Integrating informative hypotheses into the EffectLiteR framework

Using the EffectLiteR framework, researchers can test classical null hypotheses about effects of interest via Wald and F-tests, while taking into account the stochastic nature of group sizes. This paper aims at extending EffectLiteR to test informative hypotheses, assuming for example that the average effect of a new treatment is greater than the average effect of an old treatment, which in turn is greater than zero. We present a simulated data example to show two methodological novelties. First, we illustrate how to use the Fbar- and generalized linear Wald test to assess informative hypotheses. While the classical test did not reach significance, the informative test correctly rejected the null hypothesis, indicating the need to take into account the order of the treatment groups. Second, we demonstrate how to account for stochastic group sizes in informative hypotheses using the generalized non-linear Wald statistic. The paper concludes with a short data example.

Aetiology and risk factors of bacterial gastroenteritis among febrile outpatients at the Dschang District Hospital, West Region of Cameroon: a cross-sectional study

ObjectivesTo investigate the bacterial aetiologies and associated risk factors of gastroenteritis among typhoid suspected cases.DesignCross-sectional study.SettingThis study was conducted at Dschang District Hospital of the Menoua Division, West Region of Cameroon, between April–November 2019 and June 2020.ParticipantsParticipants aged ≥2 years (mean 34±18.77 years) and of both sex suspected of having typhoid fever were included, while non-suspected typhoid cases were excluded. Self-reported sociodemographic and health information at recruitment was obtained from 556 participants.MethodsCollected stool samples were examined macroscopically and microscopically and subjected to culture. After culture, Gram staining was performed, followed by biochemical testing and characterisation using the Analytical Profile Index (API-20E) test kit.Interventions’No intervention was done during the period of study.Outcome measuresWe identified bacterial causing gastroenteritis, and associated risk factors calculated using binary regression, adjusting for sociodemographic and health variables.ResultsOf 556 patients, 74.28% tested positive for gastroenteritis. Among pathogens responsible for gastroenteritis, Escherichia coli was found to be the main cause (21.1%), followed by Salmonella typhi (10.4%), Citrobacter diversus (8.2%), and Proteus mirabilis (8.2%), Proteus vulgaris (7.3%), whereas Citrobacter spp and Yersinia enterocolitica were less represented among pathogens causing the disease among patients. A significant difference (p=0.002) was observed between abdominal pain and all the micro-organisms isolated from the patients. Patients having primary level of education were significantly associated (p=0.017; 3.163 (95% CI 1.228 to 8.147)) with the prevalence of gastroenteritis. Consumption of beverages (Wald statistic: 4.823; OR: 2.471; 95% CI (1.102 to 5.539); p=0.028), use of modern toilet (Wald statistic: 4.471; OR: 1.723; 95% CI (1.041 to 2.852); p=0.034) were strongly associated with gastroenteritis and rearing of bird (Wald statistic: 4.880; OR: 0.560; 95% CI (0.335 to 0.937); p=0.027), was found to be protective.ConclusionAcute bacterial gastroenteritis is a significant cause of morbidity in Dschang, with the prevalence of 74.28%. Many pathogens accounted for gastroenteritis, and E. coli (21.1%) could be a major cause, followed by S. typhi (10.4%), C. diversus (8.2%), P. mirabilis (8.2%), P. vulgaris (7.3%), whereas Citrobacter spp and Y. enterocolitica were less represented. Gastroenteritis was highly associated with primary level of education, consumption of beverages, use of modern toilet while rearing of birds was unexpectedly found to be protective against Gastroenteritis. Further characterisation is planned.

Intracerebral Iron Accumulation may be Associated with Secondary Brain Injury in Patients with Poor Grade Subarachnoid Hemorrhage

Background The amount of intracranial blood is a strong predictor of poor outcome after subarachnoid hemorrhage (SAH). Here, we aimed to measure iron concentrations in the cerebral white matter, using the cerebral microdialysis (CMD) technique, and to associate iron levels with the local metabolic profile, complications, and functional outcome. Methods For the observational cohort study, 36 patients with consecutive poor grade SAH (Hunt & Hess grade of 4 or 5, Glasgow Coma Scale Score ≤ 8) undergoing multimodal neuromonitoring were analyzed for brain metabolic changes, including CMD iron levels quantified by graphite furnace atomic absorption spectrometry. The study time encompassed 14 days after admission. Statistical analysis was performed using generalized estimating equations. Results Patients were admitted in a poor clinical grade (n = 26, 72%) or deteriorated within 24 h (n = 10, 28%). The median blood volume in the subarachnoid space was high (SAH sum score = 26, interquartile range 20–28). Initial CMD iron was 44 µg/L (25–65 µg/L), which significantly decreased to a level of 25 µg/L (14–30 µg/L) at day 4 and then constantly increased over the remaining neuromonitoring days (p < 0.01). A higher intraventricular hemorrhage sum score (≥ 5) was associated with higher CMD iron levels (Wald-statistic = 4.1, df = 1, p = 0.04) but not with the hemorrhage load in the subarachnoid space (p = 0.8). In patients developing vasospasm, the CMD iron load was higher, compared with patients without vasospasm (Wald-statistic = 4.1, degree of freedom = 1, p = 0.04), which was not true for delayed cerebral infarction (p = 0.4). Higher iron concentrations in the brain extracellular fluid (34 µg/L, 36–56 µg/L vs. 23 µg/L, 15–37 µg/L) were associated with mitochondrial dysfunction (CMD lactate to pyruvate ratio > 30 and CMD-pyruvate > 70 µM/L, p < 0.001). Brain extracellular iron load was not associated with functional outcome after 3 months (p > 0.5). Conclusions This study suggests that iron accumulates in the cerebral white matter in patients with poor grade SAH. These findings may support trials aiming to scavenger brain extracellular iron based on the hypothesis that iron-mediated neurotoxicity may contribute to acute and secondary brain injury following SAH.

A power approximation for the Kenward and Roger Wald test in the linear mixed model

We derive a noncentral F power approximation for the Kenward and Roger test. We use a method of moments approach to form an approximate distribution for the Kenward and Roger scaled Wald statistic, under the alternative. The result depends on the approximate moments of the unscaled Wald statistic. Via Monte Carlo simulation, we demonstrate that the new power approximation is accurate for cluster randomized trials and longitudinal study designs. The method retains accuracy for small sample sizes, even in the presence of missing data. We illustrate the method with a power calculation for an unbalanced group-randomized trial in oral cancer prevention.

Duration of Surgery and Learning Curve Affect Rotator Cuff Repair Retear Rates: A Post Hoc Analysis of 1600 Cases

Background: Arthroscopic rotator cuff repair can be quite complex and time consuming, particularly early in the surgeon’s learning curve. Hypothesis: Patients who have undergone rotator cuff repair with shorter operative times will be less likely to have a rotator cuff retear at 6 months postoperatively. Study Design: Case-control study; Level of evidence, 3. Methods: This study was an analysis of data from 1600 consecutive patients (670 partial-thickness and 930 full-thickness tears) who had rotator cuff repair performed by a single surgeon utilizing an arthroscopic, single-row, knotless inverted mattress suture anchor technique. All patients underwent ultrasound at 6 months postoperatively to determine repair integrity. Moving average analysis was performed for the variables of operative time and case number to evaluate the surgeon’s learning curve. Results: For early cases, the mean operative time was approximately 35 minutes. After approximately 450 cases, the operative time plateaued at approximately 20 minutes. The mean operative time for the cohort (±SEM) was 22 ± 0.3 minutes, and the mean retear rate was 13%. Increased operative time was associated with a retear ( r = 0.18; P < .001). Multiple logistic regression analysis revealed that the variables with the most independent effect on retears were larger tear size (Wald statistic = 36; P < .001), lower case number (ie, less surgeon experience) (Wald statistic = 28; P < .001), older patient age (Wald statistic = 23; P < .001), full-thickness tears (Wald statistic = 13; P < .001), and lower surgeon-rated repair quality (Wald statistic = 8; P = .004). Operative time was not a significant independent factor contributing to retears. Conclusion: Operative time and rotator cuff retear rates decreased as surgical team experience increased. The hypothesis of this study, however, was not supported. The reduced retear rate was not related to a reduction in operative time per se but rather to improved surgical team experience and patient factors, such as improved healing with smaller tears in younger patients.

The Impact of Physical Activity and Function on Falls in Assisted Living Residents

The purpose of this study was to describe physical activity and function of older adults in assisted living communities and test the association between moderate and vigorous activity and falls. This study used baseline data from 393 participants from the first two cohorts in the Function-Focused Care in Assisted Living Using the Evidence Integration Triangle study. The majority of participants were female (N = 276, 70%) and White (N = 383, 97%) with a mean age of 87 years (SD = 7). Controlling for age, cognition, gender, setting, and function, the time spent in moderate or vigorous levels of physical activity was associated with having a fall in the prior 4 months. Those who engaged in more moderate physical activity were 0.6% less likely to have a fall (OR = 0.994, Wald statistic = 5.54, p = .02), and those who engaged in more vigorous activity were 2% less likely to have a fall (OR = 0.980, Wald statistic = 3.88, p = .05).

The Induced Smoothed lasso: A practical framework for hypothesis testing in high dimensional regression

This paper focuses on hypothesis testing in lasso regression, when one is interested in judging statistical significance for the regression coefficients in the regression equation involving a lot of covariates. To get reliable p-values, we propose a new lasso-type estimator relying on the idea of induced smoothing which allows to obtain appropriate covariance matrix and Wald statistic relatively easily. Some simulation experiments reveal that our approach exhibits good performance when contrasted with the recent inferential tools in the lasso framework. Two real data analyses are presented to illustrate the proposed framework in practice.

Exclusive Breastfeeding Knowledge, Intention to Practice and Predictors among Primiparous Women in Enugu South-East, Nigeria

Breastfeeding is considered as the most complete nutritional source for infants because breast milk contains the essential carbohydrates, fats, proteins, and immunological factors needed for infants to thrive and resist infection in the formative first year of life. Knowledge of exclusive breastfeeding (EBF) among women is essential when promoting optimal breastfeeding practices. This cross-sectional descriptive survey assessed knowledge and intention to practice EBF and its associated factors during pregnancy among primiparous women in selected communities in Enugu State, Nigeria. A total population study that applied inclusion criteria was used to recruit 201 primiparous mothers attending their third trimester antenatal care from selected health facilities in rural and urban communities in Enugu State. A researcher-developed questionnaire was used to collect data on participants’ knowledge and intention to practice EBF. Descriptive statistics of frequency, percentage, mean, and standard deviation were used to summarize categorical and continuous variables while Chi-square and Wald statistic tests predicted demographic data associated with knowledge status and intention to practice EBF of the participants. More than half (58.7%) of primiparous mothers had inadequate knowledge of EBF and only 62.7% had intention to exclusively breastfeed for 4–6 months. The Chi-square test result showed significant difference in the participants’ place of residence (p = 0.024), EBF knowledge sources (p = 0.001), and EBF knowledge. The Wald statistic in Logistic regression model indicated the coefficient of age (p = 0.026), educational attainment (p = 0.046), EBF knowledge (p = 0.016), and sources of information about EBF (p = 0.027) to be significant predictors of good intention to practise EBF. Poor EBF knowledge and intention to practice in this population may be improved by combining facility-based and in-house methods of breastfeeding counseling, education, and support especially to intending and expectant mothers. Further studies are needed to be done using the multiparous women as well as assessing the effects of in-house EBF supportive-educative intervention to improve breastfeeding outcomes.

Wald Statistics in high-dimensional PCA

In this study, we consider PCA for Gaussian observations X1, …, Xn with covariance Σ = ∑iλiPi in the ’effective rank’ setting with model complexity governed by r(Σ) ≔ tr(Σ)∕∥Σ∥. We prove a Berry-Essen type bound for a Wald Statistic of the spectral projector $\hat P_r$. This can be used to construct non-asymptotic goodness of fit tests and confidence ellipsoids for spectral projectors Pr. Using higher order pertubation theory we are able to show that our Theorem remains valid even when $\mathbf{r}(\Sigma) \gg \sqrt{n}$.

Test procedure and sample size determination for a proportion study using a double-sampling scheme with two fallible classifiers

Double sampling is usually applied to collect necessary information for situations in which an infallible classifier is available for validating a subset of the sample that has already been classified by a fallible classifier. Inference procedures have previously been developed based on the partially validated data obtained by the double-sampling process. However, it could happen in practice that such infallible classifier or gold standard does not exist. In this article, we consider the case in which both classifiers are fallible and propose asymptotic and approximate unconditional test procedures based on six test statistics for a population proportion and five approximate sample size formulas based on the recommended test procedures under two models. Our results suggest that both asymptotic and approximate unconditional procedures based on the score statistic perform satisfactorily for small to large sample sizes and are highly recommended. When sample size is moderate or large, asymptotic procedures based on the Wald statistic with the variance being estimated under the null hypothesis, likelihood rate statistic, log- and logit-transformation statistics based on both models generally perform well and are hence recommended. The approximate unconditional procedures based on the log-transformation statistic under Model I, Wald statistic with the variance being estimated under the null hypothesis, log- and logit-transformation statistics under Model II are recommended when sample size is small. In general, sample size formulae based on the Wald statistic with the variance being estimated under the null hypothesis, likelihood rate statistic and score statistic are recommended in practical applications. The applicability of the proposed methods is illustrated by a real-data example.