Binomial Logistic Regression for Binary Outcomes

2021 ◽  
pp. 101-126
Author(s):  
Keith McNulty
Keyword(s):  
Logistic Regression ◽  
Binary Outcomes ◽  
Binomial Logistic Regression

scholarly journals COVID-19-Related Changes in Perceived Household Food Waste in the United States: A Cross-Sectional Descriptive Study

2021 ◽  
Vol 18 (3) ◽  
pp. 1104
Author(s):  
Kelly Cosgrove ◽  
Maricarmen Vizcaino ◽  
Christopher Wharton
Keyword(s):  
Logistic Regression ◽  
Food Waste ◽  
The United States ◽  
Food Groups ◽  
Related Factors ◽  
Cross Sectional ◽  
Household Level ◽  
Binomial Logistic Regression ◽  
Household Food ◽  
Fresh Vegetables

Food waste contributes to adverse environmental and economic outcomes, and substantial food waste occurs at the household level in the US. This study explored perceived household food waste changes during the COVID-19 pandemic and related factors. A total of 946 survey responses from primary household food purchasers were analyzed. Demographic, COVID-19-related household change, and household food waste data were collected in October 2020. Wilcoxon signed-rank was used to assess differences in perceived food waste. A hierarchical binomial logistic regression analysis was conducted to examine whether COVID-19-related lifestyle disruptions and food-related behavior changes increased the likelihood of household food waste. A binomial logistic regression was conducted to explore the contribution of different food groups to the likelihood of increased food waste. Perceived food waste, assessed as the estimated percent of food wasted, decreased significantly during the pandemic (z = −7.47, p < 0.001). Food stockpiling was identified as a predictor of increased overall food waste during the pandemic, and wasting fresh vegetables and frozen foods increased the odds of increased food waste. The results indicate the need to provide education and resources related to food stockpiling and the management of specific food groups during periods of disruption to reduce food waste.

Alternative models and randomization techniques for Bayesian response-adaptive randomization with binary outcomes

2021 ◽  
pp. 174077452110101
Author(s):  
Jennifer Proper ◽  
John Connett ◽  
Thomas Murray
Keyword(s):  
Logistic Regression ◽  
Sample Size ◽  
Error Rate ◽  
Adaptive Design ◽  
Type I Error ◽  
Probability Model ◽  
Binary Outcomes ◽  
Type I ◽  
Operating Characteristics ◽  
Type I Error Rate

Background: Bayesian response-adaptive designs, which data adaptively alter the allocation ratio in favor of the better performing treatment, are often criticized for engendering a non-trivial probability of a subject imbalance in favor of the inferior treatment, inflating type I error rate, and increasing sample size requirements. The implementation of these designs using the Thompson sampling methods has generally assumed a simple beta-binomial probability model in the literature; however, the effect of these choices on the resulting design operating characteristics relative to other reasonable alternatives has not been fully examined. Motivated by the Advanced R2 Eperfusion STrategies for Refractory Cardiac Arrest trial, we posit that a logistic probability model coupled with an urn or permuted block randomization method will alleviate some of the practical limitations engendered by the conventional implementation of a two-arm Bayesian response-adaptive design with binary outcomes. In this article, we discuss up to what extent this solution works and when it does not. Methods: A computer simulation study was performed to evaluate the relative merits of a Bayesian response-adaptive design for the Advanced R2 Eperfusion STrategies for Refractory Cardiac Arrest trial using the Thompson sampling methods based on a logistic regression probability model coupled with either an urn or permuted block randomization method that limits deviations from the evolving target allocation ratio. The different implementations of the response-adaptive design were evaluated for type I error rate control across various null response rates and power, among other performance metrics. Results: The logistic regression probability model engenders smaller average sample sizes with similar power, better control over type I error rate, and more favorable treatment arm sample size distributions than the conventional beta-binomial probability model, and designs using the alternative randomization methods have a negligible chance of a sample size imbalance in the wrong direction. Conclusion: Pairing the logistic regression probability model with either of the alternative randomization methods results in a much improved response-adaptive design in regard to important operating characteristics, including type I error rate control and the risk of a sample size imbalance in favor of the inferior treatment.

scholarly journals Evaluating the role for renal biopsy in T1 and T2 renal masses: A single-centre study

2018 ◽  
Vol 12 (5) ◽  
pp. E226-30 ◽  
Author(s):  
Dylan Hoare ◽  
Howard Evans ◽  
Heidi Richards ◽  
Rahim Samji
Keyword(s):  
Logistic Regression ◽  
Adverse Events ◽  
Diagnostic Accuracy ◽  
Renal Biopsy ◽  
Predictive Value ◽  
Lesion Size ◽  
Renal Masses ◽  
Ct Guided ◽  
Binomial Logistic Regression ◽  
T1 And T2

Introduction: Once used primarily in the identification of renal metastasis and lymphomas, various urological bodies are now adopting an expanded role for the renal biopsy. We sought to evaluate the role of the renal biopsy in a Canadian context, focusing on associated adverse events, radiographic burden, and diagnostic accuracy.Methods: This retrospective review incorporated all patients undergoing ultrasound (US)/computed tomography (CT)-guided biopsies for T1 and T2 renal masses. There were no age or lesion size limitations. The primary outcome of interest was the correlation between initial biopsy and final surgical pathology. A binomial logistic regression analysis was conducted to determine any confounding factors. Secondary outcomes included the accuracy of tumour cell typing, grading, the safety profile, and radiographic burden associated with these patients.Results: A total of 148 patients satisfied inclusion criteria for this study. Mean age and lesions size at detection were 60.9 years (±12.4) and 3.6 cm (±2.0), respectively. Most renal masses were identified with US (52.7%) or CT (44.6%). Three patients (2.0%) experienced adverse events of note. Eighty-six patients (58.1%) proceeded to radical/partial nephrectomy. Our biopsies held a diagnostic accuracy of 90.7% (sensitivity 96.2%, specificity 87.5%, positive predictive value 98.7%, negative predictive value 70.0%, kappa 0.752, p<0.0005). Binomial logistic regression revealed that age, lesion size, number of radiographic tests, time to biopsy, and modality of biopsy (US/CT) had no influence on the diagnostic accuracy of biopsies.Conclusions: Renal biopsies are safe, feasible, and diagnostic. Their role should be expanded in the routine evaluation of T1 and T2 renal masses.

A National Cancer Database (NCDB) analysis of the impact of nonmetastatic colon cancer sidedness on survival.

2021 ◽  
Vol 39 (3_suppl) ◽  
pp. 139-139
Author(s):  
Deven Patel ◽  
Timothy DiPeri ◽  
Brian Cox ◽  
Andrew Eugene Hendifar ◽  
Arsen Osipov ◽  
...  
Keyword(s):  
Colon Cancer ◽  
Logistic Regression ◽  
Adjuvant Chemotherapy ◽  
Large Population ◽  
Stage I ◽  
Prognostic Impact ◽  
Metastatic Colon Cancer ◽  
Prognostic Features ◽  
Binomial Logistic Regression ◽  
The Impact

139 Background: Differences in embryological origin and tumor biology distinguish right-sided colon cancer (RCC) from left-sided colon cancer (LCC). Previous studies characterizing the prognostic impact of colon cancer laterality on clinical outcomes in non-metastatic colon cancer have been conflicting, thus closer examination is needed. Methods: Using the NCDB, patients with stage I-III colon cancer between 2004-2014 were stratified according to tumor location; RCC vs. LCC. Patient (pt) and tumor characteristics were compared in univariate analysis, survival (OS) was estimated by Kaplan-Meier (KM) curves and Cox proportional hazards modeling. Binomial logistic regression analysis was utilized to identify variables associated with colon cancer laterality. Results: Of the 342,735 pts who met inclusion criteria, 210,343 (61.4%) were diagnosed with RCC, and 132,392 (38.6%) with LCC. Pts with RCC were older (mean 71.6 vs. 66.4 years, p< 0.001) and predominantly female (65% vs. 35%, p< 0.001) compared to those with LCC. A trend towards poorer OS was seen in pts with RCC (mean 91.0 mos [95% CI: 90.2-91.8]) compared to LCC (112.2 mos [95% CI: 110.9-113.6]) in unadjusted analysis. On Cox multivariable adjusted analyses there was a significant but minimal impact on OS and laterality (hazard ratio or HR [LCC as ref] 0.978, 95% CI 0.967-0.989 p< 0.0001). Multiple unadjusted KM survival analyses showed RCC with T4 disease, high-grade, LVI/PNI, positive margins, N0-N2 disease, tumor deposits, and receipt of adjuvant chemotherapy had poorer OS than those features in LCC (all p < 0.0001). Binomial logistic regression showed RCCs were significantly more likely to be higher grade (odds ratio or OR 2.024) and MSI-H (OR 2.010) with trends (nonsignificant) towards more likely having N1-2 positive disease, LVI, less receipt of adjuvant chemotherapy, and fewer tumor deposits. Conclusions: The impact of sidedness on prognosis in stage I-III colon cancer is complex. In this large, population-based study, RCC tends to be associated with more adverse prognostic features than LCC. More investigation into the biologic differences between RCC and LCC is warranted and how they impact phenotype and survival.

Occurrence of Toxoplasma gondii antibodies and associated risk factors in women in selected districts of Punjab province, Pakistan

Parasitology ◽  
2020 ◽  
Vol 147 (10) ◽  
pp. 1133-1139
Author(s):  
Shahzad Ali ◽  
Zona Amjad ◽  
Tahir Mahmood Khan ◽  
Abdul Maalik ◽  
Anam Iftikhar ◽  
...  
Keyword(s):  
Risk Factors ◽  
Logistic Regression ◽  
Toxoplasma Gondii ◽  
Pregnant Women ◽  
Congenital Toxoplasmosis ◽  
Sampling Location ◽  
Serum Samples ◽  
Cross Sectional ◽  
Binomial Logistic Regression ◽  
Punjab Province

AbstractToxoplasmosis is a parasitic zoonotic disease caused by Toxoplasma (T.) gondii. Limited data are available on the occurrence of T. gondii in women especially pregnant women in Pakistan. The present study aimed to determine the occurrence and risk factors associated with T. gondii in pregnant and non-pregnant women in Punjab Province, Pakistan. A cross-sectional study was conducted and 593 samples were collected from pregnant (n = 293) and non-pregnant (n = 300) women of District Headquarter Hospitals of Chiniot, Faisalabad, Jhang and Okara, Pakistan. Data related to demographic parameters and risk factors were collected using a pretested questionnaire on blood sampling day. Serum samples were screened for antibodies (IgG) against T. gondii using ELISA. A univariant and binomial logistic regression was applied to estimate the association between seropositive and explanatory variables considering the 95% confidence interval. P value ⩽0.05 was considered statistically significant for all analysis. Out of 593, 44 (7.42%) women were seropositive for T. gondii IgG antibodies. Occupation, age, sampling location, socioeconomic status, contact with cat, pregnancy status and trimester of pregnancy were significantly associated with seropositivity for T. gondii antibodies. Location and trimester of pregnancy were identified as potential risk factors for T. gondii seropositivity based on binomial logistic regression. Toxoplasma gondii is prevalent in pregnant and non-pregnant women. Therefore, now a necessitated awareness is required to instruct the individuals about these infectious diseases (toxoplasmosis) and their control strategies to maintain the health of human population. Moreover, health awareness among public can help the minimization of T. gondii infection during pregnancy and subsequent risk of congenital toxoplasmosis.

scholarly journals The Association of Demographic Characteristics and Food Choice Motives with the Consumption of Functional Foods in Emerging Adults

Nutrients ◽  
2020 ◽  
Vol 12 (9) ◽  
pp. 2582 ◽  
Author(s):  
Lieke Vorage ◽  
Nicola Wiseman ◽  
Joana Graca ◽  
Neil Harris
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Emerging Adults ◽  
Food Choice ◽  
Functional Food ◽  
Logistic Regression Model ◽  
Functional Foods ◽  
Demographic Characteristics ◽  
Binomial Logistic Regression ◽  
Food Choice Motives

The functional food market is one of the fastest growing segments of the global food industry. The aims of this study were to understand the association of demographic characteristics and food choice motives (FCMs) with (a) attitudes toward functional foods and (b) consumption of functional foods in Australian emerging adults. Data were collected through a paper-based and online questionnaire completed by 370 young adults aged between 17 and 29 years. A binomial logistic regression was used to determine the association between demographic characteristics and FCMs with attitudes towards functional foods. The logistic regression model was statistically significant at χ2(11) = 48.310 (p < 0.001) and explained 18.1% of the variance in attitude towards functional food. Of the several predictors, only the FCMs natural content and weight control were statistically significant. A binomial logistic regression was also used to determine the association between demographic characteristics and FCMs with the consumption of functional foods. The logistic regression model was statistically significant at χ2(9) = 37.499 (p < 0.001) and explained 14.1% of the variance in functional food consumption. Of the eight predictors, three were statistically significant: living situation, natural content and health. Findings highlight that when targeting emerging adults, functional food companies could benefit from promoting the natural and health properties of their products. Furthermore, consumption can be increased by targeting the parents of emerging adults and by designing functional foods that attract emerging adults interested in controlling weight.

Stability of a Model Explaining Selected Extramusical Influences on Solo and Small-Ensemble Festival Ratings

2005 ◽  
Vol 53 (4) ◽  
pp. 358-374 ◽  
Author(s):  
Martin J. Bergee ◽  
Claude R. Westfall
Keyword(s):  
Logistic Regression ◽  
School Size ◽  
Time Of Day ◽  
Multinomial Model ◽  
Data Set ◽  
Binomial Logistic Regression ◽  
Small Ensemble ◽  
Daily Attendance ◽  
The Stability ◽  
Average Daily Attendance

This is the third study in a line of inquiry whose purpose has been to develop a theoretical model of selected extramusical variables' influence on solo and small-ensemble festival ratings. Authors of the second of these (Bergee & McWhirter, 2005) had used binomial logistic regression as the basis for their model formulation strategy. Their final model included as statistically significant variables time of day (morning/afternoon), type of event (solo/ensemble), performing medium (vocal/instrumental), school size classification (Larger/smaller), district level of expenditure per average daily attendance (high/middle/low), and type of event by performing medium interaction. For the present study, we examined the stability of their model for a different data set (the following year's ratings) by means of a similar but modified strategy. Among other modifications, we used multinomial instead of binomial logistic regression. Utimately, the present study's model converged strongly on Bergee and Mc Whirter's preliminary one. Time of day, type of event, school size, district expenditure per average daily attendance, geographical district (metropolitan/nonmetropolitan), and the time of day by geographical district interaction contributed significantly to the present study's multinomial model. Theoretical modeling thus far suggests that performing as a soloist later in the day and entering from a large, metropolitan-area, relatively high-expenditure school serve as success influences. The multinomial model showed a gradation of influences from ratings of I through II to < III.

Sign in / Sign up

Export Citation Format

Share Document