Tag: statistics

Int J Colorectal Dis. 2022 Nov 11. doi: 10.1007/s00384-022-04274-9. Online ahead of print.

ABSTRACT

PURPOSE: To investigate whether body mass index (BMI) is a risk factor for inadequate bowel preparation in elective colonoscopy. The null hypothesis being BMI does not affect bowel preparation adequacy.

METHODS: A retrospective cohort study of all participants with complete medical records who had an elective colonoscopy was conducted across three tertiary teaching hospitals in Perth, Western Australia, from January 2016 to July 2019. Participants were separated into BMI subgroups of healthy weight, overweight and obese (≥ 30 kg/m²). Data were extracted from medical records, colonoscopy and histopathology reports and were analysed using SPSS v.27.

RESULTS: Of the 1082 cases analysed, 52.7% (n = 570) were male. The median age was 61 (range 18-85 years). The median BMI was 27.8 (range 20-52). The median procedure time is 28 (range 2-69 min). Routine follow-up was the clinical indication for 65% of colonoscopy procedures undertaken during the study period. Multivariate logistic regression, controlled for statistically insignificant confounders of age, type of bowel preparation agent, grade of the endoscopist, the indication for procedure and year of procedure, showed that being obese was significantly and independently associated with inadequate bowel preparation (OR 2.0, 95% CI (1.4-2.9) p < 0.001). Another significant factor was male (OR 1.6, 95% CI (1.2-2.1) p = 0.002).

CONCLUSION: This study shows that obese patients are more likely to have inadequate bowel preparation at colonoscopy. Given the increased complication rates and health care costs associated with repeating colonoscopies and the increased risk of colorectal cancer in obese patients, it may be worth tailoring a more extensive bowel preparation regimen to ensure adequate visualisation of the colonic mucosa on the first attempt.

PMID:36357734 | DOI:10.1007/s00384-022-04274-9

Nat Microbiol. 2022 Nov 10. doi: 10.1038/s41564-022-01262-1. Online ahead of print.

ABSTRACT

Measuring immune correlates of disease acquisition and protection in the context of a clinical trial is a prerequisite for improved vaccine design. We analysed binding and neutralizing antibody measurements 4 weeks post vaccination as correlates of risk of moderate to severe-critical COVID-19 through 83 d post vaccination in the phase 3, double-blind placebo-controlled phase of ENSEMBLE, an international randomized efficacy trial of a single dose of Ad26.COV2.S. We also evaluated correlates of protection in the trial cohort. Of the three antibody immune markers we measured, we found most support for 50% inhibitory dilution (ID₅₀) neutralizing antibody titre as a correlate of risk and of protection. The outcome hazard ratio was 0.49 (95% confidence interval 0.29, 0.81; P = 0.006) per 10-fold increase in ID₅₀; vaccine efficacy was 60% (43%, 72%) at non-quantifiable ID₅₀ (<2.7 IU₅₀ ml^-1) and increased to 89% (78%, 96%) at ID₅₀ = 96.3 IU₅₀ ml^-1. Comparison of the vaccine efficacy by ID₅₀ titre curves for ENSEMBLE-US, the COVE trial of the mRNA-1273 vaccine and the COV002-UK trial of the AZD1222 vaccine supported the ID₅₀ titre as a correlate of protection across trials and vaccine types.

PMID:36357712 | DOI:10.1038/s41564-022-01262-1

Ann Surg Oncol. 2022 Nov 10. doi: 10.1245/s10434-022-12756-x. Online ahead of print.

ABSTRACT

INTRODUCTION: Right-side hepatectomy (RH) is used in oncological resection for perihilar cholangiocarcinoma (PHC); however, the decision between performing left-side hepatectomy (LH) or RH is still controversial. We compared surgical and oncologic outcomes of LH and RH in PHC type II or IV where either hepatectomy was expected to have a negative margin.

METHODS: From 2001 to 2020, 99 patients underwent major liver resection for type II or IV PHC. Patients with unilateral vascular invasion, unilateral tumor growth, and atrophy of unilateral liver were excluded. Preoperative characteristics, perioperative, and long-term outcomes were compared between the remaining RH and LH patients.

RESULTS: After excluding 47 cases with side predominance, the RH group (n = 29) and LH group (n = 23) were compared. Clinical characteristics and disease severity did not differ between the groups. Portal vein embolization (RH: 48.3% vs. LH: 0.0%, p < 0.001) and days from diagnosis to operation (RH: 31.0 ± 16.2 vs. LH: 18.8 ± 13.4, p = 0.006) were significantly higher in the RH group. The RH group had statistically higher rate of postoperative hepatic failure (RH: 55.2% vs. LH: 21.7%, p = 0.015) and a higher mortality rate that was not significant (RH: 13.8% vs. LH: 0%, p = 0.120). The R0 resection rate (RH: 72.4% vs. LH: 78.3%, p = 0.629), median disease-free (p = 0.620), and overall (p = 0.487) survival did not differ between groups. R1 resection and lymph node metastasis were significant risk factors for disease-free survival in multivariate analysis.

CONCLUSIONS: In type II or type IV PHC where either LH or RH was feasible, LH provided a shorter period of preoperative preparation, lower postoperative hepatic failure rate, similar R0 rate, and comparable long-term outcomes. LH should be considered a reasonable option in type II or IV PHC.

PMID:36357701 | DOI:10.1245/s10434-022-12756-x

Nat Methods. 2022 Nov 10. doi: 10.1038/s41592-022-01666-1. Online ahead of print.

ABSTRACT

RNA modifications such as m6A methylation form an additional layer of complexity in the transcriptome. Nanopore direct RNA sequencing can capture this information in the raw current signal for each RNA molecule, enabling the detection of RNA modifications using supervised machine learning. However, experimental approaches provide only site-level training data, whereas the modification status for each single RNA molecule is missing. Here we present m6Anet, a neural-network-based method that leverages the multiple instance learning framework to specifically handle missing read-level modification labels in site-level training data. m6Anet outperforms existing computational methods, shows similar accuracy as experimental approaches, and generalizes with high accuracy to different cell lines and species without retraining model parameters. In addition, we demonstrate that m6Anet captures the underlying read-level stoichiometry, which can be used to approximate differences in modification rates. Overall, m6Anet offers a tool to capture the transcriptome-wide identification and quantification of m6A from a single run of direct RNA sequencing.

PMID:36357692 | DOI:10.1038/s41592-022-01666-1

World J Pediatr. 2022 Nov 10. doi: 10.1007/s12519-022-00635-0. Online ahead of print.

ABSTRACT

BACKGROUND: Bronchopulmonary dysplasia (BPD) is a common chronic lung disease in extremely preterm neonates. The outcome and clinical burden vary dramatically according to severity. Although some prediction tools for BPD exist, they seldom pay attention to disease severity and are based on populations in developed countries. This study aimed to develop machine learning prediction models for BPD severity based on selected clinical factors in a Chinese population.

METHODS: In this retrospective, single-center study, we included patients with a gestational age < 32 weeks who were diagnosed with BPD in our neonatal intensive care unit from 2016 to 2020. We collected their clinical information during the maternal, birth and early postnatal periods. Risk factors were selected through univariable and ordinal logistic regression analyses. Prediction models based on logistic regression (LR), gradient boosting decision tree, XGBoost (XGB) and random forest (RF) models were implemented and assessed by the area under the receiver operating characteristic curve (AUC).

RESULTS: We ultimately included 471 patients (279 mild, 147 moderate, and 45 severe cases). On ordinal logistic regression, gestational diabetes mellitus, initial fraction of inspiration O₂ value, invasive ventilation, acidosis, hypochloremia, C-reactive protein level, patent ductus arteriosus and Gram-negative respiratory culture were independent risk factors for BPD severity. All the XGB, LR and RF models (AUC = 0.85, 0.86 and 0.84, respectively) all had good performance.

CONCLUSIONS: We found risk factors for BPD severity in our population and developed machine learning models based on them. The models have good performance and can be used to aid in predicting BPD severity in the Chinese population.

PMID:36357648 | DOI:10.1007/s12519-022-00635-0

Int Urol Nephrol. 2022 Nov 11. doi: 10.1007/s11255-022-03409-9. Online ahead of print.

ABSTRACT

PURPOSE: Plasma atherogenic index (PAI) was shown to be positively correlated with the presence of malignity in patients with suspicious findings for renal cell cancer and colon cancer in reported studies. In this study, we aimed to evaluate whether there is an association with the presence of malignity in patients PI-RADS 3 prostate lesions and PAI.

METHODS: This retrospective study reviewed the data of 139 patients who underwent transrectal ultrasonography-guided systematic and cognitive fusion prostate biopsy for PI-RADS 3 lesions in multiparametric magnetic resonance imaging. The patients were divided to two groups as malign (n = 33) and benign (n = 106). The association between age, body mass index, comorbidities, smoking status, prostate-specific antigen (PSA), PSA density, free/total PSA, prostate weight, lesion diameter, triglyceride value, high-density lipoprotein-cholesterol value, PAI value data and presence of malignity were investigated by descriptive, multivariate and receiver-operating characteristic (ROC) analysis.

RESULTS: PSA, PSAD, lesion diameter and PAI value were statistically significantly higher in the malignant group compared to the benign group, and the free/total PSA ratio was lower. In multivariate logistic regression analysis, PSA > 9.9 ng/ml, free/total PSA < 12.1%, lesion diameter > 13.5 mm and PAI > 0.13 were identified as independent risk factors for presence of prostate malignancy.

CONCLUSION: PAI was found to be a predictive parameter for prostate cancer in PI-RADS 3 prostate lesions. Our study can open new thoughts about PAI as metric to assess the prostate cancer risk.

PMID:36357644 | DOI:10.1007/s11255-022-03409-9

Neurosurg Rev. 2022 Nov 11. doi: 10.1007/s10143-022-01889-y. Online ahead of print.

ABSTRACT

The literature has had some conflicting evidence regarding the effective management of lumbar spondylolisthesis (LS). Herein, we review active and prospective clinical trials to identify the emerging trends for the management of LS. A systematic search was conducted utilizing the NIH Clinical Trials database using the search term “lumbar spondylolisthesis” on February 2, 2022. Currently active and prospective clinical trials for LS were included and analyzed. All statistical analyses were performed on R 4.1.2. We identified 37 clinical trials. Nearly half the trials (n = 18, 48.6%) include novel technologies; 6 (16.2%) are comparing surgical approaches, of which 4 (67%) include decompression alone versus decompression with instrumented fusion; 6 (16.2%) are evaluating perioperative pain management protocols, of which 3 (50%) include bupivacaine or ropivacaine; 3 (8.1%) are evaluating alternative medicines in LS; 2 (5.4%) are observational studies about the natural history of LS; 1 (2.7%) involves surgical infection prophylaxis; and 1 (2.7%) is evaluating AK1320 microspheres. The 18 trials involving novel technologies include 3D-printed titanium cages (n = 3, 16.7%), interbody implants (n = 4, 22.2%), bone graft materials (n = 4, 22.2%), and miscellaneous intraoperative devices (n = 7, 38.9%). The top 3 outcomes measured were Oswestry Disability Index (n = 28, 75.7%), visual analog scale (n = 21, 56.7%), and postoperative radiographs (n = 16, 43.2%). Patient-reported outcome measures (PROMs) were included in 34 (91.9%) trials, while 23 (62.2%) trials included lumbar spine imaging. LS can often require a multifaceted approach. Novel technologies and utilization of PROMs appear to be a significant emerging trend in LS management.

PMID:36357642 | DOI:10.1007/s10143-022-01889-y

Phys Eng Sci Med. 2022 Nov 10. doi: 10.1007/s13246-022-01196-2. Online ahead of print.

ABSTRACT

The study aims to implement a convolutional neural network framework that uses the 18F-FDG PET modality of brain imaging to detect multiple stages of dementia, including Early Mild Cognitive Impairment (EMCI) and Late Mild Cognitive Impairment (LMCI), and Alzheimer’s disease (AD) from Cognitively Normal (CN), and assess the results. 18F-FDG PET imaging modality for brain were procured from Alzheimer’s disease neuroimaging initiative’s (ADNI) repository. The ResNet50V2 model layers were utilised for feature extraction, with the final convolutional layers fine-tuned for this dataset’s multi-classification objectives. Multiple metrics and feature maps were utilized to scrutinize and evaluate the model’s statistical and qualitative inference. The multi-classification model achieved an overarching accuracy of 98.44% and Area under the receiver operating characteristic curve of 95% on the testing set. Feature maps aided in deducing finer aspects of the model’s overall operation. This framework helped classifying from the 18F-FDG PET brain images, the subtypes of Mild Cognitive Impairment (MCI) which include EMCI, LMCI, from AD, CN groups and achieved an all-inclusive sensitivity of 94% and specificity of 95% respectively.

PMID:36357627 | DOI:10.1007/s13246-022-01196-2

Eur Arch Otorhinolaryngol. 2022 Nov 10. doi: 10.1007/s00405-022-07730-y. Online ahead of print.

ABSTRACT

BACKGROUND: Several reports examined the survival of laryngeal cancer (LC) patients, most of these studies only focused on the prognosis of the disease, and just a small number of studies examined non-cancer-related causes of death. The objective of the current study is to investigate and quantify the most common causes of deaths following LC diagnosis.

METHODS: The data of 44,028 patient with LC in the United States diagnosed between 2000 and 2018 were retrieved from the Surveillance, Epidemiology, and End Results (SEER) program and analyzed. We stratified LC patients according to various demographic and clinical parameters and calculated standardized mortality ratios (SMRs) for all causes of death.

RESULTS: Over the follow-up period, 25,407 (57.7%) deaths were reported. The highest fatalities (11,121; 43.8%) occurred within 1-5 years following LC diagnosis. Non-cancer causes of death is the leading cause of death (8945; 35.2%), followed by deaths due to laryngeal cancer (8,705; 34.3%), then other cancers deaths (7757; 30.5%). The most common non-cancer causes of death were heart diseases (N = 2953; SMR 4.42), followed by other non-cancer causes of death (N = 1512; SMR 3.93), chronic obstructive pulmonary diseases (N = 1420; SMR 4.90), then cerebrovascular diseases (N = 547; SMR 4.28). Compared to the general population, LC patients had a statistically significant higher risk of death from all reported causes.

CONCLUSIONS: Non-cancer causes of death is the leading cause of death in LC patients, exceeding deaths attributed to LC itself. These findings provide important insight into how LC survivors should be counselled regarding future health risks.

PMID:36357608 | DOI:10.1007/s00405-022-07730-y

Intern Emerg Med. 2022 Nov 10. doi: 10.1007/s11739-022-03124-4. Online ahead of print.

ABSTRACT

To date, we do not know if the excess of the body mass index (BMI) improves or worsens the outcomes in colorectal cancer treatment, and the correlation between BMI and prognosis remains unclear. A recent study in vitro showed a significant negative correlation between BMI and Cetuximab-induced antibody-dependent cellular cytotoxicity. On these bases, we tried to analyze the potential correlation between BMI and survival in patients affected by metastatic colorectal cancer (mCRC) and treated with Cetuximab. Retrospective data were collected from 132 patients affected by mCRC treated with Cetuximab in monotherapy or association with chemotherapy between January 2007 and October 2019. The cohort of patients was divided into different groups according to the World Health Organization (WHO) BMI classification: underweight (BMI < 18.59), normal weight (BMI 18.5-24.9,) overweight (BMI 25-29.9), and obese (BMI > 30), and we observed the influence of BMI on survival and treatment response. Patients with BMI ≥ 25 had statistically significantly better survival than patients BMI < 25 (19 vs 10 months, p = 0.025). Dividing the sample into the four WHO BMI categories, the best survival rates were seen in the overweight and obese subgroups (18 and 26 months respectively, p < 0.01). The multivariate analysis confirmed BMI as the only parameter able to influence survival. No correlation between BMI and treatment response was seen between BMI ≥ 25 and BMI ≤ 24 groups (p = 0.14). Our experience suggests that mild obese and overweight patients treated with Cetuximab could experience a better survival. We also observed that among normal weight, overweight, and mild obese patients, there is a better response to immunochemotherapy in comparison with underweight patients, but this difference does not reach a significative statistical value.

PMID:36357605 | DOI:10.1007/s11739-022-03124-4