Odds Ratio or Prevalence Ratio? An Overview of Reported Statistical Methods and Appropriateness of Interpretations in Cross-sectional Studies with Dichotomous Outcomes in Veterinary Medicine

Martinez, Brayan Alexander Fonseca; Leotti, Vanessa Bielefeldt; Silva, Gustavo S.; Nunes, Luciana Neves; Machado, Gustavo; Corbellini, Luís Gustavo

doi:10.3389/fvets.2017.00193

Cited by 142 publications

(118 citation statements)

References 44 publications

(75 reference statements)

Supporting

Mentioning

114

Contrasting

Unclassified

Order By: Relevance

“…Binomial models were preferred because they provide prevalence ratios directly. Secondly unlike the alternative logistic regression log binomial models do not overestimate their coefficients when the outcome of interest is a common occurrence[28, 29] although they also have problems of lack of convergence[19]. In the few times non-convergence occurred we used modified Poisson regression which solves the problem but it is not also perfect since it produces inconsistent variances[19].…”

Section: Methodsmentioning

confidence: 99%

Alcohol consumption, obesity and hypertension: Relationship patterns along different age groups in Uganda

Tumwesigye

Mutungi

Bahendeka

et al. 2019

Preprint

View full text Add to dashboard Cite

IntroductionUganda is experiencing a significant increase in the prevalence of non-communicable diseases including hypertension and obesity. Frequent alcohol use is also highly prevalent in Uganda and is a key risk factor for both hypertension and obesity. This study determines the trends of frequent alcohol consumption, hypertension and obesity across different age groups, and the extent to which alcohol consumption affects the two.MethodsThe data were extracted from the 2014 National Non-communicable Diseases Risk Factor Survey (N=3,987) conducted among adults aged 18 to 69 years. Hypertension was defined as systolic blood pressure ≥140mmHG or diastolic blood pressure ≥90. Obesity was defined as body mass index >30 kg/m2. Frequent alcohol consumption was defined as alcohol use 3 or more times a week. Multivariable log binomial regression analysis was carried out for each of the two outcome variables against age group and controlled for frequency of alcohol consumption and few other independent factors. Non-parametric tests were used to compare trends of prevalence ratios across age groups. Modified Poisson regression was use in few instances when the model failed to converge.ResultsThe results showed increasing trend in the prevalence of hypertension and frequent alcohol consumption but a declining trend for obesity along different age groups (p<0.01). Frequency of alcohol consumption did not significantly modify the age group-hypertension and age group-obesity relationships although the effect was significant with ungrouped age. There was significance in difference of fitted lines for hypertension prevalence ratios between frequent drinkers and mild drinkers and between abstainers and frequent drinkers. Alcohol consumption did not have any significant effect on obesity-age group relationship.ConclusionThe results call for more research to understand the effect of alcohol on the hypertension-age relationship, and the obesity-age relationship. Why prevalence ratios for hypertension decline among those who take alcohol most frequently is another issue that needs further research.

show abstract

Section: Methodsmentioning

confidence: 99%

Alcohol consumption, obesity and hypertension: Relationship patterns along different age groups in Uganda

Tumwesigye

Mutungi

Bahendeka

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…We examined the associations between unmet needs and each independent variable by calculating the prevalence ratio (PR) using a log-binomial regression model which is preferable for the outcome variable with high prevalence. 43 Variables with p-value of ≤0.25 in the bivariate analysis were considered in the multivariable analysis and adjusted PR (APR) along with 95% CIs were estimated and a p-value <0.05 was used to declare the statistical significance. Log-likelihood ratio test and Akaike's and Bayesian information criterion were used to select the final model.…”

Section: Data Management and Analysismentioning

confidence: 99%

<p>Unmet Need for Contraception Among Young Married Women in Eastern Ethiopia</p>

Dingeta

Oljira

Worku

et al. 2019

OAJC

View full text Add to dashboard Cite

Purpose: Addressing the contraceptive needs of young married women is critical to improve their health and well-being. In patriarchal societies, young married women are under intense pressure to demonstrate their fecundity. Therefore, research that specifically address the needs for contraception of young married women have been generally given less emphasis in Ethiopia. This study assessed the extent of unmet needs for contraception and its associated factors among young married women in Eastern Ethiopia. Methods: A community-based cross-sectional study was conducted among young married women (14-24 years of age) in Eastern Ethiopia. Data were collected using a structured questionnaire. The prevalence ratio (PR) with 95% confidence intervals (CIs) was calculated, and factors associated with unmet needs for contraception were identified using log-binomial regression statistical model. Results: Among 2933 young women interviewed, the unmet need for contraception was 1014 (34.6%; 95% CI, 32.9%-36.4%). The prevalence of unmet needs for contraception decreased with increased young women's household decision-making autonomy score (APR= 0. 76; CI=0.62-0.94). Exposure to family planning (FP) information during the last 12 months (APR= 1.24; CI=1.1-1.42), age 18 or more years (APR=1.25; CI=1.04-1.5), multiparty (APR= 1. 9; CI=1.7-2.1) and desire to have 5 children or lesser than 5 children (APR= 1. 2; CI=1.06-1.32) were associated with higher prevalence of unmet needs for contraception. Conclusion: One-third of married young women had unmet needs for contraception. More efforts to empower women to make decisions that affect their own life and providing appropriate family planning information are necessary to reduce the burden of unmet needs among young married women in rural settings in Ethiopia.

show abstract

“…The output of the model was the Prevalence Ratio (PR). A log‐binomial regression was chosen owing to the high frequency of occurrence of mediastinal lymphoma, because using logistic regression would overestimate the PR . The variable age was categorized as ‘young’ (less than 36 months), ‘adult’ (more or equal to 36 months and less than 96 months) and ‘aged’ (more than 96 months).…”

Section: Methodsmentioning

confidence: 99%

Epidemiological, pathological and immunohistochemical aspects of 125 cases of feline lymphoma in Southern Brazil

Leite-Filho

Panziera

Bandinelli

et al. 2019

Vet Comparative Oncology

View full text Add to dashboard Cite

A retrospective study compiling cases of feline lymphoma diagnosed during 12 years (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016) in Southern Brazil was performed. A total of 125 cases of lymphoma diagnosed in cats were reviewed, and information including age, breed, sex and tumour topography were collected. FeLV and FIV immunohistochemical tests were performed, as well as immunophenotyping of lymphomas. The alimentary form represented the most common presentation (42/125), followed by mediastinal lymphoma (35/125). Out of 125 cases, 79 presented positive retroviral immunostaining in tumour tissue (52 FeLV alone, 14 FIV alone and 13 presented FIV and FeLV co-infections), 66/125 of the cases were of T-cell origin and 59/125 of the cases were of Bcell origin. The median age of cats with T-cell lymphoma was 120 months (10-240 months), and 60 months (6-204 months) for cats with B-cell lymphoma. The most frequent alimentary tumour presentation was the enteropathy-associated T-cell lymphoma (type 1), and the major type of mediastinal tumour observed was diffuse large B-cell lymphoma. Considering only mediastinal and alimentary lymphomas (n = 77), the prevalence of mediastinal lymphoma in FeLV-positive cats was 2.21 times higher than the prevalence of this type of tumour in FeLV-negative cats (P = .036).Mediastinal lymphoma was more frequently observed in younger cats, and the prevalence of mediastinal tumours in these animals was 3.06 times higher than the prevalence of this tumour form in old cats (P = .0125). The present study indicates that retroviral infections still play an important role in the development of feline lymphomas in southern Brazil.

show abstract

Odds Ratio or Prevalence Ratio? An Overview of Reported Statistical Methods and Appropriateness of Interpretations in Cross-sectional Studies with Dichotomous Outcomes in Veterinary Medicine

Cited by 142 publications

References 44 publications

Alcohol consumption, obesity and hypertension: Relationship patterns along different age groups in Uganda

Alcohol consumption, obesity and hypertension: Relationship patterns along different age groups in Uganda

<p>Unmet Need for Contraception Among Young Married Women in Eastern Ethiopia</p>

Epidemiological, pathological and immunohistochemical aspects of 125 cases of feline lymphoma in Southern Brazil

Contact Info

Product

Resources

About