Split-sample Method Research Articles

T2062 Proton Pump Inhibitors and Transplantation Are Important Risk Factors for the Development of Clostridium difficile Associated Disease in Hospitalized Patients

Background and Aims: Acute gastroenteritis (GE) is an important risk factor for the development of irritable bowel syndrome (IBS). We utilized observational data from the Walkerton Health Study (WHS) to develop and validate a risk score for post-infectious (PI) IBS. Methods: Model derivation and validation were based on a split sample method from a cohort of patients with exposure to acute gastroenteritis (GE) (n=1368). Study participants were randomly assigned to the derivation and validation cohorts in a 1:1 ratio. Using the derivation cohort, univariate and multivariable logistic regression were used to identify risk factors associated with IBS. The risk model was validated in the validation cohort. Overall model performance was assessed using area under the receiver operating curve (ROC). The risk score was developed using multivariable regression coefficients obtained from the derivation set and validated in the validation set. Classification and Regression Tree (CART) modeling was used to determine the cut off values for the level of risk (high, intermediate and low) based on the total score. Results: Nine variables were identified as important predictors of IBS (gender, age<60, longer duration of diarrhea, increased stool frequency, abdominal cramping, bloody stools, weight loss, fever and psychological disorders (anxiety and depression)). The discriminatory power of the risk model based on the area under ROC was 0.70 and was similar in the validation set. The risk score model showed good accuracy in both the derivation and validation sets and was able to distinguish between cohorts at low, intermediate and high risk for developing PI-IBS. Percentages of patients with PI-IBS in the low, intermediate and high risk were 10%, 35% and 60% in the derivation cohort and 17%, 36% and 62% in the validation cohort. Conclusion: A simple risk tool that uses demographics and symptoms of acute GE can predict which patients with acute GE are at risk of developing PI-IBS. This tool may be used clinically to assess risk and guide treatment.

An alternative deterministic method for the spatial interpolation of water retention parameters

Abstract. There are two approaches available for mapping water retention parameters over the study area using a spatial interpolation method. (1) Retention models can be first fitted to retention curves available at sampling locations prior to interpolating model parameters over the study area (the FI approach). (2) Retention data points can first be interpolated over the study area before retention model parameters are fitted (the IF approach). The current study compares the performance of these two approaches in representing the spatial distribution of water retention curves. Standard geostatistical interpolation methods, i.e., ordinary kriging and indicator kriging, were used. The data used in this study were obtained from the Las Cruces trench site database, which contains water retention data for 448 soil samples. Three standard water retention models, i.e., Brooks and Corey (BC), van Genuchten (VG), and Kosugi (KSG), were considered. For each model, standard validation procedures, i.e., leave-one-out cross-validation and split-sample methods were used to estimate the uncertainty of the parameters at each sampling location, allowing for the computation of prediction errors (mean absolute error and mean error). The results show that the IF approach significantly lowered mean absolute errors for the VG model, while also reducing them moderately for the KSG and BC models. In addition, the IF approach resulted in less bias than the FI approach, except when the BC model was used in the split-sample approach. Overall, IF outperforms FI for all three retention models in describing the spatial distribution of retention parameters.

Thyroglobulin and anti-thyroglobulin assays in thyroid cancer monitoring

Objectives To compare four methods for thyroglobulin (Tg) quantitation and three methods for detection and quantitation of thyroglobulin antibodies (Tg–Ab). We also wanted to explore the premise that thyroglobulin antibodies, as determined by commercially available assays, interfere with thyroglobulin assays in a predictable way. Methods Split sample method comparisons were run on all the methods for both the thyroglobulin and anti-thyroglobulin assays. In addition to this, samples from patients that had disseminated thyroid cancer but had low serum thyroglobulin concentrations and high thyroglobulin antibodies were further studied. These studies involved doing recovery studies (or antibody inhibition studies). Results There was good agreement between methods for quantitation of thyroglobulin with slopes ranging from 0.77 to 1.23 although closer agreement was expected as the assays are all calibrated to the same reference standard (CRM 457). The situation for the thyroglobulin antibody assays is significantly worse, and the rate of antibody positivity ranged from 9–21% in this group of patients although there was agreement in only 6%. Different reference standards are used for the Tg–Ab assays we investigated. The Tg–Ab data did not lend itself to traditional linear regression analysis as the data showed wide scatter. Conclusions There is good agreement between the four thyroglobulin assays compared in this study. The linear regression analysis shows that there is proportional error present between the methods that is greater than 50%. This study is unable to demonstrate any difference in assay values based on the amount of anti-thyrogobulin present in the specimen. The agreement between different anti-thyroglobulin assays is very poor. This finding is very problematic since it makes it difficult to generalize any literature reports of interference. All the thyroglobulin assays appear to be suitable for monitoring patients with thyroid cancer, provided that the differences in calibration are taken into account. Differences in calibration between different assays need to be taken into account when changing assays. Conversely, the anti-thyroglobulin assays are virtually useless since there appears to be very little agreement between the three assays studied and no evidence of assay interference in the measurements of thyroglobulin.

Testing the stability of the benefit transfer function for discrete choice contingent valuation data

This paper examines the stability of the benefit transfer function across 42 recreational forests in the British Isles. A working definition of reliable function transfer is put forward, and a suitable statistical test is provided. A novel split sample method is used to test the sensitivity of the models’ log-likelihood values to the removal of contingent valuation (CV) responses collected at individual forest sites. We find that a stable function improves our measure of transfer reliability, but not by much. We conclude that, in empirical studies on transferability, considerations of function stability are secondary to the availability and quality of site attribute data. Modellers’ can study the advantages of transfer function stability vis-à-vis the value of additional information on recreation site attributes.

Individual patient diagnosis of AD and FTD via high-dimensional pattern classification of MRI

The purpose of this study is to determine the diagnostic accuracy of MRI-based high-dimensional pattern classification in differentiating between patients with Alzheimer's disease (AD), Frontotemporal Dementia (FTD), and healthy controls, on an individual patient basis. MRI scans of 37 patients with AD and 37 age-matched cognitively normal elderly individuals, as well as 12 patients with FTD and 12 age-matched cognitively normal elderly individuals, were analyzed using voxel-based analysis and high-dimensional pattern classification. Diagnostic sensitivity and specificity of spatial patterns of regional brain atrophy found to be characteristic of AD and FTD were determined via cross-validation and via split-sample methods. Complex spatial patterns of relatively reduced brain volumes were identified, including temporal, orbitofrontal, parietal and cingulate regions, which were predominantly characteristic of either AD or FTD. These patterns provided 100% diagnostic accuracy, when used to separate AD or FTD from healthy controls. The ability to correctly distinguish AD from FTD averaged 84.3%. All estimates of diagnostic accuracy were determined via cross-validation. In conclusion, AD- and FTD-specific patterns of brain atrophy can be detected with high accuracy using high-dimensional pattern classification of MRI scans obtained in a typical clinical setting.

Neural networks for broad-band evaluation of complex permittivity using a coaxial discontinuity

The aim of this study is to determine the complex permittivity of dielectric materials using a coaxial discontinuity and the combination of neural networks (NN) with the finite element method. Two types of measurement cells are used. One is for solid samples and the other one for liquids. Data sets used to train neural networks are created using the finite element method. The number of hidden neurons of the NN is determined by the split-sample method. The designed NN are used for the estimation of the permittivity of several materials and their results compared with the ones obtained with a gradient inversion method.

Vascular Cell Adhesion Molecule 1 Predicts Cancer-Free Survival in Clear Cell Renal Carcinoma Patients

Vascular cell adhesion molecule 1 (VCAM1) is a cell surface glycoprotein implicated in various pathophysiologic conditions. We measured VCAM1 expression levels in tumor tissues and evaluated its significance and prognostic use in renal cell carcinoma (RCC). We used real-time quantitative PCR to examine the VCAM1 expression levels of a total of 485 sporadic renal tumors, including 429 clear cell, 21 papillary, 17 chromophobe, 11 oncocytomas, and 7 collecting duct carcinomas. We retrospectively examined the relationship of this expression to various clinicopathologic variables and the von Hippel-Lindau alteration status. We evaluated its significance with respect to patient survival rates using the Cox regression model combined with the split-sample method. Compared with normal kidney samples (n = 43), VCAM1 was significantly up-regulated in clear cell RCC and papillary RCC, whereas it was down-regulated in chromophobe RCC and oncocytoma. In clear cell RCC, VCAM1 expression levels were apparently high in patients asymptomatic at presentation and in patients with small tumor size, low-stage, low-grade, microvascular invasion-negative, and von Hippel-Lindau alteration-positive tumors. Univariate analyses showed that VCAM1 high expression is strongly associated with better outcomes in clear cell and papillary RCCs. Further, Cox multivariate analysis models combined with the split-sample method revealed that this association is significant only in cancer-free survival for patients with clear cell RCC after curative surgical resection. VCAM1 expression levels were found to be histologically subtype specific in renal tumors. Determination of the VCAM1 expression level as a biomarker can provide useful prognostic information for patients with clear cell RCC.

Split sample comparison of ThinPrep and conventional smears in endoscopic retrograde cholangiopancreatography‐guided pancreatic fine‐needle aspirations

Fine-needle aspiration (FNA) of pancreatic lesions is a common procedure to establish a tissue diagnosis before chemotherapy or surgery. In this study, the authors attempt to compare the diagnostic value of the ThinPrep (TP) method with conventional smears (CSs) in samples obtained by endoscopic retrograde cholangiopancreatography (ERCP)-guided pancreatic FNAs. Material obtained, prospectively, from ERCP-guided pancreatic FNAs was split to prepare CSs (2-5 slides) first, the remainder being rinsed in PreservCyte, and in the laboratory, 1 TP slide was prepared. The diagnostic categories of unsatisfactory, benign, reactive, suspicious for malignancy, and malignant were compared. Fifty-one pancreatic FNAs prepared by split sample method yielded the following results: TP yielded unsatisfactory, 6 cases; benign, 3 cases; reactive, 5 cases; suspicious for malignancy, 11 cases; and malignant, 26 cases; in contrast, CS yielded unsatisfactory, 13 cases; benign, 4 cases; reactive, 3 cases; suspicious for malignancy, 13 cases; and malignant, 18 cases. Histological follow-up was available in 21 cases (reactive, 8 cases; suspicious for malignancy, 1 case, and malignant, 12 cases). The foregoing data indicate a higher sensitivity in detection of pancreatic adenocarcinoma by the TP method (TP, 91% vs. 58% CS) with equivalent specificity (100%). In addition, TP provides better preservation and cytological detail.

Methods for categorizing a prognostic variable in a multivariable setting.

The literature is filled with examples of categorization of a continuous prognostic variable in a univariable setting followed by the addition of this categorical variable to an existing multivariable model. Typically, an "optimal" cutpoint for a new prognostic variable is obtained through a systematic search relating the variable to the outcome in an univariable manner. The corresponding categorical variable is then fitted in a multivariable model along with other already established prognostic covariates to assess the additional value of the new variable. This prompts the question whether the cutpoint search should have been performed in the same multivariable setting where it will ultimately be used. In this paper, we extend the univariable cutpoint search methods (split-sample approach and two-fold cross-validation approach) to the multivariable setting using -2 x log-likelihood statistic as the correlative measure. A Monte Carlo simulation study demonstrates that both methods are more efficient in detecting the true cutpoint and in estimating the effect size under the multivariable setting as opposed to the univariable setting. The cross-validation method performs better than the split-sample method in univariable as well as multivariable scenarios. For the cross-validation method in the multivariable setting, there is still a substantial loss of power when a cutpoint model is used in cases where there is a continuous relationship between the covariate and the outcome. An example is provided to illustrate the value of the multivariable cross-validation approach.

Performance of Monolayered Cervical Smears in a Gynecology Outpatient Setting in Kuwait

To compare ThinPrep (TP) Papanicolaou smears (Cytyc Corp., Box-borough, Massachusetts, U.S.A.) with matching conventional Papanicolaou (CP) smears for specimen adequacy, cytologic quality, diagnostic accuracy and screening time. In this prospective study of 1,024 women a split-sample, matched-pair design in favor of CP slides based on single-blind criteria was followed with a smear on a glass slide for CP and the remaining material collected in Preserv-Cyt solution (Cytyc) for a TP smear. A Cytobrush (Medscand, Hollywood, Florida, U.S.A.) was used for smear preparation for CP. TP smears were processed in ThinPrep 2000 (Cytyc). Smears were stained with Papanicolaou stain and were interpreted according to the Bethesda system. The number of satisfactory but limited (SBL) cases with TP were 77 (7.5%) as compared to 127 (12.4%) with the CP method. This reduction in SBL smears with the TP method and consequent increase in satisfactory smears were highly significant (P < .001) by McNemar's test. As regards unsatisfactory smears in discordant pairs, although the number of unsatisfactory smears was higher with TP (41 cases) as against CP (27 cases), the difference was not statistically significant (P < .05). The split-sample method showed a high correlation between the CP and TP diagnoses. TP smears had a significant advantage over CP smears in the reduction in the number of ASCUS and AGUS cases (14 vs. 29) (P < .05) and increased the pickup rate of LSIL, 6 vs. 1. Time taken to screen the TP smears was half that of CP smears. No cases of LSIL or HSIL were missed on TP smears. The liquid-based processor significantly improved the adequacy and quality of smears, resulting in fewer recall cases for SBL smears, leading to more definitive diagnoses in atypical cases, increasing the pickup rate of LSILs and reducing the screening time. A machine handling multiple specimens automatically would decrease cost and be an asset to a cytopathology laboratory.

Comparison of three polymerase chain reaction methods for routine detection of bovine herpesvirus 1 DNA in fresh bull semen

Five bulls were inoculated intrapreputially with Bovineherpes virus 1 (BHV 1), in order to compare the relative sensitivity of three polymerase chain reaction (PCR) assays for routine diagnosis of fresh bovine semen for the presence of BHV 1 Semen was collected twice a week up to 107 days post-infection (dpi). To reactivate latent virus, the bulls were treated with dexamethasone from 44 until 48 dpi. All samples were examined before and after cryopreservation treatment using a standard virus isolation (VI) method and three PCR assays: PCR A, PCR B and PCR C. PCR A and PCR C used an internal control plasmid DNA template and PCR B used the split sample method in order to control for false negative results. Of the 149 fresh semen samples that were tested, PCR A detected 45 positive, PCR B detected 39 positive and PCR C detected 66 positive, while virus was isolated from 22 samples. Of the 149 samples treated by cryopreservation, the virus was isolated from 13 samples and PCR C was positive in 21 samples. The results demonstrate that all three PCR assays are more sensitive than virus isolation, particularly during the later phases of infection.

Impact of prompt-response and quantity premiums on direct mail response rates

Conventional wisdom assumes that a carefully planned premium offer will increase the response rate to a direct marketing offer. For this study, a direct mail program was designed in order to measure the impact of two premiums—one for prompt response and another for larger quantity ordered. Twenty thousand pieces were mailed using a random split half sampling method, and responses to each offer were carefully monitored. The results were unexpected and provide important insights for the development of optimum direct marketing strategies. The quantity premium had an insignificant impact on order rate, while the prompt-response premium actually resulted in a reduced order rate.

Evaluation of the DuPont aca amylase method.

The dilution/activity curve for the DuPont aca kinetic amylase (EC 3.2.1.1) method is linear to approximately six times the upper limit of the expected range. Total reaction time on the aca is 4.3 min; however, there is a nonlinear rate of change in absorbance for at least the first 8 min, which is consistent with a lag phase. The within-run and day-to-day coefficients of variation were 2.4% at 100 U/L and 3.6% at 114 U/L, respectively. There was no interference by marked hemolysis (hemoglobin 4 g/L) or by glucose concentrations of 10--50 g/L. Comparisons by the split-sample method between the Amylochrome procedure and the aca procedure gave correlation coefficients of 0.992 for serum and 0.989 for urine. A second comparison, between the AutoAnalyzer I method and the aca, gave correlation coefficients of 0.998 for serum and 0.909 for urine. The expected range for serum was 18--106 U/L and for urine 2--33 U/2 h.

An Appraisal of the Split-Sample Method in Artificial Breeding Studies

Summary A comparison of residual variation derived from split samples and from alternate-week samples indicates a gain in efficiency of 43.5% through use of split samples in semen treatment experiments.