Statistics From A (Agreement) to Z (z Score): A Guide to Interpreting Common Measures of Association, Agreement, Diagnostic Accuracy, Effect Size, Heterogeneity, and Reliability in Medical Research

Background/Aims: In a linguistically diverse country such as India, challenges remain with regard to diagnosis of early cognitive decline among the elderly, with no prior attempts made to simultaneously validate a comprehensive battery of tests across domains in multiple languages. This study aimed to determine the utility of the Indian Council of Medical Research-Neurocognitive Tool Box (ICMR-NCTB) in the diagnosis of mild cognitive impairment (MCI) and its vascular subtype (VaMCI) in 5 Indian languages. Methods: Literate subjects from 5 centers across the country were recruited using a uniform process, and all subjects were classified based on clinical evaluations and a gold standard test protocol into normal cognition, MCI, and VaMCI. Following adaptation and harmonization of the ICMR-NCTB across 5 different Indian languages into a composite Z score, its test performance against standards, including sensitivity and specificity of the instrument as well as of its subcomponents in diagnosis of MCI, was evaluated in age and education unmatched and matched groups. Results: Variability in sensitivity-specificity estimates was noted between languages when a total of 991 controls and 205 patients with MCI (157 MCI and 48 VaMCI) were compared due to a significant impact of age, education, and language. Data from a total of 506 controls, 144 patients with MCI, and 46 patients with VaMCI who were age- and education-matched were compared. Post hoc analysis after correction for multiple comparisons revealed better performance in controls relative to all-cause MCI. An optimum composite Z-score of −0.541 achieved a sensitivity of 81.1% and a specificity of 88.8% for diagnosis of all-cause MCI, with a high specificity for diagnosis of VaMCI. Using combinations of multiple-domain 2 test subcomponents retained a sensitivity and specificity of >80% for diagnosis of MCI. Conclusions: The ICMR-NCTB is a “first of its kind” approach at harmonizing neuropsychological tests across 5 Indian languages for the diagnosis of MCI due to vascular and other etiologies. Utilizing multiple-domain subcomponents also retains the validity of this instrument, making it a valuable tool in MCI research in multilingual settings.

Download Full-text

European Coal and Steel Community

International Organization ◽

10.1017/s0020818300031039 ◽

1955 ◽

Vol 9 (4) ◽

pp. 595-596

Keyword(s):

Medical Research ◽

The Other ◽

Living Standard ◽

Association Agreement ◽

Housing Projects ◽

The Third ◽

The United Kingdom ◽

The Common ◽

Foreign Ministers ◽

Opening Address

Common AssemblyThe third ordinary session of the Common Assembly of the European Coal and Steel Community (ECSC) reconvened from June 21 to 24, 1955. In his opening address Mr. Rene Mayer, newly elected president of the High Authority of the ECSC, expressed general agreement with the policy resolutions passed by the Common Assembly at its May meeting, and specifically mentioned resolutions on cartels, on implementation of the association agreement with the United Kingdom, and on action to improve the living standard of the workers in the ECSC countries, in regard to which he announced that the High Authority would shortly conclude two loans to finance new workers' housing projects–one of $4,000,000 in Belgium and the other of just over $4,000,000 in Germany. In addition, $300,000 were to be allocated for medical research. In commenting on the Messina conference of foreign ministers, Mr. Mayer said that the High Authority welcomed the decision to explore means of extending the single market. At the same time, he warned the Assembly to expect resistance to changes as the Community developed.

Download Full-text

A group-specific prior distribution for effect-size heterogeneity in meta-analysis

Behavior Research Methods ◽

10.3758/s13428-020-01382-8 ◽

2020 ◽

Vol 52 (5) ◽

pp. 2020-2030

Author(s):

Christopher G. Thompson ◽

Betsy Jane Becker

Keyword(s):

Effect Size ◽

Prior Distribution ◽

Meta Analysis ◽

Size Heterogeneity

Download Full-text

MON-673 Use of Metabolic Syndrome Severity to Assess Treatment with Vitamin-E and Pioglitazone for Non-Alcoholic Steatohepatitis

Journal of the Endocrine Society ◽

10.1210/jendso/bvaa046.515 ◽

2020 ◽

Vol 4 (Supplement_1) ◽

Author(s):

Mark Daniel DeBoer ◽

Jasmine A Mack ◽

Matthew J Gurka

Keyword(s):

Metabolic Syndrome ◽

Vitamin E ◽

Successful Treatment ◽

Effect Size ◽

Continuous Measure ◽

Z Score ◽

Metabolic Derangement ◽

Alcoholic Fatty Liver ◽

Alcoholic Steatohepatitis ◽

The Metabolic Syndrome

Abstract BACKGROUND: Non-alcoholic steatohepatitis (NASH) represents inflammatory and fibrotic changes in the setting of non-alcoholic fatty liver disease (NAFLD) and can progress to cirrhosis. While clinical management of NASH has proven difficult, the Pioglitazone, Vitamin E or Placebo NASH study (PIVENS) demonstrated that treatment with either pioglitazone or vitamin-E increased odds of NASH-resolution. NASH is strongly associated with insulin resistance and the metabolic syndrome (MetS) as both a predictor and an outcome, though this has only been studied using dichotomous MetS criteria (e.g. ATP-III). We previously formulated a sex- and race/ethnicity-specific MetS severity Z-score (MetS-Z) that serves as a continuous measure of metabolic dysregulation. We hypothesized: 1) there would be a decrease in severity of MetS over the course of intervention in PIVENS and 2) the degree of decrease in MetS-Z early in the course of treatment would be a predictor of future NASH resolution. METHODS: Participants in PIVENS (n=201) had biopsy-confirmed NASH at baseline and were randomized to receive pioglitazone, vitamin E or placebo for 96 weeks, when they received repeat biopsy to assess for NASH resolution. We compared levels of MetS-Z and its standardized effect size (the absolute observed difference in MetS-Z for an individual divided by the overall baseline standard deviation of MetS-Z) at baseline, 48 weeks and 96 weeks and used logistic regression to determine associations between baseline MetS and the change in MetS from 0–48 weeks on ultimate NASH resolution—both overall and by intervention group. RESULTS: During the 96 weeks of intervention, 73 participants (363%) exhibited NASH resolution. Baseline MetS-Z was inversely associated with odds of NASH resolution, such that those with higher MetS severity at baseline were less likely to experience NASH resolution (odds ratio [OR] per 1-SD of MetS-Z-score: 0.54, 95% confidence interval [CI] 0.33,0.88). Of the three intervention groups, the decrease in MetS-Z during initial 48 weeks of intervention was greatest for pioglitazone treatment (effect-size: -0.31, CI -0.15,-0.48). During treatment with vitamin E, those with significant 48-week changes in MetS-Z tended to be those with vs. without ultimate NASH resolution (-0.18 vs. -0.05). In the group overall, 48-week change in MetS-Z was inversely associated with NASH resolution (OR of per 1-SD change: 0.56, CI 0.35,0.88). CONCLUSION: Individuals with more severe metabolic derangement at baseline were less likely to exhibit NASH resolution, suggesting that individuals may have a threshold of MetS-severity beyond which successful treatment is unlikely. As hypothesized, a decrease in MetS-Z over time was associated with improved odds of NASH resolution. As an integrated marker of metabolic abnormalities, MetS-Z may be a new way non-invasively follow for successful treatment of NASH.

Download Full-text

Twenty-year experience of examining biopsies of signal lymph nodes in breast cancer

Tumors of female reproductive system ◽

10.17650/1994-4098-2020-16-1-12-20 ◽

2020 ◽

Vol 16 (1) ◽

pp. 12-20 ◽

Cited By ~ 1

Author(s):

V. F. Semiglazov ◽

P. V. Krivorotko ◽

E. K. Zhiltsova ◽

S. V. Kanaev ◽

E. S. Trufanova ◽

...

Keyword(s):

Breast Cancer ◽

Lymph Nodes ◽

Diagnostic Accuracy ◽

Medical Research ◽

Systemic Therapy ◽

False Positive ◽

Ministry Of Health ◽

National Medical ◽

Neoadjuvant Systemic Therapy ◽

Oncology Center

Biopsy of signal (sentinel) lymph nodes (LN) has been performed at the N.N. Petrov National Medical Research Oncology Center, Ministry of Health of Russia, for almost 20 years. In the first few years, contrast-visual method (1 % blue isosulfan and triphenyl methane control) was used in 640 patients with early (T1–2N0M0) breast cancer. In 150 patients from this cohort, standard axillary dissection was performed irrespectively of the results of signal (sentinel) biopsy. The rate of false positive responses varied between 4.6 and 6.6 %. Since 2012, radioisotope method of visualization of signal LN with intratumor administration of 99mТс-technephyte colloid particles has been used (in 708 patients with T1–3N0M0 breast cancer). This type of signal LN biopsy had the following diagnostic characteristics: sensitivity was 58.9 %, specificity was 96.2 %, diagnostic accuracy was 87.1 %.In parallel with this study, in 2016 a study of diagnostic accuracy and safety of biopsy of axillary LN after neoadjuvant systemic therapy was started. The study included 263 patients with T1N1–3M0, T2–3N0–3M0, T4N0–1M0 breast cancer. To evaluate clinical status of axillary LN, ultrasound, single-photon emission computed tomography, mammography at baseline and after completion of neoadjuvant chemotherapy ± targeted therapy (trastuzumab) were performed. In some patients, in the recent years a double method of signal LN labeling (radioisotope and fluorescent methods) was used.In patients with baseline cN+-status, the rate of false positive signal LN biopsy conclusions was 13.6 %, in patients with baseline cN0-status it was 7.7 %.The study of double contrast of axillary LN and targeted label of metastatic LN prior to neoadjuvant systemic therapy continues. In total, various modifications of biopsy of signal LN were performed in 2,000 patients with breast cancer.The study protocol was approved by the biomedical ethics committee of N.N. Petrov National Medical Research Oncology Center, Ministry of Health of Russia.All patients gave written informed consent to participate in the study.

Download Full-text

Analysis of Bayesian posterior significance and effect size indices for the two-sample t-test to support reproducible medical research

BMC Medical Research Methodology ◽

10.1186/s12874-020-00968-2 ◽

2020 ◽

Vol 20 (1) ◽

Cited By ~ 4

Author(s):

Riko Kelter

Keyword(s):

Medical Research ◽

Effect Size ◽

T Test

Download Full-text

A Robust Effect Size Index

Psychometrika ◽

10.1007/s11336-020-09698-2 ◽

2020 ◽

Vol 85 (1) ◽

pp. 232-246 ◽

Cited By ~ 5

Author(s):

Simon Vandekar ◽

Ran Tao ◽

Jeffrey Blume

Keyword(s):

Sample Size ◽

Effect Size ◽

Parametric Models ◽

Association Strength ◽

Measures Of Association ◽

Finite Samples ◽

Wide Range ◽

Size Index ◽

Effect Size Index ◽

Log Odds

AbstractEffect size indices are useful tools in study design and reporting because they are unitless measures of association strength that do not depend on sample size. Existing effect size indices are developed for particular parametric models or population parameters. Here, we propose a robust effect size index based on M-estimators. This approach yields an index that is very generalizable because it is unitless across a wide range of models. We demonstrate that the new index is a function of Cohen’s d, $$R^2$$ R 2 , and standardized log odds ratio when each of the parametric models is correctly specified. We show that existing effect size estimators are biased when the parametric models are incorrect (e.g., under unknown heteroskedasticity). We provide simple formulas to compute power and sample size and use simulations to assess the bias and standard error of the effect size estimator in finite samples. Because the new index is invariant across models, it has the potential to make communication and comprehension of effect size uniform across the behavioral sciences.

Download Full-text