube is a well-known social sharing networking service. As more and more people join social media and become everyday users, brands have also increased their online engagement. However, it is still unclear how to effectively measure value and return on advertising using social media. As of 2021, more than 31 million YouTube channels around the globe have been opened. In this paper, we consider YouTube advertising to check its effectiveness and benefits gained. Certain statistical tools are adopted to measure the extent of advertising benefits and their correlation in creating effective advertising campaigns on YouTube. Simple linear regression analysis is performed on the data representing the YouTube advertising budget of a company and the sales data of that company. Furthermore, we develop a new statistical distribution to provide the best description of the YouTube advertising data. The result of this research shows that YouTube is an effective medium for advertising and has a strong relationship with sales.

Impact of Facebook and Newspaper Advertising on Sales: A Comparative Study of Online and Print Media

Marketing means the strategies and tactics an organization undertakes for attracting consumers to promote the buying or selling of a product or service. Active marketing is about receiving messages from potential buyers to create ways to influence their purchasing decisions. Advertising is one of the most prominent marketing strategies to promote products to consumers. It is well known that advertisement has a significant impact on the sale of certain goods or services. In this paper, we consider two mediums of advertisement, such as Facebook (which is an online medium) and Newspaper (which is a printed medium). We consider a dataset representing the advertising budget (in hundreds of US dollars) of an electronic company and the sales of that company. We apply the quantitative research approach, and the data which are used in this research are secondary data. For analysis purposes, we consider a statistical tool called simple linear regression modeling. To check the significance of the advertising on sale, definite statistical tests are applied. Based on the findings of this research, it is observed that advertising has a significant impact on sales. It is also showed that spending money on advertising through Facebook has better sales than newspapers. The finding of this research shows that the use of computer-based technologies and online mediums has a brighter future for advertising. Furthermore, a new statistical model is introduced using the Z family approach. The proposed model is very interesting and possesses heavy-tailed properties. Finally, the applicability of the proposed model is illustrated by considering the financial dataset.

Development and Validation a Nomogram for Predicting Overall Survival in Patients With Intrahepatic Cholangiocarcinoma

Background: This study aims to establish an effective nomogram to predict the overall survival of patients with intrahepatic cholangiocarcinoma (ICC).

Patients and Methods: Data used to build the nomogram comes from the Surveillance, Epidemiology, and End Results (SEER) database. Patients diagnosed with ICC between 2005 and 2016 were retrospectively collected. Prediction accuracy and discrimination ability of the nomogram was evaluated by concordance index (C-index) and calibration curve. The area under receiver operating characteristic (ROC) curve (AUC) and decision curve analysis (DCA) were used to compare the precision of the 1-, 3-, and 5-year survival of the nomogram with 8th American Joint Commission on Cancer (AJCC) tumor–node–metastasis (TNM) staging system. Finally, it was verified in a prospective study of patients diagnosed with ICC in the Second Affiliated Hospital of Nanchang University from 2013 to 2020 by bootstrap resampling.

Result: The study contains two parts of data; we establish a nomogram using external data, and we conducted internal verification and external verification. The nomogram that we have established has good calibration, with a concordance index (C-index) of 0.75 (95% CI, 0.74–0.76) for overall survival (OS) prediction. The AUC value of the nomogram predicting 1-, 3-, and 5-year OS rates were 0.821, 0.828, and 0.836, which were higher than those of the 8th AJCC TNM staging systems. The calibration curve for the probability of survival between prediction by nomogram and actual observation shows good agreement. The nomogram showed better accuracy than the 8th edition AJCC TNM staging.

Conclusion: The nomogram established can provide a more accurate prognosis for patients with intrahepatic cholangiocarcinoma.

Preoperative prognostic nutritional index as an independent prognostic factor for resected ampulla of Vater cancer

Introduction

Prognostic nutritional index (PNI) reflects the nutritional and immunologic status of the patients. The clinical application of PNI is already well-known in various kinds of solid tumors. However, there is no study investigating the relationship between PNI and oncological outcome of the resected ampulla of Vater (AoV) cancer.

Materials and methods

From January 2005 to December 2012, the medical records of patients who underwent pancreaticoduodenectomy for pathologically confirmed AoV cancer were retrospectively reviewed. Long-term oncological outcomes were compared according to the preoperative PNI value.

Result

A total of 118 patients were enrolled in this study. The preoperative PNI was 46.13±6.63, while the mean disease-free survival was 43.88 months and the mean disease-specific survival was 55.3 months. In the multivariate Cox analysis, initial CA19-9 (p = 0.0399), lymphovascular invasion (p = 0.0031), AJCC 8^th N-stage (p = 0.0018), and preoperative PNI (p = 0.0081) were identified as significant prognostic factors for resected AoV cancer. The disease-specific survival was better in the high preoperative PNI group (≤48.85: 40.77 months vs. >48.85: 68.05 months, p = 0.0015). A highly accurate nomogram was developed based on four clinical components to predict the 1, 3, and 5-year disease-specific survival probability (C-index 0.8169, 0.8426, and 0.8233, respectively).

Conclusion

In resected AoV cancer, preoperative PNI can play a significant role as an independent prognostic factor for predicting disease-specific survival.

Health disparities between binary and non binary trans people: A community-driven survey

Background: Genderqueer and nonbinary () people have remained largely invisible in health research. Previous research shows worse outcomes on health indicators for trans people when compared with cisgender controls, but the differences between binary trans and GQNB individuals are inconclusive. Aims: To compare overall health and well-being of GQNB people with controls of trans men and trans women, taking into account the impact of the additive effect of their socio-economic position, as well as their current need for gender affirming medical interventions. Methods: A community-driven survey was conducted in 2016 in five countries (Georgia, Poland, Serbia, Spain, and Sweden). Self-reported health and general well-being were analysed for differences between binary trans and GQNB respondents. The effects of multiple control variables (age, economic situation, educational level, belonging to an ethnic, religious, sexual or ability minority group, sex assigned at birth) as well as the current need for gender affirming medical interventions were controlled for. Results: The sample consisted of 853 respondents aged 16 and older, with 254 trans women (29.8%), 369 trans men (43.2%), and 230 GQNB people (26%). GQNB respondents showed significantly worse self-reported health and worse general well-being in comparison to binary trans respondents. Additional negative impacts of having a lower educational level, having more economic stress, and belonging to a disability minority group were found. Being in need of gender affirming medical interventions contributed significantly to worse self-reported health, whereas being younger contributed to worse general well-being. Discussion: In understanding health disparities between binary trans and GQNB people, it is necessary to take into account the additive effect of multiple socio-economic positions, and the current need for gender affirming medical interventions. The high proportion of GQNB respondents who report worse health outcomes highlights the need for policy makers and health-care providers in creating nonbinary-inclusive environments.

A systematic literature review and classification of knowledge discovery in traditional medicine

INTRODUCTION AND OBJECTIVE

Despite the importance of machine learning methods application in traditional medicine there is a no systematic literature review and a classification for this field. This is the first comprehensive literature review of the application of data mining methods in traditional medicine.

METHOD

We reviewed 5 database between 2000 to 2017 based on the Kitchenham systematic review methodology. 502 articles were identified and reviewed for their relevance to application of machine learning methods in traditional medicine, 42 selected papers were classified and categorized on four dimension; 1) application domain of data mining techniques in traditional medicine; 2) the data mining methods most frequently used in traditional medicine; 3) main strength and limitation of data mining techniques in traditional medicine; 4) the performance evaluation methods in data mining methods in traditional medicine.

RESULT

The result obtained showed that main application domain of data mining techniques in traditional medicine was related to syndrome differentiation. Bayesian Networks (BNs), Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs) were recognized as being the methods most frequently applied in traditional medicine. Furthermore, each data mining techniques has its own strength and limitations when applied in traditional medicine. Single scaler methods were frequently used for performance evaluation of data mining methods.

CONCLUSION

Machine learning methods have become an important research field in traditional medicine. Our research provides information about this methods by examining the related articles.

Predictive Validity of CRUSADE, ACTION and ACUITY-HORIZONS Bleeding Risk Scores in Chinese Patients With ST-Segment Elevation Myocardial Infarction

BACKGROUND

The CRUSADE, ACTION and ACUITY-HORIZONS bleeding scores have been derived using Caucasian patients, and little is known about which has the better predictive ability in Chinese patients, especially for patients with STEMI.Methods and Results:We retrospectively analyzed 2,208 consecutive STEMI patients undergoing primary PCI (PPCI). Major bleeding events were defined according to Bleeding Academic Research Consortium criteria (type 3 or 5). Predictive ability of the 3 scores was assessed using logistic regression and AUC. Unadjusted HR for 1-year death were determined on Cox proportional hazard modeling. The major bleeding rate was 2.4%. The AUC of the CRUSADE, ACTION and ACUTIY-HORIZONS models was 0.88 (95% CI: 0.84-0.92), 0.90 (95% CI: 0.87-0.94), and 0.78 (95% CI: 0.87-0.94). The calibration of the ACUTIY-HORIZONS model was not acceptable overall, or in the subgroup of access site (P<0.05). In the high-risk category, 1-year mortality was approximately 4-7-fold greater than in the low-risk category (CRUSADE: HR, 7.27; 95% CI: 3.30-16.02, P<0.001; ACTION: HR, 7.13; 95% CI: 2.19-15.41, P<0.001; ACUITY-HORIZONS: HR, 4.06; 95% CI: 1.62-10.16; P=0.003).

CONCLUSIONS

The CRUSADE and ACTION scores have greater predictive ability for in-hospital major bleeding than the ACUITY-HORIZONS risk score in Chinese STEMI patients undergoing PPCI. Mortality would increase with the transition from low- to high-risk category in 1 year.

FARMS: A New Algorithm for Variable Selection

Large datasets including an extensive number of covariates are generated these days in many different situations, for instance, in detailed genetic studies of outbreed human populations or in complex analyses of immune responses to different infections. Aiming at informing clinical interventions or vaccine design, methods for variable selection identifying those variables with the optimal prediction performance for a specific outcome are crucial. However, testing for all potential subsets of variables is not feasible and alternatives to existing methods are needed. Here, we describe a new method to handle such complex datasets, referred to as FARMS, that combines forward and all subsets regression for model selection. We apply FARMS to a host genetic and immunological dataset of over 800 individuals from Lima (Peru) and Durban (South Africa) who were HIV infected and tested for antiviral immune responses. This dataset includes more than 500 explanatory variables: around 400 variables with information on HIV immune reactivity and around 100 individual genetic characteristics. We have implemented FARMS in R statistical language and we showed that FARMS is fast and outcompetes other comparable commonly used approaches, thus providing a new tool for the thorough analysis of complex datasets without the need for massive computational infrastructure.