-
Scientific Reports Jul 2023Among diseases, cancer exhibits the fastest global spread, presenting a substantial challenge for patients, their families, and the communities they belong to. This...
Among diseases, cancer exhibits the fastest global spread, presenting a substantial challenge for patients, their families, and the communities they belong to. This paper is devoted to modeling such a disease as a special case. A newly proposed distribution called the binomial-discrete Erlang-truncated exponential (BDETE) is introduced. The BDETE is a mixture of binomial distribution with the number of trials (parameter [Formula: see text]) taken after a discrete Erlang-truncated exponential distribution. A comprehensive mathematical treatment of the proposed distribution and expressions of its density, cumulative distribution function, survival function, failure rate function, Quantile function, moment generating function, Shannon entropy, order statistics, and stress-strength reliability, are provided. The distribution's parameters are estimated using the maximum likelihood method. Two real-world lifetime count data sets from the cancer disease, both of which are right-skewed and over-dispersed, are fitted using the proposed BDETE distribution to evaluate its efficacy and viability. We expect the findings to become standard works in probability theory and its related fields.
Topics: Humans; Reproducibility of Results; Statistical Distributions; Entropy; Neoplasms
PubMed: 37507433
DOI: 10.1038/s41598-023-38709-2 -
Infectious Disease Modelling Dec 2023Accurately estimating the effective reproduction number is crucial for characterizing the transmissibility of infectious diseases to optimize interventions and responses...
Accurately estimating the effective reproduction number is crucial for characterizing the transmissibility of infectious diseases to optimize interventions and responses during epidemic outbreaks. In this study, we improve the estimation of the effective reproduction number through two main approaches. First, we derive a discrete model to represent a time series of case counts and propose an estimation method based on this framework. We also conduct numerical experiments to demonstrate the effectiveness of the proposed discretization scheme. By doing so, we enhance the accuracy of approximating the underlying epidemic process compared to previous methods, even when the counting period is similar to the mean generation time of an infectious disease. Second, we employ a negative binomial distribution to model the variability of count data to accommodate overdispersion. Specifically, given that observed incidence counts follow a negative binomial distribution, the posterior distribution of secondary infections is obtained as a Dirichlet multinomial distribution. With this formulation, we establish posterior uncertainty bounds for the effective reproduction number. Finally, we demonstrate the effectiveness of the proposed method using incidence data from the COVID-19 pandemic.
PubMed: 37701756
DOI: 10.1016/j.idm.2023.08.006 -
PLoS Medicine Jul 2023Women with psychiatric diagnoses are at increased risk of preterm birth (PTB), with potential life-long impact on offspring health. Less is known about the risk of PTB...
BACKGROUND
Women with psychiatric diagnoses are at increased risk of preterm birth (PTB), with potential life-long impact on offspring health. Less is known about the risk of PTB in offspring of fathers with psychiatric diagnoses, and for couples where both parents were diagnosed. In a nationwide birth cohort, we examined the association between psychiatric history in fathers, mothers, and both parents and gestational age.
METHODS AND FINDINGS
We included all infants live-born to Nordic parents in 1997 to 2016 in Sweden. Psychiatric diagnoses were obtained from the National Patient Register. Data on gestational age were retrieved from the Medical Birth Register. Associations between parental psychiatric history and PTB were quantified by relative risk (RR) and two-sided 95% confidence intervals (CIs) from log-binomial regressions, by psychiatric disorders overall and by diagnostic categories. We extended the analysis beyond PTB by calculating risks over the whole distribution of gestational age, including "early term" (37 to 38 weeks). Among the 1,488,920 infants born throughout the study period, 1,268,507 were born to parents without a psychiatric diagnosis, of whom 73,094 (5.8%) were born preterm. 4,597 of 73,500 (6.3%) infants were born preterm to fathers with a psychiatric diagnosis, 8,917 of 122,611 (7.3%) infants were born preterm to mothers with a pscyhiatric diagnosis, and 2,026 of 24,302 (8.3%) infants were born preterm to both parents with a pscyhiatric diagnosis. We observed a shift towards earlier gestational age in offspring of parents with psychiatric history. The risks of PTB associated with paternal and maternal psychiatric diagnoses were similar for different psychiatric disorders. The risks for PTB were estimated at RR 1.12 (95% CI [1.08, 1.15] p < 0.001) for paternal diagnoses, at RR 1.31 (95% CI [1.28, 1.34] p < 0.001) for maternal diagnoses, and at RR 1.52 (95% CI [1.46, 1.59] p < 0.001) when both parents were diagnosed with any psychiatric disorder, compared to when neither parent had a psychiatric diagnosis. Stress-related disorders were associated with the highest risks of PTB with corresponding RRs estimated at 1.23 (95% CI [1.16, 1.31] p < 0.001) for a psychiatry history in fathers, at 1.47 (95% CI [1.42, 1.53] p < 0.001) for mothers, and at 1.90 (95% CI [1.64, 2.20] p < 0.001) for both parents. The risks for early term were similar to PTB. Co-occurring diagnoses from different diagnostic categories increased risk; for fathers: RR 1.10 (95% CI [1.07, 1.13] p < 0.001), 1.15 (95% CI [1.09, 1.21] p < 0.001), and 1.33 (95% CI [1.23, 1.43] p < 0.001), for diagnoses in 1, 2, and ≥3 categories; for mothers: RR 1.25 (95% CI [1.22, 1.28] p < 0.001), 1.39 (95% CI [1.34, 1.44] p < 0.001) and 1.65 (95% CI [1.56, 1.74] p < 0.001). Despite the large sample size, statistical precision was limited in subgroups, mainly where both parents had specific psychiatric subtypes. Pathophysiology and genetics underlying different psychiatric diagnoses can be heterogeneous.
CONCLUSIONS
Paternal and maternal psychiatric history were associated with a shift to earlier gestational age and increased risk of births before full term. The risk consistently increased when fathers had a positive history of different psychiatric disorders, increased further when mothers were diagnosed and was highest when both parents were diagnosed.
Topics: Male; Infant; Infant, Newborn; Humans; Female; Sweden; Premature Birth; Term Birth; Fathers; Mothers; Risk Factors
PubMed: 37471291
DOI: 10.1371/journal.pmed.1004256 -
Trials Sep 2023Two characteristics of commonly used outcomes in medical research are zero inflation and non-negative integers; examples include the number of hospital admissions or... (Randomized Controlled Trial)
Randomized Controlled Trial
Evaluation of negative binomial and zero-inflated negative binomial models for the analysis of zero-inflated count data: application to the telemedicine for children with medical complexity trial.
BACKGROUND
Two characteristics of commonly used outcomes in medical research are zero inflation and non-negative integers; examples include the number of hospital admissions or emergency department visits, where the majority of patients will have zero counts. Zero-inflated regression models were devised to analyze this type of data. However, the performance of zero-inflated regression models or the properties of data best suited for these analyses have not been thoroughly investigated.
METHODS
We conducted a simulation study to evaluate the performance of two generalized linear models, negative binomial and zero-inflated negative binomial, for analyzing zero-inflated count data. Simulation scenarios assumed a randomized controlled trial design and varied the true underlying distribution, sample size, and rate of zero inflation. We compared the models in terms of bias, mean squared error, and coverage. Additionally, we used logistic regression to determine which data properties are most important for predicting the best-fitting model.
RESULTS
We first found that, regardless of the rate of zero inflation, there was little difference between the conventional negative binomial and its zero-inflated counterpart in terms of bias of the marginal treatment group coefficient. Second, even when the outcome was simulated from a zero-inflated distribution, a negative binomial model was favored above its ZI counterpart in terms of the Akaike Information Criterion. Third, the mean and skewness of the non-zero part of the data were stronger predictors of model preference than the percentage of zero counts. These results were not affected by the sample size, which ranged from 60 to 800.
CONCLUSIONS
We recommend that the rate of zero inflation and overdispersion in the outcome should not be the sole and main justification for choosing zero-inflated regression models. Investigators should also consider other data characteristics when choosing a model for count data. In addition, if the performance of the NB and ZINB regression models is reasonably comparable even with ZI outcomes, we advocate the use of the NB regression model due to its clear and straightforward interpretation of the results.
Topics: Humans; Child; Models, Statistical; Computer Simulation; Linear Models; Telemedicine; Bias
PubMed: 37752579
DOI: 10.1186/s13063-023-07648-8 -
BMC Infectious Diseases Jun 2023Acute encephalitis syndrome (AES) differs in its spatio-temporal distribution in Vietnam with the highest incidence seen during the summer months in the northern...
BACKGROUND
Acute encephalitis syndrome (AES) differs in its spatio-temporal distribution in Vietnam with the highest incidence seen during the summer months in the northern provinces. AES has multiple aetiologies, and the cause remains unknown in many cases. While vector-borne disease such as Japanese encephalitis and dengue virus and non-vector-borne diseases such as influenza and enterovirus show evidence of seasonality, associations with climate variables and the spatio-temporal distribution in Vietnam differs between these. The aim of this study was therefore to understand the spatio-temporal distribution of, and risk factors for AES in Vietnam to help hypothesise the aetiology.
METHODS
The number of monthly cases per province for AES, meningitis and diseases including dengue fever; influenza-like-illness (ILI); hand, foot, and mouth disease (HFMD); and Streptococcus suis were obtained from the General Department for Preventive Medicine (GDPM) from 1998-2016. Covariates including climate, normalized difference vegetation index (NDVI), elevation, the number of pigs, socio-demographics, JEV vaccination coverage and the number of hospitals were also collected. Spatio-temporal multivariable mixed-effects negative binomial Bayesian models with an outcome of the number of cases of AES, a combination of the covariates and harmonic terms to determine the magnitude of seasonality were developed.
RESULTS
The national monthly incidence of AES declined by 63.3% over the study period. However, incidence increased in some provinces, particularly in the Northwest region. In northern Vietnam, the incidence peaked in the summer months in contrast to the southern provinces where incidence remained relatively constant throughout the year. The incidence of meningitis, ILI and S. suis infection; temperature, relative humidity with no lag, NDVI at a lag of one month, and the number of pigs per 100,000 population were positively associated with the number of cases of AES in all models in which these covariates were included.
CONCLUSIONS
The positive correlation of AES with temperature and humidity suggest that a number of cases may be due to vector-borne diseases, suggesting a need to focus on vaccination campaigns. However, further surveillance and research are recommended to investigate other possible aetiologies such as S. suis or Orientia tsutsugamushi.
Topics: Animals; Swine; Humans; Acute Febrile Encephalopathy; Vietnam; Bayes Theorem; Influenza, Human; Climate
PubMed: 37312047
DOI: 10.1186/s12879-023-08300-1 -
Scientific Reports Jan 2024The spatio-temporal distribution of COVID-19 across India's states and union territories is not uniform, and the reasons for the heterogeneous spread are unclear....
The spatio-temporal distribution of COVID-19 across India's states and union territories is not uniform, and the reasons for the heterogeneous spread are unclear. Identifying the space-time trends and underlying indicators influencing COVID-19 epidemiology at micro-administrative units (districts) will help guide public health strategies. The district-wise daily COVID-19 data of cases and deaths from February 2020 to August 2021 (COVID-19 waves-I and II) for the entire country were downloaded and curated from public databases. The COVID-19 data normalized with the projected population (2020) and used for space-time trend analysis shows the states/districts in southern India are the worst hit. Coastal districts and districts adjoining large urban regions of Mumbai, Chennai, Bengaluru, Goa, and New Delhi experienced > 50,001 cases per million population. Negative binomial regression analysis with 21 independent variables (identified through multicollinearity analysis, with VIF < 10) covering demography, socio-economic status, environment, and health was carried out for wave-I, wave-II, and total (wave-I and wave-II) cases and deaths. It shows wealth index, derived from household amenities datasets, has a high positive risk ratio (RR) with COVID-19 cases (RR: 3.577; 95% CI: 2.062-6.205) and deaths (RR: 2.477; 95% CI: 1.361-4.506) across the districts. Furthermore, socio-economic factors such as literacy rate, health services, other workers' rate, alcohol use in men, tobacco use in women, overweight/obese women, and rainfall have a positive RR and are significantly associated with COVID-19 cases/deaths at the district level. These positively associated variables are highly interconnected in COVID-19 hotspot districts. Among these, the wealth index, literacy rate, and health services, the key indices of socio-economic development within a state, are some of the significant indicators associated with COVID-19 epidemiology in India. The identification of district-level space-time trends and indicators associated with COVID-19 would help policymakers devise strategies and guidelines during public health emergencies.
Topics: Male; Humans; Female; COVID-19; India; Family Characteristics
PubMed: 38167962
DOI: 10.1038/s41598-023-50363-2 -
Vaccine Aug 2023In December 2020 the U.S. began a massive COVID-19 vaccination campaign, an action that researchers felt could catalyze inequalities in COVID-19 vaccination utilization....
BACKGROUND
In December 2020 the U.S. began a massive COVID-19 vaccination campaign, an action that researchers felt could catalyze inequalities in COVID-19 vaccination utilization. While vaccines have the potential to be accessible regardless of social status, the objective of this study was to examine how and when socioeconomic status (SES) and racial/ethnic inequalities would emerge in vaccination distribution.
METHODS
Population vaccination rates reported at the county level by the Centers for Disease Control and Prevention across 46 states on 3/30/2021. Correlates included SES, the share of the population who were Black, Hispanic, Female, or aged ≥65 years, and urbanicity (thousands of residents per square mile). Multivariable-adjusted analyses relied on zero-inflated negative binomial regression to estimate the odds of providing any vaccine, and vaccination rate ratios (aVRR) comparing the distribution rate for vaccinations across the U.S.
RESULTS
Across the U.S., 16.3 % of adults and 37.9 % of adults aged 65 and older were vaccinated in lower SES counties, while 20.45 % of all adults and 48.15 % of adults aged 65 and older were vaccinated in higher SES counties. Inequalities emerged after 41 days, when < 2 % of Americans were vaccinated. Multivariable-adjusted analyses revealed that higher SES was associated with improved vaccination distribution (aVRR = 1.127, [1.100-1.155], p < 1E-06), while increases in the percent reporting Black or Hispanic race/ethnicity was associated with lower vaccination distribution (aVRR = 0.998, [0.996-0.999], p = 1.03E-04).
CONCLUSIONS
Social inequalities in COVID-19 vaccines reflect an inefficient and inequitable distribution of these technologies. Future efforts to improve health should recognize the central role of social factors in impacting vaccine delivery.
Topics: Adult; Female; Humans; Black or African American; COVID-19; COVID-19 Vaccines; Socioeconomic Factors; United States; Vaccination; Vaccines
PubMed: 37460352
DOI: 10.1016/j.vaccine.2023.07.022 -
PLoS Computational Biology Feb 2024Outbreaks of emerging and zoonotic infections represent a substantial threat to human health and well-being. These outbreaks tend to be characterised by highly...
Outbreaks of emerging and zoonotic infections represent a substantial threat to human health and well-being. These outbreaks tend to be characterised by highly stochastic transmission dynamics with intense variation in transmission potential between cases. The negative binomial distribution is commonly used as a model for transmission in the early stages of an epidemic as it has a natural interpretation as the convolution of a Poisson contact process and a gamma-distributed infectivity. In this study we expand upon the negative binomial model by introducing a beta-Poisson mixture model in which infectious individuals make contacts at the points of a Poisson process and then transmit infection along these contacts with a beta-distributed probability. We show that the negative binomial distribution is a limit case of this model, as is the zero-inflated Poisson distribution obtained by combining a Poisson-distributed contact process with an additional failure probability. We assess the beta-Poisson model's applicability by fitting it to secondary case distributions (the distribution of the number of subsequent cases generated by a single case) estimated from outbreaks covering a range of pathogens and geographical settings. We find that while the beta-Poisson mixture can achieve a closer to fit to data than the negative binomial distribution, it is consistently outperformed by the negative binomial in terms of Akaike Information Criterion, making it a suboptimal choice on parsimonious grounds. The beta-Poisson performs similarly to the negative binomial model in its ability to capture features of the secondary case distribution such as overdispersion, prevalence of superspreaders, and the probability of a case generating zero subsequent cases. Despite this possible shortcoming, the beta-Poisson distribution may still be of interest in the context of intervention modelling since its structure allows for the simulation of measures which change contact structures while leaving individual-level infectivity unchanged, and vice-versa.
Topics: Humans; Models, Statistical; Computer Simulation; Poisson Distribution; Binomial Distribution; Disease Outbreaks
PubMed: 38330050
DOI: 10.1371/journal.pcbi.1011856 -
American Journal of Obstetrics &... Aug 2023With increasing cancer incidence and survival rates, the prevalence of maternal cancer and its effect on adverse birth outcomes are important for prenatal care and...
BACKGROUND
With increasing cancer incidence and survival rates, the prevalence of maternal cancer and its effect on adverse birth outcomes are important for prenatal care and oncology management. However, the effects of different types of cancer at different gestational stages have not been widely reported.
OBJECTIVE
This study aimed to describe the epidemiologic characteristics of pregnancy-associated cancers (during and 1 year after pregnancy) and evaluate the association between adverse birth outcomes and maternal cancers.
METHODS
Of 983,162 cases, a history of maternal cancer, including pregestational cancer, pregnancy-associated cancer, and subsequent cancer, was identified in 16,475 cases using a health information network. The incidence and 95% confidence interval of pregnancy-associated cancer were calculated with the Poisson distribution. The adjusted risk ratio with 95% confidence interval of the association between adverse birth outcomes and maternal cancer were estimated using the multilevel log-binomial model.
RESULTS
A total of 38,295 offspring were born to mothers with a cancer history. Of these, 2583 (6.75%) were exposed to pregnancy-associated cancer, 30,706 (80.18%) had a subsequent cancer diagnosis, and 5006 (13.07%) were exposed to pregestational cancer. The incidence of pregnancy-associated cancer was 2.63 per 1000 pregnancies (95% confidence interval, 2.53‰-2.73‰), with cancer of the thyroid (1.15‰), breast (0.25‰), and female reproductive organs (0.23‰) being the most common cancer types. The increased risks of preterm birth and low birthweight were significantly associated with cancer diagnosed during the second and third trimester of pregnancy, whereas increased risks of birth defects (adjusted risk ratio, 1.48; 95% confidence interval, 1.08-2.04) were associated with cancer diagnosed in the first trimester. Increased risks of preterm birth (adjusted risk ratio, 1.16; 95% confidence interval, 1.02-1.32), low birthweight (adjusted risk ratio, 1.24; 95% confidence interval, 1.07-1.44), and birth defects (adjusted risk ratio, 1.22; 95% confidence interval, 1.10-1.35) were observed in thyroid cancer survivors.
CONCLUSION
Careful monitoring of fetal growth should be implemented for women diagnosed with cancer in the second and third trimester to ensure timely delivery and balance the benefits of neonatal health and cancer treatment. The higher incidence of thyroid cancer and increased risk of adverse birth outcomes among thyroid cancer survivors suggested that the regular thyroid function monitoring and regulation of thyroid hormone levels are important in maintaining pregnancy and promoting fetal development among thyroid cancer survivors before and during pregnancy.
Topics: Pregnancy; Humans; Infant, Newborn; Female; Birth Weight; Cohort Studies; Premature Birth; Pregnancy Complications; Infant, Low Birth Weight; Neoplasms
PubMed: 37245606
DOI: 10.1016/j.ajogmf.2023.101036 -
Malaria Journal Nov 2023Geostatistical analysis of health data is increasingly used to model spatial variation in malaria prevalence, burden, and other metrics. Traditional inference methods...
BACKGROUND
Geostatistical analysis of health data is increasingly used to model spatial variation in malaria prevalence, burden, and other metrics. Traditional inference methods for geostatistical modelling are notoriously computationally intensive, motivating the development of newer, approximate methods for geostatistical analysis or, more broadly, computational modelling of spatial processes. The appeal of faster methods is particularly great as the size of the region and number of spatial locations being modelled increases.
METHODS
This work presents an applied comparison of four proposed 'fast' computational methods for spatial modelling and the software provided to implement them-Integrated Nested Laplace Approximation (INLA), tree boosting with Gaussian processes and mixed effect models (GPBoost), Fixed Rank Kriging (FRK) and Spatial Random Forests (SpRF). The four methods are illustrated by estimating malaria prevalence on two different spatial scales-country and continent. The performance of the four methods is compared on these data in terms of accuracy, computation time, and ease of implementation.
RESULTS
Two of these methods-SpRF and GPBoost-do not scale well as the data size increases, and so are likely to be infeasible for larger-scale analysis problems. The two remaining methods-INLA and FRK-do scale well computationally, however the resulting model fits are very sensitive to the user's modelling assumptions and parameter choices. The binomial observation distribution commonly used for disease prevalence mapping with INLA fails to account for small-scale overdispersion present in the malaria prevalence data, which can lead to poor predictions. Selection of an appropriate alternative such as the Beta-binomial distribution is required to produce a reliable model fit. The small-scale random effect term in FRK overcomes this pitfall, but FRK model estimates are very reliant on providing a sufficient number and appropriate configuration of basis functions. Unfortunately the computation time for FRK increases rapidly with increasing basis resolution.
CONCLUSIONS
INLA and FRK both enable scalable geostatistical modelling of malaria prevalence data. However care must be taken when using both methods to assess the fit of the model to data and plausibility of predictions, in order to select appropriate model assumptions and parameters.
Topics: Humans; Models, Statistical; Computer Simulation; Software; Spatial Analysis; Malaria; Bayes Theorem
PubMed: 37990242
DOI: 10.1186/s12936-023-04760-7