Firth-Type Penalized Methods of the Modified Poisson and Least-Squares Regression Analyses for Binary Outcomes.

The modified Poisson and least-squares regression analyses for binary outcomes have been widely used as effective multivariable analysis methods to provide risk ratio and risk difference estimates in clinical and epidemiological studies. However, there is no certain evidence that assessed their operating characteristics under small and sparse data settings and no effective methods have been proposed for these regression analyses to address this issue. In this article, we show that the modified Poisson regression provides seriously biased estimates under small and sparse data settings. In addition, the modified least-squares regression provides unbiased estimates under these settings. We further show that the ordinary robust variance estimators for both of the methods have certain biases under situations that involve small or moderate sample sizes. To address these issues, we propose the Firth-type penalized methods for the modified Poisson and least-squares regressions. The adjustment methods lead to a more accurate and stable risk ratio estimator under small and sparse data settings, although the risk difference estimator is not invariant. In addition, to improve the inferences of the effect measures, we provide an improved robust variance estimator for these regression analyses. We conducted extensive simulation studies to assess the performances of the proposed methods under real-world conditions and found that the accuracies of the point and interval estimations were markedly improved by the proposed methods. We illustrate the effectiveness of these methods by applying them to a clinical study of epilepsy.

Healthcare-seeking behavior for diarrhea in under-five children and associated factors in sub-Saharan Africa: a multilevel robust Poisson regression model.

Background: Diarrhea is one of the leading causes of child death in sub-Saharan Africa (SSA). Children with diarrhea who do not receive medical advice or treatment are at high risk of poor health outcomes and increased mortality. Prompt and adequate treatment is essential to mitigate these risks. However, studies that have been conducted on the factors influencing healthcare-seeking behavior (HSB) for diarrhea in under-five children in SSA are scarce. Therefore, the purpose of this research was to determine the variables related to HSB for diarrhea in children under the age of five. Methods: A secondary data analysis was conducted on the most recent data from the Demographic and Health Surveys in 35 SSA countries. The study included a total weighted sample of 51,791 children under the age of five with diarrhea. We presented the adjusted prevalence ratio and the 95% confidence interval in the multivariable multilevel robust Poisson regression analysis to show the statistical significance and strength of the association between HSB and its determinants. Results: The pooled prevalence of HSB for diarrhea in under-five children was 58.71% (95%CI: 55.39 to 62.04). Factors found to be associated with HSB included maternal age, education and working status, antenatal care visits, postnatal checkups for the child, wasting, distance to a health facility, SSA region, and country income level. Conclusion: More than 40% of under-five children with diarrhea in SSA did not receive medical advice or treatment. To improve healthcare-seeking behavior, effective health policy interventions are necessary. These include enhancing the education and employment status of mothers, promoting regular antenatal and postnatal care visits, building health facilities in close proximity, and raising awareness in the community about the importance of seeking healthcare services for malnourished children.

Logistic Regression: Limitations in the Estimation of Measures of Association with Binary Health Outcomes.

INTRODUCTION: Logistic regression models are frequently used to estimate measures of association between an exposure, health determinant or intervention, and a binary outcome. However, when the outcome is frequent (> 10%), model estimates for relative risks and prevalence ratios might be biased. Despite the availability of several alternatives, many still rely on these models, and a consensus is yet to be reached. We aimed to compare the estimation and goodness-of-fit of logistic, log-binomial and robust Poisson regression models, in cross-sectional studies involving frequent binary outcomes. METHODS: Two cross-sectional studies were conducted. Study 1 was a nationally representative study on the impact of air pollution on mental health. Study 2 was a local study on immigrants' access to urgent healthcare services. Odds ratios (OR) were obtained through logistic regression, and prevalence ratios (PR) through log-binomial and robust Poisson regression models. Confidence intervals (CI), their ranges, and standard-errors (SE) were also computed, along with models' relative goodness-of-fit through Akaike Information Criterion (AIC), when applicable. RESULTS: In Study 1, the OR (95% CI) was 1.015 (0.970 - 1.063), while the PR (95% CI) obtained through the robust Poisson mode was 1.012 (0.979 - 1.045). The log-binomial regression model did not converge in this study. In Study 2, the OR (95% CI) was 1.584 (1.026 - 2.446), the PR (95% CI) for the log-binomial model was 1.217 (0.978 - 1.515), and 1.130 (1.013 - 1.261) for the robust Poisson model. The 95% CI, their ranges, and the SE of the OR were higher than those of the PR, in both studies. However, in Study 2, the AIC value was lower for the logistic regression model. CONCLUSION: The odds ratio overestimated PR with wider 95% CI and higher SE. The overestimation was greater as the outcome of the study became more prevalent, in line with previous studies. In Study 2, the logistic regression was the model with the best fit, illustrating the need to consider multiple criteria when selecting the most appropriate statistical model for each study. Employing logistic regression models by default might lead to misinterpretations. Robust Poisson models are viable alternatives in cross-sectional studies with frequent binary outcomes, avoiding the non-convergence of log-binomial models.

Changepoint detection on daily home activity pattern: a sliced Poisson process method.

The problem of health and care of people is being revolutionized. An important component of that revolution is disease prevention and health improvement from home. A natural approach to the health problem is monitoring changes in people's behavior or activities. These changes can be indicators of potential health problems. However, due to a person's daily pattern, changes will be observed throughout each day, with, eg, an increase of events around meal times and fewer events during the night. We do not wish to detect such within-day changes but rather changes in the daily behavior pattern from one day to the next. To this end, we assume the set of event times within a given day as a single observation. We model this observation as the realization of an inhomogeneous Poisson process where the rate function can vary with the time of day. Then, we propose to detect changes in the sequence of inhomogeneous Poisson processes. This approach is appropriate for many phenomena, particularly for home activity data. Our methodology is evaluated on simulated data. Overall, our approach uses local change information to detect changes across days. At the same time, it allows us to visualize and interpret the results, changes, and trends over time, allowing the detection of potential health decline.

Genetic parameters of behavior traits of beef cattle classified using wearable devices.

With the development of wearable devices, it is now possible to monitor livestock behavior 24 h a day. In this study, we estimated the genetic parameters of the daily duration of six behaviors (feeding, moving, lying, standing, ruminating while lying, and ruminating while standing) in beef cattle, automatically classified using wearable devices. The devices were attached to 332 Japanese beef cattle at two stations for approximately 5 months. We compared repeatability, Poisson regression, and random regression models using the deviance information criterion. Poisson regression models were selected for all traits at each station, probably because of the non-normal distribution of the phenotypes. The heritability estimates by the Poisson regression models were moderate at each station: 0.67 and 0.68 for feeding, 0.68 and 0.53 for moving, 0.47 and 0.55 for lying, 0.45 and 0.40 for standing, 0.51 and 0.59 for ruminating while lying, and 0.37 and 0.45 for ruminating while standing. The genetic correlations between these traits were all negative at both stations, whereas the residual correlations showed different directions depending on the station. Although validation studies with larger populations are needed to confirm these findings, this study provides fundamental knowledge of the genetic basis of daily behavior in beef cattle.

[Numerical study on structural design and mechanical analysis of anti-migration tracheal stent with non-uniform Poisson's ratio].

Stent migration is one of the common complications after tracheal stent implantation. The causes of stent migration include size mismatch between the stent and the trachea, physiological movement of the trachea, and so on. In order to solve the above problems, this study designed a non-uniform Poisson ratio tracheal stent by combining the size and structure of the trachea and the physiological movement of the trachea to improve the migration of the stent, meanwhile ensuring the support of the stent. In this study, the stent corresponding to cartilage was constructed with negative Poisson's ratio, and the stent corresponding to the circular connective tissue and muscular membrane was constructed with positive Poisson's ratio. And four kinds of non-uniform Poisson's ratio tracheal stents with different link lengths and negative Poisson's ratio were designed. Then, this paper numerically simulated the expansion and rebound process of the stent after implantation to observe the support of the stent, and further simulated the stretch movement of the trachea to calculate the diameter changes of the stent corresponding to different negative Poisson's ratio structures. The axial migration of the stent was recorded by applying different respiratory pressure to the wall of the tracheal wall to evaluate whether the stent has anti-migration effect. The research results show that the non-uniform Poisson ratio stent with connecting rod length of 3 mm has the largest diameter expansion in the negative Poisson ratio section when the trachea was stretched. Compared with the positive Poisson's ratio structure, the axial migration during vigorous breathing was reduced from 0.024 mm to 0.012 mm. The negative Poisson's ratio structure of the non-uniform Poisson's ratio stent designed in this study did not fail in the tracheal expansion effect. Compared with the traditional stent, the non-uniform Poisson's ratio tracheal stent has an anti-migration effect under the normal movement of the trachea while ensuring the support force of the stent.

Hepatitis A vaccination coverage survey in 24-month-old children living in Brazilian capitals, 2020.

OBJECTIVE: To estimate hepatitis A vaccination coverage in 24-month-old children and identify factors associated with non-vaccination. METHODS: This was a survey involving a sample stratified by socioeconomic strata in capital cities (2020-2022), with coverage estimates and 95% confidence intervals (95%CI), the factor analysis was performed using the prevalence ratio (PR) by means of Poisson regression. RESULTS: Among 31,001 children, hepatitis A coverage was 88.1% (95%CI 86.8;89.2). Regarding socioeconomic strata (A/B), the variable immigrant parents/guardians was associated with non-vaccination (PR = 1.91; 95%CI 1.09;3.37); in strata C/D, children of Asian race/skin color (PR = 4.69; 95%CI 2.30;9.57), fourth-born child or later (PR = 1.68; 95%CI 1.06;2 .66), not attending daycare/nursery (PR = 1.67; 95%CI 1.24;2.24) and mother with paid work (PR = 1.42; 95%CI 1.16;1.74) were associated with non-vaccination. CONCLUSION: Hepatitis A coverage was below the target (95%), suggesting that specificities of social strata should be taken into consideration. MAIN RESULTS: Hepatitis A vaccination coverage was 88%. Non-vaccination was greater in children with immigrant guardians (strata A/B); of Asian race/skin color, fourth-born child or later, those not attending daycare/nursery and mother with paid work (C/D strata). IMPLICATIONS FOR SERVICES: The results of this study contributed to the Ministry of Health and Health Departments in monitoring vaccination coverage and identifying factors that may negatively impact hepatitis A vaccination coverage. PERSPECTIVES: Further research is needed on the impact of migration on hepatitis A vaccination and vaccination in general. Health managers should be attentive to the different factors affecting vaccination among social strata.

SPLANG-a synthetic poisson-lognormal-based abundance and network generative model for microbial interaction inference algorithms.

Microbes are pervasive and their interaction with each other and the environment can impact fields as diverse as health and agriculture. While network inference and related algorithms that use abundance data from pyrosequencing can infer microbial interaction networks, the ambiguity surrounding the actual underlying networks hampers the validation of these algorithms. This study introduces a generative model to synthesize both the underlying interactive network and observable abundance data, serving as a test bed for the existing and future network inference algorithms. We tested our generative model with four typical network inference algorithms; our results suggest that none of these algorithms demonstrate adequate accuracy for inferring ecologies of non-commensalistic species, either mutualistic or competitive. We further explored the potential for predictability by combining existing algorithms with an oracle algorithm built by fusing the results of several existing algorithms. The oracle algorithm reveals promising improvements in predictability, although it falls short when applied to networks characterized by dense interspecies taxa interactions. Our work underscores the need for the continued development and validation of algorithms to unravel the intricacies of microbial interaction networks.

Utilization of dental services by preschool children: prevalence and associated factors.

Dental associations worldwide recommend that the first dental visit should take place before 12 months of age; however, preschoolers' utilization of dental services remains low. The aim of this study was to assess the prevalence of, and factors associated with, dental services utilization among children aged 1 to 3 years. This was a cross-sectional study carried out in the city of Diamantina, MG, Brazil, and involved a sample of 308 child-mother pairs. Mothers completed a questionnaire addressing sociodemographic and economic aspects of the family and characteristics pertaining to their child's oral health. The clinical assessment of the children included dental caries, trauma, malocclusion, and mucosal changes. Analysis of the data comprised statistical description, application of the chi-square test, and Poisson's regression analysis. Among the children studied, 39.6% had attended at least one dental visit in their lifetime. Children whose families had a greater number of members relying on the family's income (PR = 1.40, 95%CI:1.04 -1.89, p = 0.028) and those with moderate/extensive dental caries (Codes 3-6 of the ICDAS; PR = 1.44, 95%CI: 1.08 -1.93, p = 0.014) exhibited a higher prevalence of dental services utilization. In conclusion, the prevalence of dental services utilization among children aged 1 to 3 years was low, and was associated with a greater number of family members relying on the family's income, and with the occurrence of moderate/extensive dental caries.

A Poisson regression approach for assessing morbidity risk and determinants among under five children in Nigeria.

In this paper, we have provided more insights on the relationship between under five morbidity in Nigeria and some background characteristics using a Poisson regression model and the most recent 2018 NDHS data on Acute Respiratory Infection (ARI), diarrhoea and fever. Some of our results are that children 36-47 months old have the highest risk of ARI [OR = 1.45; CI (1.31,1.60)] while children less than 6 months old have the lowest risk of ARI [OR = 0.14; CI (0.11,0.17)]. The prevalence of diarrhoea is generally high among children under 48-59 months old but highest among children 6-11 months old [OR = 4.34; CI (3.69,5.09)]. Compared to children 48-59 months old, children in all other age categories except 24-34 months old have a high risk of fever [OR = 0.95; CI (0.73,1.24)]. ARI is more prevalent among female children [OR = 8.88; CI (8.02,9.82)] while diarrhoea [OR = 21.75; (19.10,24.76)] and fever [OR = 4.78; CI (4.31,5.32)] are more prevalent among male children. Children in urban areas are more likely to suffer ARI [OR = 9.49; CI (8.31,10.85)] while children in rural areas are more likely to suffer both diarrhoea [OR = 21.75; CI (19.10,24.76)] and fever [OR = 4.90; CI (4.26,5.63)]. Children in the South-South have the highest risk of ARI [OR = 4.03; CI (3.65,4.454)] while children in the North Central have the lowest risk of ARI [OR = 1.55; CI (1.38,1.74)] and highest risk of diarrhoea [OR = 3.34; CI (2.30,5.11)]. Children in the Northeast have the highest risk of fever [OR = 1.30; CI (1.14,1.48)]. In the Northcentral region, Kogi state has the highest prevalence of fever [OR = 2.27; CI (1.62,3.17)], while Benue state has the lowest [OR = 0.35; CI (0.20,0.60)]. Children in Abuja state face similar risks of fever and diarrhoea [OR = 0.84; CI (0.55,1.27)], with the risk of diarrhoea in Abuja being comparable to that in Plateau state [OR = 1.57; CI (0.92,2.70)]. Nasarawa state records the highest incidence of diarrhoea in the Northcentral [OR = 5.12; CI (3.03,8.65)], whereas Kogi state reports the lowest [OR = 0.29; CI (0.16,0.53)]. In the Northeast, Borno state has the highest rate of fever [OR = 3.28; CI (2.80,3.84)], and Bauchi state the lowest [OR = 0.38; CI (0.29,0.50)]. In Adamawa state, the risks of fever and diarrhoea are nearly equivalent [OR = 1.17; CI (0.97,1.41)], and the risk of fever there is similar to that in Taraba state [OR = 0.92; CI (0.75,1.12)]. Diarrhoea is most prevalent in Yobe state [OR = 3.17; CI (2.37,4.23)] and least prevalent in Borno state [OR = 0.26; CI (0.20,0.33)]. In the Northwest, the risk of fever is similarly high in Zamfara and Kebbi states [OR = 1.04; CI (0.93,1.17)], with Kastina state showing the lowest risk [OR = 0.39; CI (0.34,0.46)]. Children in Zamfara state experience notably different risks of fever and diarrhoea [OR = 0.07; CI (0.05,0.10)]. Kaduna state reports the highest incidence of diarrhoea [OR = 21.88; CI (15.54,30.82)], while Kano state has the lowest [OR = 2.50; CI (1.73,3.63)]. In the Southeast, Imo state leads in fever incidence [OR = 8.20; CI (5.61,11.98)], while Anambra state has the lowest [OR = 0.40; CI (0.21,0.78)]. In Abia state, the risk of fever is comparable to that in Enugu state [OR = 1.03; CI (0.63,1.71)], but the risks of fever and diarrhoea in Abia differ significantly [OR = 2.67; CI (1.75,4.06)]. Abia state also has the highest diarrhoea rate in the Southeast [OR = 2.67; CI (1.75,4.06)], with Ebonyi state having the lowest [OR = 0.05; CI (0.03,0.09)]. In the South-South region, Bayelsa and Edo states have similar risks of fever [OR = 1.28; CI (0.84,1.95)], with Akwa Ibom state reporting the highest fever rate [OR = 4.62; CI (3.27,6.52)] and Delta state the lowest [OR = 0.08; CI (0.02,0.25)]. Children in Bayelsa state face distinctly different risks of fever and diarrhoea [OR = 0.56; CI (0.34,0.95)]. Rivers state shows the highest incidence of diarrhoea in the South-South [OR = 10.50; CI (4.78,23.06)], while Akwa Ibom state has the lowest [OR = 0.30; CI (0.15,0.57)]. In the Southwest, Lagos and Osun states have similar risks of fever [OR = 1.00; CI (0.59,1.69)], with Ogun state experiencing the highest incidence [OR = 3.47; CI (2.28,5.28)] and Oyo state the lowest [OR = 0.18; CI (0.07,0.46)]. In Lagos state, the risks of fever and diarrhoea are comparable [OR = 0.96; CI (0.57,1.64)], and the risk of diarrhoea is similar to those in Ekiti, Ogun, and Ondo states. Oyo state has the highest diarrhoea rate in the Southwest [OR = 10.99; CI (3.81,31.67)], with Ogun state reporting the lowest [OR = 0.77; CI (0.42,1.42)]. Children of mothers with more than secondary education are significantly less likely to suffer ARI [OR = 0.35; CI (0.29,0.42)], whereas children of mothers without any education run a higher risk of diarrhoea [OR = 2.12; CI (1.89,2.38)] and fever [OR = 2.61; CI (2.34,2.91)]. Our analysis also indicated that household wealth quintile is a significant determinant of morbidity. The results in this paper could help the government and non-governmental agencies to focus and target intervention programs for ARI, diarrhoea and fever on the most vulnerable and risky under five groups and populations in Nigeria.

An inverse lomax-uniform poisson distribution and joint modeling of repeatedly measured and time-to-event data in the health sectors.

Methodological developments in different sectors like health, biomedical and biological areas are the recent burning issue in the statistical literature. The approach of implementing declining hazard function which is obtained by compounding truncated Poisson distribution and a lifetime distribution is a special concern in a few studies. In this paper we proposed a newly introduced distribution called inverse Lomax-Uniform Poisson distribution mostly applied in the health, biomedical, biological, and related sectors. Some basic statistical properties of the distribution are discussed. The capability of the model is checked by comparing it with six potential models by using a practical real data set. Based on the comparison techniques, the newly proposed model out performs all its counterparts. The simulation study is also conducted. Furthermore, the joint modelling of repeatedly measured and time-to-vent processes is discussed in detail with the real data set in the health sector.

Performance of a negative binomial-GLM in spatial scan statistic: a case study of low-birth weights in Pakistan.

Spatial cluster analyses of health events are useful for enabling targeted interventions. Spatial scan statistic is the stateof- the-art method for this kind of analysis and the Poisson Generalized Linear Model (GLM) approach to the spatial scan statistic can be used for count data for spatial cluster detection with covariate adjustment. However, its use for modelling is limited due to data over-dispersion. A Generalized Linear Mixed Model (GLMM) has recently been proposed for modelling this kind of over-dispersion by incorporating random effects to model area-specific intrinsic variation not explained by other covariates in the model. However, these random effects may exhibit a geographical correlation, which may lead to a potential spatial cluster being undetected. To handle the over-dispersion in the count data, this study aimed to evaluate the performance of a negative binomial- GLM in spatial scan statistic on real-world data of low birth weights in Khyber-Pakhtunkhwa Province, Pakistan, 2019. The results were compared with the Poisson-GLM and GLMM, showing that the negative binomial-GLM is an ideal choice for spatial scan statistic in the presence of over-dispersed data. With a covariate (maternal anaemia) adjustment, the negative binomial-GLMbased spatial scan statistic detected one significant cluster covering Dir lower district. Without the covariate adjustment, it detected two clusters, each covering one district. The district of Peshawar was seen as the most likely cluster and Battagram as the secondary cluster. However, none of the clusters were detected by GLMM spatial scan statistic, which might be due to the spatial correlation of the random effects in GLMM.

Association between bullying at school and tooth loss among 15-19-year-olds from southern Brazil.

This study aimed to investigate the association between bullying at school and tooth loss in southern Brazilian adolescents. This population-based cross-sectional study included a representative sample of 15-19-year-old students attending high schools in Santa Maria, southern Brazil. Data on sociodemographic and behavioral variables were collected through questionnaires. Contextual data on bullying at school was provided by educational institutions (bullying episodes in the previous year: 'no,' 'sometimes,' or 'often'). Tooth loss was clinically assessed by the M component of the DMFT index, modeled as a discrete variable. Multilevel Poisson regression was used, and rate ratios (RR) and 95% confidence intervals (CI) were estimated. The prevalence of tooth loss was 9.2% (95%CI = 7.5-10.8). Adolescents who attended the schools where bullying events often occurred had 0.39 (95%CI = 0.33-0.45) missing teeth, on average, in contrast to an average of 0.14 (95%CI = 0.08-0.19) among those whose schools did not experience bullying in the previous year. After adjusting for important cofactors, the contextual variable of bullying at school remained significantly associated with the study outcome. Adolescents who attended schools where bullying frequently occurred were 2.49-fold more likely to have an additional missing tooth than those whose school did not experience bullying in the previous year (RR = 2.49, 95%CI = 1.37-4.51, p = 0.003). In conclusion, the frequent bullying episodes at school were associated with more permanent teeth lost due to caries in this population. Hence, improving the school environment may improve the oral health of adolescents.

Unifying mortality forecasting model: an investigation of the COM-Poisson distribution in the GAS model for improved projections.

Forecasting mortality rates is crucial for evaluating life insurance company solvency, especially amid disruptions caused by phenomena like COVID-19. The Lee-Carter model is commonly employed in mortality modelling; however, extensions that can encompass count data with diverse distributions, such as the Generalized Autoregressive Score (GAS) model utilizing the COM-Poisson distribution, exhibit potential for enhancing time-to-event forecasting accuracy. Using mortality data from 29 countries, this research evaluates various distributions and determines that the COM-Poisson model surpasses the Poisson, binomial, and negative binomial distributions in forecasting mortality rates. The one-step forecasting capability of the GAS model offers distinct advantages, while the COM-Poisson distribution demonstrates enhanced flexibility and versatility by accommodating various distributions, including Poisson and negative binomial. Ultimately, the study determines that the COM-Poisson GAS model is an effective instrument for examining time series data on mortality rates, particularly when facing time-varying parameters and non-conventional data distributions.

Application of longitudinal multilevel zero inflated Poisson regression in modeling of infectious diseases among infants in Ethiopia.

BACKGROUND: In sub-Saharan African countries, preventable and manageable diseases such as diarrhea and acute respiratory infections still claim the lives of children. Hence, this study aims to estimate the rate of change in the log expected number of days a child suffers from Diarrhea (NOD) and flu/common cold (NOF) among children aged 6 to 11 months at the baseline of the study. METHODOLOGY: This study used secondary data which exhibit a longitudinal and multilevel structure. Based on the results of exploratory analysis, a multilevel zero-inflated Poisson regression model with a rate of change in the log expected NOD and NOF described by a quadratic trend was proposed to efficiently analyze both outcomes accounting for correlation between observations and individuals through random effects. Furthermore, residual plots were used to assess the goodness of fit of the model. RESULTS: Considering subject and cluster-specific random effects, the results revealed a quadratic trend in the rate of change of the log expected NOD. Initially, low dose iron Micronutrient Powder (MNP) users exhibited a higher rate of change compared to non-users, but this trend reversed over time. Similarly, the log expected NOF decreased for children who used MNP and exclusively breastfed for six months, in comparison to their counterparts. In addition, the odds of not having flu decreased with each two-week increment for MNP users, as compared to non-MNP users. Furthermore, an increase in NOD resulted in an increase in the log expected NOF. Region and exclusive breastfeeding also have a significant relationships with both NOD and NOF. CONCLUSION: The findings of this study underscore the importance of commencing analysis of data generated from a study with exploratory analysis. The study highlights the critical role of promoting EBF for the first six months and supporting children with additional food after six months to reduce the burden of infectious diseases.

Automated Poisson regression exposure-response analysis for binary outcomes with PoissonERM.

PoissonERM is an R package used to conduct exposure-response (ER) analysis on binary outcomes for establishing the relationship between exposure and the occurrence of adverse events (AE). While Poisson regression could be implemented with glm(), PoissonERM provides a simple way to semi-automate the entire analysis and generate an abbreviated report as an R markdown (Rmd) file that includes the essential analysis details with brief conclusions. PoissonERM processes the provided data set using the information from the user's control script and generates summary tables/figures for the exposure metrics, covariates, and event counts of each endpoint (each type of AE). After checking the incidence rate of each AE, the correlation, and missing values in each covariate, an exposure-response model is developed for each endpoint based on the provided specifications. PoissonERM has the flexibility to incorporate and compare multiple scale transformations in its modeling. The best exposure metric is selected based on a univariate model's p-value or deviance ( Δ D ) as specified. If a covariate search is specified in the control script, the final model is developed using backward elimination. PoissonERM identifies and avoids highly correlated covariates in the final model development of each endpoint. Predicting event incidence rates using external (simulated) exposure metric data is an additional functionality in PoissonERM, which is useful to understand the event occurrence associated with certain dose regimens. The summary outputs of the cleaned data, model developments, and predictions are saved in the working folder and can be compiled into a HTML report using Rmd.

Comparison of univariate and bivariate Poisson regression methods in the analysis of determinants of female schooling and fertility in Malawi.

Recent research has established existence of a correlation between women's education and fertility, suggesting that they share similar risk factors. However, in many studies, the two variables were analysed separately, which could bias the conclusions by undermining the apparent correlations of such paired outcomes. In this article, the univariate and bivariate Poisson regression models were applied to nationally representative sample of 24,562 women from the 2015-16 Malawi demographic and health survey to examine the risk factors of women's education levels and fertility. The R software version 4.1.2 was used for the analyses. The results showed that estimates from the bivariate Poisson model were consistent with those obtained from the separate univariate Poisson models. The sizes of estimates of coefficients, their standard errors, p-values, and directions were comparable in both bivariate and univariate Poisson models. Using either the univariate or bivariate Poisson model, it was found that the age of a woman at first sexual experience, her current age, household wealth index, and contraceptive usage were significantly associated with both the woman's schooling and fertility. The study further revealed that ethnicity, religion, and region of residence impacted education level only and not fertility. Similarly, marital status and occupation impacted fertility only and not education. The study also found that higher education levels were linked to a lower number of children, with a strong negative correlation of -0.62 between the two variables. The study recommends using bivariate Poisson regression for analysing paired count response data, when there is an apparent covariance between the outcome variables. The results suggest that efforts by policymakers to achieve the desired women's sexual and reproductive health in sub-Saharan Africa should be intertwined with improving women's and girls' education attainment in the region.

A simulation study of the performance of statistical models for count outcomes with excessive zeros.

BACKGROUND: Outcome measures that are count variables with excessive zeros are common in health behaviors research. Examples include the number of standard drinks consumed or alcohol-related problems experienced over time. There is a lack of empirical data about the relative performance of prevailing statistical models for assessing the efficacy of interventions when outcomes are zero-inflated, particularly compared with recently developed marginalized count regression approaches for such data. METHODS: The current simulation study examined five commonly used approaches for analyzing count outcomes, including two linear models (with outcomes on raw and log-transformed scales, respectively) and three prevailing count distribution-based models (ie, Poisson, negative binomial, and zero-inflated Poisson (ZIP) models). We also considered the marginalized zero-inflated Poisson (MZIP) model, a novel alternative that estimates the overall effects on the population mean while adjusting for zero-inflation. Motivated by alcohol misuse prevention trials, extensive simulations were conducted to evaluate and compare the statistical power and Type I error rate of the statistical models and approaches across data conditions that varied in sample size ( N = 100 $$ N=100 $$ to 500), zero rate (0.2 to 0.8), and intervention effect sizes. RESULTS: Under zero-inflation, the Poisson model failed to control the Type I error rate, resulting in higher than expected false positive results. When the intervention effects on the zero (vs. non-zero) and count parts were in the same direction, the MZIP model had the highest statistical power, followed by the linear model with outcomes on the raw scale, negative binomial model, and ZIP model. The performance of the linear model with a log-transformed outcome variable was unsatisfactory. CONCLUSIONS: The MZIP model demonstrated better statistical properties in detecting true intervention effects and controlling false positive results for zero-inflated count outcomes. This MZIP model may serve as an appealing analytical approach to evaluating overall intervention effects in studies with count outcomes marked by excessive zeros.

Multivariate Poisson cokriging: A geostatistical model for health count data.

Multivariate disease mapping is important for public health research, as it provides insights into spatial patterns of health outcomes. Geostatistical methods that are widely used for mapping spatially correlated health data encounter challenges when dealing with spatial count data. These include heterogeneity, zero-inflated distributions and unreliable estimation, and lead to difficulties when estimating spatial dependence and poor predictions. Variability in population sizes further complicates risk estimation from the counts. This study introduces multivariate Poisson cokriging for predicting and filtering out disease risk. Pairwise correlations between the target variable and multiple ancillary variables are included. By means of a simulation experiment and an application to human immunodeficiency virus incidence and sexually transmitted diseases data in Pennsylvania, we demonstrate accurate disease risk estimation that captures fine-scale variation. This method is compared with ordinary Poisson kriging in prediction and smoothing. Results of the simulation study show a reduction in the mean square prediction error when utilizing auxiliary correlated variables, with mean square prediction error values decreasing by up to 50%. This gain is further evident in the real data analysis, where Poisson cokriging yields a 74% drop in mean square prediction error relative to Poisson kriging, underscoring the value of incorporating secondary information. The findings of this work stress on the potential of Poisson cokriging in disease mapping and surveillance, offering richer risk predictions, better representation of spatial interdependencies, and identification of high-risk and low-risk areas.

rtestim: Time-varying reproduction number estimation with trend filtering.

To understand the transmissibility and spread of infectious diseases, epidemiologists turn to estimates of the instantaneous reproduction number. While many estimation approaches exist, their utility may be limited. Challenges of surveillance data collection, model assumptions that are unverifiable with data alone, and computationally inefficient frameworks are critical limitations for many existing approaches. We propose a discrete spline-based approach that solves a convex optimization problem-Poisson trend filtering-using the proximal Newton method. It produces a locally adaptive estimator for instantaneous reproduction number estimation with heterogeneous smoothness. Our methodology remains accurate even under some process misspecifications and is computationally efficient, even for large-scale data. The implementation is easily accessible in a lightweight R package rtestim.