Forward Stepwise Binary Logistic Regression Research Articles

Advanced gastrointestinal stromal tumor: reliable classification of imatinib plasma trough concentration via machine learning

AimPatients with advanced gastrointestinal stromal tumors (GISTs) exhibiting an imatinib plasma trough concentration (IM Cmin) under 1100 ng/ml may show a reduced drug response rate, leading to the suggestion of monitoring for IM Cmin. Consequently, the objective of this research was to create a customized IM Cmin classification model for patients with advanced GISTs from China.MethodsInitial data and laboratory indicators from patients with advanced GISTs were gathered, and the above information was segmented into a training set, validation set, and testing set in a 6:2:2 ratio. Key variables associated with IM Cmin were identified to construct the classification model using the least absolute shrinkage and selection operator (LASSO) regression and forward stepwise binary logistic regression. Within the training and validation sets, nine ML classification models were constructed via the resampling method and underwent comparison through the Brier scores, the areas under the receiver-operating characteristic curve (AUROC), the decision curve, and the precision-recall (AUPR) curve to determine the most suitable model for this dataset. Two methods of internal validation were used to assess the most suitable model's classification performance: tenfold cross-validation and random split-sample validation (test set), and the value of the test set AUROC was used to evaluate the model's classification performance.ResultsSix key variables (gender, daily IM dose, metastatic site, red blood cell count, platelet count, and percentage of neutrophils) were ultimately selected to construct the classification model. In the validation set, it is found by comparison that the Extreme Gradient Boosting (XGBoost) model has the largest AUROC, the lowest Brier score, the largest area under the decision curve, and the largest AUPR value. Furthermore, as evaluated via internal verification, it also performed well in the test set (AUROC = 0.725).ConclusionFor patients with advanced GISTs who receive IM, initial data and laboratory indicators could be used to accurately estimate whether the IM Cmin is below 1100 ng/ml. The XGBoost model may stand a chance to assist clinicians in directing the administration of IM.

Both reduced ovarian reserve and severe semen alterations are overrepresented in couples seeking assisted reproductive technology treatment for the first time: a cross-sectional study

BackgroundOnce a mate choice decision has been made, couples that fail to reach a live birth in natural and/or intrauterine insemination (IUI) cycles will likely visit fertility clinics seeking assisted reproductive technology (ART) treatment. During the more or less prolonged period of infertility experienced, those couples with mild/moderate reproductive anomalies would have advantage over couples displaying more severe reproductive alterations in achieving a natural or IUI conception. Thus, we can expect to find a progressive increase in the proportion of couples with more severe reproductive anomalies as duration of infertility rises. In this study, we aim to ascertain whether there is an association between male and female infertility diagnoses and duration of infertility in couples seeking ART treatment for the first time.MethodsA cross-sectional analysis of 1383 infertile couples that sought ART treatment for the first time. Forward-stepwise binary logistic regression analyses were applied to calculate exponentiated regression coefficients.ResultsMen suffering from any combination of oligo-, astheno-, and teratozoospermia (ACOAT) exhibited higher odds of having a duration of infertility > 2 years compared with non-ACOAT men [odds ratio (95% confidence interval): 1.340 (1.030–1.744)]. Women from ACOAT couples displaying a duration of infertility > 2 years presented shorter menstrual cycles (P ≤ 0.047) and lower antral follicular count (AFC) values (P ≤ 0.008) and serum anti-Müllerian hormone (AMH) levels (P ≤ 0.007) than women from non-ACOAT couples exhibiting > 2 years of infertility. Likewise, AFC values (P ≤ 0.013) and serum AMH levels (P ≤ 0.001) were decreased when compared with women from ACOAT couples displaying ≤ 2 years of infertility. A relative low but significant percentage of ACOAT couples displaying > 2 years of infertility stood out for their smoking habits.ConclusionsCouples consisting of ACOAT men and women with a relative low ovarian reserve are overrepresented in couples seeking ART treatment for the first time after experiencing > 2 years of infertility. This outcome leads us to develop a general hypothesis proposing that the origin of couple’s infertility is a consequence of a process of positive assortative mating shaped by sexual selection forces.

Impact of pandemic COVID-19 lockdown on penetrating keratoplasty patients.

To study the impact of the COVID-19 lockdown on the regular follow-up of keratoplasty patients. This retrospective interventional case series included 30 patients who had immunological corneal endothelial rejection out of 190 patients who came for post-PKP follow-up between September 15, 2019, and September 30, 2020. The demographics, primary diagnosis, surgical technique, time of presentation, recovery of graft, associated ocular problems, and visual acuity at 1 month were analyzed. Forward stepwise (likelihood ratio) binary logistic regression was used to find significant variables. The study population had 19 males (63.33%) and 11 females (36.67%). The mean age of the study group was 42.83 ± 18.89 (8-80) years. Of 30 patients, 19 (63.3%) presented before and 11 (36.7%) after the COVID-19 lockdown. Overall, 23 (77%) showed a reversal of graft rejection. Logistic regression showed that preoperative indications, large-sized grafts, and deep corneal vascularization were significant risk factors for non-resolution of graft rejection. It was noted that patients who presented to the hospital late had poor recovery (P = 0.002). The delay in the presentation was a significant risk factor for non-resolution of graft rejection (P < 0.01). Z-test for proportions revealed that the difference in the non-resolution of rejection on immediate or delayed treatment in patients presenting during lockdown (P = 0.002) was significant. This article is to highlight the impact of the COVID-19 lockdown on graft rejection recovery of PKP patients due to delays in follow-up. Early treatment helps in the recovery of graft transparency and the reversal of immunological graft rejection. Also, primary diagnosis, deep vascularization, and large-sized grafts were significant risk factors for non-resolution of graft rejection.

Prevalence and predictors of medication non-adherence in children with inflammatory bowel disease in China: A cross-sectional study.

Background: Non-adherence to physician-prescribed medications, especially oral medications, is common in children with inflammatory bowel disease (IBD), and medication non-adherence is associated with poorer outcomes in IBD. Therefore, understanding and improving medication adherence in children with IBD is critical for optimizing treatment and improving treatment outcomes. Despite the relatively high prevalence of IBD in children in China, to date, very little is known about medication adherence in these patients. Objective: The aim of this study was to investigate the prevalence of medication non-adherence and its risk factors in children with IBD in China to provide a basis for developing adherence improvement strategies. Methods: A cross-sectional design was employed. Children (aged <18years) with IBD who visited the Children's Hospital, Zhejiang University School of Medicine, from September 2020 to December 2021 were included. Data were collected by a general information questionnaire, the 4-item Medication Adherence Report Scale (MARS-4) and Crohn's and Colitis Knowledge (CCKNOW) questionnaire. Subsequently, forward stepwise binary logistic regression analysis was performed to determine independent predictors of medication non-adherence. Results: A total of 119children were included in this study. The results showed that 33 (27.73%) and 86 (72.27%) children had poor and good medication adherence, respectively. Of these, 83 (69.75%) had forgotten to take their medications sometimes, often, or always. On binary logistic regression, we found that the incidence of medication non-adherence in children with IBD course of 3years and above [OR 4.82 (95%CI: 1.47-15.88); p = 0.01] was significantly higher than that in children with course of 3 months to 1year, whereas children with higher parental CCKNOW scores [OR 0.77 (95%CI: 0.67-0.88); p = 0.00] had significantly lower incidence of medication non-adherence than those with lower parental CCKNOW scores, and the results of the correlation between parental knowledge scores of the four categories and children's medication adherence showed that drug knowledge scores (r = 0.36, p < 0.00) and complication knowledge scores (r = 0.24, p = 0.01) were positively correlated with medication adherence. Conclusion: Poor medication adherence in children with IBD in China was common, and forgetting to take medication was the main barrier. Longer disease duration (3years and above) in children could act as a risk factor for medication adherence, whereas higher level of parental knowledge about IBD could act as a protective factor, and one interesting novel finding was that the level of parental knowledge about drug and complication were significantly correlated with medication adherence in children with IBD. Our results may inform on the design and implementation of medication adherence interventions for children with IBD.

Evaluation of the relationship between dermatoglyphics and mandibular third molar impaction: A cross-sectional study.

Background: Dermatoglyphics can be utilised in clinical settings to identify those who are more likely to have impacted teeth. Additionally, dermatoglyphics looks to have potential as a non-invasive diagnostic method for predicting the presence or absence of an impacted tooth. The goal of this study was to look at the most common dermatoglyphic pattern in people who had or didn't have an impacted mandibular third molar teeth and see if there was a dermatoglyphic signature. Methods: A cross-sectional study with 180 participants was conducted (90 cases and 90 controls). The rolling impression technique was used to apply blue duplicating ink to participants' fingertips, which was then recorded. There was an increase in the frequency of the whorl-plain pattern in the right-hand ring finger (60%; p=0.028) and left-hand little finger (33.3%; p=0.009), as well as the loop-ulnar pattern in the right-hand middle finger (74.4%; p=0.024) in individuals with a predisposition to the presence of impacted teeth. Results: The left-hand little finger was found to be the most predictive for impaction in a forward stepwise binary logistic regression analysis. Conclusions: Dermatoglyphics could be used as a non-invasive sign to predict whether or not a tooth is affected. Its value comes in early detection, which helps to avoid the surgical problems that come with symptomatic extraction of an impacted tooth.

Evaluating a Global Assessment Measure Created by Standardized Patients for the Multiple Mini Interview in Medical School Admissions: Mixed Methods Study.

BackgroundStandardized patients (SPs) are essential stakeholders in the multiple mini interviews (MMIs) that are increasingly used to assess medical school applicants’ interpersonal skills. However, there is little evidence for their inclusion in the development of instruments.ObjectiveThis study aimed to describe the process and evaluate the impact of having SPs co-design and cocreate a global measurement question that assesses medical school applicants’ readiness for medical school and acceptance status.MethodsThis study used an exploratory, sequential, and mixed methods study design. First, we evaluated the initial MMI program and determined the next quality improvement steps. Second, we held a collaborative workshop with SPs to codevelop the assessment question and response options. Third, we evaluated the created question and the additional MMI rubric items through statistical tests based on 1084 applicants’ data from 3 cohorts of applicants starting in the 2018-2019 academic year. The internal reliability of the MMI was measured using a Cronbach α test, and its prediction of admission status was tested using a forward stepwise binary logistic regression.ResultsProgram evaluation indicated the need for an additional quantitative question to assess applicant readiness for medical school. In total, 3 simulation specialists, 2 researchers, and 21 SPs participated in a workshop leading to a final global assessment question and responses. The Cronbach α’s were >0.8 overall and in each cohort year. The final stepwise logistic model for all cohorts combined was statistically significant (P<.001), explained 9.2% (R2) of the variance in acceptance status, and correctly classified 65.5% (637/972) of cases. The final model consisted of 3 variables: empathy, rank of readiness, and opening the encounter.ConclusionsThe collaborative nature of this project between stakeholders, including nonacademics and researchers, was vital for the success of this project. The SP-created question had a significant impact on the final model predicting acceptance to medical school. This finding indicates that SPs bring a critical perspective that can improve the process of evaluating medical school applicants.

Longitudinal screening for depression and anxiety in prostate cancer (PC) and association with disease and treatment factors.

5023 Background: Untreated depression and anxiety are associated with worse outcomes in patients with cancer. Despite recommendations for longitudinal screening, many patients are only assessed at the start of care. Men with PC often experience many phases of disease or treatment over a span of years, and androgen deprivation therapy (ADT) is associated with mood changes and depression. How depressive or anxiety symptoms fluctuate in men with PC, influenced by disease and treatment factors, is not well-described. Methods: Men with ≥1 Urology or Medical Oncology clinic visit for PC in the prior 6 months were emailed the PHQ-9 and GAD-7 depression and anxiety screening tools every 60 days; a score of ≥10 (moderate to severe symptoms) on either was considered a positive screen. Baseline characteristics and disease/treatment changes (PSA, radiographic, or biopsy progression, treatment change or start, or discontinuation of treatment due to lack of efficacy or toxicity) were collected by survey and chart review. We report early findings of factors associated with a positive screen or change in screening status with χ2 and forward stepwise binary logistic regression (model inputs: receipt of ADT or disease/treatment change during study, and variables previously associated with depression or anxiety: age, race, marital status, education, income, history of psychiatric disorder, use of psychoactive medication, time since diagnosis, and localized, biochemically recurrent, or metastatic disease). Results: From 6/2021-12/2021, 201 men enrolled. At baseline, 50.7% had localized, 18.9% biochemically recurrent, and 30.3% metastatic disease; 40.8% were on ADT; 30.8% had a history of psychiatric disorder (22.9% depression, 19.9% anxiety, 9.0% other); and 24.9% were on psychoactive medication (19.9% antidepressant, 8.5% anxiolytic, 2.0% antipsychotic). 184 men completed at least 2 screens with mean follow-up 6.5 months (SD 1.3). 32 men (15.9%) screened positive at least once (15.4% PHQ-9, 4.5% GAD-7), of which half (N = 16) initially screened negative and later positive. Changing from a negative to positive screen was more likely when a disease/treatment change occurred during the study (18.3% vs 4.5%, p = 0.003). A higher proportion of men on ADT screened positive, especially if newly started during the study or in the 60 days preceding (35.7% new ADT vs 24.7% continuing ADT vs 8.0% no ADT, p = 0.002). In fully adjusted multivariable analyses, factors associated with a positive screen were history of psychiatric disorder (OR 6.3, 95% CI 2.6-15.4, p &lt; 0.001), receipt of ADT (OR 3.8, 95% CI 1.5-9.5, p = 0.005), and lower income bracket (OR 1.7, 95% CI 1.3-2.5, p = 0.002). Conclusions: Longitudinal screening for depression and anxiety in PC identifies men who initially screen negative. Symptoms are associated with ADT and disease or treatment changes, which may inform optimal screening practices.

Nomogram and Machine Learning Models Predict 1-Year Mortality Risk in Patients With Sepsis-Induced Cardiorenal Syndrome

ObjectiveEarly prediction of long-term outcomes in patients with sepsis-induced cardiorenal syndrome (CRS) remains a great challenge in clinical practice. Herein, we aimed to construct a nomogram and machine learning model for predicting the 1-year mortality risk in patients with sepsis-induced CRS.MethodsThis retrospective study enrolled 340 patients diagnosed with sepsis-induced CRS in Shanghai Tongji Hospital between January 2015 and May 2019, as a discovery cohort. Two predictive models, the nomogram and machine learning model, were used to predict 1-year mortality. The prognostic variables used to develop the nomogram were identified based on a forward stepwise binary logistic regression, and the predictive ability of the nomogram was evaluated by the areas under the receiver operating characteristic curve (AUC) and the calibration curve. Meanwhile, machine learning (ML) techniques, such as support vector machine, random forest (RF), and gradient boosted decision tree, were assessed mainly by accuracy and AUC. Feature ranking analysis was performed using the ML algorithm. Both nomogram and ML models were externally validated by an independent cohort of 103 patients diagnosed with sepsis-induced CRS between June 2019 and December 2020.ResultsAge, sequential sepsis-related organ failure score (SOFA), serum myoglobin (MYO), vasopressor use, and mechanical ventilation were identified as independent risk factors for 1-year mortality in the nomogram predictive model. In the discovery cohort, the nomogram yielded higher AUC for predicting mortality than did the SOFA score (0.855 [95% CI: 0.815–0.895] vs. 0.756 [95% CI: 0.705–0.808]). For ML, the model developed by RF showed the highest accuracy (0.765) and AUC (0.854). In feature ranking analysis, factors such as age, MYO, SOFA score, vasopressor use, and baseline serum creatinine were identified as important features affecting 1-year prognosis. Moreover, the nomogram and RF model both performed well in external validation, with an AUC of 0.877 and 0.863, respectively.ConclusionOur nomogram and ML models showed that age, SOFA score, serum MYO levels, and the use of vasopressors during hospitalization were the main factors influencing the risk of long-term mortality. Our models may serve as useful tools for assessing long-term prognosis in patients with sepsis-induced CRS.

Racial/Ethnic Differences in the Application and Receipt of Services to Address Social Needs: Impact of COVID-19.

Black and Hispanic Americans are disproportionately affected by COVID-19, which impacts their social needs. The objective of this study was to examine differences in white, Black, and Hispanic adults' application and receipt of services to address their social needs during COVID-19. Utilizing weeks 1, 2, and 3 from the COVID Impact Survey, descriptive statistics analyzed covariates and the 12 social services by participants' race/ethnicity. Unweighted frequencies and weighted percentages were computed for the services score by race/ethnicity. Forward stepwise binary logistic regression analyses examined the relationship between services needed and race/ethnicity. All analyses were conducted using STATA MP 14. Among 20 533 participants, unemployment insurance was a common service participants reported applying for or trying to apply for. Compared with white participants, Hispanic participants had higher adjusted odds of needing unemployment insurance services (adjusted odds ratio [AOR] = 1.58; 95% confidence interval [CI], 1.18-2.11). Black (AOR = 3.25; 95% CI, 2.49-4.25) and Hispanic (AOR = 1.55; 95% CI, 1.14-2.10) participants had higher adjusted odds of needing Supplemental Nutrition Assistance Program services than white participants. It is important for Black and Hispanic Americans to have access to these social services. Research and evaluation studies are warranted to inform policies that sustain/modify social services for future use. These studies must include a representative sample of Black and Hispanic Americans.

A new parameter for predict the clinical outcome of patients with COVID-19 pneumonia: The direct/total bilirubin ratio.

AimAn urgent need to define clinical and laboratory parameters to predict progression to the severe and lethal forms of Coronavirus Disease 2019 (COVID‐19). To investigate the direct/total bilirubin ratio (D/TBil), as a novel parameter, to predict the poor survival of COVID‐19 Pneumonia.MethodsThe clinical characteristics and laboratory parameters of hospitalised COVID‐19 pneumonia patients were analysed from 20 March to August 1, 2020, in a tertiary hospital, retrospectively. All remarkable variables were selected for a forward stepwise binary logistic regression analysis to define the independent risk factors for mortality.Results537 (248 women and 289 men) patients were separated into two groups for analysis: survivors vs deceased. The mean age of the deceased group was statistically significantly higher than the survivor group 72 (30‐92) years vs 50 (18‐97) years (P < .001). D/TBil, age, gender, hypertension and neutrophil‐to‐lymphocyte ratio (NLR) variables contributed significantly to the binary logistic regression model. The mortality risk increased 14.6 times in patients with D/TBil > 0.5, and 2.4 times in patients with NLR > 4.ConclusionD/TBil > 0.5 was associated with a novel parameter to poor survival of COVID‐19 on admission. Also, the combination of age, gender, the presence of hypertension, D/TBil and NLR contributed significantly to predicting the poor survival of COVID‐19.

Adverse birth outcomes and the risk of dental caries at age 3years.

Since birth outcome is associated with maternal and newborn health, it can be a predictor of the future health of the child. To investigate the association between adverse birth outcomes and dental caries. The present registry-based cohort study included all children born in 2000-2003, residing in Stockholm County, Sweden and who received a dental examination at the age of 3years (n=74748). National registries supplied data on socioeconomic conditions, maternal health, maternal health behavior, and birth outcomes. Forward stepwise binary logistic regression was performed to determine predictors of caries experience in the 3-year-olds. Of the subjects, 6.0% had caries experience (decayed, extracted, and filled teeth [deft]≥1), 5.6% were born preterm (<37weeks); 2.2% were born small (SGA) and 3.7% large (LGA) for gestational age. Of the studied adverse birth outcomes, only SGA was significantly associated with caries experience at 3years of age, and only for mothers who had refrained from smoking during pregnancy. SGA is associated with caries experience in 3-year-old children; however, this relationship occurs only in mothers who did not smoke during pregnancy.

Trust in Physicians and Regional Brain Volumes: A Population-Based Study.

Trust, a fundamental part of human interpersonal relationships, is known to be associated with specific brain regions and demographic characteristics. Level of trust in medical professionals can alter population health outcomes and influence the nature of the doctor-patient relationship. This study utilized structural magnetic resonance imaging (MRI) and trust data from the Dallas Heart Study (DHS), a large community-based study, to determine brain regions associated with degree of trust in physicians and the medical profession. The first phase of the DHS was conducted from 2000 to 2002 and the second phase from 2007 to 2009. The MRI data were analyzed as part of the DHS using automated FreeSurfer software. Forward stepwise binary logistic regression was performed to investigate the association between measures of trust and bilateral brain region volumes and thickness followed by confirmatory multiple regressions of significant brain regions. A total of 1,596 participants were included in the final analysis. Left caudal anterior cingulate cortex (ACC) thickness was inversely correlated with trust of physicians (P < .01). There were no significant associations between trust in physicians and age, race or ethnicity, or education. The ACC is an integral part of the salience network, the brain network responsible for communication and social behavior. Trust in physicians did not appear to be influenced by demographic characteristics. The findings suggest there are neuroanatomical correlates of trust in physicians.

Prediction of gender employing mandibular measurements: Application of Cone beam computerized tomography

The identification of skeletal remains is of fundamental value in medico-legal fields. The skeletal components most often investigated for gender determination are the pelvis and skull, with the mandible especially in cases of destroyed bodies. The purpose of this study was to define the validity of mandibular measures taken through cone beam computerized tomography in sex determination and to establish a formula for such differentiation. The sample consisted of 100 cone beam computerized tomography (CBCT). The CBCT images were analyzed for five measurements; the distance from the inferior border of the mandibular body to the highest point of the alveolar crest ( IBM-HAC), the distance between the inferior margins of the mental foramen to the inferior border of mandibular body (IMMF-IBM),the distance between the superior margins of the mental foramen to the highest point of the alveolar crest (SMMF-HAC),the distance between the superior margins of the mental foramen to the inferior border of the mandible (SMMF-IBM) ,and volume of mental foramen. (VMF). Data analysis was achieved using SPSS version 22. Forward stepwise binary logistic regression analysis was accomplished to predict the probability of identifying male based on one or more independent variables. The logistic regression for the different varies revealed only the distance between the superior margins of the mental foramen to the inferior border of the mandible (SMMF-IBM), and volume of mental foramen. (VMF) as the significant contributors to the regression model. The logistic regression model was statistically significant and able to 75% of the difference in sex and accurately categorized all cases. Based on the logistic regression model, an equation for prediction of sex was derived. In conclusion, the current study demonstrates the probability of sex determination employing mandibular measurements taken thought CBCT that is proved as a superior tool if compared with the other radiological techniques.

Circulating microRNA-92a level predicts acute coronary syndrome in diabetic patients with coronary heart disease

PurposeThis study was designed to explore the value of monitoring miR-92a in T2DM patients with coronary heart disease (CHD).Materials and methods40 ACS patients with prior history of CHD and diabetes while the onset time of diabetes preceded that of CHD by more than 2 years were enrolled as the DACS group(diabetic ACS group). 40 ACS subjects who had had a definite diagnosis of CHD for more than 2 years with no history of T2DM were recuited as the CACS group(chronic CHD with ACS group). All enrolled subjects from DACS and CACS group came from an emergency basis and diagnosed with ACS by coronary angiography. Another 68 age- and sex-matched volunteers with chronic stable CHD without diabetes history were assigned as the control group (CHD group). We examined the serum levels of miR-92a and analyzed their correlations with blood pressure, glucose level, and lipid level.ResultsThe levels of miR-92a were significantly elevated in the DACS group compared with those of the CACS and CHD groups. Multivariate analysis showed that miR-92a, systolic blood pressure (SBP), and glycosylated hemoglobin (HbA1c) were significantly related to ACS events in patients with T2DM. Forward stepwise binary logistic regression analysis identified miR-92a as an independent predictive factor for ACS events in the patients with T2DM.ConclusionAn elevated circulating miR-92a level was associated with an increased risk of ACS in CHD patients with T2DM. Thus the level of miR-92a, especially combined with elevated SBP and HbA1c, may be helpful in the detection of ACS in patients with T2DM.

Preoperative risk factors for massive blood loss in adrenalectomy for pheochromocytoma.

BackgroundThis retrospective analysis of patients who underwent adrenalectomy for pheochromocytoma aimed to determine preoperative risk factors for intraoperative massive blood loss. Preoperative identification of patients at high-risk of massive blood loss may be helpful in anesthesia management and preoperative preparation.Materials and MethodsThe study involved data of 268 patients who had undergone pheochromocytoma surgery at the Peking Union Medical College Hospital between January 1, 2013 and October 31, 2016. For analysis, the patients were grouped according to intraoperative blood loss: ≥ 20% of estimated blood volume (group A, n = 38) and < 20% of estimated blood volume (group B, n = 230). Perioperative characteristics were compared between the two groups. Significant variables were selected for a forward stepwise binary logistic regression analysis to determine the independent risk factors for massive blood loss.ResultsThe two groups showed significant differences in tumor location, tumor size, operative approach, preoperative 24-hour urine level of total noradrenaline, preoperative hemoglobin concentration, phenoxybenzamine maximum daily dose, preoperative preparation time, intraoperative urine volume, crystalloid and colloidal fluid volumes, allogeneic red blood cell transfusion, plasma and autologous blood transfusion volumes, incidence of prolonged hypotension, postoperative drainage volume, lowest and discharge hemoglobin concentrations, length of stay in intensive care unit and length of postoperative hospitalization. Binary logistic regression analysis indicated increased risk of intraoperative massive blood loss in subjects with tumors proximal to vessels or other organs (odds ratio (OR): 4.227), with tumors ≥ 5 cm (OR: 7.321), or with preoperative preparation time of ≤ 14 days (OR: 17.747).ConclusionsTumors proximal to vessels and other organs or with maximum diameter of ≥ 5 cm (as shown by preoperative radiographic evidence), and preoperative preparation time of ≤ 14 days were independent risk factors of intraoperative massive blood loss in patients treated with adrenalectomy for pheochromocytoma.

Developing Lebanon's fire danger forecast

Lebanon lacks a system for forecasting wildfire danger with the combined use of dynamic weather forecasts and comprehensive fire risk information (i.e., hazard and vulnerability). Accordingly, the aim of this study was to develop a new national fire danger forecast system for use as an operational fire management tool in Lebanon. The specific objectives were to 1) produce an advanced fire risk map of Lebanon, 2) customize an index for use in fire danger forecasting, and 3) develop an automated system for daily fire danger forecasts. Producing an advanced fire risk map involved the combined use of collected fire risk data and predicted wildfire occurrence. Forward stepwise binary logistic regression analysis of 24 socio-economic and biophysical variables was initially adopted to predict wildfire occurrence. Geographic object-based image analysis (GEOBIA) approach was employed to map the different classes of fire risk. Consequently, the customization of an index for fire danger forecasting involved the use of both fire risk and Fire Weather Index (FWI) data using also the GEOBIA approach. Overall, modeling socio-economic and biophysical factors correctly predicted an average of 85.7% of fire occurrences between 1998 and 2012 across the country. In addition, a comparison of FWI forecasts and fire occurrence (i.e., observed fire events) during the fire season of 2015 resulted in sensitivity (76.2%) and specificity (18.6%) estimates. Finally, automated daily fire danger forecasts are expected to serve as essential tools for both scientific and operational purposes.

Exploring relationships between grid cell size and accuracy for debris-flow susceptibility models: a test in the Giampilieri catchment (Sicily, Italy)

Debris flows are among the most hazardous phenomena in nature, requiring the preparation of susceptibility models in order to cope with this severe threat. The aim of this research was to verify whether a grid cell-based susceptibility model was capable of predicting the debris-flow initiation sites in the Giampilieri catchment (10 km2), which was hit by a storm on the 1st October 2009, resulting in more than one thousand landslides. This kind of event is to be considered as recurrent in the area as attested by historical data. Therefore, predictive models have been prepared by using forward stepwise binary logistic regression (BLR), a landslide inventory and a set of geo-environmental attributes as predictors. In particular, the effects produced in the quality of the predictive models by changing the grid cell size (2, 4, 16 and 32 m) have been explored in terms of predictive performance, robustness, importance and role of the selected predictors. The results generally attested for high predictive performances of the 2, 8 and 16 m model sets (AUROC > 0.8), with the latter producing slightly better predictions and the 32 m showing the worst yet still acceptable performance and the lowest robustness. As regards the predictors, although all the 4 sets of models share a common group (topographic attributes, outcropping lithology and land use), the similarity resulted higher between the 8 and 16 m sets. The research demonstrates that no meaningful loss in the predictive performance arises by adopting a coarser cell size for the mapping unit. However, the largest adopted cell size resulted in marginally worse model performance, with AUROC slightly below 0.8 and error rates above 0.3.

Estimation of Odds and Probability Risk Factors in Experiencing Heart Attack

The Heart Attack has produced an alarming rate of deaths and considered as one of the most deadly events for persons with hypertension, diabetes, with family history, a smoker and alcoholic. The study aims to produce a model using the Forward Stepwise Binary Logistic Regression Method for estimating the odds and probabilities when conceded with the given situations, and also the comparisons of the odds in which the factors are arranged by combinations of the significant variables that are included in the equation of every circumstances. Combination of covariates, gender aside, where the outcome of the study indicates that men are two times more likely to have heart attack than women. Hypertension and Diabetes, two covariates giving two of the highest odds among the factors, is 35 times more likely to trigger a heart attack than a person with none. Smoking plus the two mentioned covariates gives 114 times more. Alcohol added, has the ratio of 210 to 1 which is harshly a huge number. Summing up all the risks of the factors considered, including the family history as the last covariate, it shows the ratio 352 times that is, than a person with no complication nor any harmful vices.

Water sources as reservoirs of Vibrio cholerae O1 and non-O1 strains in Bepanda, Douala (Cameroon): relationship between isolation and physico-chemical factors.

BackgroundCholera has been endemic in Douala since 1971. Most outbreaks start from Bepanda, an overcrowded neighbourhood with poor hygiene and sanitary conditions. We investigated water sources in Bepanda as reservoirs of Vibrio cholerae, the causative agent of cholera, determined its antibiotic susceptibility and some physico-chemical characteristics that could maintain the endemicity of this organism in Bepanda.MethodsThree hundred and eighteen water samples collected from 45 wells, 8 taps and 1 stream from February to July 2009 were analyzed for V. cholerae using standard methods. Isolates were characterized morphologically, biochemically and serologically. The disc diffusion technique was employed to investigate antibiotic susceptibility. Differences in prevalence of organism between seasons were analysed. Correlation strength and direction of association between physico-chemical parameters and occurrence of V. cholerae was analyzed using the Kendall tau_b non-parametric correlation. This was further confirmed with the forward-stepwise binary logistic regression.ResultsEighty-seven (27.4%) samples were positive for V. cholerae. Isolation was highest from wells. The organism was isolated in the rainy season and dry season but the frequency of isolation was significantly higher (χ2 = 7.009, df = 1, P = 0.008) in the rainy season. Of the 96 confirmed V. cholerae isolates, 32 (33.3%) belonged to serogroup O1 and 64 (66.6%) were serogroup non-O1/non-O139. Isolates from tap (municipal water) were non-O1/non-O139 strains. Salinity had a significant positive correlation with isolation in the dry season (+0.267, P = 0.015) and rainy season (+0.223, P = 0.028). The forward-stepwise method of binary logistic regression indicated that as pH (Wald = 11.753, df = 1), P = 0.001) increased, odds of isolation of V. cholerae also increased (B = 1.297, S.E = 0.378, Exp(B) = 3.657). All isolates were sensitive to ciprofloxacin and ofloxacin. Multi-drug resistance was predominant among the non-O1/non-O139 isolates.ConclusionV. cholerae was found in wells and stream in both seasons. Cholera will continue to be a health threat in Bepanda if intervention measures to prevent outbreak are not implemented. Continuous monitoring of water sources in this and other cholera high-risk areas in Cameroon is necessary, for a better preparedness and control of cholera.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2334-14-421) contains supplementary material, which is available to authorized users.

Decline in platelet count as a prognostic marker in critically ill children

Objectives: To study the variation in platelet count in critically ill children and correlate its association with their outcomes Methods: This was a prospective cross-sectional analysis of 150 critically ill children admitted in PICU of tertiary care hospital over period of 1 year. Laboratory data was collected with daily platelet counts from day of admission till death or discharge whichever was earlier in all patients. The study population was grouped as thrombocytopenic (platelet count < 150 χ 10 /L) and nonthrombocytopenic. They were compared with each other with respect to laboratory parameters and risk factors. Survivor and non survivors were compared with variation in platelet count. Decline in platelet count was correlated with mortality. Chi-square test, median test, ROC curve and forward stepwise binary logistic regression was used for statistical analysis. Results: Forty eight (32%) children had thrombocytopenia. They had significantly higher mortality [14(29.16%) vs 14(1372%)] and bleeding tendency [13(28.08%) vs 4(3.92%) than non- thrombocytopenic children. Admission thrombocytopenia was not found to be risk factor for mortality. Though survivors and non-survivors had decline in platelet count in first four days, non-survivors had significantly higher drop. Platelet counts decline >30% at 72 hours was independent risk factor (odds ratio 4.126) for mortality with high sensitivity(91.7%) and specificity (84.4%) with area under ROC curve of 0.898 which was associated with PRISMII. Conclusion: Decline in platelet count can be used as prognostic marker of poor outcome in critically ill children.