Bivariate Random-effects Meta-analysis of Sensitivity and Specificity with SAS PROC GLIMMIX

2010 ◽  
Vol 49 (01) ◽  
pp. 54-64 ◽  
Author(s):  
J. Menke
Keyword(s):  
Sensitivity And Specificity ◽  
Random Effects ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Meta Analysis ◽  
Empirical Evaluation ◽  
Test Accuracy ◽  
Random Effects Model ◽  
Operating Characteristics ◽  
Meta Analyses

Summary Objectives: Meta-analysis allows to summarize pooled sensitivities and specificities from several primary diagnostic test accuracy studies. Often these pooled estimates are indirectly obtained from a hierarchical summary receiver operating characteristics (HSROC) analysis. This article presents a generalized linear random-effects model with the new SAS PROC GLIMMIX that obtains the pooled estimates for sensitivity and specificity directly. Methods: Firstly, the formula of the bivariate random-effects model is presented in context with the literature. Then its implementation with the new SAS PROC GLIMMIX is empirically evaluated in comparison to the indirect HSROC approach, utilizing the published 2 x 2 count data of 50 meta-analyses. Results: According to the empirical evaluation the meta-analytic results from the bivariate GLIMMIX approach are nearly identical to the results from the indirect HSROC approach. Conclusions: A generalized linear mixed model with PROC GLIMMIX offers a straightforward method for bivariate random-effects meta-analysis of sensitivity and specificity.

scholarly journals On estimating a constrained bivariate random effects model for meta-analysis of test accuracy studies

2022 ◽  
pp. 096228022110651
Author(s):  
Mohammed Baragilly ◽  
Brian Harvey Willis
Keyword(s):  
Sensitivity And Specificity ◽  
Random Effects ◽  
Coverage Probability ◽  
Mean Squared Error ◽  
Meta Analysis ◽  
Test Accuracy ◽  
Random Effects Model ◽  
Mahalanobis Distances ◽  
Constrained Model ◽  
True Values

Tailored meta-analysis uses setting-specific knowledge for the test positive rate and disease prevalence to constrain the possible values for a test's sensitivity and specificity. The constrained region is used to select those studies relevant to the setting for meta-analysis using an unconstrained bivariate random effects model (BRM). However, sometimes there may be no studies to aggregate, or the summary estimate may lie outside the plausible or “applicable” region. Potentially these shortcomings may be overcome by incorporating the constraints in the BRM to produce a constrained model. Using a penalised likelihood approach we developed an optimisation algorithm based on co-ordinate ascent and Newton-Raphson iteration to fit a constrained bivariate random effects model (CBRM) for meta-analysis. Using numerical examples based on simulation studies and real datasets we compared its performance with the BRM in terms of bias, mean squared error and coverage probability. We also determined the ‘closeness’ of the estimates to their true values using the Euclidian and Mahalanobis distances. The CBRM produced estimates which in the majority of cases had lower absolute mean bias and greater coverage probability than the BRM. The estimated sensitivities and specificity for the CBRM were, in general, closer to the true values than the BRM. For the two real datasets, the CBRM produced estimates which were in the applicable region in contrast to the BRM. When combining setting-specific data with test accuracy meta-analysis, a constrained model is more likely to yield a plausible estimate for the sensitivity and specificity in the practice setting than an unconstrained model.

Systematic Review of Second Primary Oropharyngeal Cancers in Patients With p16+ Oropharyngeal Cancer

2020 ◽  
pp. 019459982095117
Author(s):  
Craig A. Bollig ◽  
David S. Lee ◽  
Angela L. Mazul ◽  
Katelyn Stepan ◽  
Sidharth V. Puram ◽  
...  
Keyword(s):  
Surgical Treatment ◽  
Random Effects ◽  
Meta Analysis ◽  
Second Primary ◽  
Random Effects Model ◽  
Treatment Information ◽  
Pooled Prevalence ◽  
Alive With Disease ◽  
Medical Librarian ◽  
Meta Analyses

Objective To systematically review the literature to determine the prevalence and clinical outcomes of second primary oropharyngeal squamous cell carcinoma (OPSCC). Data Sources Search strategies created with a medical librarian were implemented using multiple databases in October 2019. Review Methods The population of interest included adults age >18 years with a p16+ or human papillomavirus-positive OPSCC. The outcome was a synchronous or metachronous second primary OPSCC. Inclusion and exclusion criteria were designed to capture all study designs. In total, 685 records were identified by the search strategy. Two reviewers independently performed the review, extracted data, and performed a quality assessment. Primary Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. A random-effects model was used for the meta-analysis. Results A total of 2470 patients with 35 second primary OPSCCs from 15 studies were identified. The pooled prevalence of second primary OPSCC was 1.4% (range, 0%-14.3%). In the random-effects model, the prevalence was estimated at 1.3% (95% CI, 0.7%-2.3%; P = .51, I2 = 52%). Of the 30 patients with treatment information, 26 (86.7%) received surgical treatment, while 4 (13.3%) underwent nonsurgical therapy. Of the 29 patients with available survival information, 22 (75.9%) had no evidence of disease at last follow-up, 5 (17.2%) ultimately died of disease, and 2 (6.9%) were alive with disease. Conclusion Overall, the rate of second primary OPSCC in patients with an index p16+ OPSCC is low, and most patients are successfully treated. Insufficient evidence currently exists to recommend routine elective tonsillectomy during surgical treatment of p16+ OPSCC.

scholarly journals High burden of acute kidney injury in COVID-19 pandemic: systematic review and meta-analysis

2020 ◽  
pp. jclinpath-2020-207023
Author(s):  
Camila Barbosa Oliveira ◽  
Camilla Albertina Dantas Lima ◽  
Gisele Vajgel ◽  
Antonio Victor Campos Coelho ◽  
Paula Sandrin-Garcia
Keyword(s):  
Systematic Review ◽  
Acute Kidney Injury ◽  
Critically Ill ◽  
Random Effects ◽  
Meta Analysis ◽  
Kidney Injury ◽  
Random Effects Model ◽  
Risk Of Death ◽  
Meta Analyses ◽  
The Impact

AimsHospitalised patients with COVID-19 have a variable incidence of acute kidney injury (AKI) according to studies from different nationalities. The present systematic review and meta-analysis describes the incidence of AKI, need for renal replacement therapy (RRT) and mortality among patients with COVID-19-associated AKI.MethodsWe systematically searched electronic database PubMed, SCOPUS and Web of Science to identify English articles published until 25 May 2020. In case of significant heterogeneity, the meta-analyses were conducted assuming a random-effects model.ResultsFrom 746 screened publications, we selected 21 observational studies with 15 536 patients with COVID-19 for random-effects model meta-analyses. The overall incidence of AKI was 12.3% (95% CI 7.3% to 20.0%) and 77% of patients with AKI were critically ill (95% CI 58.9% to 89.0%). The mortality among patients with AKI was 67% (95% CI 39.8% to 86.2%) and the risk of death was 13 times higher compared with patients without AKI (OR=13.3; 95% CI 6.1 to 29.2). Patients with COVID-19-associated AKI needed for RRT in 23.4% of cases (95% CI 12.6% to 39.4%) and those cases had high mortality (89%–100%).ConclusionThe present study evidenced an incidence of COVID-19-associated AKI higher than previous meta-analysis. The majority of patients affected by AKI were critically ill and mortality rate among AKI cases was high. Thus, it is extremely important for health systems to be aware about the impact of AKI on patients’ outcomes in order to establish proper screening, prevention of additional damage to the kidneys and adequate renal support when needed.

Surgical Outcomes of Endoscopic Versus Open Resection for Primary Sinonasal Malignancy: A Meta-analysis

2019 ◽  
Vol 33 (5) ◽  
pp. 608-616 ◽  
Author(s):  
Victor M. Lu ◽  
Krishnan Ravindran ◽  
Kevin Phan ◽  
Jamie J. Van Gompel ◽  
Timothy R. Smith ◽  
...  
Keyword(s):  
Random Effects ◽  
Surgical Outcomes ◽  
Meta Analysis ◽  
Random Effects Model ◽  
Limited Evidence ◽  
Open Resection ◽  
Sinonasal Malignancies ◽  
Meta Analyses ◽  
Risk Ratios ◽  
Indirect Comparisons

Background Endoscopic resection (ER) for uncommon sinonasal malignancies (SNMs) has been reported to confer superior surgical outcomes compared to open resection (OR) based on indirect comparisons of limited evidence. Objective The aim of this study was to pool all direct comparative studies in the literature to validate this potential superior association. Methods Systematic searches of 7 electronic databases from their inception to April 2019 were conducted following Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines. There were 1001 articles identified for screening. Outcomes of interest were pooled as risk ratios (RRs) and mean difference (MD) and analyzed using a random-effects model. Results There were 10 studies included in this meta-analysis, with 900 SNM patients in total where ER and OR were utilized in 399 (44%) and 501 (56%) cases, respectively. Compared to OR, random-effects (RE) modeling indicated ER resulted in statistically comparable complications (RR = 0.68; P-effect = .12) and recurrence (RR = 0.84; P-effect = .35). ER was associated with significantly shorter length of stay (LOS) compared to OR (MD = −2.9 days; P-effect <.01). Conclusions The use of ER to manage SNM was associated with significantly favorable reduction in LOS compared to OR. However, with respect to other surgical outcomes and recurrence, the current literature does not indicate either ER or OR as statistically superior. Therefore, until greater validation of these associations can be proven, expectations that ER for SNMs confers superior surgical outcomes compared to OR should be tempered.

scholarly journals Bivariate Random Effects Meta-Analysis of Diagnostic Studies Using Generalized Linear Mixed Models

2009 ◽  
Vol 30 (4) ◽  
pp. 499-508 ◽  
Author(s):  
Haitao Chu ◽  
Hongfei Guo ◽  
Yijie Zhou
Keyword(s):  
Sensitivity And Specificity ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Random Effect ◽  
Point Estimation ◽  
Test Accuracy ◽  
Logit Transformation ◽  
Link Functions ◽  
Special Cases ◽  
The Impact

Bivariate random effect models are currently one of the main methods recommended to synthesize diagnostic test accuracy studies. However, only the logit transformation on sensitivity and specificity has been previously considered in the literature. In this article, the authors consider a bivariate generalized linear mixed model to jointly model the sensitivities and specificities, and they discuss the estimation of the summary receiver operating characteristic curve (ROC) and the area under the ROC curve (AUC). As the special cases of this model, the authors discuss the commonly used logit, probit, and complementary log-log transformations. To evaluate the impact of misspecification of the link functions on the estimation, they present 2 case studies and a set of simulation studies. Their study suggests that point estimation of the median sensitivity and specificity and AUC is relatively robust to the misspecification of the link functions. However, the misspecification of link functions has a noticeable impact on the standard error estimation and the 95% confidence interval coverage, which emphasizes the importance of choosing an appropriate link function to make statistical inference.

Robust bivariate random-effects model for accommodating outlying and influential studies in meta-analysis of diagnostic test accuracy studies

2020 ◽  
Vol 29 (11) ◽  
pp. 3308-3325
Author(s):  
Zelalem F Negeri ◽  
Joseph Beyene
Keyword(s):  
Diagnostic Test ◽  
Random Effects ◽  
Meta Analysis ◽  
Diagnostic Test Accuracy ◽  
Marginal Model ◽  
Test Accuracy ◽  
Random Effects Model ◽  
Random Effects Models ◽  
Proposed Model ◽  
Standard Models

Due to the inevitable inter-study correlation between test sensitivity (Se) and test specificity (Sp), mostly because of threshold variability, hierarchical or bivariate random-effects models are widely used to perform a meta-analysis of diagnostic test accuracy studies. Conventionally, these models assume that the random-effects follow the bivariate normal distribution. However, the inference made using the well-established bivariate random-effects models, when outlying and influential studies are present, may lead to misleading conclusions, since outlying or influential studies can extremely influence parameter estimates due to their disproportional weight. Therefore, we developed a new robust bivariate random-effects model that accommodates outlying and influential observations and gives robust statistical inference by down-weighting the effect of outlying and influential studies. The marginal model and the Monte Carlo expectation-maximization algorithm for our proposed model have been derived. A simulation study has been carried out to validate the proposed method and compare it against the standard methods. Regardless of the parameters varied in our simulations, the proposed model produced robust point estimates of Se and Sp compared to the standard models. Moreover, our proposed model resulted in precise estimates as it yielded the narrowest confidence intervals. The proposed model also generated a similar point and interval estimates of Se and Sp as the standard models when there are no outlying and influential studies. Two published meta-analyses have also been used to illustrate the methods.

scholarly journals The impact of studies with no events in both arms on meta-analysis of rare events: a simulation study using generalized linear mixed model

2021 ◽  
Author(s):  
Chang Xu ◽  
Lifeng Lin
Keyword(s):  
Effect Size ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Mean Squared Error ◽  
Meta Analysis ◽  
Unbiased Estimation ◽  
Double Zero ◽  
Squared Error ◽  
Meta Analyses ◽  
The Impact

AbstractObjectiveThe common approach to meta-analysis with double-zero studies is to remove such studies. Our previous work has confirmed that exclusion of these studies may impact the results. In this study, we undertook extensive simulations to investigate how the results of meta-analyses would be impacted in relation to the proportion of such studies.MethodsTwo standard generalized linear mixed models (GLMMs) were employed for the meta-analysis. The statistical properties of the two GLMMs were first examined in terms of percentage bias, mean squared error, and coverage. We then repeated all the meta-analyses after excluding double-zero studies. Direction of estimated effects and p-values for including against excluding double-zero studies were compared in nine ascending groups classified by the proportion of double-zero studies within a meta-analysis.ResultsBased on 50,000 simulated meta-analyses, the two GLMMs almost achieved unbiased estimation and reasonable coverage in most of the situations. When excluding double-zero studies, 0.00% to 4.47% of the meta-analyses changed the direction of effect size, and 0.61% to 8.78% changed direction of the significance of p-value. When the proportion of double-zero studies increased in a meta-analysis, the probability of the effect size changed the direction increased; when the proportion was about 40% to 60%, it has the largest impact on the change of p-values.ConclusionDouble-zero studies can impact the results of meta-analysis and excluding them may be problematic. The impact of such studies on meta-analysis varies by the proportion of such studies within a meta-analysis.

scholarly journals Efficacy of Corticosteroids in Patients with SARS, MERS and COVID-19: A Systematic Review and Meta-Analysis

2020 ◽  
Vol 9 (8) ◽  
pp. 2392 ◽  
Author(s):  
Keum Hwa Lee ◽  
Sojung Yoon ◽  
Gwang Hun Jeong ◽  
Jong Yeob Kim ◽  
Young Joo Han ◽  
...  
Keyword(s):  
Systematic Review ◽  
Random Effects ◽  
Meta Analysis ◽  
Random Effect Model ◽  
Control Group ◽  
Random Effects Model ◽  
Number Of Patients ◽  
Significant Difference ◽  
Meta Analyses ◽  
Coronavirus Infections

(1) Background: The use of corticosteroids in critical coronavirus infections, including severe acute respiratory syndrome (SARS), Middle East Respiratory Syndrome (MERS), or Coronavirus disease 2019 (COVID-19), has been controversial. However, a meta-analysis on the efficacy of steroids in treating these coronavirus infections is lacking. (2) Purpose: We assessed a methodological criticism on the quality of previous published meta-analyses and the risk of misleading conclusions with important therapeutic consequences. We also examined the evidence of the efficacy of corticosteroids in reducing mortality in SARS, MERS and COVID-19. (3) Methods: PubMed, MEDLINE, Embase, and Web of Science were used to identify studies published until 25 April 2020, that reported associations between steroid use and mortality in treating SARS/MERS/COVID-19. Two investigators screened and extracted data independently. Searches were restricted to studies on humans, and articles that did not report the exact number of patients in each group or data on mortality were excluded. We calculated odds ratios (ORs) or hazard ratios (HRs) under the fixed- and random-effect model. (4) Results: Eight articles (4051 patients) were eligible for inclusion. Among these selected studies, 3416 patients were diagnosed with SARS, 360 patients with MERS, and 275 with COVID-19; 60.3% patients were administered steroids. The meta-analyses including all studies showed no differences overall in terms of mortality (OR 1.152, 95% CI 0.631–2.101 in the random effects model, p = 0.645). However, this conclusion might be biased, because, in some studies, the patients in the steroid group had more severe symptoms than those in the control group. In contrast, when the meta-analysis was performed restricting only to studies that used appropriate adjustment (e.g., time, disease severity), there was a significant difference between the two groups (HR 0.378, 95% CI 0.221–0.646 in the random effects model, p < 0.0001). Although there was no difference in mortality when steroids were used in severe cases, there was a difference among the group with more underlying diseases (OR 3.133, 95% CI 1.670–5.877, p < 0.001). (5) Conclusions: To our knowledge, this study is the first comprehensive systematic review and meta-analysis providing the most accurate evidence on the effect of steroids in coronavirus infections. If not contraindicated, and in the absence of side effects, the use of steroids should be considered in coronavirus infection including COVID-19.

Exploring the Relationships Between Problem Gambling and ADHD: A Meta-Analysis

2016 ◽  
Vol 23 (12) ◽  
pp. 1427-1437 ◽  
Author(s):  
Jennifer Theule ◽  
Kylee E. Hurl ◽  
Kristene Cheung ◽  
Michelle Ward ◽  
Brenna Henrikson
Keyword(s):  
Confidence Interval ◽  
Problem Gambling ◽  
Random Effects ◽  
Meta Analysis ◽  
Random Effects Model ◽  
Adhd Symptoms ◽  
Weighted Mean ◽  
Problem Gamblers ◽  
Gambling Severity ◽  
Meta Analyses

Objective: At present, there are inconsistencies in the literature pertaining to the association between ADHD and problem gambling. This study utilized meta-analytic techniques to clarify the association between symptoms of problem gambling and symptoms of ADHD. Method: Several meta-analyses were conducted using a random effects model. PsycINFO, PubMed, ProQuest Dissertations & Theses, and Google Scholar were searched for relevant studies. Results: The weighted mean correlation between ADHD symptomology and gambling severity was r = .17, 95% confidence interval (CI) = [0.12, 0.22], p < .001. Mean age of the sample was the only moderator to approach significance, with greater age being linked to a stronger relationship between symptoms of ADHD and gambling severity. Conclusion: Clinicians needs to be cognizant of the greater risk of ADHD symptoms when working with problem gamblers and vice versa.

Sign in / Sign up

Export Citation Format

Share Document