Category: Statistics

Stat Med. 2021 Apr 11. doi: 10.1002/sim.8950. Online ahead of print.

ABSTRACT

We consider the non-trivial problem of estimating a health cost repartition among diseases from patients’ hospital stays’ global costs in the presence of multimorbidity, that is, when the patients may suffer from more than one disease. The problem is even harder in the presence of interactions among the disease costs, that is, when the costs of having, for example, two diseases simultaneously do not match the sum of the basic costs of having each disease alone, generating an extra cost which might be either positive or negative. In such a situation, there might be no “true solution” and the choice of the method to be used to solve the problem will depend on how one wishes to allocate the extra costs among the diseases. In this article, we study mathematically how different methods proceed in this regard, namely ordinary least squares (OLS), generalized linear models (GLM), and an iterative proportional repartition (IPR) algorithm, in a simple case with only two diseases. It turned out that only IPR allowed to retrieve the total costs and the unambiguous solution that one would have in a setting without interaction, that is, when no extra cost has to be allocated, while OLS and GLM may produce some negative health costs. Also, contrary to OLS, IPR is taking into account the basic costs of the diseases for the allocation of the extra cost. We conclude that IPR seems to be the most natural method to solve the problem, at least among those considered.

PMID:33843071 | DOI:10.1002/sim.8950

Health Care Manag Sci. 2021 Apr 12. doi: 10.1007/s10729-021-09561-5. Online ahead of print.

ABSTRACT

Demand for Personal Protective Equipment (PPE) such as surgical masks, gloves, and gowns has increased significantly since the onset of the COVID-19 pandemic. In hospital settings, both medical staff and patients are required to wear PPE. As these facilities resume regular operations, staff will be required to wear PPE at all times while additional PPE will be mandated during medical procedures. This will put increased pressure on hospitals which have had problems predicting PPE usage and sourcing its supply. To meet this challenge, we propose an approach to predict demand for PPE. Specifically, we model the admission of patients to a medical department using multiple independent [Formula: see text] queues. Each queue represents a class of patients with similar treatment plans and hospital length-of-stay. By estimating the total workload of each class, we derive closed-form estimates for the expected amount of PPE required over a specified time horizon using current PPE guidelines. We apply our approach to a data set of 22,039 patients admitted to the general internal medicine department at St. Michael’s hospital in Toronto, Canada from April 2010 to November 2019. We find that gloves and surgical masks represent approximately 90% of predicted PPE usage. We also find that while demand for gloves is driven entirely by patient-practitioner interactions, 86% of the predicted demand for surgical masks can be attributed to the requirement that medical practitioners will need to wear them when not interacting with patients.

PMID:33843005 | DOI:10.1007/s10729-021-09561-5

Virchows Arch. 2021 Apr 12. doi: 10.1007/s00428-021-03090-w. Online ahead of print.

ABSTRACT

Tumor budding scoring guidelines from the International Tumor Budding Consensus Conference (ITBCC) for colorectal cancer propose three groups: BD1 (0-4 buds/0.785 mm²), BD2 (5-9 buds/0.785 mm²), and BD3 (10 or more buds/0.785 mm²). Here, we investigate whether a fourth scoring category, namely zero buds, may have additional clinical relevance. The number of tumor buds/0.785 mm² was scored in 959 cases. Those with zero tumor buds were considered BD0, while a new BD1 category of 1-4 buds was proposed. Associations of both scoring approaches with clinicopathological features were analyzed. Conventional ITBCC scoring showed expected associations with unfavorable histopathological prognostic factors. In total, 111/959 (11.6%) were BD0. A significant difference was found when BD0 was compared statistically to BD1 (1-4 buds) for pT, TNM, tumor grade, and lymphatic, venous, and perineural invasion (p < 0.01, all). Tumors with BD0 occur relatively frequently and contribute additional information on tumor behavior. BD0 should be considered for subsequent ITBCC guidelines.

PMID:33843013 | DOI:10.1007/s00428-021-03090-w

Int J Implant Dent. 2021 Apr 12;7(1):27. doi: 10.1186/s40729-021-00310-5.

ABSTRACT

BACKGROUND: This study aimed to investigate the effects of systemic omeprazole treatment on the osseointegration of titanium implants.

MATERIAL AND METHODS: After surgical insertion of titanium implants into the metaphyseal part of rats’ both right and left tibial bones, the animals were randomly divided into three equal groups: control (n = 8), omeprazole dosage-1 (n = 8) (OME-1), and omeprazole dosage-2 (n = 8) (OME-2) and totally 48 implants were surgically integrated. The rats in the control group received no treatment during the four-week postoperative experimental period. In the OME-1 and OME-2 groups, the rats received omeprazole in doses of 5 and 10 mg/kg, respectively, every 3 days for 4 weeks. After the experimental period, the rats were euthanized. One rat died in each group and the study was completed with seven rats in each group. Blood serum was collected for biochemical analysis, and the implants and surrounding bone tissue were used for biomechanical reverse-torque analysis. In the biomechanical analysis, implants that were not properly placed and were not osseointegrated were excluded from the evaluation.

RESULTS: One-way analysis of variance and Tukey’s honestly significant difference test and Student’s t test were used for statistical analysis. The reverse-torque test (control (n = 9), OME-1 (N = 7), and OME-2 (n = 7)) analysis of biochemical parameters (alkaline phosphatase, calcium, phosphorus, aspartate aminotransferase, alanine amino transferase, urea, and creatinine) revealed no significant differences between the groups (control (n = 7), OME-1 (N = 7), and OME-2 (n = 7)) (P > 0.05).

CONCLUSIONS: Omeprazole had no biomechanical or biochemical effects on the osseointegration process of titanium implants.

PMID:33843027 | DOI:10.1186/s40729-021-00310-5

Sleep. 2021 Apr 12:zsab094. doi: 10.1093/sleep/zsab094. Online ahead of print.

ABSTRACT

STUDY OBJECTIVES: To evaluate interrater reliability for artefact correction in the context of semi-automated quantification of rapid eye movement (REM) sleep without atonia (RWA) in the mentalis and flexor digitorum superficialis (FDS) muscles.

METHODS: We included video-polysomnographies of 14 subjects with apnea-hypopnea-index in REM sleep (AHIREM)<15/h and 11 subjects with AHIREM≥15/h. Eight subjects had isolated REM sleep behavior disorder. A validated algorithm (www.osg.be) automatically scored phasic and “any” EMG activity in the mentalis muscle, and phasic EMG activity in the FDS muscles. Four independent expert scorers performed artefact correction according to the SINBAR (Sleep Innsbruck Barcelona) recommendations. Interrater reliability for artefact correction was computed with B-statistics. The variability across scorers of four RWA indices (phasic mentalis, “any” mentalis, phasic FDS and SINBAR – i.e. “any” mentalis and/or phasic FDS – EMG activity indices) was computed. With Friedman tests we compared B-statistics obtained for mentalis and FDS muscles, and the variability of the RWA indices. Influence of AHIREM and RBD diagnosis on the RWA indices variability was evaluated with linear regressions.

RESULTS: Interrater reliability for artefact correction was higher in the FDS than in the mentalis muscle (p<0.001). Phasic FDS activity was minimally affected by artefacts. Accordingly, the phasic FDS EMG activity index had the lowest variability across scorers (p<0.001). Variability across scorers of the RWA indices including the mentalis muscle increased with AHIREM and was independent from RBD diagnosis.

CONCLUSIONS: Due to the consistently found low number of artefacts, phasic FDS activity is a reliable measure of RWA.

PMID:33842971 | DOI:10.1093/sleep/zsab094

Eur J Trauma Emerg Surg. 2021 Apr 11. doi: 10.1007/s00068-021-01640-0. Online ahead of print.

ABSTRACT

INTRODUCTION: Dementia is common in patients with hip fractures and is strongly associated with increased postoperative mortality. The choice of surgical intervention for displaced femoral neck fractures (dFNF) in patients with dementia has been a matter of debate. This study aims to investigate how short- and long-term mortality differs between those who have been operated with hemiarthroplasty or pins/screws.

METHODS: All patients with dementia and dFNF, i.e., Garden III and IV, who underwent primary emergency hip fracture surgery, with either hemiarthroplasty or pins/screws, in Sweden between Jan 1, 2008 and Dec 31, 2017 were eligible for inclusion in the current study. Patients were divided into two groups based on the surgical intervention: hemiarthroplasty and pins/screws. The primary outcome of interest was 30-day postoperative mortality, and the secondary outcome was 1-year postoperative mortality. Poisson and Cox regression analyses were performed both before and after propensity score matching.

RESULTS: A total of 9394 cases met the inclusion criteria; 84% received hemiarthroplasty and 16% received pins/screws. In the unmatched analysis, the adjusted incidence rate ratio (IRR) for 30-day postoperative mortality was not affected by the chosen surgical method (adj. IRR 0.96, CI 95% 0.83-1.12, p = 0.629). After propensity score matching, similar results were observed with no difference in 30-day postoperative mortality (adj. IRR 0.89, CI 95% 0.74-1.09, p = 0.286). There was a statistically significant decrease in the risk of 1-year postoperative mortality in the hemiarthroplasty group compared to the pins/screws group, both before and after propensity score matching.

CONCLUSION: This study could not demonstrate any difference in 30-day mortality in patients with dementia and dFNFs when comparing hemiarthroplasty with pins/screws. Patients that received hemiarthroplasties did, however, have a lower risk of 1-year postoperative mortality.

PMID:33842982 | DOI:10.1007/s00068-021-01640-0

IBRO Neurosci Rep. 2021 Feb 18;10:178-185. doi: 10.1016/j.ibneur.2021.02.007. eCollection 2021 Jun.

ABSTRACT

Current assessments of recovery following spinal cord injury (SCI) focus on clinical outcome measures. These assessments bear an inherent risk of bias, emphasizing the need for more reliable prognostic biomarkers to measure SCI severity. This study evaluated fluid biomarkers as an objective tool to aid with prognosticating outcomes following SCI. Using a ¹H nuclear magnetic resonance (NMR)-based quantitative metabolomics approach of urine samples, the objectives were to determine (a) if alterations in metabolic profiles reflect the extent of recovery of individual SCI patients, (b) whether changes in urine metabolites correlate to patient outcomes, and (c) whether biological pathway analysis reflects mechanisms of neural damage and repair. An inception cohort exploratory pilot study collected morning urine samples from male SCI patients (n=6) following injury and again at 6-months post-injury. A 700 MHz Bruker Avance III HD NMR spectrometer was used to acquire the metabolic signatures of urine samples, which were used to derive metabolic pathways. Multivariate statistical analyses were used to identify changes in metabolic signatures, which were correlated to clinical outcomes in the Spinal Cord Independence Measure (SCIM). Among SCI-induced metabolic changes, biomarkers which significantly correlated to patient SCIM scores included caffeine (R = -0.76, p < 0.01), 3-hydroxymandelic acid (R= -0.85, p < 0.001), L-valine (R = 0.90, p < 0.001; R = -0.64, p < 0.05), and N-methylhydantoin (R = -0.90, p < 0.001). The most affected pathway was purine metabolism. These findings indicate that urinary metabolites reflect SCI lesion severity and recovery and provide potentially prognostic biomarkers of SCI outcome in precision medicine approaches.

PMID:33842921 | PMC:PMC8020035 | DOI:10.1016/j.ibneur.2021.02.007

J Radiat Res. 2021 Apr 12:rrab016. doi: 10.1093/jrr/rrab016. Online ahead of print.

ABSTRACT

Non-homologous end joining is one of the main pathways for DNA double-strand break (DSB) repair and is also implicated in V(D)J recombination in immune system. Therefore, mutations in non-homologous end-joining (NHEJ) proteins were found to be associated with immunodeficiency in human as well as in model animals. Several human patients with mutations in XRCC4 were reported to exhibit microcephaly and growth defects, but unexpectedly showed normal immune function. Here, to evaluate the functionality of these disease-associated mutations of XRCC4 in terms of radiosensitivity, we generated stable transfectants expressing these mutants in XRCC4-deficient murine M10 cells and measured their radiosensitivity by colony formation assay. V83_S105del, R225X and D254Mfs*68 were expressed at a similar level to wild-type XRCC4, while W43R, R161Q and R275X were expressed at even higher level than wild-type XRCC4. The expression levels of DNA ligase IV in the transfectants with these mutants were comparable to that in the wild-type XRCC4 transfectant. The V83S_S105del transfectant and, to a lesser extent, D254Mfs*68 transfectant, showed substantially increased radiosensitivity compared to the wild-type XRCC4 transfectant. The W43R, R161Q, R225X and R275X transfectants showed a slight but statistically significant increase in radiosensitivity compared to the wild-type XRCC4 transfectant. When expressed as fusion proteins with Green fluorescent protein (GFP), R225X, R275X and D254Mfs*68 localized to the cytoplasm, whereas other mutants localized to the nucleus. These results collectively indicated that the defects of XRCC4 in patients might be mainly due to insufficiency in protein quantity and impaired functionality, underscoring the importance of XRCC4’s DSB repair function in normal development.

PMID:33842963 | DOI:10.1093/jrr/rrab016

Gates Open Res. 2021 Feb 24;5:24. doi: 10.12688/gatesopenres.13211.1. eCollection 2021.

ABSTRACT

The global Family Planning Estimation model (FPEM) combines a Bayesian hierarchical model with country-specific time trends to yield estimates of contraceptive prevalence and unmet need for family planning for countries worldwide. In this paper, we introduce the R package fpemlocal that carries out the estimation of family planning indicators for a single population, for example, for a single country or smaller area. In this implementation of FPEM, all non-population-specific parameters are fixed at outcomes obtained in a prior global FPEM run. The development of this model was motivated by the demand for computational efficiency, without loss of model accuracy, when estimates and projections from FPEM were needed only for a single country. We present use cases to produce estimates for a single population of women by union status or all women based on package-provided data bases and user-specified data. We also explain how to aggregate estimates across multiple populations. The R package forms the basis of the Track20 Family Planning Estimation Tool to monitor trends in family planning indicators for the FP2020 initiative. Fpemlocal is available from: https://github.com/AlkemaLab/fpemlocal.

PMID:33842844 | PMC:PMC8008158 | DOI:10.12688/gatesopenres.13211.1

Radiol Artif Intell. 2020 Nov 11;3(1):e200047. doi: 10.1148/ryai.2020200047. eCollection 2021 Jan.

ABSTRACT

PURPOSE: To generate and assess an algorithm combining eye tracking and speech recognition to extract brain lesion location labels automatically for deep learning (DL).

MATERIALS AND METHODS: In this retrospective study, 700 two-dimensional brain tumor MRI scans from the Brain Tumor Segmentation database were clinically interpreted. For each image, a single radiologist dictated a standard phrase describing the lesion into a microphone, simulating clinical interpretation. Eye-tracking data were recorded simultaneously. Using speech recognition, gaze points corresponding to each lesion were obtained. Lesion locations were used to train a keypoint detection convolutional neural network to find new lesions. A network was trained to localize lesions for an independent test set of 85 images. The statistical measure to evaluate our method was percent accuracy.

RESULTS: Eye tracking with speech recognition was 92% accurate in labeling lesion locations from the training dataset, thereby demonstrating that fully simulated interpretation can yield reliable tumor location labels. These labels became those that were used to train the DL network. The detection network trained on these labels predicted lesion location of a separate testing set with 85% accuracy.

CONCLUSION: The DL network was able to locate brain tumors on the basis of training data that were labeled automatically from simulated clinical image interpretation.© RSNA, 2020.

PMID:33842890 | PMC:PMC7845782 | DOI:10.1148/ryai.2020200047