A benefit risk approach in cutoff determination for diagnostic tests.

A crucial step in the design of a diagnostic test is determining the cutoff point, the threshold which separates a negative measurement from a positive one. The results of a diagnostic test have clinical consequences: only when disease is accurately detected, proper treatments be administered, and vice versa. Benefit-Risk (BR) analysis should be used to determine the optimal cutoff point that optimizes the consequence. Quantitative BR analysis requires measurable benefit and risk and a function, e.g., linear or ratio, to combine all the components. When BR corresponding to the four possible diagnostic test outcomes are all scaled in units of risk resulting from an untreated disease, we propose a net BR (linear BR) equation as a function of diagnostic parameters, disease prevalence, benefit of correct diagnosis and risk of false diagnostic results. Optimal cutoff of a diagnostic test can be obtained using this function. Comparison of diagnostic tests based on their benefit and risk of tests is also discussed. Use of this function is illustrated with a biosensor rapid antigen test for SARS-CoV-2.

A multicenter study evaluation of the ClearLLab 10C panels.

Multiparameter flow cytometry plays an important role in the diagnosis, staging, and monitoring of patients with a suspected hematological malignancy. The ClearLLab 10C Panels consist of four reagent panels (B-Lineage Tube, T-Lineage Tube, and 2 Myeloid Lineage Tubes), each consisting of 10 color/10 antibody conjugates utilizing Beckman Coulters proprietary dry format optimized for investigating patients with suspected leukemia or lymphoma. A multicenter study was conducted to evaluate the performance of the ClearLLab 10C Panels for qualitative assessment of normal versus abnormal phenotype in peripheral blood, bone marrow, and lymph node samples with suspected hematological malignancies. ClearLLab 10C was compared to laboratory developed tests (LDTs) and final clinical diagnosis. Four clinical sites were used to enroll patient's spent specimens (n = 453); three laboratories in North America and one in Europe. Of the 453 specimens, 198 had no malignancy and 255 contained an abnormal population. The diagnostic accuracy of the ClearLLab 10C Panels was achieved with sensitivity of 96% and specificity of 95% with respect to patient final clinical diagnosis. The agreement of phenotyping between ClearLLab10C Panels and LDTs was 98%. Any differences noted between ClearLLab 10C and LDT were due to either the presence of populations below the level of detection, the lack of clinical information provided to the evaluators, or marker(s) not present in these panels. Overall, the ClearLLab 10C demonstrated excellent agreement to LDTs and diagnosis. These four reagent panels can be adopted by individual laboratories to assess the presence or absence of malignancy.

Precise Identification of Cell and Tissue Features Important for Histopathologic Diagnosis by a Whole Slide Imaging System.

BACKGROUND: Previous studies have demonstrated the noninferiority of pathologists' interpretation of whole slide images (WSIs) compared to microscopic slides in diagnostic surgical pathology; however, to our knowledge, no published studies have tested analytical precision of an entire WSI system. METHODS: In this study, five pathologists at three locations tested intra-system, inter-system/site, and intra- and inter-pathologist precision of the Aperio AT2 DX System (Leica Biosystems, Vista, CA, USA). Sixty-nine microscopic slides containing 23 different morphologic features suggested by the Digital Pathology Association as important to diagnostic pathology were identified and scanned. Each of 202 unique fields of view (FOVs) had 1-3 defined morphologic features, and each feature was represented in three different tissues. For intra-system precision, each site scanned 23 slides at three different times and one pathologist interpreted all FOVs. For inter-system/site precision, all 69 slides were scanned once at each of three sites, and FOVs from each site were read by one pathologist. To test intra- and inter-pathologist precision, all 69 slides were scanned at one site, FOVs were saved in three different orientations, and the FOVs were transferred to a different site. Three different pathologists then interpreted FOVs from all 69 slides. Wildcard (unscored) slides and washout intervals were included in each study. Agreement estimates with 95% confidence intervals were calculated. RESULTS: Combined precision from all three studies, representing 606 FOVs in each of the three studies, showed overall intra-system agreement of 97.9%; inter-system/site agreement was 96%, intra-pathologist agreement was 95%, and inter-pathologist agreement was 94.2%. CONCLUSIONS: Pathologists using the Aperio AT2 DX System identified histopathological features with high precision, providing increased confidence in using WSI for primary diagnosis in surgical pathology.

Digital Whole Slide Imaging Compared With Light Microscopy for Primary Diagnosis in Surgical Pathology.

CONTEXT.­: The adoption of digital capture of pathology slides as whole slide images (WSI) for educational and research applications has proven utility. OBJECTIVE.­: To compare pathologists' primary diagnoses derived from WSI versus the standard microscope. Because WSIs differ in format and method of observation compared with the current standard glass slide microscopy, this study is critical to potential clinical adoption of digital pathology. DESIGN.­: The study enrolled a total of 2045 cases enriched for more difficult diagnostic categories and represented as 5849 slides were curated and provided for diagnosis by a team of 19 reading pathologists separately as WSI or as glass slides viewed by light microscope. Cases were reviewed by each pathologist in both modalities in randomized order with a minimum 31-day washout between modality reads for each case. Each diagnosis was compared with the original clinical reference diagnosis by an independent central adjudication review. RESULTS.­: The overall major discrepancy rates were 3.64% for WSI review and 3.20% for manual slide review diagnosis methods, a difference of 0.44% (95% CI, -0.15 to 1.03). The time to review a case averaged 5.20 minutes for WSI and 4.95 minutes for glass slides. There was no specific subset of diagnostic category that showed higher rates of modality-specific discrepancy, though some categories showed greater discrepancy than others in both modalities. CONCLUSIONS.­: WSIs are noninferior to traditional glass slides for primary diagnosis in anatomic pathology.

Monocyte Distribution Width: A Novel Indicator of Sepsis-2 and Sepsis-3 in High-Risk Emergency Department Patients.

OBJECTIVES: Most septic patients are initially encountered in the emergency department where sepsis recognition is often delayed, in part due to the lack of effective biomarkers. This study evaluated the diagnostic accuracy of peripheral blood monocyte distribution width alone and in combination with WBC count for early sepsis detection in the emergency department. DESIGN: An Institutional Review Board approved, blinded, observational, prospective cohort study conducted between April 2017 and January 2018. SETTING: Subjects were enrolled from emergency departments at three U.S. academic centers. PATIENTS: Adult patients, 18-89 years, with complete blood count performed upon presentation to the emergency department, and who remained hospitalized for at least 12 hours. A total of 2,212 patients were screened, of whom 2,158 subjects were enrolled and categorized per Sepsis-2 criteria, such as controls (n = 1,088), systemic inflammatory response syndrome (n = 441), infection (n = 244), and sepsis (n = 385), and Sepsis-3 criteria, such as control (n = 1,529), infection (n = 386), and sepsis (n = 243). INTERVENTIONS: The primary outcome determined whether an monocyte distribution width of greater than 20.0 U, alone or in combination with WBC, improves early sepsis detection by Sepsis-2 criteria. Secondary endpoints determined monocyte distribution width performance for Sepsis-3 detection. MEASUREMENTS AND MAIN RESULTS: Monocyte distribution width greater than 20.0 U distinguished sepsis from all other conditions based on either Sepsis-2 criteria (area under the curve, 0.79; 95% CI, 0.76-0.82) or Sepsis-3 criteria (area under the curve, 0.73; 95% CI, 0.69-0.76). The negative predictive values for monocyte distribution width less than or equal to 20 U for Sepsis-2 and Sepsis-3 were 93% and 94%, respectively. Monocyte distribution width greater than 20.0 U combined with an abnormal WBC further improved Sepsis-2 detection (area under the curve, 0.85; 95% CI, 0.83-0.88) and as reflected by likelihood ratio and added value analyses. Normal WBC and monocyte distribution width inferred a six-fold lower sepsis probability. CONCLUSIONS: An monocyte distribution width value of greater than 20.0 U is effective for sepsis detection, based on either Sepsis-2 criteria or Sepsis-3 criteria, during the initial emergency department encounter. In tandem with WBC, monocyte distribution width is further predicted to enhance medical decision making during early sepsis management in the emergency department.

Improved Early Detection of Sepsis in the ED With a Novel Monocyte Distribution Width Biomarker.

BACKGROUND: Sepsis most often presents to the ED, and delayed detection is harmful. WBC count is often used to detect sepsis, but changes in WBC count size also correspond to sepsis. We sought to determine if volume increases of circulating immune cells add value to the WBC count for early sepsis detection in the ED. METHODS: A blinded, prospective cohort study was conducted in two different ED populations within a large academic hospital. RESULTS: Neutrophil and monocyte volume parameters were measured in conjunction with routine CBC testing on a UniCel DxH 800 analyzer at the time of ED admission and were evaluated for the detection of sepsis. There were 1,320 subjects in the ED consecutively enrolled and categorized as control subjects (n = 879) and those with systemic inflammatory response syndrome (SIRS) (n = 203), infection (n = 140), or sepsis (n = 98). Compared with other parameters, monocyte distribution width (MDW) best discriminated sepsis from all other conditions (area under the curve [AUC], 0.79; 95% CI, 0.73-0.84; sensitivity, 0.77; specificity, 0.73; MDW threshold, 20.50), sepsis from SIRS (AUC, 0.74; 95% CI, 0.67-0.84), and severe sepsis from noninfected patients in the ED (AUC, 0.88; 95% CI, 0.75-0.99; negative predictive value, 99%). The added value of MDW to WBC count was statistically significant (AUC, 0.89 for MDW + WBC vs 0.81 for WBC alone; P < .01); a decision curve analysis also showed improved performance compared with WBC count alone. CONCLUSIONS: The incorporation of MDW with WBC count is shown in this prospective cohort study to improve detection of sepsis compared with WBC count alone at the time of admission in the ED. TRIAL REGISTRY: ClinicalTrials.gov; No.: NCT02232750; URL: www.clinicaltrials.gov.

Standardization of whole blood immune phenotype monitoring for clinical trials: panels and methods from the ONE study.

BACKGROUND: Immune monitoring by flow cytometry is a fast and highly informative way of studying the effects of novel therapeutics aimed at reducing transplant rejection or treating autoimmune diseases. The ONE Study consortium has recently initiated a series of clinical trials aimed at using different cell therapies to promote tolerance to renal allografts. To compare the effectiveness of different cell therapies, the consortium developed a robust immune monitoring strategy, including procedures for whole blood (WB) leukocyte subset profiling by flow cytometry. METHODS: Six leukocyte profiling panels computing 7- to 9-surface marker antigens for monitoring the major leukocyte subsets as well as characteristics of T cell, B cell, and dendritic cell (DC) subsets were designed. The precision and variability of these panels were estimated. The assay was standardized within eight international laboratories using Flow-Set Pro beads for mean fluorescence intensity target definition and the flow cytometer setup procedure. Standardization was demonstrated by performing inter-site comparisons. RESULTS: Optimized methods for sample collection, storage, preparation, and analysis were established, including protocols for gating target subsets. WB specimen age testing demonstrated that staining must be performed within 4 hours of sample collection to keep variability low, meaning less than or equal to 10% for the majority of defined leukocyte subsets. Inter-site comparisons between all participating centers testing shipped normal WB revealed good precision, with a variability of 0.05% to 30% between sites. Intra-assay analyses revealed a variability of 0.05% to 20% for the majority of subpopulations. This was dependent on the frequency of the particular subset, with smaller subsets showing higher variability. The intra-assay variability performance defined limits of quantitation (LoQ) for subsets, which will be the basis for assessing statistically significant differences achieved by the different cell therapies. CONCLUSIONS: Local performance and central analysis of the ONE Study flow cytometry panel yields acceptable variability in a standardized assay at multiple international sites. These panels and procedures with WB allow unmanipulated analysis of changes in absolute cell numbers of leukocyte subsets in single- or multicenter clinical trials. Accordingly, we propose the ONE Study panel may be adopted as a standardized method for monitoring patients in clinical trials enrolling transplant patients, particularly trials of novel tolerance promoting therapies, to facilitate fair and meaningful comparisons between trials.

In-use stability modeling.

Experimental design and modeling of in-use stability testing are presented in this paper. In-use open container degradation is considered in terms of time open container or/and the number of instances that the same container is used. Degradation is estimated based on two models, the fixed and the general model. The fixed model estimates in-use degradation for those fixed time points of closed container where the in-used experimental data is collected. The general model estimates in-use degradation for any time point of closed container using the estimated relationship between closed container time and the degradation rate of open container. Data for in-use open container stability does not have to be collected at a closed container time of interest to estimate in-use degradation at this time point as long as this point is within the range of the experiment. Stability of the product in terms of drift from the initial time to the time of interest is calculated as the sum of closed and in-use open containers drifts.

Comparison of the ATS versus EU Mini Wright peak flow meter in normal volunteers.

OBJECTIVES: The American Thoracic Society (ATS) and European Union (EU) have precise and accurate Mini Wright peak flow meters. The purpose of this investigation was to compare both 1) for accuracy using a pneumotachometer, 2) in volunteers to determine whether they are interchangeable, and 3) to spirometrically predicted peak flows. METHODS: Lab testing: A pneumotachometer was connected in series with each peak flow meter and varying flows pushed through both meters for comparison. Human subjects: Nonsmoking adult volunteers did three standing peak flows. The order of peak flow meter used was random. The best of three efforts was used for analysis. The t-test, concordance correlation coefficient (CCC), Deming regression, and Bland-Altman plot were the analytic strategies used to determine agreement. Peak flow results were compared to spirometrically predicted values. RESULTS: Fifty-seven volunteers, average age 37 ± 12 years and mean BMI 24.9 ± 2.5 years, were included. The average peak flows were different at 541 ± 114 and 526 ± 112 L/min for the ATS and EU meters, respectively (p < .01). Both peak flow meter values were significantly different than spirometrically predicted values of 483 ± 86 L/min (p < .01). The CCC was 0.98 (0.97-0.99) and regression revealed a slope and y-intercept consistent with 1 and 0, respectively. The Bland-Altman plot revealed no increase in scatter of values over the range of peak flows versus the difference with a mean bias of 15 ± 15 L/min. Laboratory testing revealed that the ATS and EU peak flow meters read 3.0 ± 2.1% above and -2.0 ± 1.5% below the comparison pneumotachometer, respectively. The pneumotachometer comparison was significantly different for both meters at p < .01, paired t-test. CONCLUSIONS: The ATS peak flow meter reads 2.8% higher than the EU peak flow meter across a range of flows. Both meters have similar accuracy with a different bias compared with a pneumotachometer. Finally, both peak flow meters read slightly and significantly higher than spirometrically derived peak flows. Therefore, the peak flow meters are not interchangeable and both may obtain slightly higher values than those determined using current spirometrically derived prediction equations.

Modeling the effects of storage temperature excursions on shelf life.

Excursions from storage condition requirements may affect product performance and stability. The effects of temperature excursion on stability depend on the amount of time that a product is subjected to these conditions, temperature level, and activation energy. Both time at elevated temperature and the temperature level can be directly measured, while activation energy needs to be estimated from the accelerated stability tests. Coulter Clenz reagent degradation information is used to demonstrate the effects of temperature excursions. The stability of the product is affected by any excursion, but Coulter Clenz will not lose all of its stability for excursion of up to 30 days at 35 degrees C and 20 days at 40 degrees C. Temperature excursion for up to 20 days at 40 degrees C will reduce the stability of a product that has activation energy in the range of 26-30kcalmol(-1) approximately by 5-7 months. Products with lower activation energy will have a significantly lower reduction in stability. The effects of excursions on shelf life performance are less severe when lower level of risk is implemented to establish the claimed shelf life. The proposed model can effectively predict temperature excursion if used within the scope of a product performance and its characteristics.

Uncertainty of measurement and error in stability studies.

It is required that shelf life be determined based on the lower limit of the confidence interval of the estimate from the stability tests. Simulations indicate that a 1-year prediction of shelf life will have approximately 1 month of error. However, this is product specific and is related to the uncertainty of measurement and experimental design. Factors associated with product and experimental design, such as degradation rate, number of time points, implementing a full versus a reduced design, etc., can significantly affect the error of shelf life. Uncertainty in measurement is positively correlated to the amount of error through the manufacturing lot-to-lot variability, precision of the analytical method and calibrator. Experimental design can control random variability and actually can reduce error by increasing number of lots and replicates in stability tests. The decision on the number of lots and replicates will be a balancing act between the uncertainty of the measurement, design and other practical considerations.

Total lung capacity: single breath methane dilution versus plethysmography in normals.

OBJECTIVE AND BACKGROUND: Methane is an inert tracer gas used to obtain TLC estimates during single breath diffusion capacity (DL(CO)) measurements. The aim of this study was to assess the accuracy of methane dilution TLC in normal subjects undergoing single breath diffusion capacity measurements using plethysmography as the gold standard comparison method. METHODS: Fifty non-smoking adults underwent lung function testing. Total lung volume was obtained by both plethysmography and methane dilution during a single breath DL(CO) measurement. Deming regression and the concordance correlation coefficient, r(ccc), were used to determine agreement between methods for TLC. Bias was the mean difference between methods and limits of agreement were the mean difference between methods +/- 1.96 (SD). All values are mean +/- SD unless otherwise stated. RESULTS: Plethysmography and methane dilution TLC values were not significantly different. The r(ccc) was 0.87 (95% confidence interval (CI) 0.78-0.92). Deming regression revealed a slope of 0.93 (P = 0.17, H(o): beta = 1.0; 95% CI 0.84-1.03) and a y-intercept of 0.20 (P = 0.39, H(o): alpha = 0; 95% CI -0.27-0.70). The bias was 0.11 L favouring plethysmography. Limits of agreement varied as 0.11 +/- 0.92 L. CONCLUSIONS: There is statistical agreement between methods suggesting the average TLC by methane could substitute for plethysmography in normals at the population level. At the individual level, a normal methane dilution value indicates a normal TLC whereas values below the normal range should be validated using plethysmography.

Estimation of bias between 2 analytical methods at clinically important ranges.

An approach for estimating bias between 2 analytical methods at different clinical ranges is introduced in this article. The approach models replicated data obtained from the reference and the test method in terms of repeatability and trueness bias. The latter can be partitioned into constant and proportional bias. The approach is based on maximum likelihood estimation and can accommodate normal as well as Poisson and binomial distributions that apply to hematology applications and/or other laboratory methods that count particles per unit of volume and/or time. A full spectrum of statistical inference in the form of confidence intervals for each estimate as well as any statistical hypothesis testing is provided. At the same time these estimates can be practically interpreted and related to any clinical important range or decision point. We recommend this approach as an alternative to the National Committee for Clinical Laboratory Standards (NCCLS) EP9-A2 approach in cases where the application of the NCCLS standard is not appropriate.

Whole blood fixation and permeabilization protocol with red blood cell lysis for flow cytometry of intracellular phosphorylated epitopes in leukocyte subpopulations.

BACKGROUND: Previous studies of intracellular expression of phospho-epitopes in human leukocytes using flow cytometry have used erythrocyte removal or lysis before fixation. Because many of the phospho-epitopes of interest are part of signaling networks that respond to the environment and turn over rapidly, the interval and manipulations used to eliminate erythrocytes from samples have the potential to introduce artifacts. We report a procedure to fix samples containing red blood cells with formaldehyde and then remove erythrocytes by lysis. Detection of phospho-Thr 202/Tyr 204-p44/42 extracellular-regulated kinase (ERK) after phorbol ester acetate (PMA) stimulation was used as a model to measure phospho-epitopes in leukocyte populations in whole blood. METHODS: Normal blood samples were activated with PMA followed by formaldehyde fixation and subsequent treatments with detergents and protein denaturants. The effects of each treatment were monitored by light scatter, selected CD expression intensity, and phosphorylated ERK (pERK) expression. RESULTS: Red cells could be lysed using 0.1% Triton X-100 after brief fixation of whole blood with 2% or 4% formaldehyde. Light scatter improved as a function of formaldehyde concentration and inversely with MeOH concentration. CD3 signal intensity increased when MeOH concentration was reduced. The ratio of pERK immunofluorescence in PMA-stimulated versus nonstimulated (control) samples was highest with high MeOH (90%) and lowest without MeOH treatment. This pattern is consistent with epitope unmasking by alcohol. The pERK epitope could also be unmasked by treatment with high salt, urea, acid, or heat, but none of these produced the level of unmasking of MeOH and each of these was associated with degradation of light scatter and CD3 staining intensity. The final procedure employed 4% formaldehyde, 0.1% Triton X-100, followed by 50% methanol denaturation. Samples prepared in this way demonstrated good preservation of light scatter and surface immunophenotypic patterns, similar to those obtained using a commercial whole blood/red blood cell lysing system (Q-Prep) and an acceptable PMA-stimulated pERK signal (essentially 100% of CD3+ cells that are pERK positive). CONCLUSIONS: Brief fixation of whole blood in 4% formaldehyde followed by treatment with Triton X-100 results in erythrocyte lysis and leukocyte light scatter and immunophenotypic features equivalent to those of other commercial lysis reagents. Intracellular pERK staining is significantly improved by treatment with methanol, but levels of MeOH above 50% degrade light scatter and CD3 expression. This protocol (formaldehyde/Triton X-100/MeOH) circumvents potential artifactual changes in phospho-epitopes due to removal of erythrocytes or erythrocyte lysis followed by fixation, and results in a pERK signal that resolves positive from negative cell populations.

Establishing tolerance levels for customer complaints.

Customer complaints data are usually expressed as counts for a period of time and are governed by a Poisson process. This process is stationary when the number of complaints is constant, while a change in these numbers would indicate a potential change in the product performance. In this paper we describe an approach for establishing the maximum tolerance level for the number of complaints received within a month. Tolerance level is based on a relatively stable period of time when the Poisson process is stationary. A change-point analysis is performed to the complaints data that exhibit large changes to partition the relatively stable period from the problematic period. Examples that illustrate this approach are provided.

Bias estimation in method comparison studies.

A test and a reference analytical method are usually compared for agreement based on paired data obtained from several independent subjects. Bias between two methods can be classified as constant and proportional. In this article, we provide an approach for maximum likelihood estimation of total bias between two methods and partitioning it into constant and proportional bias for each subject. Normal, binomial, or Poisson distribution are the conditional distributions of the response variable that we have considered here, whereas subjects are considered to be random sample from a normally distributed population. Real data on blood cell counts and hemoglobin are used for demonstration. The estimate of biases can be used to test different statistical hypotheses and/or for graphical interpretation of the agreement. The partitioning of total biases in terms of constant and proportional gives an insight on the sources of disagreement between two methods and helps designers and manufacturer's define a remedial strategy.

Determining shelf life by comparing degradations at elevated temperatures.

The degradation of a biological product follows a specific pattern that depends on the kinetics of the chemical reaction. For most biological and pharmaceutical products, accelerated stability tests are preferred to establish shelf life. We describe a new approach for accelerated stability tests, for situations in which the Arrhenius equation is not appropriate. This approach consists of estimating stability at elevated temperatures and comparing these results with the stability estimates for a similar product with a known shelf life. In this article, "Test" refers to Immuno-Trol Low Cells, and "Control" (the product with a known shelf life) refers to Immuno-Trol Cells. The degradation rates and stabilities at elevated temperatures of three antigens of the Test are estimated and compared to their respective Control values. Most of the degradation occurs at the beginning of the experimental period and then slows down until it levels off to form a plateau at the minimum level. Both Control and Test showed similar degradation patterns at three elevated temperatures, indicating that they both have the same mechanism of degradation. Thus, it is expected that they will degrade similarly at storage temperature and have nonsignificantly different shelf lives. The approach for accelerated stability testing discussed here is applicable to situations in which the Arrhenius equation is not appropriate, and the chemical properties of both the Test and Control products are similar.

Linearity evaluation of analytical methods that count particles.

Linearity evaluation of an analytical method is important for both manufacturers of diagnostic devices and laboratory users. Some of the statistical assumptions for estimation and testing in linear regression are violated in analytical methods that count particles per unit of volume or/and time, leading to potential erroneous evaluation of linearity. The objective of this paper is to provide an approach for evaluating linearity in these cases. The number of counts for each concentration level has a Poisson probability distribution that is linear, second-, or higher-order polynomial function of the concentration. Maximum likelihood approach is used to estimate the parameters of the models. Deviance of a particular model and the likelihood ratio test are used to test for linearity. An evaluation of linearity of an analytical method in multiple experiments is also described. No particular changes to the standard testing protocols and data collections are necessary. There are several statistical software packages that can perform the calculations. Formulas and SAS codes presented in this article can also assist in estimation and statistical testing.

Accelerated stability model for predicting shelf-life.

Second- and higher-order degradation reactions sometimes cannot be approximated with linear or exponential relationships and need to be appropriately modeled. Events above the COULTER HmX Analyzer white blood cell (WBC) counting threshold were recorded for the HmX PAK reagent system stored at five elevated temperatures. An accelerated stability model for a second-degree polynomial degradation pattern was used. The shelf-life of the reagent, along with 95% lower bound confidence intervals, is predicted using the same pattern of degradation as well as the Arrhenius approximation. Experiments indicated that the degradation of the HmX PAK reagent occurred in two phases, the lag phase and the degradation phase, in all tested temperatures. The phase durations are temperature-dependent, and the Arrhenius approximation is appropriate (P=0.639). The degradation of the reagent during the lag phase was experimentally undetectable. Changes of the reagent were nonsignificant for a predicted period of 164 days at 25 degrees C. The rate of degradation increased significantly later on during the degradation phase. The lower bound of the 95% confidence interval of this prediction indicated that it would take at least 326 days before the HmX PAK reagent would have any performance issue related to aging at storage temperature.