Statistical Equivalence Tests Research Articles

Outcomes of myeloma cast nephropathy in the era of anti-CD38 monoclonal antibody-based frontline therapy: A retrospective cohort study.

7555 Background: Myeloma cast nephropathy (MCN) has been a well-known negative predictive marker in newly diagnosed multiple myeloma (NDMM). Anti-CD38+ monoclonal antibodies (mAb) have improved outcomes, but little is known about their impact in patients with MCN. We performed a retrospective cohort study to investigate the outcomes of patients with MCN in the era of anti-CD38+ mAb-based frontline therapy. Methods: 115 NDDM patients received frontline anti-CD38+ mAb from 11/15/18 to 1/24/23. MCN was defined as evidence of light chain casts on biopsy, serum creatinine of >2mg/dL, or 2021 CKD-EPI eGFR of <40mL/min/1.73m2; ≥1g/d proteinuria; and involved FLC ≥50mg/dL. 23 had MCN; 92 were contemporary controls. We obtained data regarding clinical course. Results: 6 MCN patients needed hemodialysis (HD) at diagnosis. MCN patients had similar R2-ISS to controls. Median proteinuria in MCN patients was 5.7g/d more than in controls; serum creatinine (Cr) was 3.2mg/dL higher; and median hemoglobin was 1.5g/dL less. More MCN patients received quadruplet regimens (86.9% vs 62.0%, p = 0.010); similar proportions were transplanted. At 1, 3, and 6 months, serum Cr was higher in MCN patients (p < 0.001, p = 0.002, p = 0.002). Urine protein:creatinine ratios trended toward difference at 1 month (2.04 vs 0.15g/g MCN vs controls, p = 0.061) but were similar at 3 and 6 months. Response rates at 6 months did not differ. 4 MCN patients still required HD after 6 months. Median follow-up was 23.0-23.5 months. 4 MCN patients died, 1 due to progressive disease. No significant difference in disease-free survival (DFS; HR = 0.58, 95% CI 0.20-1.64, p = 0.295), time to next treatment (TTNT; HR = 0.53, 95% CI 0.20-1.39, p = 0.191), or overall survival (OS; HR 0.610, 95% CI 0.19-1.98, p = 0.406) was seen at any point between MCN and control cohorts. 1 year DFS, TTNT, and OS were: 87.0% MCN, 94.5% control; 77.2% MCN, 91.7% control; 87.9% MCN, 96.1% control. By statistical equivalence testing, MCN DFS and TTNT were within 6 months of control DFS and TTNT (p = 0.038, p = 0.032); MCN OS was within 5 months of control OS (p = 0.034). Conclusions: Patients with NDMM and MCN who receive upfront anti-CD38+ mAb therapy experience prolonged survival compared to prior findings. Anti-CD38+ mAb may substantially reduce, though not fully eliminate, the negative survival impact of MCN. Multiple mechanisms, including time to proteinuria resolution, may be involved. Further study is warranted. [Table: see text]

Perspective on statistical power and equivalence tests.

The use of both sexes or genders should be considered in experimental design, analysis, and reporting. Since there is no requirement to double the sample size or to have sufficient power to study sex differences, challenges for the statistical analysis can arise. In this article, we focus on the topics of statistical power and ways to increase this power. We also discuss the choice of an appropriate design and statistical method and include a separate section on equivalence tests needed to show the absence of a relevant difference.

Hormonal intrauterine devices and heat exchange during exercise.

Synthetic progestins in oral contraceptives are thought to blunt heat dissipation by reducing skin blood flow and sweating. However, whether progestin-releasing intrauterine devices (IUDs) modulate heat loss during exercise-heat stress is unknown. We used direct calorimetry to measure whole-body total (dry + evaporative) heat loss in young, physically active women (mean (SD); aged 24(4)years, 39.3(5.3) ml/kg/min) with (IUD; n=19) and without (Control; n=17) IUDs in the follicular and luteal phases of the menstrual cycle during light- and moderate-intensity exercise at fixed rates of heat production (∼175 and ∼275W/m2 ) in 30°C, ∼21% relative humidity. Between-group and -phase differences were evaluated using traditional hypothesis testing and statistical equivalence testing within pre-determined bounds (±11W/m2 ; difference required to elicit a ±0.3°C difference in core temperature over 1h) in each exercise bout. Whole-body total heat loss was statistically equivalent between groups within ±11Wm-2 (IUD-Control [90% CIs]; Light: -2 [-8, 5] W/m2 , P=0.007; Moderate: 0 [-6, 6] W/m2 , P=0.002), as were dry and evaporative heat loss (P ≤ 0.023), except for evaporative heat loss during moderate-intensity exercise (equivalence: P=0.063, difference: P=0.647). Whole-body total and evaporative heat loss were not different between phases (P≥0.267), but dry heat loss was 3 [95% CIs: 1, 5] W/m2 greater in the luteal phase (P≤0.022). Despite this, all whole-body heat loss outcomes were equivalent between phases (P≤0.003). These findings expand our understanding of the factors that modulate heat exchange in women and provide valuable mechanistic insight of the role of endogenous and exogenous female sex hormones in thermoregulation. KEY POINTS: Progestin released by hormonal intrauterine devices (IUDs) may negatively impact heat dissipation during exercise by blunting skin blood flow and sweating. However, the influence of IUDs on thermoregulation has not previously been assessed. We used direct calorimetry to show that IUD users and non-users display statistically equivalent whole-body dry and evaporative heat loss, body heat storage and oesophageal temperature during moderate- and high-intensity exercise in a warm, dry environment, indicating that IUDs do not appear to compromise exercise thermoregulation. However, within IUD users and non-users, dry heat loss was increased and body heat storage and oesophageal temperature were reduced in the luteal compared to the follicular phase of the menstrual cycle, though these effects were small and unlikely to be practically meaningful. Together, these findings expand our understanding of the factors that modulate heat exchange in women and have important practical implications for the design of future studies of exercise thermoregulation.

Comparison of the agar contact method and the wet-dry double swabbing method for determining the total viable bacterial count on pig carcass surfaces

A practical and representative sampling method for microbiological examination of the slaughter process is useful for identifying abattoir-specific risk factors within the pig slaughter line. The aim of this study was to examine the suitability of an agar contact method (ACM), where the agar was homogenized before the microbiological processing, in comparison with the wet-dry double swabbing method (WDSM) for quantitative determination of total viable counts (TVC) on pig skin surfaces. In our experimental trial, pig skin pieces were artificially contaminated at 2 levels (3 log and 7 log cfu/ml) with a suspension of bacteria species commonly found on pig skin and cultivated in vitro. Within our field trial, pig carcasses were investigated at pre-chilling in an abattoir under standard processing conditions. For both sampling methods, TVC was determined, and statistical equivalence tests were calculated. Linear regression models showed the similarity of the sampling methods, with coefficient of determination (R2) > 90% and slope parameters of nearly 1 for both trials separately. Statistically significant equivalence between the 2 sampling methods was proven in both trials (with p < 0.0001 within an equivalence range of ± 0.5 log cfu/ml, respectively). The field trial revealed TVC on carcass surfaces sometimes at or below the lower detection limit for the ACM, while TVC from all carcasses were able to be determined by WDSM. Overall, low contamination levels were less reliably detectable by ACM than by WDSM. The ACM can be seen as an additional and suitable sampling procedure for pig skin and can contribute to the identification of abattoir specific risk factors for investigations of the hygienic status at process stages along the pig slaughter line.

Multivariate equivalence testing for food safety assessment.

Products for food and feed derived from genetically modified (GM) crops are only allowed on the market when they are deemed to be safe for human health and the environment. The European Food Safety Authority (EFSA) performs safety assessment including a comparative approach: the compositional characteristics of a GM genotype are compared to those of reference genotypes that have a history of safe use. Statistical equivalence tests are used to carry out such a comparative assessment. These tests are univariate and therefore only consider one measured variable at a time. Phenotypic data, however, often comprise measurements on multiple variables that must be integrated to arrive at a single decision on acceptance in the regulatory process. The surge of modern molecular phenotyping platforms further challenges this integration, due to the large number of characteristics measured on the plants. This paper presents a new multivariate equivalence test that naturally extends a recently proposed univariate equivalence test and allows to assess equivalence across all variables simultaneously. The proposed test is illustrated on plant compositional data from a field study on maize grain and on untargeted metabolomic data of potato tubers, while its performance is assessed on simulated data.

Positron driven high-field terahertz waves via dielectric wakefield interaction

Advanced acceleration methods based on wakefields generated by high energy electron bunches passing through dielectric-based structures have demonstrated $>$GV/m fields, paving the first steps on a path to applications such as future compact linear colliders. For a collider scenario, it is desirable that, in contrast to plasmas, wakefields in dielectrics do not behave differently for positron and electron bunches. In this Letter, we present measurements of large amplitude fields excited by positron bunches with collider-relevant parameters (energy 20 GeV, and $0.7 \times 10^{10}$ particles per bunch) in a 0.4 THz, cylindrically symmetric dielectric structure. Interferometric measurements of emitted coherent Cerenkov radiation permit spectral characterization of the positron-generated wakefields, which are compared to those excited by electron bunches. Statistical equivalence tests are incorporated to show the charge-sign invariance of the induced wakefield spectra. Transverse effects on positron beams resulting from off-axis excitation are examined and found to be consistent with the known linear response of the DWA system. The results are supported by numerical simulations and demonstrate high-gradient wakefield excitation in dielectrics for positron beams.

Statistical Equivalence Testing of Higher-Order Protein Structures with Differential Hydrogen Exchange-Mass Spectrometry (HX-MS).

Hydrogen exchange-mass spectrometry (HX-MS) is widely recognized for its potential utility for establishing the equivalence of the higher-order structures of proteins, particularly in comparability and similarity contexts. However, recent progress in the statistical analysis of HX-MS data has instead placed an emphasis on significance testing to identify regions of proteins where there are significant differences in HX between two or more protein states. In the cases involving assessment of similarity or equivalence of the higher-order structure of different protein samples (e.g., biosimilars), significance testing of HX-MS data is unsuitable. To meet this need, we have adapted the univariate two one-sided test (TOST) equivalence testing method for HX-MS data. Equivalence acceptance criteria were determined using maximum deviations from randomized resampling of truly equivalent samples to define hybrid equivalence criteria (maximum deviation of true equivalents, MDTE). Application of the TOST-MDTE test on differential HX-MS measurements of wild-type and mutated maltose-binding proteins demonstrates that the equivalence testing method was fit-for-purpose. Three infliximab biosimilars (Remsima, Renflexis, and Inflectra) were found to be equivalent to their Remicade reference product based on differential HX-MS measurements, while 5% deglycosylated NIST mAb was not statistically equivalent to the unmodified NIST mAb reference.

The HoloLens Augmented Reality System Provides Valid Measures of Gait Performance in Healthy Adults

Biomechanical measures are the gold standard in the assessment of gait in healthy and chronic disease populations. Augmented reality (AR) systems represent an opportunity to evaluate human movement under more realistic and interactive conditions. A barrier to integrating AR into healthcare is the unknown accuracy of systems in the quantification of human movement. This project aimed to determine the accuracy of the HoloLens relative to three-dimensional motion capture (MoCap) in quantifying gait. Ten healthy adults completed nine walking trials (n = 3 for slow, medium, and fast speed, respectively). Outcome measures included: cumulative walking distance, number of steps, step length, and speed. Statistical equivalence testing, using an a priori threshold of five percent, confirmed biomechanical measures derived from the HoloLens was equivalent to MoCap. Cumulative walking distance from the HoloLens was within 1.5-2.1% of the MoCap system for all walking speeds. Difference between systems in terms of movement accuracy was less than 3.7 cm across trials. Equivalence in outcomes makes the HoloLens appropriate for the quantification of frequently used gait variables to characterize walking performance. Future AR applications have the potential to deliver digital therapeutics to patient populations under more real-world conditions and monitor performance using objective and quantitative outcomes.

Mechanistic model of pH and excipient concentration during ultrafiltration and diafiltration processes of therapeutic antibodies.

Ultrafiltration and diafiltration (UF/DF) unit operations are widely used for the manufacture of therapeutic antibodies to control drug substance protein concentration, pH, and excipient properties. During UF/DF, molecular interactions and volume exclusion effects often lead to substantial differences in pH and excipient concentrations between the diafiltration buffer and final UF/DF pool. These differences complicate the design process beyond simply specifying a buffer with the desired drug substance pH and excipient conditions. This article describes a UF/DF process model which dynamically and accurately simulates UF/DF retentate pool pH and excipient conditions throughout the UF/DF process. This multiscale model accounts for microscopic descriptions of ion-protein charge interactions using the Poisson-Boltzmann equation as well as macroscopic descriptions of volume exclusion and mass transfer. Model predictions of the final UF/DF pool properties were experimentally verified through comparisons to design of experiment (DoE) data from four monoclonal antibody (mAb) processes, each with differing formulations and UF/DF operating conditions. Additionally, model simulations of the retentate pool properties throughout the UF/DF process were verified for two mAb processes through comparisons to experimental data collected at intermediate process points. Model results were qualified, using statistical equivalence tests, against the outputs from large-scale GMP runs which confirmed that the model accurately captures large-scale process performance. Finally, the model was applied toward the simulation of process scenarios beyond those examined experimentally. These in-silico experiments demonstrate the model's capability as a tool for augmented process design and it is potential to reduce the extent of UF/DF laboratory experiments.

Brain Tumor Surgery is Safe in Octogenarians and Nonagenarians: A Single-Surgeon 741 Patient Series

Elderly patients with surgically accessible brain tumors are often not offered clinically indicated brain tumor surgery (BTS) because of to assumptions of greater risk for perioperative morbidity and mortality. Because brain tumor incidence is highest in the geriatric population, and because the global population is aging, accurate understanding of BTS risk in elderly patients is critical. We aimed to compare safety of BTS in elderly patients with younger counterparts to better understand the risk-benefit profile of BTS for elderly patients. Retrospective cohort study of young (20-29 years), senior (60-79 years), and elderly (80+ years) patients who underwent BTS with a single neurosurgeon. Differences between pre- and postoperative modified Rankin score (ΔmRS), length of hospitalization (LOH), complication rate, and 30-day readmission rates (30DRR) were recorded. A total of 741 patients (83 elderly, 570 senior, and 88 young) were identified. No significant difference in preoperative mRS between different age groups, χ2=0.269, P= 0.874. Elderly complication rate was 6.0%, not significantly different from young (4.5%, P= 0.667) or senior (7.2%, P= 0.696) complication rate. Elderly LOH was 1.93 ± SD 0.176 days; not significantly different from young (3.01 ± 0.384 days, P= 0.081) or senior (2.47 ± 0.144 days, P= 0.881). Statistical equivalence testing showed with 95% confidence that there was equivalence in ΔmRS among age groups. Elderly patients did not have significantly different ΔmRS, LOH, 30DRR, or complication rates after BTS compared with younger counterparts. Therefore, in healthy patients, advanced age alone should not prevent patients from being offered BTS.

Evaluation of novel organosilane modifications of paper spray mass spectrometry substrates for analyzing polar compounds

Paper spray mass spectrometry (PSMS) is currently used in different analytical fields, but less effort has been made so far to use PSMS for highly polar compounds. Such analytes usually show poor performance in PSMS due to their high affinity for common paper substrates in addition to high matrix effects. In this study, strategies for hydrophobic modifications of commercially available paper substrates using ten different organosilanes were developed. The modified substrates were generated, characterized, and applied for PSMS analysis of polar toxins. By using the modified paper, PSMS performance of some of the toxins could be considerably increased, especially for orellanine, showing a more than 80-fold signal enhancement when substrates modified with chlorotrimethylsilane were used. For other toxins like ricinine, only small beneficial effects could be shown on PSMS performance when using modified substrates. Statistical equivalence tests showed sufficient ruggedness of the developed procedures also compared to conventional substrates. Thus, further systematic development of paper substrates modified with organosilane derivatives based on the presented study for application in PSMS should be encouraged.

Establishing Statistical Equivalence of Data from Different Sampling Approaches for Assessment of Bacterial Phenotypic Antimicrobial Resistance.

ABSTRACTTo assess phenotypic bacterial antimicrobial resistance (AMR) in different strata (e.g., host populations, environmental areas, manure, or sewage effluents) for epidemiological purposes, isolates of target bacteria can be obtained from a stratum using various sample types. Also, different sample processing methods can be applied. The MIC of each target antimicrobial drug for each isolate is measured. Statistical equivalence testing of the MIC data for the isolates allows evaluation of whether different sample types or sample processing methods yield equivalent estimates of the bacterial antimicrobial susceptibility in the stratum. We demonstrate this approach on the antimicrobial susceptibility estimates for (i) nontyphoidal Salmonella spp. from ground or trimmed meat versus cecal content samples of cattle in processing plants in 2013-2014 and (ii) nontyphoidal Salmonella spp. from urine, fecal, and blood human samples in 2015 (U.S. National Antimicrobial Resistance Monitoring System data). We found that the sample types for cattle yielded nonequivalent susceptibility estimates for several antimicrobial drug classes and thus may gauge distinct subpopulations of salmonellae. The quinolone and fluoroquinolone susceptibility estimates for nontyphoidal salmonellae from human blood are nonequivalent to those from urine or feces, conjecturally due to the fluoroquinolone (ciprofloxacin) use to treat infections caused by nontyphoidal salmonellae. We also demonstrate statistical equivalence testing for comparing sample processing methods for fecal samples (culturing one versus multiple aliquots per sample) to assess AMR in fecal Escherichia coli. These methods yield equivalent results, except for tetracyclines. Importantly, statistical equivalence testing provides the MIC difference at which the data from two sample types or sample processing methods differ statistically. Data users (e.g., microbiologists and epidemiologists) may then interpret practical relevance of the difference.IMPORTANCE Bacterial antimicrobial resistance (AMR) needs to be assessed in different populations or strata for the purposes of surveillance and determination of the efficacy of interventions to halt AMR dissemination. To assess phenotypic antimicrobial susceptibility, isolates of target bacteria can be obtained from a stratum using different sample types or employing different sample processing methods in the laboratory. The MIC of each target antimicrobial drug for each of the isolates is measured, yielding the MIC distribution across the isolates from each sample type or sample processing method. We describe statistical equivalence testing for the MIC data for evaluating whether two sample types or sample processing methods yield equivalent estimates of the bacterial phenotypic antimicrobial susceptibility in the stratum. This includes estimating the MIC difference at which the data from the two approaches differ statistically. Data users (e.g., microbiologists, epidemiologists, and public health professionals) can then interpret whether that present difference is practically relevant.

A Primer on the Use of Equivalence Testing for Evaluating Measurement Agreement.

Statistical equivalence testing is more appropriate than conventional tests of difference to assess the validity of physical activity (PA) measures. This article presents the underlying principles of equivalence testing and gives three examples from PA and fitness assessment research. The three examples illustrate different uses of equivalence tests. Example 1 uses PA data to evaluate an activity monitor's equivalence to a known criterion. Example 2 illustrates the equivalence of two field-based measures of physical fitness with no known reference method. Example 3 uses regression to evaluate an activity monitor's equivalence across a suite of 23 activities. The examples illustrate the appropriate reporting and interpretation of results from equivalence tests. In the first example, the mean criterion measure is significantly within ±15% of the mean PA monitor. The mean difference is 0.18 METs and the 90% confidence interval of -0.15 to 0.52 is inside the equivalence region of -0.65 to 0.65. In the second example, we chose to define equivalence for these two measures as a ratio of mean values between 0.98 and 1.02. The estimated ratio of mean V˙O2 values is 0.99, which is significantly (P = 0.007) inside the equivalence region. In the third example, the PA monitor is not equivalent to the criterion across the suite of activities. The estimated regression intercept and slope are -1.23 and 1.06. Neither confidence interval is within the suggested regression equivalence regions. When the study goal is to show similarity between methods, equivalence testing is more appropriate than traditional statistical tests of differences (e.g., ANOVA and t-tests).

Neurogenomic Profiling Reveals Distinct Gene Expression Profiles Between Brain Parts That Are Consistent in Ophthalmotilapia Cichlids.

The detection of external and internal cues alters gene expression in the brain which in turn may affect neural networks that underly behavioral responses. Previous studies have shown that gene expression profiles differ between major brain regions within individuals and between species with different morphologies, cognitive abilities and/or behaviors. A detailed description of gene expression in all macroanatomical brain regions and in species with similar morphologies and behaviors is however lacking. Here, we dissected the brain of two cichlid species into six macroanatomical regions. Ophthalmotilapia nasuta and O. ventralis have similar morphology and behavior and occasionally hybridize in the wild. We use 3′ mRNA sequencing and a stage-wise statistical testing procedure to identify differential gene expression between females that were kept in a social setting with other females. Our results show that gene expression differs substantially between all six brain parts within species: out of 11,577 assessed genes, 8,748 are differentially expressed (DE) in at least one brain part compared to the average expression of the other brain parts. At most 16% of these DE genes have |log2FC| significantly higher than two. Functional differences between brain parts were consistent between species. The majority (61–79%) of genes that are DE in a particular brain part were shared between both species. Only 32 genes show significant differences in fold change across brain parts between species. These genes are mainly linked to transport, transmembrane transport, transcription (and its regulation) and signal transduction. Moreover, statistical equivalence testing reveals that within each comparison, on average 89% of the genes show an equivalent fold change between both species. The pronounced differences in gene expression between brain parts and the conserved patterns between closely related species with similar morphologies and behavior suggest that unraveling the interactions between genes and behavior will benefit from neurogenomic profiling of distinct brain regions.

Statistical Equivalence Testing Approaches for Mantel–Haenszel DIF Analysis

The null hypothesis test used in differential item functioning (DIF) detection tests for a subgroup difference in item-level performance—if the null hypothesis of “no DIF” is rejected, the item is flagged for DIF. Conversely, an item is kept in the test form if there is insufficient evidence of DIF. We present frequentist and empirical Bayes approaches for implementing statistical equivalence testing for DIF using the Mantel–Haenszel (MH) DIF statistic. With these approaches, rejection of the null hypothesis of “DIF” allows the conclusion of statistical equivalence, a more stringent criterion for keeping items. In other words, the roles of the null and alternative hypotheses are interchanged in order to have positive evidence that the DIF of an item is small. A simulation study compares the equivalence testing approaches to the traditional MH DIF detection method with the Educational Testing Service classification system. We illustrate the methods with item response data from the 2012 Programme for International Student Assessment.

Arthroscopic Labral Base Repair in the Hip: 5-Year Minimum Clinical Outcomes

Background: Arthroscopic labral base repair (LBR) in the hip is a previously described technique designed to restore the native functional anatomy of the labrum by reproducing its seal against the femoral head. LBR has been shown to have good short-term outcomes. Hypothesis/Purpose: The purpose was to evaluate clinical outcomes of an LBR cohort with a minimum 5-year follow-up. It was hypothesized that patients who underwent LBR would continue to have significant improvement from their preoperative scores and maintain scores similar to their 2-year outcomes. Study Design: Case series; Level of evidence, 4. Methods: Data for patients undergoing primary hip arthroscopic surgery with LBR from February 2008 to May 2011 with a minimum 5-year follow-up were prospectively collected and retrospectively reviewed. Patients with preoperative Tonnis osteoarthritis grade ≥2, previous hip conditions (slipped capital femoral epiphysis, avascular necrosis, Legg-Calv-Perthes disease), severe dysplasia (lateral center-edge angle <18°), or previous ipsilateral hip surgery were excluded. Statistical equivalence tests evaluated patient-reported outcomes (PROs) including the modified Harris Hip Score (mHHS), Non-Arthritic Hip Score (NAHS), Hip Outcome Score–Sport-Specific Subscale (HOS-SSS), visual analog scale (VAS) for pain, and patient satisfaction (0-10 scale; 10 = very satisfied). Results: Of the 70 patients (74 hips) who met inclusion and exclusion criteria, 60 (85.7%) patients (64 hips) were available at a minimum 5-year follow-up. All PRO scores significantly improved from preoperative values with a mean follow-up of 67.8 ± 7.4 months (range, 60.0-89.7 months). The mean mHHS increased from 64.4 ±13.8 to 85.3 ± 17.7 (P < .001), the mean NAHS from 63.7 ± 17.0 to 87.0 ± 14.7 (P < .001), and the mean HOS-SSS from 47.1 ± 23.2 to 76.5 ± 25.9 (P < .001). The mean VAS score decreased from 5.9 ± 2.4 to 2.0 ± 2.1 (P < .001). The mean patient satisfaction score was 8.1 ± 2.0. The improvement in PRO scores was sustained from 2- to 5-year follow-up. At 2 and 5 years, survivorship rates were 96.9% and 90.6%, respectively, and the respective secondary arthroscopic surgery rates were 10.9% (7/64) and 17.2% (11/64). Conclusion: At a minimum 5-year follow-up, arthroscopic LBR continued to be a successful procedure and valid technique based on 3 PROs, the VAS, patient satisfaction, and survivorship. Significantly improved outcome scores were maintained compared with preoperative values and showed no signs of deterioration from the 2-year scores. The long-term survivorship of hip arthroscopic surgery has yet to be determined; however, these midterm results demonstrate the rates of additional procedures (both secondary arthroscopic surgery and conversion to total hip arthroplasty), that may be necessary after 2 years.

Improved Statistical Methods for Evaluation of Stability of In Vitro Diagnostic Reagents

ABSTRACTIn vitro diagnostic (IVD) reagent stability is typically evaluated using regression analysis of measurand drift across time following CLSI guideline EP25-A. The corresponding stability duration establishment has several limitations. The stability duration conclusion is based on a two-stage acceptance criteria using the p-value of the regression slope followed by the 95% confidence interval (CI) on the fitted regression line if the p-value <0.05. This analysis technique is based on traditional statistical hypothesis testing; however, the statistical equivalence testing framework better represents the goals of IVD reagent stability evaluation. The resulting stability duration CI does not achieve 95% coverage probability and the statistical properties of the estimated stability duration are substantially impacted by presence of common variance components not accounted for during power analysis and by not accounting for variability in the baseline estimate. The current proposal based on the equivalence testing framework uses a one-stage acceptance criteria based on the 95% CI for proportional measurand drift derived from the regression fit. The proposed methodology was applied to automated immunoassay data (Akbas, Budd, and Klee 2016). Monte Carlo simulation studies are presented to illustrate the improved statistical properties of the current proposal along with an example power analysis for study design. Supplementary materials for this article are available online.

Integral Dose-Based Inverse Optimization May Reduce Side Effects in Radiotherapy of Prostate Carcinoma.

The purpose of this work is to apply a novel inverse optimization approach, based on utilization of quantitative imaging information in the optimization function, to prostate carcinoma. This new inverse optimization algorithm relies upon quantitative information derived from computed tomography (CT) imaging studies. The Hounsfield numbers of the CT voxels are converted to physical density, which in turn is used to calculate voxel mass and the corresponding integral dose, by summation over the product of dose and mass in each dose voxel. This integral dose is used for plan optimization through its global minimization. The optimization results are compared to the optimization results derived from most commonly used dose-volume-based inverse optimization, where objective functions are formed as summation over all dose voxels of the squared differences between voxel doses and user specified doses. The data from 25 prostate plans were optimized with dose-volume histogram (DVH) and integral dose (energy) minimization objective functions. The results obtained with the energy- and DVH-based optimization schemes were studied through commonly used dosimetric indices (DIs). Statistical equivalence tests were further performed to establish population-based significance results. Both DVH- and energy-based plans for each case were normalized so that 95% of the planning target volume receives the prescription dose. The average differences for the rectum and bladder DIs ranged from 1.6 to 25%, where the energy-based quantities were lower. For both femoral heads, the energy-based optimization-derived doses were lower on average by 32%. The statistical tests demonstrated that the significant differences in the tallied dose indices range from 2.7% to more than 50% for rectum, bladder, and femoral heads. For majority of the clinically relevant dosimetric quantities, energy-based inverse optimization performs better than the standard of care DVH-based optimization in prostate carcinoma. The population averaged statistically significant differences range from ~3 to ~50%. Therefore, this newly proposed optimization approach, incorporating explicitly quantitative imaging information in the inverse optimization function, holds potential for further reduction of complication rates in prostate cancer.

Statistical Approaches to Assess Biosimilarity from Analytical Data.

Protein therapeutics have unique critical quality attributes (CQAs) that define their purity, potency, and safety. The analytical methods used to assess CQAs must be able to distinguish clinically meaningful differences in comparator products, and the most important CQAs should be evaluated with the most statistical rigor. High-risk CQA measurements assess the most important attributes that directly impact the clinical mechanism of action or have known implications for safety, while the moderate- to low-risk characteristics may have a lower direct impact and thereby may have a broader range to establish similarity. Statistical equivalence testing is applied for high-risk CQA measurements to establish the degree of similarity (e.g., highly similar fingerprint, highly similar, or similar) of selected attributes. Notably, some high-risk CQAs (e.g., primary sequence or disulfide bonding) are qualitative (e.g., the same as the originator or not the same) and therefore not amenable to equivalence testing. For biosimilars, an important step is the acquisition of a sufficient number of unique originator drug product lots to measure the variability in the originator drug manufacturing process and provide sufficient statistical power for the analytical data comparisons. Together, these analytical evaluations, along with PK/PD and safety data (immunogenicity), provide the data necessary to determine if the totality of the evidence warrants a designation of biosimilarity and subsequent licensure for marketing in the USA. In this paper, a case study approach is used to provide examples of analytical similarity exercises and the appropriateness of statistical approaches for the example data.

Pooling Morphometric Estimates: A Statistical Equivalence Approach.

Changes in hardware or image-processing settings are a common issue for large multicenter studies. To pool MRI data acquired under these changed conditions, it is necessary to demonstrate that the changes do not affect MRI-based measurements. In these circumstances, classical inference testing is inappropriate because it is designed to detect differences, not prove similarity. We used a method known as statistical equivalence testing to address this limitation. Equivalence testing was carried out on 3 datasets: (1) cortical thickness and automated hippocampal volume estimates obtained from healthy individuals imaged using different multichannel head coils; (2) manual hippocampal volumetry obtained using two readers; and (3) corpus callosum area estimates obtained using an automated method with manual cleanup carried out by two readers. Equivalence testing was carried out using the "two one-sided tests" (TOST) approach. Power analyses of the TOST were used to estimate sample sizes required for well-powered equivalence testing analyses. Mean and standard deviation estimates from the automated hippocampal volume dataset were used to carry out an example power analysis. Cortical thickness values were found to be equivalent over 61% of the cortex when different head coils were used (q < .05, false discovery rate correction). Automated hippocampal volume estimates obtained using the same two coils were statistically equivalent (TOST P = 4.28 × 10(-15) ). Manual hippocampal volume estimates obtained using two readers were not statistically equivalent (TOST P = .97). The use of different readers to carry out limited correction of automated corpus callosum segmentations yielded equivalent area estimates (TOST P = 1.28 × 10(-14) ). Power analysis of simulated and automated hippocampal volume data demonstrated that the equivalence margin affects the number of subjects required for well-powered equivalence tests. We have presented a statistical method for determining if morphometric measures obtained under variable conditions can be pooled. The equivalence testing technique is applicable for analyses in which experimental conditions vary over the course of the study.