Unmatched Dataset Research Articles

Effects of Matching on Evaluation of Accuracy, Fairness, and Fairness Impossibility in AI-ML Systems

“Matching” procedures in statistics involve construction of datasets with similar covariates between compared groups. Matching has recently been proposed as a means of addressing fairness impossibility (i.e. inconsistency of fairness metrics) in AI and ML systems: Beigang argues on conceptual grounds that, when matched rather than unmatched datasets are analyzed, the tradeoff between the fairness metrics equalized odds (EO) and positive predictive value (PPV) will be reduced. Here we evaluate matching as a practical rather than merely conceptual approach to reducing fairness impossibility. As a case study we conduct pre-match and post-match analyses on the well-known COMPAS dataset from Broward Co., Florida, 2013-2014. We then reflect on what these results suggest about effects of matching on (a) accuracy estimates, (b) fairness estimates, and (c) difference between fairness estimates – that is, the extent to which matching reduces “fairness impossibility” in practice. We conclude that matching is a promising tool for improving evaluations on all three fronts, but faces problems due to potential biases introduced by matching procedures themselves, as well as limited power under conditions extremely common to ML evaluation contexts such as non-independent variables and relevance of hidden variables.

Perceived rhythmic regularity is greater for song than speech: examining acoustic correlates of rhythmic regularity in speech and song.

Rhythm is a key feature of music and language, but the way rhythm unfolds within each domain differs. Music induces perception of a beat, a regular repeating pulse spaced by roughly equal durations, whereas speech does not have the same isochronous framework. Although rhythmic regularity is a defining feature of music and language, it is difficult to derive acoustic indices of the differences in rhythmic regularity between domains. The current study examined whether participants could provide subjective ratings of rhythmic regularity for acoustically matched (syllable-, tempo-, and contour-matched) and acoustically unmatched (varying in tempo, syllable number, semantics, and contour) exemplars of speech and song. We used subjective ratings to index the presence or absence of an underlying beat and correlated ratings with stimulus features to identify acoustic metrics of regularity. Experiment 1 highlighted that ratings based on the term "rhythmic regularity" did not result in consistent definitions of regularity across participants, with opposite ratings for participants who adopted a beat-based definition (song greater than speech), a normal-prosody definition (speech greater than song), or an unclear definition (no difference). Experiment 2 defined rhythmic regularity as how easy it would be to tap or clap to the utterances. Participants rated song as easier to clap or tap to than speech for both acoustically matched and unmatched datasets. Subjective regularity ratings from Experiment 2 illustrated that stimuli with longer syllable durations and with less spectral flux were rated as more rhythmically regular across domains. Our findings demonstrate that rhythmic regularity distinguishes speech from song and several key acoustic features can be used to predict listeners' perception of rhythmic regularity within and across domains as well.

Integration of single-cell RNA-Seq and CyTOF data characterises heterogeneity of rare cell subpopulations

Background: The simultaneous measurement of cellular proteins and transcriptomes of single cell data has become an exciting new possibility with the advent of highly multiplexed multi-omics methodologies. However, mass cytometry (CyTOF) is a well-established, affordable technique for the analysis of proteomic data, which is well suited for the discovery and characterisation of very rare subpopulations of cells with a wealth of publicly available datasets. Methods: We present and evaluate the multimodal integration of single cell RNA-Seq and CyTOF datasets coming from both matched and unmatched samples, using two publicly available datasets. Results: We demonstrate that the integration of well annotated CyTOF data with single cell RNA sequencing can aid in the identification and annotation of cell populations with high accuracy. Furthermore, we show that the integration can provide imputed measurements of protein markers which are comparable to the current gold standard of antibody derived tags (ADT) from CITE-Seq for both matched and unmatched datasets. Using this methodology, we identify and transcriptionally characterise a rare subpopulation of CD11c positive B cells in high resolution using publicly available data and we unravel its heterogeneity in a single cell setting without the need to sort the cells in advance, in a manner which had not been previously possible. Conclusions: This approach provides the framework for using available proteomic and transcriptomic datasets in a unified and unbiased fashion to assist ongoing and future studies of cellular characterisation and biomarker identification.

Integration of single-cell RNA-Seq and CyTOF data characterises heterogeneity of rare cell subpopulations

Background: The simultaneous measurement of cellular proteins and transcriptomes of single cell data has become an exciting new possibility with the advent of highly multiplexed multi-omics methodologies. However, mass cytometry (CyTOF) is a well-established, affordable technique for the analysis of proteomic data, which is well suited for the discovery and characterisation of very rare subpopulations of cells with a wealth of publicly available datasets. Methods: We present and evaluate the multimodal integration of single cell RNA-Seq and CyTOF datasets coming from both matched and unmatched samples, using two publicly available datasets. Results: We demonstrate that the integration of well annotated CyTOF data with single cell RNA sequencing can aid in the identification and annotation of cell populations with high accuracy. Furthermore, we show that the integration can provide imputed measurements of protein markers which are comparable to the current gold standard of antibody derived tags (ADT) from CITE-Seq for both matched and unmatched datasets. Using this methodology, we identify and transcriptionally characterise a rare subpopulation of CD11c positive B cells in high resolution using publicly available data and we unravel its heterogeneity in a single cell setting without the need to sort the cells in advance, in a manner which had not been previously possible. Conclusions: This approach provides the framework for using available proteomic and transcriptomic datasets in a unified and unbiased fashion to assist ongoing and future studies of cellular characterisation and biomarker identification.

P07.11.A Impact of awake craniotomy on extent-of-resection and performance in glioma: a retrospective, propensity-score matched cohort study

Abstract Background The goal of awake craniotomy is to safely optimize the extent-of-resection in patients with glioma. Lower postoperative tumor volume is associated with better overall survival, emphasizing the importance of optimal surgical resection. Awake craniotomy causes less postoperative neurological deficits than surgery under general anesthesia, however, it is unclear whether awake craniotomy leads to lower postoperative tumor volume compared with general anesthesia in presumed lower grade glioma. Methods Retrospective, matched cohort study in patients with astrocytoma, IDH-mutant grade 2 and 3, oligodendroglioma, IDH-mutant, and 1p/19q-codeleted grade 2 and 3, and low-grade glioma, IDH-wildtype (now designated glioblastoma, IDH-wildtype), who underwent resection with awake craniotomy or under general anesthesia between 2003 and 2021 at Erasmus MC Brain Tumor Center. Pre- and postoperative tumor volumes were measured by semi-automatic 3D MRI-segmentation. First, we performed a multivariate logistic regression to assess which factors predicted selection for awake craniotomy. Thereafter, matching based on propensity score was attempted. Outcome variables were postoperative tumor volume, resection percentage and Karnofsky Performance Status (KPS) 3 months after surgery. Results We identified 181 awake craniotomy-patients and 135 general anesthesia-patients. Awake craniotomy-patients were younger, in better condition, more often male, with tumors more often in eloquent areas, in the left side of the brain and non-contrast enhancing. When performing matching without replacement, only 68 awake craniotomy-patients could be matched with 68 general anesthesia-patients, underscoring the imbalance in the dataset. Matching with replacement yielded a matched dataset of 181 awake craniotomy-patients with 60 general anesthesia-patients with adequate matching on most baseline variables, except for eloquent area (47% for awake craniotomy and 21.7% for general anesthesia, p < 0.001). In this matched dataset, median postoperative volume in awake craniotomy was 5.8 mL (IQR 0 - 92.8) vs 12.2 mL (IQR 0 - 90.1) in general anesthesia (p-value = 0.114). Resection percentages did not differ between the awake craniotomy- and general anesthesia-groups. KPS scores at 3 months after surgery did not differ between awake craniotomy and general anesthesia (p = 0.15). Conclusion Postoperative tumor volume was not lower in awake craniotomy-patients than in general anesthesia-patients. Neither were resection percentage and KPS scores at 3 months after surgery significantly different between the groups. Adequate matching was only obtainable using replacement, underscoring the risk of bias in unmatched datasets. These data should be interpreted with care, given the retrospective nature, potential residual confounding and potential lack of generalizability due to unmatched normal resection patients.

Benefits of Adding Glucagon-Like Peptide 1 Receptor Agonists to Sodium-Glucose Co-Transporter 2 Inhibitors in Diabetic Patients With Atherosclerotic Disease and Heart Failure

Sodium-glucose co-transporter 2 inhibitors (SGLT2i) reduce the risk of cardiovascular events and heart failure hospitalization (HFH) in patients with heart failure with reduced ejection fraction (HFrEF), diabetes mellitus type 2 (DM2), and atherosclerotic cardiovascular disease (ASCVD). The role of glucagon-like peptide 1 agonists (GLP1a) in these patients is unclear. We designed this study to assess if the addition of GLP1a to SGLT2i therapy improves outcomes in patients with HFrEF, DM2, and ASCVD. This was a retrospective cohort study of patients with DM2, ASCVD, and HFrEF in the national Veterans Affairs database. Patients on SGLT2i were propensity matched to patients on both SGTL2i and GLP1a. The co-primary outcomes were HFH and the composite of all-cause death, myocardial infarction, and stroke. We assessed them through a Cox regression model including unbalanced baseline characteristics. From a cohort of 5,576 patients, 343 were propensity matched to each study arm. The addition of GLP1a was associated with a 67% reduction in the 1-year risk of a composite event compared with therapy with SGLT2i (confidence interval 0.138 to 0.714, p=0.007). The risk of HFH was not significantly different between both arms (p=0.199). Sensitivity analyses in the unmatched dataset confirmed these findings. In conclusion, the addition of GLP1a to SGLT2i may reduce the risk of adverse events in patients with HFrEF who have DM2 and ASCVD, but it does not affect the risk of HFH.

Integration of single-cell RNA-Seq and CyTOF data characterises heterogeneity of rare cell subpopulations

Background: The simultaneous measurement of cellular proteins and transcriptomes of single cell data has become an exciting new possibility with the advent of highly multiplexed multi-omics methodologies. However, mass cytometry (CyTOF) is a well-established, affordable technique for the analysis of proteomic data, which is well suited for the discovery and characterisation of very rare subpopulations of cells with a wealth of publicly available datasets. Methods: We present and evaluate the multimodal integration of single cell RNA-Seq and CyTOF datasets coming from both matched and unmatched samples, using two publicly available datasets. Results: We demonstrate that the integration of well annotated CyTOF data with single cell RNA sequencing can aid in the identification and annotation of cell populations with high accuracy. Furthermore, we show that the integration can provide imputed measurements of protein markers which are comparable to the current gold standard of antibody derived tags (ADT) from CITE-Seq for both matched and unmatched datasets. Using this methodology, we identify and transcriptionally characterise a rare subpopulation of CD11c positive B cells in high resolution using publicly available data and we unravel its heterogeneity in a single cell setting without the need to sort the cells in advance, in a manner which had not been previously possible. Conclusions: This approach provides the framework for using available proteomic and transcriptomic datasets in a unified and unbiased fashion to assist ongoing and future studies of cellular characterisation and biomarker identification.

Video-evoked fMRI BOLD responses are highly consistent across different data acquisition sites.

Naturalistic imaging paradigms, in which participants view complex videos in the scanner, are increasingly used in human cognitive neuroscience. Videos evoke temporally synchronized brain responses that are similar across subjects as well as within subjects, but the reproducibility of these brain responses across different data acquisition sites has not yet been quantified. Here, we characterize the consistency of brain responses across independent samples of participants viewing the same videos in functional magnetic resonance imaging (fMRI) scanners at different sites (Indiana University and Caltech). We compared brain responses collected at these different sites for two carefully matched datasets with identical scanner models, acquisition, and preprocessing details, along with a third unmatched dataset in which these details varied. Our overall conclusion is that for matched and unmatched datasets alike, video‐evoked brain responses have high consistency across these different sites, both when compared across groups and across pairs of individuals. As one might expect, differences between sites were larger for unmatched datasets than matched datasets. Residual differences between datasets could in part reflect participant‐level variability rather than scanner‐ or data‐ related effects. Altogether our results indicate promise for the development and, critically, generalization of video fMRI studies of individual differences in healthy and clinical populations alike.

Pooled Data Analysis of the Long-Term Treatment Effects of Tolvaptan in ADPKD

IntroductionIn 1- and 3-year randomized trials, tolvaptan slowed kidney function decline in subjects with autosomal dominant polycystic kidney disease (ADPKD) at risk of rapid progression. The 3-year trial also evaluated effects on total kidney volume (TKV); slowing of TKV growth was demonstrated. Subjects were followed in open-label extension trials. To characterize longer-term effects of treatment, an analysis was conducted comparing tolvaptan-treated subjects with subjects from standard of care (SOC) ADPKD studies without tolvaptan.MethodsThis was a pooled, longitudinal analysis of data from 8 tolvaptan clinical trials and 5 studies without tolvaptan (natural history or SOC) in ADPKD. Data from subjects who participated in multiple studies were linked for longer follow-up. Outcomes were rates of change in estimated glomerular filtration rate (eGFR) and TKV over 5.5 years. To control for heterogeneity in disease characteristics between tolvaptan and SOC treatment groups, analysis populations matched for baseline demographic and disease characteristics were constructed.ResultsMatched analysis (n = 1186 in each treatment group) indicated that tolvaptan slowed annualized eGFR decline by 1.01 ml/min per 1.73 m2 (P < 0.001) versus SOC over 5.5 years. An analysis conducted on the full, unmatched data set (tolvaptan: n = 2928; SOC: n = 4189) confirmed significant reduction in annual eGFR decline. Among subjects with TKV data, TKV was significantly reduced at years 1, 3, and 5 for tolvaptan versus SOC in both matched and full data sets.ConclusionComparison of a pooled tolvaptan cohort to a pooled control cohort with ADPKD supports longer-term treatment effects of tolvaptan.

Sidedness determines clinical characteristics and survival outcomes in medullary adenocarcinoma of the colon

Colon medullary adenocarcinoma (MAC) is a rare histologic subtype. Clinical presentation and cancer outcomes of MAC, compared to colon adenocarcinoma (AC), remain incompletely described. Annual age-adjusted incidence rates were computed using Surveillance, Epidemiology, and End Results (2002–2017). A cohort analysis using the National Cancer Database (2010–2016) compared patient characteristics in an unmatched dataset and prognostic characteristics in a 1:1 matched subset. Reported annual age-adjusted incidence of MAC has significantly increased, with an average annual percent change (APC) increase of 23.8% (95% CI: 19.2–28.6); concurrent AC incidence declined (APC: − 2.8, 95% CI: − 3.1 to − 2.8). Analyses of 1018 MAC and 210,784 AC unmatched patients showed that MAC patients were more often older, female, and white, with higher disease stage, poorly-differentiated tumors, right-sided laterality, and lymphovascular invasion (all p < 0.0001). Among those with known microsatellite status, instability was more prevalent among MAC than AC patients (82% vs. 24%, p < 0.0001). Multivariate analyses of the matched dataset revealed that MAC histology was not independently associated with overall survival. However, when stratifying by laterality, left-sided MAC was associated with shorter survival when compared to right-sided MAC (HR 1.66, 95% CI 1.16–2.38) and right-sided AC (HR 1.54, 95% CI 1.12–2.12). The reported incidence of MAC is increasing, in contrast to the declining incidence of AC. MAC clinical and molecular features are distinct from AC and likely account for outcome differences. Overall, left-sided MAC was associated with the shortest OS. Molecular profiling may improve treatment guidelines for MAC.

5α-Reductase Inhibitors Are Associated with Reduced Risk of SARS-CoV-2 Infection: A Matched-Pair, Registry-Based Analysis.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has a disproportionately severe effect on men, suggesting that the androgen pathway plays a role in the disease. Studies on the effect of castration and androgen receptor blockade have been mixed, while 5α-reductase inhibitor (5ARI) use in men with COVID-19 (2019 novel coronavirus) have shown potential benefits. We assessed the association of 5ARI use on risk of community acquired SARS-CoV-2 infection. A total of 60,474 males in a prospective registry of people tested for SARS-CoV-2 between March 8, 2020 and February 15, 2021 were included. Using a matched cohort design, men using 5ARIs were matched 1:1 to non5ARI users. Independent analysis using unconditional multivariable logistic regression on the entire unmatched data set was completed for validation. Primary outcome measures were the association of 5ARI use on rates of SARS-Cov-2 positivity and disease severity. Of the men 1,079 (1.8%) reported 5ARI use and 55,100 were available for matching. The final matched cohorts included 944 men each. Mean duration of use was 60.4 months (IQR 17-84 months). Absolute risk for infection was significantly lower in 5ARI users compared to nonusers, 42.3% (399/944) vs 47.2% (446/944), respectively (absolute risk reduction [ARR] 4.9%, OR 0.81, 95% CI 0.67-0.97, p=0.026). Unconditional multivariable logistic regression analysis of the entire study cohort of 55,100 men confirmed the protective association of 5ARI use (ARR 5.3%, OR=0.877, 95% CI 0.774-0.995, p=0.042). Use of 5ARIs was not associated with disease severity. Use of 5ARIs in men without prostate cancer was associated with a reduction in community acquired SARS-CoV-2 infection.

Abstract PO-075: Kinotype to phenotype: Perturbed phosphoproteomic state predicts cancer cell growth rates in vitro

Abstract Numerous aspects of cellular signaling are regulated by the kinome - the network of over 500 protein kinases that guides and modulates information transfer throughout the cell. The key role played by both individual kinases and assemblies of kinases organized into functional subnetworks leads to kinome dysregulation being a key driver of many diseases, particularly cancer. Linked with their role in disease, the druggability of kinases has led to increased interest in the development of kinase inhibitors, with over fifty now having achieved FDA approval. These and other yet to be approved compounds have been used in a variety of contexts, including in screens to test their efficacy as potential therapeutics. Recent proteomics techniques including Kinobeads and multiplexed inhibitor beads linked with mass spectroscopy (MIB/MS) are a relatively new mechanism which affords the ability to assess the state of the protein kinome en masse. When combined with perturbation with targeted kinase inhibitors, the quantification of the dynamic response of the kinome provides a novel platform for the study of cell signaling and potential design of drug therapies. Here, we describe an approach that links the state of the kinome, or kinotype, with a downstream cancer cell phenotype, in this case, cellular growth. Integrating independent perturbation and growth response data sets, we find that the kinotype of a cell has a significant and predictive relationship with cell growth. More specifically, we characterize kinome response to perturbation via Kinobeads within models aimed at predicting cellular growth inhibition rates. Integrating 82 drugs, 237 kinases, and 380 cell lines (both malignant and nonmalignant) into a sparse GLM predicts growth rate inhibition at multiple doses of an unseen small molecule with an r^2 value of 0.624 (Bonferroni corrected p-value of 5.7 × 10^−69), even with an unmatched dataset. We further explore the use of higher-level kinase relationships, i.e., kinome subnetworks, along with manifold learning approaches as a component of these models to better ascertain the potential impact of kinome architecture in cellular response. Together, this systems view of the kinome presents potential opportunities for improved cancer disease classification, the identification of new drug targets, as well as impacting on the broader design of combination therapies for cancer treatment. Citation Format: Isaac S. Robson, Matthew E. Berginski, Shawn M. Gomez. Kinotype to phenotype: Perturbed phosphoproteomic state predicts cancer cell growth rates in vitro [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-075.

Effect of the Extent of Resection on Survival Outcome in Glioblastoma: Propensity Score Approach

Abstract Objective To evaluate the effectiveness of the extent of resection (EOR) on survival outcome using propensity score-based approaches. Materials and Methods A retrospective cohort study was performed in patients with newly diagnosed glioblastoma. Propensity score matching (PSM) and propensity score regression adjustment were used in the matched and unmatched dataset, respectively. Therefore, the Kaplan-Meier survival curve and Cox's regression analyses were performed to determine the effect of the EOR on survival outcomes. Results One hundred and sixty-eight patients were included for analyzes. The total tumor resection in the unmatched dataset was 22.6% of all cases. Using PSM, incomplete tumor resection had an unfavorable survival outcome when compared with total tumor resection (hazard ratio (HR) 2.92, 95% confidence interval [CI] 1.72–4.94). Additionally, biopsy and partial tumor resection were significantly associated with poor prognosis when compared with total tumor resection using propensity score regression adjustment (HR of biopsy 1.89, 95%CI 1.13–3.16 and HR of partial resection 1.89, 95%CI 1.28–2.80). Conclusions Patients with total tumor resection tend to have a more favorable prognosis than patients with partial tumor resection. The propensity score-based analysis is an alternative approach to evaluate the effect of an intervention that has limitations to perform a randomized controlled trial.

A Longitudinal Analysis of Job Satisfaction During a Recession in the Netherlands

Between 2008 and 2013, the Netherlands was confronted by a severe recession. This recession may have affected the job satisfaction of workers. Currently, little is known about how job satisfaction changes during a recession. To investigate the effect of the 2008–2013 recession on job satisfaction in the Netherlands, and to assess how job satisfaction changed over time. Longitudinal data from six waves of a national panel in the Netherlands are used to investigate the effects. These data capture the periods before, during and after the recession. A Blinder–Oaxaca decomposition technique is used to decompose the ordinal outcome variable job satisfaction. Subsequent waves are compared, which results in five comparison groups. Workers who participated in subsequent waves are matched to assess their job satisfaction over time. Cross-sectional associations are analyzed using the entire unmatched dataset. Workers became more satisfied with their job during the recession. After the recession ended, average job satisfaction decreased again. Both unmatched and matched analyses indicated only changes in job level affecting job satisfaction. The coefficient of education had a small effect cross-sectionally. The level of education and industry had a small effect longitudinally. However, these effects were not robust. Job satisfaction decreased before the recession commenced but increased during the recession. After the recession, job satisfaction decreased again. An increase in job satisfaction during the recession may be explained by a change in the composition of workers with respect to job level, instead of by the effect of predictors.

Galaxy formation and evolution science in the era of the Large Synoptic Survey Telescope

The field of galaxy formation and evolution synthesizes the physics of baryons and dark matter to describe the origin of systems such as the Milky Way and the enormous diversity of the galaxy population. The broad variation in possible formation histories and the wide range of cosmic environments make large statistical samples of galaxies essential for identifying the important physical mechanisms that govern their formation. Starting in the early 2020s, the Large Synoptic Survey Telescope (LSST) will provide an unmatched dataset for galaxy evolution studies by observing the entire southern sky in ultraviolet, optical and near-infrared wavelengths, producing multi-epoch digital images over a 10-year nominal mission that when summed will provide the deepest, wide-angle view of our Universe ever assembled. Here, we discuss the importance of LSST for deepening our understanding of galaxy formation and evolution over cosmic time. We present some outstanding problems in the field that LSST will address, and we present a roadmap of some preparatory research efforts required to make effective use of the LSST dataset for galaxy formation science. The Large Synoptic Survey Telescope (LSST), an upcoming astronomical survey, will deeply observe the entire southern sky in a broad range of colours. We present the LSST science opportunities and technical challenges in the field of galaxy formation and evolution.

Role of pelvic chemoradiation therapy in patients with initially metastatic anal canal cancer: A National Cancer Database review.

Although the management of localized anal canal squamous cell carcinomas is well established, the role of pelvic chemoradiation (CRT) in the treatment of patients presenting with synchronous metastatic (stage IV) disease is poorly defined. This study used a national cancer database to compare the overall survival (OS) rates of patients with synchronous metastatic disease receiving CRT to the pelvis and patients treated with chemotherapy (CT) alone. This study included adult patients with anal canal squamous cell carcinomas presenting with synchronous metastases diagnosed from 2004 to 2012. Multiple imputation and 2:1 propensity score matching were used to create a matched data set for testing. The proportional hazards model was used to estimate the hazard ratio (HR) for the effect of the treatment group on OS. With only patients in the matched data set, the OS of the treatment groups was estimated with the Kaplan-Meier method by treatment group. This study started with an unmatched data set of 978 patients, and 582 patients were selected for the matched data set: 388 in the CRT group and 194 in the CT-alone group. The HR for the group effect was 0.75 (95% confidence interval [CI], 0.61-0.92; P=.006). The median OS was 21.1months in the CRT group (95% CI, 17.4-24.0months) and 14.6months in the CT group (95% CI, 12.2-18.4months). The corresponding 5-year OS rates were 23% (95% CI, 18%-28%) and 14% (95% CI, 7%-21%), respectively. In this large series analyzing OS in patients with stage IV anal cancer, CRT was associated with improved OS in comparison with CT alone. Because of the lack of prospective data in this setting, this evidence will help to guide treatment approaches in this group of patients.

Long-term Survival Outcomes of Primary Breast Cancer in Women With or Without Preoperative Magnetic Resonance Imaging: A Matched Cohort Study

AimsTo investigate whether preoperative magnetic resonance imaging (MRI) in patients with primary breast cancer is predictive of disease-free (DFS) and overall survival and to determine the prognostic factors indicating survival. Materials and methodsThis retrospective study was approved by the institutional review board and the requirement for informed consent was waived. From 2009 to 2010, 828 women with primary breast cancer and preoperative MRI were matched with 1613 women without such imaging. Patients were matched with regards to 25 patient and tumour-related covariates. A Cox proportional hazards model was used to investigate the time to recurrence and to estimate the hazard ratio for preoperative MRI. Log-rank tests and Cox proportional hazards survival analysis were carried out on total recurrence DFS and overall survival in the unmatched datasets. ResultsIn total, 799 matched pairs were available for survival analysis. The MRI group showed a tendency towards better survival outcome; however, there were no significant differences in DFS and overall survival. Age at diagnosis (DFS hazard ratio = 0.98; overall survival hazard ratio = 1.04), larger tumour size (DFS hazard ratio = 1.01; overall survival hazard ratio = 1.02), triple negative breast cancer (DFS hazard ratio = 2.64; overall survival hazard ratio = 3.44) and the presence of lymphovascular invasion (DFS hazard ratio = 2.12; overall survival hazard ratio = 2.70) were independent significant variables for worse DFS and overall survival. ConclusionPreoperative MRI did not result in an improvement in a patient's outcome. Age at diagnosis, tumour size, molecular subtype and lymphovascular invasion were significant independent factors affecting both DFS and overall survival.

Risk of cardiovascular disease in Chinese patients with rheumatoid arthritis: A cross-sectional study based on hospital medical records in 10 years.

ObjectiveThough the risk of cardiovascular disease (CVD) in rheumatoid arthritis (RA) has been established in Western population, little is known about the risk in Chinese people with RA. Our objective was to estimate the risk of CVD in Chinese people with RA using hospital medical records data.MethodsThe inpatients medical record database 2005–2015 of Sichuan provincial people’s hospital was examined. All individuals with a primary diagnosis of RA were included as cases, and those of osteoarthritis (OA) were included as controls, which consisted of the unmatched dataset. Then, RA cases and OA controls were matched by sex and age at 1:1 ratio, forming the matched dataset. The morbidity of CVD (including ischemia heart disease (IHD), congestive heart failure (CHF), et al), stroke and arthrosclerosis were extracted from the database, so as the demographic data and comorbidities related to CVD. Multiple logistic regression analysis was used to estimate the risk of CVD in RA adjusted for demographics and comorbidities using the unmatched dataset. Sensitivity analysis was conducted 1) considering interaction terms between RA and comorbidities, and 2) using multivariable conditional logistic regression for the matched dataset.ResultsThe unmatched dataset comprised of 1824RA cases and 1995 OA controls and the matched dataset comprised of 1022 pairs of sex and age matched RA and OA patients. RA exhibited increased odds of prevalent CVD compared with OA, and the adjusted ORs (95%CIs) for CVD, stroke, IHD, CHF, and atherosclerosis were1.86(1.42–2.43), 1.11(0.71–1.74), 1.47(0.97–2.24), 2.09(1.03–4.22), and 2.49 (1.97–3.13), respectively, and was 2.26 (1.29–3.96)for IHD further adjusted for interaction term. The matched dataset analysis found similar results.ConclusionsChinese people with RA were approximated 2 times more likely to have CVD, IHD, CHF and atherosclerosis compared with those with OA. The findings justified the need of further longitudinal study to establish the causal-relationship between RA and CVD and to estimate the precise risk in this population.

Prestroke Antiplatelet Agents in First-Time Ischemic Stroke Are Related to Subtypes

Background: It is not well known whether prestroke antiplatelet agents (PAs) are associated with the subtypes of ischemic stroke. Methods: We screened patients in a hospital-based stroke registry. Patients who were admitted with a diagnosis of first-time ischemic stroke within 5 days of symptom onset were included. Ischemic stroke subtypes were classified in accordance with the Trial of ORG 10172 in Acute Stroke Treatment classification based on stroke mechanism: large-artery atherosclerosis (LA), cardioembolism (CE), small vessel occlusion (SVO), other determined (OC) or undetermined causes (UC). Multinomial logistic regression analyses were performed to evaluate the effect of PA on stroke subtypes before and after propensity score matching. Results: Among 3,025 patients, 748 (24.7%) were taking antiplatelet agents prior to stroke. After propensity score matching, 1,190 patients were ultimately included. The PA group was associated with strokes caused by SVO rather than LA in multinomial logistic regression of an unmatched dataset. However, multivariable analysis after propensity score matching demonstrated that PA use was associated with a higher probability of SVO and CE (OR 2.05, p < 0.001 and OR 1.62, p = 0.05, respectively) compared with LA. Conclusions: PAs were associated with specific index stroke subtypes.

Ethnicity and rehabilitation outcomes: the Needs Assessment Checklist.

Retrospective cohort study with matched samples. To investigate whether significant differences in rehabilitation outcomes exist between different ethnic groups, using the Needs Assessment Checklist (NAC). Tertiary care, spinal cord injury rehabilitation unit (National Spinal Injuries Centre), Stoke Mandeville Hospital, UK. Rehabilitation outcomes and demographic information were obtained from the NAC. Data on 461 individuals were included in the study. Analysis of variance was employed to investigate differences in rehabilitation outcomes between various ethnic groups, across both the whole cohort and matched samples. Significant differences were evident across the different rehabilitation domains when ethnicity was examined, in particular within the domains of physical healthcare and psychological well-being. Within the unmatched data set, significant differences were found to exist in 3 of the 10 rehabilitation domains, and with the matched data set within 4 of the 10 domains. The results indicated that the cohort as a whole made significant improvements from the first to the second NAC within all rehabilitation domains. Results indicate that ethnicity may have an impact on rehabilitation outcomes for individuals with spinal cord injury. Further investigation is needed to explore the nature of this relationship, and the future role of targeted interventions focusing on improving rehabilitation outcomes within the domains of physical and psychological care, in particular for individuals from different ethnic backgrounds.