-
Analyzing adverse drug reaction using statistical and machine learning methods: A systematic review.Medicine Jun 2022Adverse drug reactions (ADRs) are unintended negative drug-induced responses. Determining the association between drugs and ADRs is crucial, and several methods have...
BACKGROUND
Adverse drug reactions (ADRs) are unintended negative drug-induced responses. Determining the association between drugs and ADRs is crucial, and several methods have been proposed to demonstrate this association. This systematic review aimed to examine the analytical tools by considering original articles that utilized statistical and machine learning methods for detecting ADRs.
METHODS
A systematic literature review was conducted based on articles published between 2015 and 2020. The keywords used were statistical, machine learning, and deep learning methods for detecting ADR signals. The study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement (PRISMA) guidelines.
RESULTS
We reviewed 72 articles, of which 51 and 21 addressed statistical and machine learning methods, respectively. Electronic medical record (EMR) data were exclusively analyzed using the regression method. For FDA Adverse Event Reporting System (FAERS) data, components of the disproportionality method were preferable. DrugBank was the most used database for machine learning. Other methods accounted for the highest and supervised methods accounted for the second highest.
CONCLUSIONS
Using the 72 main articles, this review provides guidelines on which databases are frequently utilized and which analysis methods can be connected. For statistical analysis, >90% of the cases were analyzed by disproportionate or regression analysis with each spontaneous reporting system (SRS) data or electronic medical record (EMR) data; for machine learning research, however, there was a strong tendency to analyze various data combinations. Only half of the DrugBank database was occupied, and the k-nearest neighbor method accounted for the greatest proportion.
Topics: Adverse Drug Reaction Reporting Systems; Databases, Factual; Drug-Related Side Effects and Adverse Reactions; Electronic Health Records; Humans; Machine Learning
PubMed: 35758373
DOI: 10.1097/MD.0000000000029387 -
PloS One 2021Working memory is an essential cognitive skill for storing and processing limited amounts of information over short time periods. Researchers disagree about the extent... (Meta-Analysis)
Meta-Analysis
BACKGROUND AND OBJECTIVE
Working memory is an essential cognitive skill for storing and processing limited amounts of information over short time periods. Researchers disagree about the extent to which socioeconomic position affects children's working memory, yet no study has systematically synthesised the literature regarding this topic. The current review therefore aimed to investigate the relationship between socioeconomic position and working memory in children, regarding both the magnitude and the variability of the association.
METHODS
The review protocol was registered on PROSPERO and the PRISMA checklist was followed. Embase, Psycinfo and MEDLINE were comprehensively searched via Ovid from database inception until 3rd June 2021. Studies were screened by two reviewers at all stages. Studies were eligible if they included typically developing children aged 0-18 years old, with a quantitative association reported between any indicator of socioeconomic position and children's working memory task performance. Studies were synthesised using two data-synthesis methods: random effects meta-analyses and a Harvest plot.
KEY FINDINGS
The systematic review included 64 eligible studies with 37,737 individual children (aged 2 months to 18 years). Meta-analyses of 36 of these studies indicated that socioeconomic disadvantage was associated with significantly lower scores working memory measures; a finding that held across different working memory tasks, including those that predominantly tap into storage (d = 0.45; 95% CI 0.27 to 0.62) as well as those that require processing of information (d = 0.52; 0.31 to 0.72). A Harvest plot of 28 studies ineligible for meta-analyses further confirmed these findings. Finally, meta-regression analyses revealed that the association between socioeconomic position and working memory was not moderated by task modality, risk of bias, socioeconomic indicator, mean age in years, or the type of effect size.
CONCLUSION
This is the first systematic review to investigate the association between socioeconomic position and working memory in children. Socioeconomic disadvantage was associated with lower working memory ability in children, and that this association was similar across different working memory tasks. Given the strong association between working memory, learning, and academic attainment, there is a clear need to share these findings with practitioners working with children, and investigate ways to support children with difficulties in working memory.
Topics: Child; Cognition; Humans; Memory Disorders; Memory, Long-Term; Memory, Short-Term; Socioeconomic Factors
PubMed: 34855871
DOI: 10.1371/journal.pone.0260788 -
JAMA Network Open Mar 2021An increasing number of machine learning (ML)-based clinical decision support systems (CDSSs) are described in the medical literature, but this research focuses almost... (Comparative Study)
Comparative Study
IMPORTANCE
An increasing number of machine learning (ML)-based clinical decision support systems (CDSSs) are described in the medical literature, but this research focuses almost entirely on comparing CDSS directly with clinicians (human vs computer). Little is known about the outcomes of these systems when used as adjuncts to human decision-making (human vs human with computer).
OBJECTIVES
To conduct a systematic review to investigate the association between the interactive use of ML-based diagnostic CDSSs and clinician performance and to examine the extent of the CDSSs' human factors evaluation.
EVIDENCE REVIEW
A search of MEDLINE, Embase, PsycINFO, and grey literature was conducted for the period between January 1, 2010, and May 31, 2019. Peer-reviewed studies published in English comparing human clinician performance with and without interactive use of an ML-based diagnostic CDSSs were included. All metrics used to assess human performance were considered as outcomes. The risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) and Risk of Bias in Non-Randomised Studies-Intervention (ROBINS-I). Narrative summaries were produced for the main outcomes. Given the heterogeneity of medical conditions, outcomes of interest, and evaluation metrics, no meta-analysis was performed.
FINDINGS
A total of 8112 studies were initially retrieved and 5154 abstracts were screened; of these, 37 studies met the inclusion criteria. The median number of participating clinicians was 4 (interquartile range, 3-8). Of the 107 results that reported statistical significance, 54 (50%) were increased by the use of CDSSs, 4 (4%) were decreased, and 49 (46%) showed no change or an unclear change. In the subgroup of studies carried out in representative clinical settings, no association between the use of ML-based diagnostic CDSSs and improved clinician performance could be observed. Interobserver agreement was the commonly reported outcome whose change was the most strongly associated with CDSS use. Four studies (11%) reported on user feedback, and, in all but 1 case, clinicians decided to override at least some of the algorithms' recommendations. Twenty-eight studies (76%) were rated as having a high risk of bias in at least 1 of the 4 QUADAS-2 core domains, and 6 studies (16%) were considered to be at serious or critical risk of bias using ROBINS-I.
CONCLUSIONS AND RELEVANCE
This systematic review found only sparse evidence that the use of ML-based CDSSs is associated with improved clinician diagnostic performance. Most studies had a low number of participants, were at high or unclear risk of bias, and showed little or no consideration for human factors. Caution should be exercised when estimating the current potential of ML to improve human diagnostic performance, and more comprehensive evaluation should be conducted before deploying ML-based CDSSs in clinical settings. The results highlight the importance of considering supported human decisions as end points rather than merely the stand-alone CDSSs outputs.
Topics: Clinical Competence; Decision Support Systems, Clinical; Humans; Machine Learning
PubMed: 33704476
DOI: 10.1001/jamanetworkopen.2021.1276 -
The Lancet. Haematology May 2023Given the paucity of high-certainty evidence, and differences in opinion on the use of nuclear medicine for hematological malignancies, we embarked on a consensus... (Review)
Review
Given the paucity of high-certainty evidence, and differences in opinion on the use of nuclear medicine for hematological malignancies, we embarked on a consensus process involving key experts in this area. We aimed to assess consensus within a panel of experts on issues related to patient eligibility, imaging techniques, staging and response assessment, follow-up, and treatment decision-making, and to provide interim guidance by our expert consensus. We used a three-stage consensus process. First, we systematically reviewed and appraised the quality of existing evidence. Second, we generated a list of 153 statements based on the literature review to be agreed or disagreed with, with an additional statement added after the first round. Third, the 154 statements were scored by a panel of 26 experts purposively sampled from authors of published research on haematological tumours on a 1 (strongly disagree) to 9 (strongly agree) Likert scale in a two-round electronic Delphi review. The RAND and University of California Los Angeles appropriateness method was used for analysis. Between one and 14 systematic reviews were identified on each topic. All were rated as low to moderate quality. After two rounds of voting, there was consensus on 139 (90%) of 154 of the statements. There was consensus on most statements concerning the use of PET in non-Hodgkin and Hodgkin lymphoma. In multiple myeloma, more studies are required to define the optimal sequence for treatment assessment. Furthermore, nuclear medicine physicians and haematologists are awaiting consistent literature to introduce volumetric parameters, artificial intelligence, machine learning, and radiomics into routine practice.
Topics: Humans; Consensus; Nuclear Medicine; Artificial Intelligence; Hematologic Neoplasms; Molecular Imaging
PubMed: 37142345
DOI: 10.1016/S2352-3026(23)00030-3 -
JAMA Psychiatry Aug 2023Social anxiety disorder (SAD) can be adequately treated with cognitive behavioral therapy (CBT). However, there is a large gap in knowledge on factors associated with... (Meta-Analysis)
Meta-Analysis
Baseline Severity as a Moderator of the Waiting List-Controlled Association of Cognitive Behavioral Therapy With Symptom Change in Social Anxiety Disorder: A Systematic Review and Individual Patient Data Meta-analysis.
IMPORTANCE
Social anxiety disorder (SAD) can be adequately treated with cognitive behavioral therapy (CBT). However, there is a large gap in knowledge on factors associated with prognosis, and it is unclear whether symptom severity predicts response to CBT for SAD.
OBJECTIVE
To examine baseline SAD symptom severity as a moderator of the association between CBT and symptom change in patients with SAD.
DATA SOURCES
For this systematic review and individual patient data meta-analysis (IPDMA), PubMed, PsycInfo, Embase, and the Cochrane Library were searched from January 1, 1990, to January 13, 2023. Primary search topics were social anxiety disorder, cognitive behavior therapy, and randomized controlled trial.
STUDY SELECTION
Inclusion criteria were randomized clinical trials comparing CBT with being on a waiting list and using the Liebowitz Social Anxiety Scale (LSAS) in adults with a primary clinical diagnosis of SAD.
DATA EXTRACTION AND SYNTHESIS
Authors of included studies were approached to provide individual-level data. Data were extracted by pairs of authors following the Preferred Reporting Items for Systematic Reviews and Meta-analyses reporting guideline, and risk of bias was assessed using the Cochrane tool. An IPDMA was conducted using a 2-stage approach for the association of CBT with change in LSAS scores from baseline to posttreatment and for the interaction effect of baseline LSAS score by condition using random-effects models.
MAIN OUTCOMES AND MEASURES
The main outcome was the baseline to posttreatment change in symptom severity measured by the LSAS.
RESULTS
A total of 12 studies including 1246 patients with SAD (mean [SD] age, 35.3 [10.9] years; 738 [59.2%] female) were included in the meta-analysis. A waiting list-controlled association between CBT and pretreatment to posttreatment LSAS change was found (b = -20.3; 95% CI, -24.9 to -15.6; P < .001; Cohen d = -0.95; 95% CI, -1.16 to -0.73). Baseline LSAS scores moderated the differences between CBT and waiting list with respect to pretreatment to posttreatment symptom reductions (b = -0.22; 95% CI, -0.39 to -0.06; P = .009), indicating that individuals with severe symptoms had larger waiting list-controlled symptom reductions after CBT (Cohen d = -1.13 [95% CI, -1.39 to -0.88] for patients with very severe SAD; Cohen d = -0.54 [95% CI, -0.80 to -0.29] for patients with mild SAD).
CONCLUSIONS AND RELEVANCE
In this systematic review and IPDMA, higher baseline SAD symptom severity was associated with greater (absolute but not relative) symptom reductions after CBT in patients with SAD. The findings contribute to personalized care by suggesting that clinicians can confidently offer CBT to individuals with severe SAD symptoms.
Topics: Adult; Humans; Female; Male; Phobia, Social; Waiting Lists; Cognitive Behavioral Therapy; Randomized Controlled Trials as Topic
PubMed: 37256597
DOI: 10.1001/jamapsychiatry.2023.1291 -
Cancers Jun 2021Radiomics supposes an alternative non-invasive tumor characterization tool, which has experienced increased interest with the advent of more powerful computers and more... (Review)
Review
Radiomics supposes an alternative non-invasive tumor characterization tool, which has experienced increased interest with the advent of more powerful computers and more sophisticated machine learning algorithms. Nonetheless, the incorporation of radiomics in cancer clinical-decision support systems still necessitates a thorough analysis of its relationship with tumor biology. Herein, we present a systematic review focusing on the clinical evidence of radiomics as a surrogate method for tumor molecular profile characterization. An extensive literature review was conducted in PubMed, including papers on radiomics and a selected set of clinically relevant and commonly used tumor molecular markers. We summarized our findings based on different cancer entities, additionally evaluating the effect of different modalities for the prediction of biomarkers at each tumor site. Results suggest the existence of an association between the studied biomarkers and radiomics from different modalities and different tumor sites, even though a larger number of multi-center studies are required to further validate the reported outcomes.
PubMed: 34208595
DOI: 10.3390/cancers13123015 -
Journal of Imaging Dec 2023Three-dimensional human pose estimation has made significant advancements through the integration of deep learning techniques. This survey provides a comprehensive... (Review)
Review
Three-dimensional human pose estimation has made significant advancements through the integration of deep learning techniques. This survey provides a comprehensive review of recent 3D human pose estimation methods, with a focus on monocular images, videos, and multi-view cameras. Our approach stands out through a systematic literature review methodology, ensuring an up-to-date and meticulous overview. Unlike many existing surveys that categorize approaches based on learning paradigms, our survey offers a fresh perspective, delving deeper into the subject. For image-based approaches, we not only follow existing categorizations but also introduce and compare significant 2D models. Additionally, we provide a comparative analysis of these methods, enhancing the understanding of image-based pose estimation techniques. In the realm of video-based approaches, we categorize them based on the types of models used to capture inter-frame information. Furthermore, in the context of multi-person pose estimation, our survey uniquely differentiates between approaches focusing on relative poses and those addressing absolute poses. Our survey aims to serve as a pivotal resource for researchers, highlighting state-of-the-art deep learning strategies and identifying promising directions for future exploration in 3D human pose estimation.
PubMed: 38132693
DOI: 10.3390/jimaging9120275 -
PloS One 2016Our study examined the psychological outcomes associated with failed ART treatment outcomes in men and women. (Meta-Analysis)
Meta-Analysis Review
OBJECTIVE
Our study examined the psychological outcomes associated with failed ART treatment outcomes in men and women.
SEARCH STRATEGY
A systematic search for studies published between January 1980 and August 2015 was performed across seven electronic databases.
INCLUSION CRITERIA
Studies were included if they contained data on psychosocial outcomes taken pre and post ART treatment.
DATA EXTRACTION AND SYNTHESIS
A standardised form was used to extract data and was verified by two independent reviewers. Studies were meta-analysed to determine the association of depression and anxiety with ART treatment outcomes. Narrative synthesis identified factors to explain variations in the size and directions of effects and relationships explored within and between the studies.
MAIN RESULTS
Both depression and anxiety increased after a ART treatment failure with an overall pooled standardised mean difference (SMD) of 0.41 (95% CI: 0.27, 0.55) for depression and 0.21 (95% CI: 0.13, 0.29) for anxiety. In contrast, depression decreased after a successful treatment, SMD of -0.24 (95% CI: -0.37,-0.11). Both depression and anxiety decreased as time passed from ART procedure. Nonetheless, these remained higher than baseline measures in the group with the failed outcome even six months after the procedure. Studies included in the narrative synthesis also confirmed an association with negative psychological outcomes in relation to marital satisfaction and general well-being following treatment failure.
CONCLUSION
Linking ART failure and psychosocial outcomes may elucidate the experience of treatment subgroups, influence deliberations around recommendations for resource allocation and health policy and guide patient and clinician decision making.
Topics: Adaptation, Psychological; Adult; Anxiety; Depression; Female; Humans; Infertility, Female; Infertility, Male; Male; Quality of Life; Reproductive Techniques, Assisted; Treatment Failure
PubMed: 27835654
DOI: 10.1371/journal.pone.0165805 -
JAMA Dermatology Feb 2023Primary cutaneous squamous cell carcinoma is usually curable; however, a subset of patients develops poor outcomes, including local recurrence, nodal metastasis, distant... (Meta-Analysis)
Meta-Analysis
Association of Patient Risk Factors, Tumor Characteristics, and Treatment Modality With Poor Outcomes in Primary Cutaneous Squamous Cell Carcinoma: A Systematic Review and Meta-analysis.
IMPORTANCE
Primary cutaneous squamous cell carcinoma is usually curable; however, a subset of patients develops poor outcomes, including local recurrence, nodal metastasis, distant metastasis, and disease-specific death.
OBJECTIVES
To evaluate all evidence-based reports of patient risk factors and tumor characteristics associated with poor outcomes in primary cutaneous squamous cell carcinoma and to identify treatment modalities that minimize poor outcomes.
DATA SOURCES
PubMed, Embase, and SCOPUS databases were searched for studies of the topic in humans, published in the English language, from database inception through February 8, 2022.
STUDY SELECTION
Two authors independently screened the identified articles and included those that were original research with a sample size of 10 patients or more and that assessed risk factors and/or treatment modalities associated with poor outcomes among patients with primary cutaneous squamous cell carcinoma.
DATA EXTRACTION AND SYNTHESIS
Data extraction was performed by a single author, per international guidelines. The search terms, study objectives, and protocol methods were defined before study initiation. A total of 310 studies were included for full-text assessment. Owing to heterogeneity of the included studies, a random-effects model was used. Data analyses were performed from May 25 to September 15, 2022.
MAIN OUTCOMES AND MEASURES
For studies of risk factors, risk ratios and incidence proportions; and for treatment studies, incidence proportions.
RESULTS
In all, 129 studies and a total of 137 449 patients with primary cutaneous squamous cell carcinoma and 126 553 tumors were included in the meta-analysis. Several patient risk factors and tumor characteristics were associated with local recurrence, nodal metastasis, distant metastasis, disease-specific death, and all-cause death were identified. Among all factors reported by more than 1 study, the highest risks for local recurrence and disease-specific death were associated with tumor invasion beyond subcutaneous fat (risk ratio, 9.1 [95% CI, 2.8-29.2] and 10.4 [95% CI, 3.0- 36.3], respectively), and the highest risk of any metastasis was associated with perineural invasion (risk ratio, 5.0; 95% CI, 2.3-11.1). Patients who received Mohs micrographic surgery had the lowest incidence of nearly all poor outcomes; however, in some results, the 95% CIs overlapped with those of other treatment modalities.
CONCLUSIONS AND RELEVANCE
This meta-analysis identified the prognostic value of several risk factors and the effectiveness of the available treatment modalities. These findings carry important implications for the prognostication, workup, treatment, and follow-up of patients with primary cutaneous squamous cell carcinoma.
TRIAL REGISTRATION
PROSPERO Identifier: CRD42022311250.
Topics: Humans; Carcinoma, Squamous Cell; Skin Neoplasms; Prognosis; Mohs Surgery; Risk Factors
PubMed: 36576732
DOI: 10.1001/jamadermatol.2022.5508 -
Frontiers in Immunology 2023The surge in the number of publications on psoriasis has posed significant challenges for researchers in effectively managing the vast amount of information. However,...
BACKGROUND
The surge in the number of publications on psoriasis has posed significant challenges for researchers in effectively managing the vast amount of information. However, due to the lack of tools to process metadata, no comprehensive bibliometric analysis has been conducted.
OBJECTIVES
This study is to evaluate the trends and current hotspots of psoriatic research from a macroscopic perspective through a bibliometric analysis assisted by machine learning based semantic analysis.
METHODS
Publications indexed under the Medical Subject Headings (MeSH) term "Psoriasis" from 2003 to 2022 were extracted from PubMed. The generative statistical algorithm latent Dirichlet allocation (LDA) was applied to identify specific topics and trends based on abstracts. The unsupervised Louvain algorithm was used to establish a network identifying relationships between topics.
RESULTS
A total of 28,178 publications were identified. The publications were derived from 176 countries, with United States, China, and Italy being the top three countries. For the term "psoriasis", 9,183 MeSH terms appeared 337,545 times. Among them, MeSH term "Severity of illness index", "Treatment outcome", "Dermatologic agents" occur most frequently. A total of 21,928 publications were included in LDA algorithm, which identified three main areas and 50 branched topics, with "Molecular pathogenesis", "Clinical trials", and "Skin inflammation" being the most increased topics. LDA networks identified "Skin inflammation" was tightly associated with "Molecular pathogenesis" and "Biological agents". "Nail psoriasis" and "Epidemiological study" have presented as new research hotspots, and attention on topics of comorbidities, including "Cardiovascular comorbidities", "Psoriatic arthritis", "Obesity" and "Psychological disorders" have increased gradually.
CONCLUSIONS
Research on psoriasis is flourishing, with molecular pathogenesis, skin inflammation, and clinical trials being the current hotspots. The strong association between skin inflammation and biologic agents indicated the effective translation between basic research and clinical application in psoriasis. Besides, nail psoriasis, epidemiological study and comorbidities of psoriasis also draw increased attention.
Topics: Humans; United States; Psoriasis; Arthritis, Psoriatic; Bibliometrics; Dermatitis; Machine Learning; Inflammation
PubMed: 37954610
DOI: 10.3389/fimmu.2023.1272080