scholarly journals Sufficient Sample Size and Power in Multilevel Ordinal Logistic Regression Models

2016 ◽  
Vol 2016 ◽  
pp. 1-8 ◽  
Author(s):  
Sabz Ali ◽  
Amjad Ali ◽  
Sajjad Ahmad Khan ◽  
Sundas Hussain
Keyword(s):  
Multilevel Models ◽  
Outcome Variable ◽  
Variable Model ◽  
Ordered Categories ◽  
Logistic Regression Models ◽  
Ordinal Response ◽  
Cumulative Logit Model ◽  
Severe Intensity ◽  
Better Than ◽  
Ordinal Outcome

For most of the time, biomedical researchers have been dealing with ordinal outcome variable in multilevel models where patients are nested in doctors. We can justifiably apply multilevel cumulative logit model, where the outcome variable represents the mild, severe, and extremely severe intensity of diseases like malaria and typhoid in the form of ordered categories. Based on our simulation conditions, Maximum Likelihood (ML) method is better than Penalized Quasilikelihood (PQL) method in three-category ordinal outcome variable. PQL method, however, performs equally well as ML method where five-category ordinal outcome variable is used. Further, to achieve power more than 0.80, at least 50 groups are required for both ML and PQL methods of estimation. It may be pointed out that, for five-category ordinal response variable model, the power of PQL method is slightly higher than the power of ML method.

scholarly journals Bayesian models for prediction of the set-difference in volleyball

2021 ◽  
Author(s):  
Ioannis Ntzoufras ◽  
Vasilis Palaskas ◽  
Sotiris Drikos
Keyword(s):  
Logistic Regression ◽  
Multinomial Logistic Regression ◽  
Bayesian Models ◽  
Outcome Variable ◽  
Logistic Regression Models ◽  
Ordinal Response ◽  
Standard Models ◽  
Discrete Values ◽  
Using Data ◽  
Skellam Distribution

Abstract We study and develop Bayesian models for the analysis of volleyball match outcomes as recorded by the set-difference. Due to the peculiarity of the outcome variable (set-difference) which takes discrete values from $-3$ to $3$, we cannot consider standard models based on the usual Poisson or binomial assumptions used for other sports such as football/soccer. Hence, the first and foremost challenge was to build models appropriate for the set-difference of each volleyball match. Here we consider two major approaches: (a) an ordered multinomial logistic regression model and (b) a model based on a truncated version of the Skellam distribution. For the first model, we consider the set-difference as an ordinal response variable within the framework of multinomial logistic regression models. Concerning the second model, we adjust the Skellam distribution to account for the volleyball rules. We fit and compare both models with the same covariate structure as in Karlis & Ntzoufras (2003). Both models are fitted, illustrated and compared within Bayesian framework using data from both the regular season and the play-offs of the season 2016/17 of the Greek national men’s volleyball league A1.

scholarly journals The role of unintended pregnancy in internalized stigma among women living with HIV in Kenya

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Sara Chace Dwyer ◽  
Aparna Jain ◽  
Wilson Liambila ◽  
Charlotte E. Warren
Keyword(s):  
Unintended Pregnancy ◽  
Hiv Treatment ◽  
Outcome Variable ◽  
Internalized Stigma ◽  
Health Seeking ◽  
Logistic Regression Models ◽  
Stigma And Discrimination ◽  
Women Living With Hiv ◽  
Related Stigma ◽  
Living With Hiv

Abstract Background Kenya has successfully expanded HIV treatment, but HIV-related stigma and discrimination, and unintended pregnancy remain issues for many Kenyan women living with HIV. While HIV-related stigma can influence the health seeking behaviors of those living with HIV, less is known about how reproductive health outcomes influence internalized stigma among women living with HIV. Methods Baseline data only were used in this analysis and came from an implementation science study conducted in Kenya from 2015 to 2017. The analytic sample was limited to 1116 women who are living with HIV, between 18 to 44 years old, and have ever experienced a pregnancy. The outcome variable was constructed from 7 internalized stigma statements and agreement with at least 3 statements was categorized as medium/high levels of internalized stigma. Unintended pregnancy, categorized as unintended if the last pregnancy was mistimed or unwanted, was the key independent variable. Univariate and multivariate logistic regression models were used to assess the association between unintended pregnancy and internalized stigma. Associations between internalized stigma and HIV-related discrimination and violence/abuse were also explored. Results About 48% agreed with at least one internalized stigma statement and 19% agreed with at least three. Over half of women reported that their last pregnancy was unintended (59%). Within the year preceding the survey, 52% reported experiencing discrimination and 41% reported experiencing violence or abuse due to their HIV status. Women whose last pregnancy was unintended were 1.6 times (95% CI 1.2–2.3) more likely to have medium/high levels of internalized stigma compared to those whose pregnancy was wanted at the time, adjusting for respondents’ characteristics, experiences of discrimination, and experiences of violence and abuse. Women who experienced HIV-related discrimination in the past 12 months were 1.8 times (95% CI 1.3–2.6) more likely to have medium/high levels of internalized stigma compared to those who experienced no discrimination. Conclusions Results suggest that unintended pregnancy is associated with internalized stigma. Integrated HIV and FP programs in Kenya should continue to address stigma and discrimination while increasing access to comprehensive voluntary family planning services for women living with HIV.

scholarly journals The relationship between socioeconomic position and health literacy among urban and rural adults in regional China

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Wei Chen ◽  
Hongfu Ren ◽  
Na Wang ◽  
Yaqing Xiong ◽  
Fei Xu
Keyword(s):  
Health Literacy ◽  
Educational Attainment ◽  
Socioeconomic Position ◽  
Population Based ◽  
Outcome Variable ◽  
Average Income ◽  
Cross Sectional Survey ◽  
Cross Sectional ◽  
Logistic Regression Models ◽  
Rural Adults

Abstract Background To examine associations of socioeconomic position (SEP), separately indicated by education, monthly family average income (FAI) and occupation, with health literacy (HL) among adults in regional China. Methods A cross-sectional survey was conducted among urban and rural adults (aged 25–69 years) who were randomly selected, using the probability proportionate to size sampling approach, from Nanjing municipality of China during October and December of 2016. HL, the outcome variable, was assessed using the Chinese Resident Health Literacy Scale. SEP, our independent variable, was separately measured with educational attainment, monthly family average income and occupation. Logistic regression models were introduced to examine SEP-HL association with odds ratio (OR) and 95% confidence interval (CI). Results Totally, 8698 participants completed the survey. The proportion of participants with unweighted and weighted adequate HL was 18.0% (95%CI = 17.2, 18.8%) and 19.9% (95%CI = 16.6, 23.6%), respectively, in this study. After adjustment for possible confounding factors, each SEP indicator was in significantly positive relation to both unweighted and weight HL level. Participants who obtained 13+ and 10–12 years educational attainment, respectively, had 2.41 (95%CI = 1.60, 3.64) and 1.68 (95%CI = 1.23, 2.29) times odds to record weighted adequate HL compared to their counterparts who were with 0–9 years education. Subjects within upper (OR = 1.92, 95%CI = 1.24, 2.98) and middle FAI tertile (OR = 1.59, 95%CI = 1.19, 2.13), respectively, were more likely to report weighted adequate HL relative to those who were within lower FAI tertile. White collars were more likely to have weighted adequate HL (OR = 1.33, 95%CI = 1.09, 1.61) than blue collars. Conclusions Each of education, FAI and occupation was positively associated with health literacy among urban and rural adults in China. The findings have important implications that different SEP indicators can be used to identify vulnerable residents in population-based health literacy promotion campaigns.

scholarly journals AutoSense Model for Word Sense Induction

2019 ◽  
Vol 33 ◽  
pp. 6212-6219 ◽  
Author(s):  
Reinald Kim Amplayo ◽  
Seung-won Hwang ◽  
Min Song
Keyword(s):  
Latent Variable ◽  
Word Sense ◽  
Name Disambiguation ◽  
Variable Model ◽  
Fine Grained ◽  
Word Sense Induction ◽  
Author Name Disambiguation ◽  
Competing Models ◽  
Word Senses ◽  
Better Than

Word sense induction (WSI), or the task of automatically discovering multiple senses or meanings of a word, has three main challenges: domain adaptability, novel sense detection, and sense granularity flexibility. While current latent variable models are known to solve the first two challenges, they are not flexible to different word sense granularities, which differ very much among words, from aardvark with one sense, to play with over 50 senses. Current models either require hyperparameter tuning or nonparametric induction of the number of senses, which we find both to be ineffective. Thus, we aim to eliminate these requirements and solve the sense granularity problem by proposing AutoSense, a latent variable model based on two observations: (1) senses are represented as a distribution over topics, and (2) senses generate pairings between the target word and its neighboring word. These observations alleviate the problem by (a) throwing garbage senses and (b) additionally inducing fine-grained word senses. Results show great improvements over the stateof-the-art models on popular WSI datasets. We also show that AutoSense is able to learn the appropriate sense granularity of a word. Finally, we apply AutoSense to the unsupervised author name disambiguation task where the sense granularity problem is more evident and show that AutoSense is evidently better than competing models. We share our data and code here: https://github.com/rktamplayo/AutoSense.

Abstract 3775: Average Blood Glucose and Mortality in Patients Hospitalized with Acute Myocardial Infarction

Circulation ◽  
2007 ◽  
Vol 116 (suppl_16) ◽  
Author(s):  
Mikhail Kosiborod ◽  
Silvio Inzucchi ◽  
Harlan M Krumholz ◽  
Lan Xiao ◽  
Phillip G Jones ◽  
...  
Keyword(s):  
Blood Glucose ◽  
Hospital Mortality ◽  
Regression Models ◽  
Independent Predictor ◽  
Laboratory Data ◽  
Logistic Regression Models ◽  
Average Blood Glucose ◽  
The Relationship ◽  
Multivariable Adjustment ◽  
Better Than

Background: Elevated blood glucose (BG) on admission is associated with higher mortality risk in patients (pts) hospitalized with AMI. However, the prognostic value of average BG, which reflects overall glycemic exposure much better than admission BG, is unknown. Furthermore, the nature of the relationship between average BG and mortality has not been determined. Methods: We evaluated a cohort of 16,871 AMI pts hospitalized from January 2000-December 2005, using Cerner Corporation’s Health Facts® database from 40 hospitals, which contains demographics, clinical and comprehensive laboratory data. Logistic regression models evaluated the nature of the relationship between mean BG during the entire AMI hospitalization and in-hospital mortality, after adjusting for multiple patient factors and confounders. Similar analyses were performed in subgroups of pts with and without diabetes (DM). Results: A J-shaped relationship was observed between mean BG and in-hospital mortality, which persisted after multivariable adjustment (Figure ). Mortality increased with each 10 mg/dL incremental rise in mean BG over >120 mg/dL, and with incremental decline in mean BG <80 mg/dL. The slope of these relationships was much steeper in pts without DM. Conclusions: Average BG during the entire AMI hospitalization is a powerful independent predictor of in-hospital mortality. Both persistent hyper- and hypoglycemia are associated with adverse prognosis. Whether strategies directed at optimizing BG control will improve survival remains to be established. Association Between Mean BG and In-Hospital Mortality After Multivariable Adjustment (Reference: Mean BG 100 to <110)

scholarly journals Do Socioeconomic Conditions Explain Ethnic Inequalities in Tooth Loss among US Adults?

2018 ◽  
Vol 28 (3) ◽  
pp. 201
Author(s):  
Fatma W. Nazer ◽  
Wael Sabbah
Keyword(s):  
Ethnic Differences ◽  
Tooth Loss ◽  
Family Income ◽  
The United States ◽  
Outcome Variable ◽  
Behavioral Risk ◽  
Cross Sectional Survey ◽  
Cross Sectional ◽  
Logistic Regression Models ◽  
Ethnic Inequalities

<p class="Pa7"><strong>Objective: </strong>To assess whether there are ethnic differences in tooth loss among adult Americans aged &lt;40 years and whether socioeconomic position attenuates these differences if they exist.</p><p class="Pa7"><strong>Methods: </strong>Data were from the 2014 Behavioral Risk Factor Surveillance System, a health-related telephone cross-sectional survey of a nationally representative sample of US adults. Tooth loss (one tooth or more) was used as the outcome variable. Ethnicity was the main explanatory variable. Family income, education and health insurance were also used in the analysis. Logistic regression models for tooth loss were constructed adjusting for demographic (age, sex, and ethnicity), socioeconomic indica­tors (income and education), health insur­ance, dental visits, smoking and diabetes.</p><p class="Pa7"><strong>Results: </strong>A total of 76,273 participants were included in the analysis. The prevalence of tooth loss was highest among Blacks (33.7%). Hispanics and other ethnic groups had a higher prevalence of tooth loss than Whites, 29.1% (95%CI: 27.7-30.6), 22.0% (95%CI: 20.3-23.8), and 20.8% (95%CI: 20.2-21.4), respectively. Blacks had odds ra­tios (OR) 1.98 (95%CI: 1.81-2.16) for tooth loss compared with Whites. After adjusting for socioeconomic positions (SEP), the rela­tionship attenuated but remained significant with OR 1.71 (95%CI: 1.55-1.90).</p><p class="Default"><strong>Conclusions: </strong>Despite recent changes in the health care system in the United States, ethnic inequalities in tooth loss still exist. Income and education partially explained ethnic differences in tooth loss among Americans aged &lt;40 years.</p><p class="Default"><em>Ethn Dis. </em>2018;28(3):201-206; doi:10.18865/ ed.28.3.201</p>

scholarly journals Greater effects of mutual cooperation and defection on subsequent cooperation in direct reciprocity games than generalized reciprocity games: Behavioral experiments and analysis using multilevel models

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0242607
Author(s):  
Yutaka Horita
Keyword(s):  
Multilevel Models ◽  
Cooperative Behavior ◽  
The Other ◽  
Behavioral Strategy ◽  
Behavioral Models ◽  
Behavioral Differences ◽  
Behavioral Experiments ◽  
Mutual Cooperation ◽  
Shift Strategy ◽  
Better Than

Reciprocity toward a partner’s cooperation is a fundamental behavioral strategy underlying human cooperation not only in interactions with familiar persons but also with strangers. However, a strategy that takes into account not only one’s partner’s previous action but also one’s own previous action—such as a win-stay lose-shift strategy or variants of reinforcement learning—has also been considered an advantageous strategy. This study investigated empirically how behavioral models can be used to explain the variances in cooperative behavior among people. To do this, we considered games involving either direct reciprocity (an iterated prisoner’s dilemma) or generalized reciprocity (a gift-giving game). Multilevel models incorporating inter-individual behavioral differences were fitted to experimental data using Bayesian inference. The results indicate that for these two types of games, a model that considers both one’s own and one’s partner’s previous actions fits the empirical data better than the other models. In the direct reciprocity game, mutual cooperation or defection—rather than relying solely on one’s partner’s previous actions—affected the increase or decrease, respectively, in subsequent cooperation. Whereas in the generalized reciprocity game, a weaker effect of mutual cooperation or defection on subsequent cooperation was observed.

scholarly journals Vaccine hesitancy among paediatric nurses: Prevalence and associated factors

PLoS ONE ◽  
2021 ◽  
Vol 16 (5) ◽  
pp. e0251735
Author(s):  
Usue Elizondo-Alzola ◽  
Mireia G. Carrasco ◽  
Laia Pinós ◽  
Camila Andrea Picchio ◽  
Cristina Rius ◽  
...  
Keyword(s):  
Whooping Cough ◽  
Vaccination Schedule ◽  
Childhood Vaccination ◽  
Vaccine Hesitancy ◽  
Outcome Variable ◽  
Bivariate Analysis ◽  
Cross Sectional ◽  
Logistic Regression Models ◽  
Perception Of Safety ◽  
Adjusted Logistic Regression

Objective This study describes the prevalence of vaccine hesitancy associated with the Catalan systematic childhood vaccination calendar and some related psychosocial determinants among paediatric primary care nurses in Barcelona (Spain). Methods Cross-sectional descriptive study. In 2017 we invited the paediatric nurses (N = 165) working in Barcelona public primary health centres with paediatric departments (N = 41) to participate. They answered a questionnaire with sociodemographic and behavioural variables: severity and perceived probability of contracting the diseases of the vaccines in the vaccination schedule; safety and protection offered by each vaccine; and beliefs, social norms, and knowledge about vaccines. Outcome variable was vaccine hesitancy, dichotomized into not hesitant (nurses who would vaccinate their own offspring), and hesitant (including those who would not vaccinate them, those who had doubts and those who would delay the administration of one or more vaccines). We performed bivariate analysis and adjusted logistic regression models. Results 83% of paediatric nurses (N = 137) agreed to participate. 67.9% had the intention to vaccinate their children of all the vaccines in the systematic schedule. 32.1% of nurses experienced vaccine hesitancy, especially about the HPV (21.9%) and varicella (17.5%) vaccines. The multivariate analysis suggests associations between hesitancy and low perception of the severity of whooping cough (aOR: 3.88; 95%CI:1.32–11.4), low perception of safety of the HPV vaccine (aOR:8.5;95%CI:1.24–57.8), the belief that vaccines are administered too early (aOR:6.09;95%CI:1.98–18.8), and not having children (aOR:4.05;95%CI:1.22–13.3). Conclusions Although most paediatric nurses had the intention to vaccinate their own children, almost one-third reported some kind of vaccine hesitancy, mainly related to doubts about HPV and varicella vaccines, as well as some misconceptions. These factors should be addressed to enhance nurses’ fundamental role in promoting vaccination to families.

scholarly journals Disparities in injury morbidity among young adults in the USA: individual and contextual determinants

2018 ◽  
Vol 72 (6) ◽  
pp. 458-464 ◽  
Author(s):  
Susanne Schmidt ◽  
P Johnelle Sparks
Keyword(s):  
Young Adults ◽  
Conceptual Framework ◽  
Regression Models ◽  
Injury Risk ◽  
College Graduates ◽  
Multilevel Models ◽  
Health Behaviours ◽  
Population Data ◽  
Multilevel Regression ◽  
Logistic Regression Models

BackgroundInjuries have been recognised as important public health concerns, particularly among adolescents and young adults. Few studies have examined injuries using a multilevel perspective that addresses individual socioeconomic status (SES) and health behaviours and local socioeconomic conditions in early adolescence. We offer a conceptual framework incorporating these various components.MethodsWe test our conceptual framework using population data from the National Longitudinal Study of Adolescent Health Wave 4 when respondents were young adults and linked them to contextual level data from when they were middle-schoolers. We use logistic and multilevel regression models to examine self-reported injury risk in young adults by sex (n=14 356).ResultsLogistic regression models showed that men were more likely to experience serious injuries than women (OR 1.75, P<0.0001), but SES and health behaviours operated differently by sex. In stratified models, men with lower education had consistently higher injury risk, while only women with some college had increased injury risk (OR 1.40, P=0.0089) than college graduates. Low household income (OR 1.54, P=0.0011) and unemployment (OR 1.50, P=0.0008) increased female injury risk, but was non-significant for men. Alcohol consumption increased injury risk for both sexes, while only female smokers had elevated injury risk (OR 1.38, P=0.0154). In multilevel models, significant county-level variation was only observed for women. Women living in disadvantaged neighbourhoods during adolescence had increased injury risk (OR 1.001, P<0.0001).ConclusionsThese findings highlight the importance of investigating mechanisms that link early-life contextual conditions to early adult SES and health behaviours and their linkage to injury risk, particularly for women.

scholarly journals The updated Charlson comorbidity index is a useful predictor of mortality in patients with Staphylococcus aureus bacteraemia

2018 ◽  
Vol 146 (16) ◽  
pp. 2122-2130 ◽  
Author(s):  
H. G. Ternavasio-de la Vega ◽  
F. Castaño-Romero ◽  
S. Ragozzino ◽  
R. Sánchez González ◽  
M. P. Vaquero-Herrero ◽  
...  
Keyword(s):  
Staphylococcus Aureus ◽  
Charlson Comorbidity Index ◽  
Regression Models ◽  
Model Fit ◽  
Logistic Regression Models ◽  
Relevant Variables ◽  
Comorbidity Index ◽  
The Mean ◽  
Related Mortality ◽  
Better Than

AbstractThe objective was to compare the performance of the updated Charlson comorbidity index (uCCI) and classical CCI (cCCI) in predicting 30-day mortality in patients with Staphylococcus aureus bacteraemia (SAB). All cases of SAB in patients aged ⩾14 years identified at the Microbiology Unit were included prospectively and followed. Comorbidity was evaluated using the cCCI and uCCI. Relevant variables associated with SAB-related mortality, along with cCCI or uCCI scores, were entered into multivariate logistic regression models. Global model fit, model calibration and predictive validity of each model were evaluated and compared. In total, 257 episodes of SAB in 239 patients were included (mean age 74 years; 65% were male). The mean cCCI and uCCI scores were 3.6 (standard deviation, 2.4) and 2.9 (2.3), respectively; 161 (63%) cases had cCCI score ⩾3 and 89 (35%) cases had uCCI score ⩾4. Sixty-five (25%) patients died within 30 days. The cCCI score was not related to mortality in any model, but uCCI score ⩾4 was an independent factor of 30-day mortality (odds ratio, 1.98; 95% confidence interval, 1.05–3.74). The uCCI is a more up-to-date, refined and parsimonious prognostic mortality score than the cCCI; it may thus serve better than the latter in the identification of patients with SAB with worse prognoses.

Sign in / Sign up

Export Citation Format

Share Document