-
Computers in Biology and Medicine Dec 2023Artificial intelligence (AI) has potential uses in healthcare including the detection of health conditions and prediction of health outcomes. Past systematic reviews had... (Meta-Analysis)
Meta-Analysis Review
BACKGROUND
Artificial intelligence (AI) has potential uses in healthcare including the detection of health conditions and prediction of health outcomes. Past systematic reviews had reviewed the accuracy of artificial neural networks (ANN) on Electrocardiogram (ECG) readings but that of other AI models on other Acute Coronary Syndrome (ACS) detection tools remains unclear.
METHODS
Nine electronic databases were searched from 2012 to 31 August 2022 including grey literature search and hand searching of references of included articles. Risk of bias was assessed by two independent reviewers using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). Test characteristics namely true positives, false positives, true negatives, and false negatives were extracted from all included articles into a 2x2 table. Study-specific estimates of sensitivity and specificity were pooled using hierarchical summary receiver operating characteristic (HSROC) model and displayed using a forest plot and HSROC curve.
RESULTS
66 studies were included in the review. A total of 518,931 patients were included whose mean ages varied from 32.62 to 70 years old. In 66 studies, the sensitivity and specificity of AI-based detection for ACS screening ranged from 64 % to 100 % and 65 %-100 %, respectively. The overall quality of evidence was low due to the inclusion of case-control studies.
CONCLUSION
Results of the study inform the potential of using AI-assisted ACS detection for accurate diagnosis and prompt treatment for ACS. Adherence to the Standards for Reporting of Diagnostic Accuracy (STARD) guideline and having more cohort studies for future Diagnostic Test Accuracy (DTA) studies are necessary to improve the quality of evidence of AI-based detection of ACS.
Topics: Humans; Adult; Middle Aged; Aged; Artificial Intelligence; Acute Coronary Syndrome; Sensitivity and Specificity; ROC Curve; Diagnostic Tests, Routine
PubMed: 37925910
DOI: 10.1016/j.compbiomed.2023.107636 -
Epilepsy & Behavior : E&B Feb 2020Among people with epilepsy, levetiracetam (LEV) can cause neuropsychiatric adverse events (NPAEs) that impact negatively on quality of life. It has been suggested that...
OBJECTIVE
Among people with epilepsy, levetiracetam (LEV) can cause neuropsychiatric adverse events (NPAEs) that impact negatively on quality of life. It has been suggested that pyridoxine can ameliorate LEV-related NPAEs. We conducted a systematic review of studies on the use of pyridoxine supplementation to relieve NPAEs associated with LEV therapy.
METHODS
The review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Medline, EMBASE, Scholar, Cochrane-CENTRAL (2000-2019), and EThOS platform were searched for studies on the use of pyridoxine in patients with LEV-related NPAEs. Proportions of patients reported to benefit from pyridoxine supplementation were tabulated, and a random-effect model meta-analysis was conducted.
RESULTS
Eleven retrospective studies/case reports and one randomized prospective study, mostly including pediatric populations, were identified. Retrospective studies, which were rated as low quality due to failure to control for bias, reported an overall improvement of NPAEs after pyridoxine supplementation in 72.5% (108/149) of patients. The proportion of patients showing improvement in a pooled analysis of the four largest retrospective studies (n = 134) was 72.1% (95% confidence interval (CI) 47.1-88.3), although there was high heterogeneity across studies (I = 82%, p < 0.01). In the only prospective trial, patients randomized to pyridoxine supplementation were more likely to show relief from NPAEs than patients not receiving supplementation (p < 0.01), but outcomes might have been affected by assessment bias.
CONCLUSION
This systematic review suggests that pyridoxine might be of benefit in relieving LEV-related NPAEs. However, the quality of the evidence is poor, and better-designed prospective studies that include quantitative as well as qualitative data are needed to define the role of pyridoxine in the management of LEV-related NPAEs.
Topics: Anticonvulsants; Diagnostic Tests, Routine; Dietary Supplements; Drug Therapy, Combination; Epilepsy; Humans; Levetiracetam; Mental Disorders; Prospective Studies; Pyridoxine; Quality of Life; Retrospective Studies; Vitamin B Complex
PubMed: 31917143
DOI: 10.1016/j.yebeh.2019.106861 -
Preventive Veterinary Medicine Aug 2016Several rapid tests for use on farm have been validated for the detection of hyperketonemia (HK) in dairy cattle, however the reported sensitivity and specificity of... (Meta-Analysis)
Meta-Analysis Review
Several rapid tests for use on farm have been validated for the detection of hyperketonemia (HK) in dairy cattle, however the reported sensitivity and specificity of each method varies and no single study has compared them all. Meta-analysis of diagnostic test accuracy is becoming more common in human medical literature but there are few veterinary examples. The objective of this work was to perform a systematic review and meta-analysis to determine the point-of-care testing method with the highest combined sensitivity and specificity, the optimal threshold for each method, and to identify gaps in the literature. A comprehensive literature search resulted in 5196 references. After removing duplicates and performing relevance screening, 23 studies were included for the qualitative synthesis and 18 for the meta-analysis. The three index tests evaluated in the meta-analysis were: the Precision Xtra(®) handheld device measuring beta-hydroxybutyrate (BHB) concentration in whole blood, and Ketostix(®) and KetoTest(®) semi-quantitative strips measuring the concentration of acetoacetate in urine and BHB in milk, respectively. The diagnostic accuracy of the 3 index tests relative to the reference standard measurement of BHB in serum or whole blood between 1.0-1.4mmol/L was compared using the hierarchical summary receiver operator characteristic (HSROC) method. Subgroup analysis was conducted for each index test to examine the accuracy at different thresholds. The impact of the reference standard threshold, the reference standard method, the prevalence of HK in the population, the primary study source and risk of bias of the primary study was explored using meta-regression. The Precision Xtra(®) device had the highest summary sensitivity in whole blood BHB at 1.2mmol/L, 94.8% (CI95%: 92.6-97.0), and specificity, 97.5% (CI95%: 96.9-98.1). The threshold employed (1.2-1.4mmol/L) did not impact the diagnostic accuracy of the test. The Ketostix(®) and KetoTest(®) strips had the highest summary sensitivity and specificity when the trace and weak positive thresholds were used, respectively. Controlling for the source of publication, HK prevalence and reference standard employed did not impact the estimated sensitivity and specificity of the tests. Including only peer-reviewed studies reduced the number of primary studies evaluating the Precision Xtra(®) by 43% and Ketostix(®) by 33%. Diagnosing HK with blood, urine or milk are valid options, however, the diagnostic inaccuracy of urine and milk should be considered when making economic and treatment decisions.
Topics: 3-Hydroxybutyric Acid; Animals; Cattle; Cattle Diseases; Diagnostic Tests, Routine; Female; Ketosis; Milk; Point-of-Care Systems; Sensitivity and Specificity
PubMed: 27435643
DOI: 10.1016/j.prevetmed.2016.06.002 -
Journal of Dairy Science Jul 2014The objective of this study was to conduct a systematic and critical appraisal of the quality of previous publications and describe diagnostic methods, diagnostic... (Review)
Review
The objective of this study was to conduct a systematic and critical appraisal of the quality of previous publications and describe diagnostic methods, diagnostic criteria and definitions, repeatability, and agreement among methods for diagnosis of vaginitis, cervicitis, endometritis, salpingitis, and oophoritis in dairy cows. Publications (n=1,600) that included the words "dairy," "cows," and at least one disease of interest were located with online search engines. In total, 51 papers were selected for comprehensive review by pairs of the authors. Only 61% (n=31) of the 51 reviewed papers provided a definition or citation for the disease or diagnostic methods studied, and only 49% (n=25) of the papers provided the data or a citation to support the test cut point used for diagnosing disease. Furthermore, a large proportion of the papers did not provide sufficient detail to allow critical assessment of the quality of design or reporting. Of 11 described diagnostic methods, only one complete methodology, i.e., vaginoscopy, was assessed for both within- and between-operator repeatability (κ=0.55-0.60 and 0.44, respectively). In the absence of a gold standard, comparisons between different tests have been undertaken. Agreement between the various diagnostic methods is at a low level. These discrepancies may indicate that these diagnostic methods assess different aspects of reproductive health and underline the importance of tying diagnostic criteria to objective measures of reproductive performance. Those studies that used a reproductive outcome to select cut points and tests have the greatest clinical utility. This approach has demonstrated, for example, that presence of (muco)purulent discharge in the vagina and an increased proportion of leukocytes in cytological preparations following uterine lavage or cytobrush sampling are associated with poorer reproductive outcomes. The lack of validated, consistent definitions and outcome variables makes comparisons of the different tests difficult. The quality of design and reporting in future publications could be improved by using checklists as a guideline. Further high-quality research based on published standards to improve study design and reporting should improve cow-side diagnostic tests. Specifically, more data on intra- and interobserver agreement are needed to evaluate test variability. Also, more studies are necessary to determine optimal cut points and time postpartum of examination.
Topics: Animals; Cattle; Cattle Diseases; Diagnostic Tests, Routine; Female; Genital Diseases, Female; Inflammation
PubMed: 24835959
DOI: 10.3168/jds.2013-7450 -
BioMed Research International 2023Schistosomiasis is causing high morbidity and significant mortality in endemic areas. Kato-Katz stool examination and urine filtration techniques are the conventional... (Meta-Analysis)
Meta-Analysis Review
INTRODUCTION
Schistosomiasis is causing high morbidity and significant mortality in endemic areas. Kato-Katz stool examination and urine filtration techniques are the conventional methods for the detection of intestinal and urinary schistosomiasis. The most appropriate diagnostic tools for the detection of schistosomiasis especially in low-prevalence settings should be used. Therefore, this study is aimed at investigating the diagnostic accuracy of and diagnostic tools in sub-Saharan Africa.
METHODS
Electronic databases such as PubMed, PubMed Central/Medline, HINARI, Scopus, EMBASE, Science Direct, Google Scholar, and Cochrane Library were reviewed. The pooled estimates and heterogeneity were determined using Midas in Stata 14.0. The diagnostic accuracy of index tests was compared using the hierarchical summary of the receiver operating characteristic (HSROC) curve in Stata 14.0.
RESULTS
Twenty-four studies consisting of 12,370 individuals were tested to evaluate the accuracy of antigen, antibody, and molecular test methods for the detection of and . The pooled estimate of sensitivity and specificity of CCA was 88% (95% CI: 83-92) and 72 (95% CI: 62-80), respectively, when it is compared with parasitological stool examination for detection. On the other hand, ELISA showed a pooled estimate of sensitivity and specificity of 95% (95% CI: 93-96) and 35% (95% CI: 21-52), respectively, for the examination of using stool examination as a reference test. With regard to , the pooled estimate of sensitivity and specificity of polymerase chain reaction was 97% (95% CI: 78-100) and 94% (95% CI: 74-99), respectively. Moreover, the sensitivity and specificity of urine CCA vary between 41-80% and 55-91%, respectively, compared to urine microscopy.
CONCLUSION
The effort of schistosomiasis elimination requires accurate case identification especially in low-intensity infections. This study showed that CCA had the highest sensitivity and moderate specificity for the diagnosis of . Similarly, the sensitivity of ELISA was excellent, but its specificity was low. The diagnostic accuracy of PCR for the detection of was excellent compared to urine microscopic examination.
Topics: Humans; Animals; Microscopy; Schistosoma mansoni; Urinalysis; Africa South of the Sahara; Diagnostic Tests, Routine
PubMed: 37621699
DOI: 10.1155/2023/3769931 -
Korean Journal of Radiology Mar 2022
-
Ageing Research Reviews Jan 2024Biomarkers are emerging as a potential tool for screening or diagnosing sarcopenia. We aimed to summarize the current evidence on the diagnostic test accuracy of... (Meta-Analysis)
Meta-Analysis Review
Biomarkers are emerging as a potential tool for screening or diagnosing sarcopenia. We aimed to summarize the current evidence on the diagnostic test accuracy of biomarkers for sarcopenia. We comprehensively searched Ovid MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials up to January 2023 and only included diagnostic test accuracy studies. We identified 32 studies with 23,840 participants (women, 58.26%) that assessed a total of 30 biomarkers. The serum creatinine to cystatin C ratio (Cr/CysC) demonstrated a pooled sensitivity ranging from 51% (95% confidence interval [CI] 44-59%) to 86% (95% CI 70-95%) and a pooled specificity ranged from 55% (95% CI 38-70%) to 76% (95% CI 63-86%) for diagnosing sarcopenia defined by five different diagnostic criteria (11 studies, 7240 participants). The aspartate aminotransferase to alanine aminotransferase ratio demonstrated a pooled sensitivity of 62% (95% CI 56-67%) and a pooled specificity of 66% (95% CI 60-72%) (3 studies, 11,146 participants). The other 28 blood biomarkers exhibited low-to-moderate diagnostic accuracy for sarcopenia regardless of the reference standards. In conclusion, none of these biomarkers are optimal for screening or diagnosing sarcopenia. Well-designed studies are needed to explore and validate novel biomarkers for sarcopenia.
Topics: Humans; Female; Sensitivity and Specificity; Sarcopenia; Biomarkers; Diagnostic Tests, Routine
PubMed: 38036104
DOI: 10.1016/j.arr.2023.102148 -
Physical Therapy Jun 2023The purpose of this study was to summarize and evaluate the research on the accuracy of provocative maneuvers to diagnose carpal tunnel syndrome (CTS). (Meta-Analysis)
Meta-Analysis
OBJECTIVE
The purpose of this study was to summarize and evaluate the research on the accuracy of provocative maneuvers to diagnose carpal tunnel syndrome (CTS).
METHODS
The MEDLINE, CINAHL, Cochrane, and Embase databases were searched, and studies that assessed the diagnostic accuracy of at least 1 provocative test for CTS were selected. Study characteristics and data about the diagnostic accuracy of the provocative tests for CTS were extracted. A random-effects meta-analysis of the sensitivity (Sn) and specificity (Sp) of the Phalen test and Tinel sign was conducted. The risk of bias (ROB) was rated using the QUADAS-2 tool.
RESULTS
Thirty-one studies that assessed 12 provocative maneuvers were included. The Phalen test and the Tinel sign were the 2 most assessed tests (in 22 and 20 studies, respectively). The ROB was unclear or low in 20 studies, and at least 1 item was rated as having high ROB in 11 studies. Based on a meta-analysis of 7 studies (604 patients), the Phalen test had a pooled Sn of 0.57 (95% CI = 0.44-0.68; range = 0.12-0.92) and a pooled Sp of 0.67 (95% CI = 0.52-0.79; range = 0.30-0.95). For the Tinel sign (7 studies, 748 patients), the pooled Sn was 0.45 (95% CI = 0.34-0.57; range = 0.17-0.97) and the pooled Sp was 0.78 (95% CI = 0.60-0.89; range = 0.40-0.92). Other provocative maneuvers were less frequently studied and had conflicting diagnostic accuracies.
CONCLUSION
Meta-analyses are imprecise but suggest that the Phalen test has moderate Sn and Sp, whereas the Tinel test has low Sn and high Sp. Clinicians should combine provocative maneuvers with sensorimotor tests, hand diagrams, and diagnostic questionnaires to achieve better overall diagnostic accuracy rather than relying on individual clinical tests.
IMPACT
Evidence of unclear and high ROB do not support the use of any single provocative maneuver for the diagnosis of CTS. Clinicians should consider a combination of noninvasive clinical diagnostic tests as the first choice for the diagnosis of CTS.
Topics: Humans; Carpal Tunnel Syndrome; Hand; Sensitivity and Specificity; Surveys and Questionnaires; Diagnostic Tests, Routine
PubMed: 37366626
DOI: 10.1093/ptj/pzad029 -
Journal of Endodontics Mar 2023The aim of this systematic review and meta-analysis was to investigate the overall accuracy of deep learning models in detecting periapical (PA) radiolucent lesions in... (Meta-Analysis)
Meta-Analysis Review
INTRODUCTION
The aim of this systematic review and meta-analysis was to investigate the overall accuracy of deep learning models in detecting periapical (PA) radiolucent lesions in dental radiographs, when compared to expert clinicians.
METHODS
Electronic databases of Medline (via PubMed), Embase (via Ovid), Scopus, Google Scholar, and arXiv were searched. Quality of eligible studies was assessed by using Quality Assessment and Diagnostic Accuracy Tool-2. Quantitative analyses were conducted using hierarchical logistic regression for meta-analyses on diagnostic accuracy. Subgroup analyses on different image modalities (PA radiographs, panoramic radiographs, and cone beam computed tomographic images) and on different deep learning tasks (classification, segmentation, object detection) were conducted. Certainty of evidence was assessed by using Grading of Recommendations Assessment, Development, and Evaluation system.
RESULTS
A total of 932 studies were screened. Eighteen studies were included in the systematic review, out of which 6 studies were selected for quantitative analyses. Six studies had low risk of bias. Twelve studies had risk of bias. Pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio of included studies (all image modalities; all tasks) were 0.925 (95% confidence interval [CI], 0.862-0.960), 0.852 (95% CI, 0.810-0.885), 6.261 (95% CI, 4.717-8.311), 0.087 (95% CI, 0.045-0.168), and 71.692 (95% CI, 29.957-171.565), respectively. No publication bias was detected (Egger's test, P = .82). Grading of Recommendations Assessment, Development and Evaluationshowed a "high" certainty of evidence for the studies included in the meta-analyses.
CONCLUSION
Compared to expert clinicians, deep learning showed highly accurate results in detecting PA radiolucent lesions in dental radiographs. Most studies had risk of bias. There was a lack of prospective studies.
Topics: Deep Learning; Cone-Beam Computed Tomography; Radiography, Panoramic; Diagnostic Tests, Routine; Sensitivity and Specificity
PubMed: 36563779
DOI: 10.1016/j.joen.2022.12.007 -
Journal of Medical Internet Research Jul 2023Deep learning (DL) prediction models hold great promise in the triage of COVID-19. (Meta-Analysis)
Meta-Analysis Review
BACKGROUND
Deep learning (DL) prediction models hold great promise in the triage of COVID-19.
OBJECTIVE
We aimed to evaluate the diagnostic test accuracy of DL prediction models for assessing and predicting the severity of COVID-19.
METHODS
We searched PubMed, Scopus, LitCovid, Embase, Ovid, and the Cochrane Library for studies published from December 1, 2019, to April 30, 2022. Studies that used DL prediction models to assess or predict COVID-19 severity were included, while those without diagnostic test accuracy analysis or severity dichotomies were excluded. QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2), PROBAST (Prediction Model Risk of Bias Assessment Tool), and funnel plots were used to estimate the bias and applicability.
RESULTS
A total of 12 retrospective studies involving 2006 patients reported the cross-sectionally assessed value of DL on COVID-19 severity. The pooled sensitivity and area under the curve were 0.92 (95% CI 0.89-0.94; I=0.00%) and 0.95 (95% CI 0.92-0.96), respectively. A total of 13 retrospective studies involving 3951 patients reported the longitudinal predictive value of DL for disease severity. The pooled sensitivity and area under the curve were 0.76 (95% CI 0.74-0.79; I=0.00%) and 0.80 (95% CI 0.76-0.83), respectively.
CONCLUSIONS
DL prediction models can help clinicians identify potentially severe cases for early triage. However, high-quality research is lacking.
TRIAL REGISTRATION
PROSPERO CRD42022329252; https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD 42022329252.
Topics: Humans; COVID-19; Deep Learning; Retrospective Studies; PubMed; Diagnostic Tests, Routine; COVID-19 Testing
PubMed: 37477951
DOI: 10.2196/46340