Assessing the dependence of sensitivity and specificity on prevalence in meta-analysis.
Ontology highlight
ABSTRACT:
SUBMITTER: Li J
PROVIDER: S-EPMC4042906 | biostudies-other | 2011 Oct
REPOSITORIES: biostudies-other
Publications
Assessing the dependence of sensitivity and specificity on prevalence in meta-analysis.
Biostatistics (Oxford, England) 20110427 4
We consider modeling the dependence of sensitivity and specificity on the disease prevalence in diagnostic accuracy studies. Many meta-analyses compare test accuracy across studies and fail to incorporate the possible connection between the accuracy measures and the prevalence. We propose a Pearson type correlation coefficient and an estimating equation-based regression framework to help understand such a practical dependence. The results we derive may then be used to better interpret the result ...[more]
Similar Datasets
Project description:BackgroundAnecdotal evidence suggests that the sensitivity and specificity of a diagnostic test may vary with disease prevalence. Our objective was to investigate the associations between disease prevalence and test sensitivity and specificity using studies of diagnostic accuracy.MethodsWe used data from 23 meta-analyses, each of which included 10-39 studies (416 total). The median prevalence per review ranged from 1% to 77%. We evaluated the effects of prevalence on sensitivity and specificity using a bivariate random-effects model for each meta-analysis, with prevalence as a covariate. We estimated the overall effect of prevalence by pooling the effects using the inverse variance method.ResultsWithin a given review, a change in prevalence from the lowest to highest value resulted in a corresponding change in sensitivity or specificity from 0 to 40 percentage points. This effect was statistically significant (p < 0.05) for either sensitivity or specificity in 8 meta-analyses (35%). Overall, specificity tended to be lower with higher disease prevalence; there was no such systematic effect for sensitivity.InterpretationThe sensitivity and specificity of a test often vary with disease prevalence; this effect is likely to be the result of mechanisms, such as patient spectrum, that affect prevalence, sensitivity and specificity. Because it may be difficult to identify such mechanisms, clinicians should use prevalence as a guide when selecting studies that most closely match their situation.
| S-EPMC3735771 | biostudies-literature
Project description:Determine age-specific infection fatality rates for COVID-19 to inform public health policies and communications that help protect vulnerable age groups. Studies of COVID-19 prevalence were collected by conducting an online search of published articles, preprints, and government reports that were publicly disseminated prior to 18 September 2020. The systematic review encompassed 113 studies, of which 27 studies (covering 34 geographical locations) satisfied the inclusion criteria and were included in the meta-analysis. Age-specific IFRs were computed using the prevalence data in conjunction with reported fatalities 4 weeks after the midpoint date of the study, reflecting typical lags in fatalities and reporting. Meta-regression procedures in Stata were used to analyze the infection fatality rate (IFR) by age. Our analysis finds a exponential relationship between age and IFR for COVID-19. The estimated age-specific IFR is very low for children and younger adults (e.g., 0.002% at age 10 and 0.01% at age 25) but increases progressively to 0.4% at age 55, 1.4% at age 65, 4.6% at age 75, and 15% at age 85. Moreover, our results indicate that about 90% of the variation in population IFR across geographical locations reflects differences in the age composition of the population and the extent to which relatively vulnerable age groups were exposed to the virus. These results indicate that COVID-19 is hazardous not only for the elderly but also for middle-aged adults, for whom the infection fatality rate is two orders of magnitude greater than the annualized risk of a fatal automobile accident and far more dangerous than seasonal influenza. Moreover, the overall IFR for COVID-19 should not be viewed as a fixed parameter but as intrinsically linked to the age-specific pattern of infections. Consequently, public health measures to mitigate infections in older adults could substantially decrease total deaths.
| S-EPMC7721859 | biostudies-literature
Project description:Connectivity of terrestrial deep biosphere microbial communities with surface communities
| PRJEB41312 | ENA
Project description:Introduction:Atrial fibrillation (AF) is the commonest arrhythmia and a major cause of stroke and health care utilization. Researchers and administrators use electronic health data to assess disease burden, quality and variance in care, value of interventions and prognosis. We performed a systematic review and meta-analysis to assess the validity of AF case definitions in administrative databases. Methods:Medline was searched from 2000 to 2018. Extracted information included sensitivity, specificity, positive and negative predictive values (PPV and NPV) for various AF case definitions. Estimates were pooled using random-effects models due to significant heterogeneity between studies. Results:We identified 24 studies, including 21 from North America or Scandinavia. Hospital, ambulatory and mixed data sources were assessed in 10, 4 and 10 studies, respectively. Nine different AF case definitions were evaluated, most based on ICD-9 or 10 codes. Twenty-two studies assessed case definitions in patients diagnosed with AF and thus could generate PPV alone. Half the studies sampled unrestricted populations including a mix of those with and without AF to assess sensitivity. Only 13 studies included ECG confirmation as a gold standard. The pooled random effects estimates were: sensitivity 80% (95% CI 72-86%); specificity 98% (96-99%); PPV 88% (82-94%); NPV 97% (94-99%). Only 3 studies reported all accuracy parameters and included rhythm monitoring in the gold standard definition. Conclusion:Relatively few studies examined sensitivity, and fewer still included rhythm monitoring in the gold standard comparison. Administrative data may fail to identify a significant proportion of patients with AF. This, in turn, may bias estimates of quality of care and prognosis.
| S-EPMC6712502 | biostudies-literature
Project description:BackgroundConventional meta-analysis based on genetic markers may be less powerful for heterogeneous samples. In this study, we introduced a new meta-analysis for 4 genomewide association studies on alcohol dependence that integrated the information of putative causal variants.MethodsA total of 12,481 subjects in 4 independent cohorts were analyzed, including 1 European American cohort (1,409 cases with alcohol dependence and 1,518 controls), 1 European Australian cohort (a total of 6,438 family subjects with 1,645 probands), 1 African American cohort from SAGE + COGA (681 cases and 508 controls), and 1 African American cohort from Yale (1,429 cases and 498 controls). The genomewide association analysis was conducted for each cohort, and then, a new meta-analysis was performed to derive the combined p-values. cis-Acting expression of quantitative locus (cis-eQTL) analysis of each risk variant in human tissues and RNA expression analysis of each risk gene in rat brain served as functional validation.ResultsIn meta-analysis of European American and European Australian cohorts, we found 10 top-ranked single nucleotide polymorphisms (SNPs) (p < 10(-6) ) that were associated with alcohol dependence. They included 6 at SERINC2 (3.1 × 10(-8) ≤ p ≤ 9.6 × 10(-8) ), 1 at STK40 (p = 1.3 × 10(-7) ), 2 at KIAA0040 (3.3 × 10(-7) ≤ p ≤ 5.2 × 10(-7) ), and 1 at IPO11 (p = 6.9 × 10(-7) ). In meta-analysis of 2 African American cohorts, we found 2 top-ranked SNPs including 1 at SLC6A11 (p = 2.7 × 10(-7) ) and 1 at CBLN2 (p = 7.4 × 10(-7) ). In meta-analysis of all 4 cohorts, we found 2 top-ranked SNPs in PTP4A1-PHF3 locus (6.0 × 10(-7) ≤ p ≤ 7.2 × 10(-7) ). In an African American cohort only, we found 1 top-ranked SNP at PLD1 (p = 8.3 × 10(-7) ; OR = 1.56). Many risk SNPs had positive cis-eQTL signals, and all these risk genes except KIAA0040 were found to express in both rat and mouse brains.ConclusionsWe found multiple genes that were significantly or suggestively associated with alcohol dependence. They are among the most appropriate for follow-up as contributors to risk for alcohol dependence.
| S-EPMC5587504 | biostudies-literature
Project description:BackgroundAtrial fibrillation (AF) is a prevalent and preventable cause of stroke and mortality.AimThis systematic review and meta-analysis aimed to investigate the sensitivity and specificity of office and out-of-office automated blood pressure (BP) devices to detect AF.MethodsDiagnostic studies, extracted from databases such as Ovid Medline and Embase, on AF detection by BP device(s), electrocardiography, and reported sensitivity and specificity, were included. Screening of abstracts and full texts, data extraction, and quality assessment were conducted independently by two investigators using Covidence software. The sensitivity and specificity of the BP devices were pooled using a random-effects model.ResultsSixteen studies including 10,158 participants were included. Only a few studies were conducted in primary care (n = 3) or with a low risk of bias (n = 5). Office BP devices, which utilised different algorithms to detect AF, had a sensitivity and specificity of 96.2 and 94%, respectively. Specificity was reduced when only one positive result was considered among consecutive BP measurements. Only a few studies (n = 3) investigated out-of-office BP. Only one study (n = 100) suggested the use of ≥79 and ≥26% of positive readings on 24-h ambulatory BP measurements to detect AF and paroxysmal AF, respectively.ConclusionsOffice BP devices can be used clinically to screen for AF in high-risk populations. Clinical trials are needed to determine the effect of AF screening using office BP devices in reducing stroke risk and mortality. Further studies are also required to guide out-of-office use of BP devices for detecting paroxysmal AF or AF.Systematic review registrationhttps://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022319541, PROSPERO CRD42022319541.
| S-EPMC9411860 | biostudies-literature
Project description:The LuxR protein of the bacterium Vibrio fischeri belongs to a family of transcriptional activators that underlie pheromone-mediated signaling by responding to acyl-homoserine lactones (-HSLs) or related molecules. V. fischeri produces two acyl-HSLs, N-3-oxo-hexanoyl-HSL (3OC6-HSL) and N-octanoyl-HSL (C8-HSL), each of which interact with LuxR to facilitate its binding to a "lux box" DNA sequence, thereby enabling LuxR to activate transcription of the lux operon responsible for bioluminescence. We have investigated the HSL sensitivity of four different variants of V. fischeri LuxR: two derived from wild-type strains ES114 and MJ1, and two derivatives of LuxRMJ1 generated by directed evolution. For each LuxR variant, we measured the bioluminescence induced by combinations of C8-HSL and 3OC6-HSL. We fit these data to a model in which the two HSLs compete with each other to form multimeric LuxR complexes that directly interact with lux to activate bioluminescence. The model reproduces the observed effects of HSL combinations on the bioluminescence responses directed by LuxR variants, including competition and non-monotonic responses to C8-HSL and 3OC6-HSL. The analysis yields robust estimates for the underlying dissociation constants and cooperativities (Hill coefficients) of the LuxR-HSL complexes and their affinities for the lux box. It also reveals significant differences in the affinities of LuxRMJ1 and LuxRES114 for 3OC6-HSL. Further, LuxRMJ1 and LuxRES114 differed sharply from LuxRs retrieved by directed evolution in the cooperativity of LuxR-HSL complex formation and the affinity of these complexes for lux. These results show how computational modeling of in vivo experimental data can provide insight into the mechanistic consequences of directed evolution.
| S-EPMC4427320 | biostudies-literature
Project description:BackgroundRecently, the tuberculosis (TB) Task Force Impact Measurement acknowledged the need to review the assumptions underlying the TB mortality estimates published annually by the World Health Organization (WHO). TB mortality is indirectly measured by multiplying estimated TB incidence with estimated case fatality ratio (CFR). We conducted a meta-analysis to estimate the TB case fatality ratio in TB patients having initiated TB treatment.MethodsWe searched for eligible studies in the PubMed and Embase databases through March 4(th) 2011 and by reference listing of relevant review articles. Main analyses included the estimation of the pooled percentages of: a) TB patients dying due to TB after having initiated TB treatment and b) TB patients dying during TB treatment. Pooled percentages were estimated using random effects regression models on the combined patient population from all studies.Main resultsWe identified 69 relevant studies of which 22 provided data on mortality due to TB and 59 provided data on mortality during TB treatment. Among HIV infected persons the pooled percentage of TB patients dying due to TB was 9.2% (95% Confidence Interval (CI): 3.7%-14.7%) and among HIV uninfected persons 3.0% (95% CI: -1.2%-7.4%) based on the results of eight and three studies respectively providing data for this analyses. The pooled percentage of TB patients dying during TB treatment was 18.8% (95% CI: 14.8%-22.8%) among HIV infected patients and 3.5% (95% CI: 2.0%-4.92%) among HIV uninfected patients based on the results of 27 and 19 studies respectively.ConclusionThe results of the literature review are useful in generating prior distributions of CFR in countries with vital registration systems and have contributed towards revised estimates of TB mortality This literature review did not provide us with all data needed for a valid estimation of TB CFR in TB patients initiating TB treatment.
| S-EPMC3124477 | biostudies-literature
Project description:Antibodies against HPV16 early proteins have been shown to be promising biomarkers for the identification of HPV-driven oropharyngeal cancer (HPV-OPC) among OPC cases in multiple studies. A systematic literature search was performed to identify original research articles comparing HPV early antigen serology with established reference methods to determine molecular HPV tumor status. Random-effects models were used to calculate summary estimates for sensitivity and specificity of HPV16 E2, E6 and E7 serology for HPV-OPC. Subgroup analyses were performed to explore heterogeneity across studies and describe variables associated with test performance. We identified n = 23 studies meeting all eligibility criteria and included these in the meta-analysis. E6 serology showed the best performance with pooled sensitivity and specificity estimates of 83.1% (95% confidence interval (CI) 72.5-90.2%) and 94.6% (95% CI 89.0-97.4%), respectively, while E2 and E7 serological assays were highly specific (E2: 92.5% (95% CI 79.1-97.6%); E7: 88.5% (95% CI 77.9-94.4%)) but moderately sensitive (E2: 67.8% (95% CI 58.9-75.6%); E7: 67.0% (95% CI 63.2-70.6%)). Subgroup analyses revealed increased pooled sensitivity for bacterially (89.9% (95% CI 84.5-93.6%)) vs. in vitro expressed E6 antigen (55.3% (95% CI 41.0-68.7%)), while both showed high specificity (95.2% (95% CI 93.0-96.7%) and 91.1% (95% CI 46.6-99.2%), respectively). Pooled specificity estimates for HPV16 E2, E6 and E7 serology were significantly lower in studies utilizing HPV DNA PCR as the only molecular reference method compared to those using a combination of any two reference methods (HPV DNA, RNA, in situ hybridization (ISH), p16 immunohistochemistry (IHC)), or histopathological reference methods (ISH or p16 IHC) as stand-alone marker. In conclusion, HPV16 E6 seropositivity is a highly sensitive and specific biomarker for HPV-OPC. However, its performance differs between serological assays and depends on molecular reference methods.
| S-EPMC8234521 | biostudies-literature
Project description:Random-effects meta-analyses of observational studies can produce biased estimates if the synthesized studies are subject to unmeasured confounding. We propose sensitivity analyses quantifying the extent to which unmeasured confounding of specified magnitude could reduce to below a certain threshold the proportion of true effect sizes that are scientifically meaningful. We also develop converse methods to estimate the strength of confounding capable of reducing the proportion of scientifically meaningful true effects to below a chosen threshold. These methods apply when a "bias factor" is assumed to be normally distributed across studies or is assessed across a range of fixed values. Our estimators are derived using recently proposed sharp bounds on confounding bias within a single study that do not make assumptions regarding the unmeasured confounders themselves or the functional form of their relationships with the exposure and outcome of interest. We provide an R package, EValue, and a free website that compute point estimates and inference and produce plots for conducting such sensitivity analyses. These methods facilitate principled use of random-effects meta-analyses of observational studies to assess the strength of causal evidence for a hypothesis.
| S-EPMC7518377 | biostudies-literature