Search

Objectives There is continuing pressure to improve the cost effectiveness of quality control (QC) for clinical laboratory testing. Risk-based approaches are promising but recent research has uncovered problems in some common methods. There is a need for improvements in risk-based methods for quality control. Methods We provide an overview of a dynamic model for assay behavior. We demonstrate the practical application of the model using simulation and compare the performance of simple Shewhart QC monitoring against Westgard rules. We also demonstrate the utility of trade-off curves for analysis of QC performance. Results Westgard rules outperform simple Shewhart control over a narrow range of the trade-off curve of false-positive and false negative risk. The risk trade-off can be visualized in terms of risk, risk vs. cost, or in terms of cost. Risk trade-off curves can be “smoothed” by log transformation. Conclusions Dynamic risk-models may provide advantages relative to static models for risk-based QC analysis.

Background Setting quality control (QC) limits involves balancing the risk of false-positive results and false-negative results. Recent approaches to QC have focused on the assessment of false-negative results. The Parvin model is the most-used model for risk analysis. The Parvin model assumes that the system makes a transition from an in-control to an out-of-control (OOC) state but makes no further transitions after moving to the OOC state. The implications of this assumption are unclear. Methods We used simulation experiments to compare the performance of QC systems based on no OOC transitions allowed (NOOCTA) vs systems where OOC transitions were allowed (OOCTA). Results The NOOCTA assumption leads to paradoxical tradeoff curves between false-positive results and false-negative results. Predictions of a false-negative result based on NOOCTA were about 10 times lower than models based on OOCTA. Conclusions The most common models for QC risk analysis underestimate false-negative results. There is a need to develop better risk-based methods for QC analysis.

The purpose of this scoping review was to identify available sources of evidence on the epidemiology of transfusion-related acute lung injury (TRALI) and whether meta-analysis on the incidence of TRALI is feasible. TRALI is a serious complication and the second leading cause of death related to blood transfusion. Estimates of the incidence of TRALI would provide a useful benchmark for research to reduce TRALI.We searched the Medline, EMBASE, and PubMed databases for publications related to the incidence of TRALI and hemovigilance. We included all studies irrespective of language or country. Both full-text articles and conference abstracts were included. Participants of the studies must all have received a blood transfusion. Reviews and case studies were excluded.We identified 427 articles or abstracts to include for review. More than half were abstracts, and the majority were published after 2010. Reported TRALI definitions varied, but only 27.2% of studies reported any definition for TRALI. TRALI rates were reported using different denominators, such as per blood unit (54.1%), patient (34.4%), and transfusion episode (14.8%). Study populations and contexts were mostly general (75.6% and 80.3%, respectively). There was also variation in study design with most being observational (90.6%) and only 13.1% of all studies used modern donor restriction policies.There was substantial variation in reporting in studies on TRALI incidence. Meta-analysis of TRALI rates may be feasible in specific circumstances where reporting is clear. Future studies should clearly report key items, such as a TRALI definition.

Non-pathogen reduction platelet bacterial risk control strategies in the US FDA guidance include at least one culture. Almost all of these strategies have a culture hold time of ≥12 h. Studies have reported time to detection (TTD) of bacterial cultures inoculated with bacteria from contaminated platelets, but these data and estimates of risk associated with detection failures have not been synthesized.We performed a literature search to identify studies reporting TTD for samples obtained from spiked platelet components. Using extracted data, regression analysis was used to estimate TTD for culture bottles at different inoculum sizes. Detection failures were defined as events in which contaminated components are transfused to a patient. We then used published data on time of transfusion (ToT) to estimate the risk of detection failures in practice.The search identified 1427 studies, of which 16 were included for analysis. TTD data were available for 16 different organisms, including 14 in aerobic cultures and 11 in anaerobic cultures. For inocula of 1 colony forming unit (CFU), the average TTD for aerobic organisms was 19.2 h while it was 24.9 h in anaerobic organisms, but there was substantial overall variation. A hold time of 12 versus 24 h had minimal effect for most organisms.TTD variation occurs between bacterial species and within a particular species. Under typical inventory management, the relative contribution of culture detection failures is much smaller than the residual risk from sampling failures. Increasing the hold period beyond 12 h has limited value.

BackgroundWe developed a theoretical framework (Precision Quality Control [PQC]) to minimize the cost of quality, but it is not known whether the method can be applied in practice.MethodsWe used data for 2 analytes, cadmium and carbohydrate-deficient transferrin (CDT), and applied the PQC framework to find the optimal control limits. These analytes were selected because they differed with respect to sigma values that are major determinants of control limits. We explored different ways to visualize the results: (a) risk trade-off (false-positive risk vs false-negative risk), (b) cost–risk trade-off (false-positive cost vs false-negative risk), and (c) cost minimization.ResultsWe were able to use the PQC limit to produce 3 different visualizations to suggest control limits. The risk-based analysis was the simplest to apply, but the most difficult to interpret. The cost vs risk method was easy to apply but was still difficult to interpret. The cost minimization method was easy to interpret but required users to declare a willingness to pay that may be difficult to estimate.ConclusionsThe PQC method can be used to find control limits that minimize the cost of quality.

To determine the impact of myeloperoxidase (MPO) and proteinase 3 (PR3) antigen-specific immunoassays in the stratification of patients at-risk for anti-neutrophil cytoplasmic antibodies (ANCA)-associated vasculitides (AAV) at diagnosis.A Medline search was conducted to identify diagnostic accuracy studies using PR3-ANCA or MPO-ANCA for the evaluation of granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA), and eosinophilic granulomatosis with polyangiitis (EGPA). Studies estimates were pooled using the bivariate method.Diagnostic accuracy varied by analyte and AAV subtype. PR3-ANCA had greater sensitivity than MPO-ANCA for GPA (74% vs 11%, p 0.001) and MPO-ANCA greater sensitivity for MPA (73% vs 7%, p 0.001). Specificities of both MPO-ANCA and PR3-ANCA were consistently high (mean 97%, range: 93-99%) for both AAV subtypes. There was insufficient data to perform meta-analysis for the diagnostic accuracy of EPGA.These results validate the use of high quality MPO-ANCA and PR3-ANCA immunoassays to screen patients at-risk for AAV as well as to categorize disease as GPA or MPA subtype. However, caution must be exercised in doing so, since some assays may not have optimal performance. Each laboratory should validate appropriate algorithms based on the tests used and testing population.

The dynamic Precision QC (PQC) model can be used to evaluate the performance of quality control (QC) monitoring systems. The model depends on inputs that describe the intrinsic shift behavior (i.e., stability) of an assay. The output of the model is a trade-off curve that shows the relationship between false negative (FN) and false positive (FP) risk events. The relationship between the inputs and outputs of this model has not yet been explored.We used Monte Carlo simulation to generate trade-off curves using the PQC. We varied the input parameters that determine assay stability (shift probability and shift size distribution) and studied the impact of these inputs on the output (i.e., the trade-off curve relating FN risk to FP risk).FN risk is sensitive to the shift probability and the width of the control limits. FN risk is sensitive to the shape of the shift size distribution when the standard deviation (SD) of the shift size distribution is relatively narrow (i.e., SD2) but is less sensitive to the width of the shift size distribution when the SD is relatively large (i.e., SD2).Practical use of the PQC model may require the estimation of the shift probability and shift size distribution.

Objective To examine the effects of Hispanic nativity on the risk and severity of hypertension relative to US-born non-Hispanic whites. Methods The analytic sample ( n = 34,007) was comprised of cross-sectional data drawn from twenty years of the National Health and Nutrition Examination Survey, 1999–2018. Results Foreign-born Hispanics aged 65 years and older had a greater risk of severe hypertension compared to non-Hispanic Whites. When examined by length of residency in the US, elderly foreign-born Hispanics with less than 10 years of residency were at greater risk of hypertension and severe hypertension, while those with 20 or more years of residency had similar risks compared to non-Hispanic Whites. Conclusion The “Hispanic Paradox” of better health despite lower socioeconomic status, was not observed in foreign-born or US-born Hispanics aged 65 years and older. Among elderly immigrants, those with fewer years of residency had the greatest hypertensive risk.

Septic reactions from platelet transfusions are one of the leading causes of transfusion-associated mortality. The FDA guidance for platelet bacterial risk control includes bacterial culture using both aerobic and anaerobic bottles. Several studies have reported false positive rates (FPR) of culture, but these data have not been summarized or influencing factors analyzed. A systematic review and meta-analysis was performed according to published guidelines to assess the false positive rate and influencing factors. Eighteen studies were included for analysis. The combined aerobic/anaerobic FPR was 2.4 events per thousand (EPT) with a prediction interval of 0.5 to 5.7, while the aerobic FPR rate was 1.0 EPT (prediction interval: 0.2-2.2) and the anaerobic rate was 1.8 EPT. Estimates were based on a total of almost 5 million units tested. The rate of false positives due to instrument error was between 0.5-1.7 EPT, while it was between 0.3-1.0 EPT for sampling contamination based on whether only aerobic, anaerobic, or aerobic/anaerobic cultures were performed. The FPR is approximately 2 to 5 times higher than the literature reported true positive rate of 0.5 EPT.

Background and objectives Septic transfusion reactions are a principal cause of transfusion-related mortality. The frequency of detectable bacterial contamination is greater in platelets compared to other blood components because platelets are stored at room temperature. Most strategies outlined in the September 2019 FDA guidance require both aerobic culture (AC) and anaerobic culture (AnC) testing. We performed a systematic review and meta-analysis in an effort to provide the best available estimate of the effectiveness of AnC. Materials and methods Our analysis was performed according to published guidelines. Broad and context-specific meta-analyses of bacterial detection rates in platelets by AnC were performed to assess the practical effectiveness of AnC as a risk control measure. Results Seven studies with a total of 1 767 014 tested platelet components were included for analysis. With exclusion of positives due to Cutibacterium/Propionibacterium species and redundancy due to AC results, AnC detected 0·06 contamination events per thousand (EPT) components tested, twofold lower than the AC (0·12 EPT). Conclusion Excluding Cutibacterium/Propionibacterium species, AnC detects occasional bacterial contamination events that are not detected by AC (~1 in 17 000 platelet components).

Background Invasive fungal infections (IFIs) are life-threatening opportunistic infections that occur in immunocompromised or critically ill people. Early detection and treatment of IFIs is essential to reduce morbidity and mortality in these populations. (1→3)-β-D-glucan (BDG) is a component of the fungal cell wall that can be detected in the serum of infected individuals. The serum BDG test is a way to quickly detect these infections and initiate treatment before they become life-threatening. Five different versions of the BDG test are commercially available: Fungitell, Glucatell, Wako, Fungitec-G, and Dynamiker Fungus. Objectives To compare the diagnostic accuracy of commercially available tests for serum BDG to detect selected invasive fungal infections (IFIs) among immunocompromised or critically ill people. Search methods We searched MEDLINE (via Ovid) and Embase (via Ovid) up to 26 June 2019. We used SCOPUS to perform a forward and backward citation search of relevant articles. We placed no restriction on language or study design. Selection criteria We included all references published on or after 1995, which is when the first commercial BDG assays became available. We considered published, peer-reviewed studies on the diagnostic test accuracy of BDG for diagnosis of fungal infections in immunocompromised people or people in intensive care that used the European Organization for Research and Treatment of Cancer (EORTC) criteria or equivalent as a reference standard. We considered all study designs (case-control, prospective consecutive cohort, and retrospective cohort studies). We excluded case studies and studies with fewer than ten participants. We also excluded animal and laboratory studies. We excluded meeting abstracts because they provided insufficient information. Data collection and analysis We followed the standard procedures outlined in the Cochrane Handbook for Diagnostic Test Accuracy Reviews. Two review authors independently screened studies, extracted data, and performed a quality assessment for each study. For each study, we created a 2 × 2 matrix and calculated sensitivity and specificity, as well as a 95% confidence interval (CI). We evaluated the quality of included studies using the Quality Assessment of Studies of Diagnostic Accuracy-Revised (QUADAS-2). We were unable to perform a meta-analysis due to considerable variation between studies, with the exception of Candida, so we have provided descriptive statistics such as receiver operating characteristics (ROCs) and forest plots by test brand to show variation in study results. Main results We included in the review 49 studies with a total of 6244 participants. About half of these studies (24/49; 49%) were conducted with people who had cancer or hematologic malignancies. Most studies (36/49; 73%) focused on the Fungitell BDG test. This was followed by Glucatell (5 studies; 10%), Wako (3 studies; 6%), Fungitec-G (3 studies; 6%), and Dynamiker (2 studies; 4%). About three-quarters of studies (79%) utilized either a prospective or a retrospective consecutive study design; the remainder used a case-control design. Based on the manufacturer's recommended cut-off levels for the Fungitell test, sensitivity ranged from 27% to 100%, and specificity from 0% to 100%. For the Glucatell assay, sensitivity ranged from 50% to 92%, and specificity ranged from 41% to 94%. Limited studies have used the Dynamiker, Wako, and Fungitec-G assays, but individual sensitivities and specificities ranged from 50% to 88%, and from 60% to 100%, respectively. Results show considerable differences between studies, even by manufacturer, which prevented a formal meta-analysis. Most studies (32/49; 65%) had no reported high risk of bias in any of the QUADAS-2 domains. The QUADAS-2 domains that had higher risk of bias included participant selection and flow and timing. Authors' conclusions We noted considerable heterogeneity between studies, and these differences precluded a formal meta-analysis. Because of wide variation in the results, it is not possible to estimate the diagnostic accuracy of the BDG test in specific settings. Future studies estimating the accuracy of BDG tests should be linked to the way the test is used in clinical practice and should clearly describe the sampling protocol and the relationship of time of testing to time of diagnosis.

Background Bacterial contamination of platelets is a problem that can lead to harmful septic transfusion reactions. The US Food and Drug Administration published a guidance in September 2019 detailing several permissible risk control strategies. Our objective was to compare the safety of each bacterial testing strategy for apheresis platelets. Study design and methods We used simulation to compare safety of the nine risk control strategies involving apheresis platelet testing. The primary outcome was the risk of exposure. An exposure event occurred if a patient received platelets exceeding a specific contamination threshold (>0, 103 , and 105 colony-forming units (CFU/mL). We generated a range of bacterial contamination scenarios (inoculum size, doubling time, lag time) and compared risk of exposure for each policy in each contamination scenario. We then computed the average risk difference over all scenarios. Results At the 0 CFU/mL exposure threshold, two-step policies that used secondary culture ranked best (all top three), while single-step 24-hour culture with 3-day expiration ranked last (ninth). This latter policy performed well (median rank of 1) at both the 103 and 105 CFU/mL thresholds, but 48-hour culture with 7-day expiration performed relatively poorly. At these higher thresholds, median ranks of two-step policies that used secondary culture were again top three. Two-step policies that used rapid testing improved at the higher (105 CFU/mL) harm threshold, with median rankings between 1 and 5. Conclusion Two-step policies that used secondary culture were generally safer than single-step policies. Performance of two-step policies that used rapid testing depended on the CFU per milliter threshold of exposure used.

Background Platelets have the highest bacterial contamination risk of all blood components, and septic transfusion reactions remain a problem. A good estimate of contamination rates could provide information about residual risk and inform optimal testing strategies. We performed a systematic review and meta-analysis of platelet contamination rates by primary culture. Study design and methods A literature search in December 2019 identified articles on platelet contamination rates using primary culture. We used meta-analysis to estimate the overall rate of contamination and meta-regression to identify heterogeneity. We studied the following sources of heterogeneity: collection method, sample volume, positivity criteria, and study date. Contamination rate estimates were obtained for apheresis (AP), platelet rich plasma (PRP), and buffy coat (BC) collection methods. Results The search identified 6102 studies, and 22 were included for meta-analysis. Among these 22 studies, there were 21 AP cohorts (4,072,022 components), 4 PRP cohorts (138,869 components), and 15 BC cohorts (1,474,679 components). The overall mean contamination rate per 1000 components was 0.51 (95% CI: 0.38-0.67) including AP (0.23, 95% CI: 0.18-0.28), PRP, (0.38, 95% CI: 0.15-0.70), and BC (1.12, 95% CI: 0.51-1.96). There was considerable variability within each collection method. Sample volume, positivity criteria, and publication year were significant sources of heterogeneity. Conclusion The bacterial contamination rate of platelets by primary culture is 1 in 1961. AP and PRP components showed a lower contamination rate than BC components. There is clinically significant between-study variability for each method. Larger sample volumes increased sensitivity, and bacterial contamination rates have decreased over time.

Background: Inherited genetic variants can modify the cancer-chemopreventive effect of aspirin. We evaluated the clinical and economic value of genotype-guided aspirin use for colorectal cancer chemoprevention in average-risk individuals. Methods: A decision analytical model compared genotype-guided aspirin use versus no genetic testing, no aspirin. The model simulated 100,000 adults ≥50 years of age with average colorectal cancer and cardiovascular disease risk. Low-dose aspirin daily starting at age 50 years was recommended only for those with a genetic test result indicating a greater reduction in colorectal cancer risk with aspirin use. The primary outcomes were quality-adjusted life-years (QALY), costs, and incremental cost-effectiveness ratio (ICER). Results: The mean cost of using genotype-guided aspirin was $187,109 with 19.922 mean QALYs compared with $186,464 with 19.912 QALYs for no genetic testing, no aspirin. Genotype-guided aspirin yielded an ICER of $66,243 per QALY gained, and was cost-effective in 58% of simulations at the $100,000 willingness-to-pay threshold. Genotype-guided aspirin was associated with 1,461 fewer polyps developed, 510 fewer colorectal cancer cases, and 181 fewer colorectal cancer-related deaths. This strategy prevented 1,078 myocardial infarctions with 1,430 gastrointestinal bleeding events, and 323 intracranial hemorrhage cases compared with no genetic testing, no aspirin. Conclusions: Genotype-guided aspirin use for colorectal cancer chemoprevention may offer a cost-effective approach for the future management of average-risk individuals. Impact: A genotype-guided aspirin strategy may prevent colorectal cancer, colorectal cancer-related deaths, and myocardial infarctions, while minimizing bleeding adverse events. This model establishes a framework for genetically-guided aspirin use for targeted chemoprevention of colorectal cancer with application toward commercial testing in this population.

Background Quality control charts (Levey Jennings Charts) are based on estimates of variation. There are two general approaches for estimating variation: those based on short-term variation and those based on long-term variation. We have observed that clinical laboratory science (CLS) tends to estimate variation using long-term variation but that most other fields use short-term variation. The objective of this study is to compare these two methods of estimating process variation, compare the accuracy of control limits generated by each method, and explore whether it would be useful for clinical laboratories to adopt methods used in other fields. Methods We conducted a literature review to compare recommendations for methods for estimation of variation in CLS with other fields. We searched textbooks for suggested methods and also searched the primary literature for references to methods associated with short-term and long-term variation. We provide theoretical results from statistics to show that, in theory, short-term estimates can differ from long-term estimates of variation. We used simulation studies to show that one can construct examples where short-term and long-term estimates of variation lead to significant differences in control limits. Finally, we show laboratory data comparing short-term and long-term estimates of variation. Results We found that practice in CLS differs from other fields. We found no references to methods based on short-term variation in CLS textbooks and only one reference in the primary literature. In contrast, standard quality control (QC) texts recommend methods based on short-term variation and the primary literature makes frequent reference to such methods. We found statistical papers that show that, in theory, estimates based on long-term variation can produce inflated estimates of process variation. We used simulation to show that such examples can be constructed. We examined 95 QC charts and found that in 93 cases, there were significant differences between short-term and long-term estimates of variation. The ratio of long-term to short-term variation was greater than 1.5 in 18% of cases. Conclusion Estimates of variation based on short-term and long-term variation can lead to significant differences in estimates. Estimates based on long-term variation are frequently larger than estimates based on short-term variation.

Objectives To determine the accuracy of Fungitell, a β-d-glucan (BDG) test, for the diagnosis of invasive fungal infection (IFI) among cancer patients. Methods For this meta-analysis, MEDLINE and EMBASE were searched for references related to BDG testing. Study quality was evaluated using QUADAS-2. Statistical analysis was performed using Stata 14. Results We screened 12,426 references and identified 189 studies for full-text review. Nineteen studies were included in the final meta-analysis. There was moderate heterogeneity between studies. Nine studies had a high risk of bias, which significantly elevated the overall specificity estimate. Restricting to only low-bias studies, the sensitivity and specificity were 80% and 63%, respectively. Conclusions The overall sensitivity and specificity of Fungitell as a diagnostic test for IFI is moderate, and there is substantial heterogeneity between studies. Limiting studies to only low-bias risk reduced heterogeneity but also lowered the overall specificity estimate.

Objectives To determine the impact of systemwide charge display on laboratory utilization. Methods This was a randomized controlled trial with a baseline period and an intervention period. Tests were randomized to a control arm or an active arm. The maximum allowable Medicare reimbursement rate was displayed for tests in the active arm during the intervention period. Total volume of tests in the active arm was compared with those in the control arm. Residents were surveyed before and after the intervention to assess charge awareness. Results Charge display had no effect on order behavior. This result held for patient type (inpatient vs outpatient) and for insurance category (commercial, government, self-pay). Residents overestimated the charges of tests both before and after the intervention. Many residents failed to notice the charge display in the computerized order entry system. Conclusions The impact of charge display depends on context. Charge display is not always effective.

BackgroundMultirules are often employed to monitor quality control (QC). The performance of multirules is usually determined by simulation and is difficult to predict. Previous studies have not provided computer code that would enable one to experiment with multirules. It would be helpful for analysts to have computer code to analyze rule performance.ObjectiveTo provide code to calculate power curves and to investigate certain properties of multirule QC.MethodsWe developed computer code in the R language to simulate multirule performance. Using simulation, we studied the incremental performance of each rule and determined the average run length and time to signal.ResultsWe provide R code for simulating multirule performance. We also provide a Microsoft Excel spreadsheet with a tabulation of results that can be used to create power curves. We found that the R4S and 10x rules add very little power to a multirule set designed to detect shifts in the mean.ConclusionQC analysts should consider using a limited-rule set.

Background Invasive fungal infection (IFI) is a life-threatening complication of allogeneic hematopoietic stem cell transplantation (HSCT) that is also associated with excess healthcare costs. Current approaches include universal antifungal prophylaxis, preemptive therapy based on biomarker surveillance, and empiric treatment initiated in response to clinical signs/symptoms. However, no study has directly compared the cost-effectiveness of these treatment strategies for an allogeneic HSCT patient population. Methods We developed a state transition model to study the impact of treatment strategies on outcomes associated with IFIs in the first 100 days following myeloablative allogeneic HSCT. We compared three treatment strategies: empiric voriconazole, preemptive voriconazole (200 mg), or prophylactic posaconazole (300 mg) for the management of IFIs. Preemptive treatment was guided by scheduled laboratory surveillance with galactomannan (GM) testing. Endpoints were cost and survival at 100 days post-HSCT. Results Empiric treatment was the least costly ($147 482) and was equally effective (85.2% survival at 100 days) as the preemptive treatment strategies. Preemptive treatments were slightly more costly than empiric treatment (GM cutoff ≥ 1.0 $147 910 and GM cutoff ≥ 0.5 $148 108). Preemptive therapy with GM cutoff ≥ 1.0 reduced anti-mold therapy by 5% when compared to empiric therapy. Posaconazole prophylaxis was the most effective (86.6% survival at 100 days) and costly ($152 240) treatment strategy with a cost of $352 415 per life saved when compared to empiric therapy. Conclusions One preemptive treatment strategy reduced overall anti-mold drug exposure but did not reduce overall costs. Prevention of IFI using posaconazole prophylaxis was the most effective treatment strategy and may be cost-effective, depending upon the willingness to pay per life saved.

To determine the accuracy of Fungitell, a β-d-glucan (BDG) test, for the diagnosis of invasive fungal infection (IFI) among cancer patients.For this meta-analysis, MEDLINE and EMBASE were searched for references related to BDG testing. Study quality was evaluated using QUADAS-2. Statistical analysis was performed using Stata 14.We screened 12,426 references and identified 189 studies for full-text review. Nineteen studies were included in the final meta-analysis. There was moderate heterogeneity between studies. Nine studies had a high risk of bias, which significantly elevated the overall specificity estimate. Restricting to only low-bias studies, the sensitivity and specificity were 80% and 63%, respectively.The overall sensitivity and specificity of Fungitell as a diagnostic test for IFI is moderate, and there is substantial heterogeneity between studies. Limiting studies to only low-bias risk reduced heterogeneity but also lowered the overall specificity estimate.

Objective: To determine the accuracy with which morphology alone can distinguish adenocarcinoma and squamous cell carcinoma in non-small cell lung cancer. Methods: We performed a systematic review and meta-analysis. Three data bases (MEDLINE, EMBASE, Scopus) were searched for studies on the diagnostic accuracy of subtyping non-small cell lung cancer. Accuracy data was abstracted and synthesized using bivariate mixed effects logistic regression as implemented in the midas package in Stata 14. Heterogeneity was assessed using the Higgins I2. Results: We included 17 studies (2,235 cases). Most studies had a low risk of bias. The pooled diagnostic accuracy for cytological diagnosis of adenocarcinoma resulted in a sensitivity of 63% (48–76%) and specificity of 95% (87–98%). The I2 values were 93 and 88% for sensitivity and specificity, respectively. The pooled diagnostic accuracy for the cytological diagnosis of squamous cell carcinoma resulted in a sensitivity of 84% (79–88%) and a specificity of 90% (84–94%). The I2 values were 69 and 86% for sensitivity and specificity, respectively. Conclusion: Accuracy varies widely by study and summary estimates do not provide a useful representation of accuracy. Squamous cell carcinoma was diagnosed more accurately than adenocarcinoma.

Objective: Rapid on-site evaluation (ROSE) has been shown to improve adequacy rates and reduce needle passes. ROSE is often performed by cytopathologists who have limited availability and may be costlier than alternatives. Several recent studies examined the use of alternative evaluators (AEs) for ROSE. A summary of this information could help inform guidelines regarding the use of AEs. The objective was to assess the accuracy of AEs compared to cytopathologists in assessing the adequacy of specimens during ROSE. Study Design: This was a systematic review and meta-analysis. Reporting and study quality were assessed using the STARD guidelines and QUADAS-2. All steps were performed independently by two evaluators. Summary estimates were obtained using the hierarchal method in Stata v14. Heterogeneity was evaluated using Higgins’ I2 statistic. Results: The systematic review identified 13 studies that were included in the meta-analysis. Summary estimates of sensitivity and specificity for AEs were 97% (95% CI: 92–99%) and 83% (95% CI: 68–92%). There was wide variation in accuracy statistics between studies (I2 = 0.99). Conclusions: AEs sometimes have accuracy that is close to cytopathologists. However, there is wide variability between studies, so it is not possible to provide a broad guideline regarding the use of AEs.

Background Appropriate test utilization is usually evaluated by adherence to published guidelines. In many cases, medical guidelines are not available. Benchmarking has been proposed as a method to identify practice variations that may represent inappropriate testing. This study investigated the use of benchmarking to identify sites with inappropriate utilization of testing for a particular analyte. Methods We used a Web-based survey to compare 2 measures of vitamin D utilization: overall testing intensity (ratio of total vitamin D orders to blood-count orders) and relative testing intensity (ratio of 1,25(OH)2D to 25(OH)D test orders). Results A total of 81 facilities contributed data. The average overall testing intensity index was 0.165, or approximately 1 vitamin D test for every 6 blood-count tests. The average relative testing intensity index was 0.055, or one 1,25(OH)2D test for every 18 of the 25(OH)D tests. Both indexes varied considerably. Conclusions Benchmarking can be used as a screening tool to identify outliers that may be associated with inappropriate test utilization. * TTG : tissue transglutaminase Ig : immunoglobulin CBCs : complete blood counts WBC : white blood cell count RTI : relative testing intensity OTI : overall testing intensity USPSTF : United States Preventive Services Task Force T3 : triiodothyronine T4 : thyroxine IQR : interquartile range

Objective The aim of this article is to determine the cost effectiveness of cell free DNA (cfDNA) as a replacement forintegrated screening using a societal cost perspective. Method This study used Monte-Carlo simulation with one-way and probabilistic sensitivity analysis. Results Cell free DNA is more effective and less costly than integrated screening. The incremental cost-effectivenessratio for cfDNA relative to the integrated test was $277955 per case detected (95th percent confidence interval $881882 to $532785). Conclusion Cell free DNA screening is a cost-effective replacement for maternal serum screening when the lifetimecosts of Down syndrome live births are considered. The adoption of cfDNA screening would save approximately$277955 for each additional case detected over integrated screening. © 2014 John Wiley & Sons, Ltd. Funding sources: NoneConflicts of interest: All of the authors are employees of ARUP laboratories, which provide a wide range of clinical laboratory tests including prenatal screening tests(both maternal serum screening and cfDNA).

BACKGROUND Reliable estimates of accuracy are important for any diagnostic test. Diagnostic accuracy studies are subject to unique sources of bias. Verification bias and classification bias are 2 sources of bias that commonly occur in diagnostic accuracy studies. Statistical methods are available to estimate the impact of these sources of bias when they occur alone. The impact of interactions when these types of bias occur together has not been investigated. METHODS We developed mathematical relationships to show the combined effect of verification bias and classification bias. A wide range of case scenarios were generated to assess the impact of bias components and interactions on total bias. RESULTS Interactions between verification bias and classification bias caused overestimation of sensitivity and underestimation of specificity. Interactions had more effect on sensitivity than specificity. Sensitivity was overestimated by at least 7% in approximately 6% of the tested scenarios. Specificity was underestimated by at least 7% in less than 0.1% of the scenarios. CONCLUSIONS Interactions between verification bias and classification bias create distortions in accuracy estimates that are greater than would be predicted from each source of bias acting independently. Cancer (Cancer Cytopathol) 2014. © 2014 American Cancer Society. Cancer (Cancer Cytopathol) 2015;123:193–201. © 2014 American Cancer Society.

To determine the impact of systemwide charge display on laboratory utilization.This was a randomized controlled trial with a baseline period and an intervention period. Tests were randomized to a control arm or an active arm. The maximum allowable Medicare reimbursement rate was displayed for tests in the active arm during the intervention period. Total volume of tests in the active arm was compared with those in the control arm. Residents were surveyed before and after the intervention to assess charge awareness.Charge display had no effect on order behavior. This result held for patient type (inpatient vs outpatient) and for insurance category (commercial, government, self-pay). Residents overestimated the charges of tests both before and after the intervention. Many residents failed to notice the charge display in the computerized order entry system.The impact of charge display depends on context. Charge display is not always effective.

Context.—Statistical literacy can be defined as understanding the statistical tests and terminology needed for the design, analysis, and conclusions of original research or laboratory testing. Little is known about the statistical literacy of clinical or anatomic pathologists. Objective.—To determine the statistical methods most commonly used in pathology studies from the literature and to assess familiarity and knowledge level of these statistical tests by pathology residents and practicing pathologists. Design.—The most frequently used statistical methods were determined by a review of 1100 research articles published in 11 pathology journals during 2015. Familiarity with statistical methods was determined by a survey of pathology trainees and practicing pathologists at 9 academic institutions in which pathologists were asked to rate their knowledge of the methods identified by the focused review of the literature. Results.—We identified 18 statistical tests that appear frequently in published pathology studies. On average, pathologists reported a knowledge level between “no knowledge” and “basic knowledge” of most statistical tests. Knowledge of tests was higher for more frequently used tests. Greater statistical knowledge was associated with a focus on clinical pathology versus anatomic pathology, having had a statistics course, having an advanced degree other than an MD degree, and publishing research. Statistical knowledge was not associated with length of pathology practice. Conclusions.—An audit of pathology literature reveals that knowledge of about 12 statistical tests would be sufficient to provide statistical literacy for pathologists. On average, most pathologists report they can interpret commonly used tests but are unable to perform them. Most pathologists indicated that they would benefit from additional statistical training.

Objectives In maternal serum screening for Down syndrome, a cutoff of 1 : 270 is often used as a decision point to recommend invasive confirmatory testing. However, it has not been established how well this or any other cutoff relates to patient preferences, that is, the values that pregnant women attach to various screening outcomes. The purpose of this study was to examine the clinical and economic tradeoffs of a wide range of risk cutoffs for the quadruple screen. Methods Screening costs and outcomes for multiple risk cutoffs were modeled using a Monte Carlo simulation. Results The optimal cutoff for maternal serum screening depends on the relative values placed by the patient on different outcomes. Total societal costs were similar across the range of cutoffs. Conclusions Given that different screening outcomes are optimized by different cutoff values, a one-size-fits-all approach may not be appropriate. © 2013 John Wiley & Sons, Ltd.

Background Biochemical prenatal screening tests are used to determine the risk of fetal aneuploidy based on the concentration of several biomarkers. The concentration of these biomarkers could be affected by preanalytical factors (PAFs) such as sample type (whole blood vs serum), storage time, and storage temperature. The impact of these factors on posttest risk is unknown. Methods Blood samples were collected from 25 pregnant patients. Each sample was divided into 24 aliquots, and each aliquot was subjected to 1 of 24 different treatments (2 sample types × 2 temperatures × 6 storage times). The impact of each PAF on calculated risk was estimated using mixed-effects regression and simulation analysis. Results PAFs were associated with statistically significant changes in concentration for some analytes. Simulation studies showed that PAFs accounted for 6% of the variation in posttest risk, and analytical imprecision accounted for 94% of the variation. We estimated that the background misclassification rate due to analytical imprecision is approximately 1.37% for trisomy 21 and 0.12% for trisomy 18. Preanalytical factors increased the probability of misclassification by 0.46% and 0.06% for trisomies 21 and 18, respectively. Conclusions Relaxing sample specifications for biochemical prenatal serum screening tests to permit analysis of serum samples stored for up to 72 h at room temperature or 4 °C as well as serum obtained from whole blood stored similarly has a small impact in calculated posttest aneuploidy risk.

Candida bloodstream infections (BSI) are associated with significant morbidity, mortality, and increased health care costs. Early treatment is essential, because delayed therapy detrimentally impacts clinical outcomes. The FDA recently approved the first culture-independent direct molecular detection method for Candida BSIs (T2Candida). The speed and sensitivity of this assay give it the potential to improve patient care, but the reagents and instrumentation are expensive. We used an analytic decision tree model to compare the cost-effectiveness of T2Candida-directed antifungal therapy (T2DT) to that of either empirical therapy (ET) or blood culture-directed therapy (BCDT). The costs included those of T2Candida testing, antifungal treatment, and hospital length of stay. The effectiveness measure was survival status at hospital discharge. T2DT was less costly and more effective than BCDT but was less costly and less effective than ET with an echinocandin (incremental cost-effectiveness ratio, $111,084 per additional survivor). One-way sensitivity analyses demonstrated that the cost-effectiveness of T2DT was highly dependent on Candida BSI prevalence and the cost of antifungal therapy and T2Candida test reagents. The use of T2DT reduced the number of unnecessarily treated patients by 98% relative to that with ET. Reduced drug exposure might lessen the possibility of drug-related adverse events and may also prevent the development of antifungal resistance or emergence of drug-resistant Candida species. The greatest benefit of T2Candida appears to be the ability to confidently withhold or stop empirical antifungal therapy in low-to-moderate-risk patients who are unlikely to benefit from treatment.

Background Non-invasive prenatal testing (NIPT) is a relatively new technology for diagnosis of fetal aneuploidies. NIPT is more accurate than conventional maternal serum screening (MSS) but is also more costly. Contingent NIPT may provide a cost-effective alternative to universal NIPT screening. Contingent screening used a two-stage process in which risk is assessed by MSS in the first stage and, based on a risk cutoff, high-risk pregnancies are referred for NIPT. The objective of this study was to (1) determine the optimum MSS risk cutoff for contingent NIPT and (2) compare the cost effectiveness of optimized contingent NIPT to universal NIPT and conventional MSS. Study Design Decision-analytic model using micro-simulation and probabilistic sensitivity analysis. We evaluated cost effectiveness from three perspectives: societal, governmental, and payer. Results From a societal perspective, universal NIPT dominated both contingent NIPT and MSS. From a government and payer perspective, contingent NIPT dominated MSS. Compared to contingent NIPT, adopting a universal NIPT would cost $203,088 for each additional case detected from a government perspective and $263,922 for each additional case detected from a payer perspective. Conclusions From a societal perspective, universal NIPT is a cost-effective alternative to MSS and contingent NIPT. When viewed from narrower perspectives, contingent NIPT is less costly than universal NIPT and provides a cost-effective alternative to MSS.

e18318 Background: Colonoscopy is the “gold standard” for colorectal cancer (CRC) screening. However, adherence rates are low and detection is not optimal. Concomitant aspirin chemoprevention is recommended by US Preventive Task Force, but bleeding complications can be limiting. Variant genotypes in aspirin metabolism can modify CRC and adenoma risk. Genotype guided aspirin (ggASA) use will identify a targeted average-risk population for maximal aspirin benefit while minimizing adverse events rates compared to the general population. We conducted a cost-effectiveness analysis (CEA) of primary chemoprevention in CRC using ggASA compared to no intervention, and colonoscopy ±general aspirin in healthy average-risk individuals. Methods: Our Markov decision analytical model consisted of 5 possible health states: no CRC/polyps, adenoma, pre-clinical CRC, CRC, and death. Model probabilities for CRC and its prevalence were estimated using SEER database and published literature. A microsimulation of 10,000 individuals aged 50-64 years was used to estimate cost-effectiveness from US payer perspective over lifetime. One way and probabilistic sensitivity analyses and model validation results will be reported in the final poster. Results: Our results suggest that compared to colonoscopy and no intervention, ggASA was associated with fewer CRC cases, and CRC-related deaths and MI cases. Compared to colonoscopy + general aspirin, ggASA was associated with fewer bleeding events, similar rates of CRC and CRC-related deaths, and fewer MI cases prevented. From a cost-effectiveness standpoint, ggASA use over a lifetime had the lowest costs and highest quality adjusted life years gained compared to other strategies, if testing costs were ignored. Once genetic testing costs exceeds $63, colonoscopy + general aspirin becomes the most cost effective strategy. Between genetic testing cost of $63-283, the costs of using ggASA per quality adjusted life year gained is below $100,000. Conclusions: Genotype-guided aspirin use precisely identifies an average-risk population, and lowers adverse events rates compared to general aspirin. The economic value of genotype-guided aspirin is dependent on the genetic testing costs.

Reliable estimates of accuracy are important for any diagnostic test. Diagnostic accuracy studies are subject to unique sources of bias. Verification bias and classification bias are 2 sources of bias that commonly occur in diagnostic accuracy studies. Statistical methods are available to estimate the impact of these sources of bias when they occur alone. The impact of interactions when these types of bias occur together has not been investigated.We developed mathematical relationships to show the combined effect of verification bias and classification bias. A wide range of case scenarios were generated to assess the impact of bias components and interactions on total bias.Interactions between verification bias and classification bias caused overestimation of sensitivity and underestimation of specificity. Interactions had more effect on sensitivity than specificity. Sensitivity was overestimated by at least 7% in approximately 6% of the tested scenarios. Specificity was underestimated by at least 7% in less than 0.1% of the scenarios.Interactions between verification bias and classification bias create distortions in accuracy estimates that are greater than would be predicted from each source of bias acting independently.

The aim of this article is to determine the cost effectiveness of cell free DNA (cfDNA) as a replacement for integrated screening using a societal cost perspective.This study used Monte-Carlo simulation with one-way and probabilistic sensitivity analysis.Cell free DNA is more effective and less costly than integrated screening. The incremental cost-effectiveness ratio for cfDNA relative to the integrated test was -$277 955 per case detected (95th percent confidence interval -$881 882 to $532 785).Cell free DNA screening is a cost-effective replacement for maternal serum screening when the lifetime costs of Down syndrome live births are considered. The adoption of cfDNA screening would save approximately $277 955 for each additional case detected over integrated screening.

Background: The effectiveness of endoscopic ultrasound-guided fine-needle aspiration increases with the number of needle passes but needle passes are also associated with increased risk of adverse events. The trade-off between needle passes and adequacy has not been well-characterized. Aims: The purpose of this study was to compare the risk-benefit tradeoff of different sampling protocols with and without rapid onsite evaluation (ROSE). Patients and Methods: We used a discrete-event simulation model to compare eight different sampling protocols. Each sampling protocol was simulated 10,000 times to obtain the average performance for each scenario. The per-pass adequacy rates, ROSE, accuracy of the assessor and sampling limits were varied to determine the impact of these factors on the number of needle passes and adequacy Results: Increasing per-class adequacy can be achieved at a cost of increased needle passes. Sampling with ROSE achieved higher adequacy with fewer needle passes than policies using a fixed number of needle passes without ROSE. Conclusions: Variable sampling policies using ROSE generally achieve greater per-case adequacy with fewer needle passes than non-ROSE sampling policies using a fixed number of passes. Refereed/Peer-reviewed

Various models and questionnaires have been developed for screening specific populations for obstructive sleep apnea (OSA) as defined by the apnea/hypopnea index (AHI); however, almost every method is based upon dichotomizing a population, and none function ideally. We evaluated the possibility of using the STOP-Bang model (SBM) to classify severity of OSA into 4 categories ranging from none to severe.Anthropomorphic data and the presence of snoring, tiredness/sleepiness, observed apneas, and hypertension were collected from 1426 patients who underwent diagnostic polysomnography. Questionnaire data for each patient was converted to the STOP-Bang equivalent with an ordinal rating of 0 to 8. Proportional odds logistic regression analysis was conducted to predict severity of sleep apnea based upon the AHI: none (AHI5/h), mild (AHI ≥ 5 to15/h), moderate (≥ 15 to30/h), and severe (AHI ≥ 30/h).Linear, curvilinear, and weighted models (R(2) = 0.245, 0.251, and 0.269, respectively) were developed that predicted AHI severity. The linear model showed a progressive increase in the probability of severe (4.4% to 81.9%) and progressive decrease in the probability of none (52.5% to 1.1%). The probability of mild or moderate OSA initially increased from 32.9% and 10.3% respectively (SBM score 0) to 39.3% (SBM score 2) and 31.8% (SBM score 4), after which there was a progressive decrease in probabilities as more patients fell into the severe category.The STOP-Bang model may be useful to categorize OSA severity, triage patients for diagnostic evaluation or exclude from harm.

Rapid on-site evaluation (ROSE) can improve adequacy rates and reduce needle passes but it also increases operational costs. It is not clear when ROSE is less costly than sampling with a fixed number of needle passes. The performance of ROSE depends on many factors, and the potential advantage of ROSE is likely to vary by context. …