Is there Bias in Alternatives to Standardized Tests? An Investigation into Content Differences in Letters of Recommendation

Individuals concerned with subgroup differences on standardized tests suggest replacing these tests with holistic evaluations of unstructured application materials, such as letters of recommendation (LORs), which they posit show less bias. We empirically investigate this proposition that LORs are bias-free, and argue that LORs might actually invite systematic, race and gender subgroup differences in the content and evaluation of LORs. We text analyzed over 37,000 LORs submitted on behalf of over 10,000 graduate school applicants. Results showed that LOR content does differ across applicants. Furthermore, we see some systematic gender, race, and gender-race intersection differences in LOR content. Content of LORs also systematically differed between degree programs (S.T.E.M. vs. non-S.T.E.M.) and degree sought (doctoral vs. masters). Finally, LOR content alone did not predict an appreciable amount of variance in offers of admission (the first barrier to increasing diversity and inclusion in graduate programs). Our results, combined with past research on LOR content bias, highlight concerns that LORs can be biased against marginalized groups. We conclude with suggestions for reducing potential bias in LOR and for increasing diversity in graduate programs.

Systematic Review and Meta-Analysis of Endurance Exercise Training Protocols for Mice

Inbred and genetically modified mice are frequently used to investigate the molecular mechanisms responsible for the beneficial adaptations to exercise training. However, published paradigms for exercise training in mice are variable, making comparisons across studies for training efficacy difficult. The purpose of this systematic review and meta-analysis was to characterize the diversity across published treadmill-based endurance exercise training protocols for mice and to identify training protocol parameters that moderate the adaptations to endurance exercise training in mice. Published studies were retrieved from PubMed and EMBASE and reviewed for the following inclusion criteria: inbred mice; inclusion of a sedentary group; and exercise training using a motorized treadmill. Fifty-eight articles met those inclusion criteria and also included a “classical” marker of training efficacy. Outcome measures included changes in exercise performance, V˙O2max, skeletal muscle oxidative enzyme activity, blood lactate levels, or exercise-induced cardiac hypertrophy. The majority of studies were conducted using male mice. Approximately 48% of studies included all information regarding exercise training protocol parameters. Meta-analysis was performed using 105 distinct training groups (i.e., EX-SED pairs). Exercise training had a significant effect on training outcomes, but with high heterogeneity (Hedges’ g=1.70, 95% CI=1.47–1.94, Tau2=1.14, I2=80.4%, prediction interval=−0.43–3.84). Heterogeneity was partially explained by subgroup differences in treadmill incline, training duration, exercise performance test type, and outcome variable. Subsequent analyses were performed on subsets of studies based on training outcome, exercise performance, or biochemical markers. Exercise training significantly improved performance outcomes (Hedges’ g=1.85, 95% CI=1.55–2.15). Subgroup differences were observed for treadmill incline, training duration, and exercise performance test protocol on improvements in performance. Biochemical markers also changed significantly with training (Hedges’ g=1.62, 95% CI=1.14–2.11). Subgroup differences were observed for strain, sex, exercise session time, and training duration. These results demonstrate there is a high degree of heterogeneity across exercise training studies in mice. Training duration had the most significant impact on training outcome. However, the magnitude of the effect of exercise training varies based on the marker used to assess training efficacy.

Endogenous Oxytocin Levels in Autism—A Meta-Analysis

Oxytocin (OT) circuitry plays a major role in the mediation of prosocial behavior. Individuals with autism spectrum disorder (ASD) are characterized by impairments in social interaction and communication and have been suggested to display deficiencies in central OT mechanisms. The current preregistered meta-analysis evaluated potential group differences in endogenous OT levels between individuals with ASD and neurotypical (NT) controls. We included 18 studies comprising a total of 1422 participants. We found that endogenous OT levels are lower in children with ASD as compared to NT controls (n = 1123; g = −0.60; p = 0.006), but this effect seems to disappear in adolescent (n = 152; g = −0.20; p = 0.53) and adult populations (n = 147; g = 0.27; p = 0.45). Secondly, while no significant subgroup differences were found in regard to sex, the group difference in OT levels of individuals with versus without ASD seems to be only present in the studies with male participants (n = 814; g = −0.44; p = 0.08) and not female participants (n = 192; g = 0.11; p = 0.47). More research that employs more homogeneous methods is necessary to investigate potential developmental changes in endogenous OT levels, both in typical and atypical development, and to explore the possible use of OT level measurement as a diagnostic marker of ASD.

A Longitudinal Study on Hematocrit Changes after Glifozins Exposure in Patients with Type 2 Diabetes

Background and Methods: Gliflozins are widely prescribed drugs in patients with type 2 diabetes. We pursue to explain abnormal increments in red cell parameters observed in this population, by means of a longitudinal study in 149 patients with a gliflozins exposure period of 12±6 months. Red cell parameters, HbA1c and other variables were recorded. Results: HbA1c fraction decreased (-0.5±1.3, 95% CI: -0.7 to -0.3, p<0.001), while mean hemoglobin (0.5±0.9, 95% CI: 0.3 to 0.6, P<0.001) and hematocrit (1.6±2.6, 95% CI: 1.2 to 2.0, P<0.001) increased. Mean (SD) hematocrit increased 2.7±1.9 in 112 patients, and decreased -1.7±1.5 in 37 (p<0.001 for subgroup differences). The larger increments in PCV were proportional to higher plasma fraction at baseline (p=0.009). Conclusion: Red cell parameters after gliflozins exposure tend to increase and may reach abnormally high thresholds in some patients with type 2 diabetes.

Imputing the Number of Responders from the Mean and Standard Deviation of CGI-Improvement in Clinical Trials Investigating Medications for Autism Spectrum Disorder

Introduction: Response to treatment, according to Clinical Global Impression-Improvement (CGI-I) scale, is an easily interpretable outcome in clinical trials of autism spectrum disorder (ASD). Yet, the CGI-I rating is sometimes reported as a continuous outcome, and converting it to dichotomous would allow meta-analysis to incorporate more evidence. Methods: Clinical trials investigating medications for ASD and presenting both dichotomous and continuous CGI-I data were included. The number of patients with at least much improvement (CGI-I ≤ 2) were imputed from the CGI-I scale, assuming an underlying normal distribution of a latent continuous score using a primary threshold θ = 2.5 instead of θ = 2, which is the original cut-off in the CGI-I scale. The original and imputed values were used to calculate responder rates and odds ratios. The performance of the imputation method was investigated with a concordance correlation coefficient (CCC), linear regression, Bland–Altman plots, and subgroup differences of summary estimates obtained from random-effects meta-analysis. Results: Data from 27 studies, 58 arms, and 1428 participants were used. The imputation method using the primary threshold (θ = 2.5) had good performance for the responder rates (CCC = 0.93 95% confidence intervals [0.86, 0.96]; β of linear regression = 1.04 [0.95, 1.13]; bias and limits of agreements = 4.32% [−8.1%, 16.74%]; no subgroup differences χ2 = 1.24, p-value = 0.266) and odds ratios (CCC = 0.91 [0.86, 0.96]; β = 0.96 [0.78, 1.14]; bias = 0.09 [−0.87, 1.04]; χ2 = 0.02, p-value = 0.894). The imputation method had poorer performance when the secondary threshold (θ = 2) was used. Discussion: Assuming a normal distribution of the CGI-I scale, the number of responders could be imputed from the mean and standard deviation and used in meta-analysis. Due to the wide limits of agreement of the imputation method, sensitivity analysis excluding studies with imputed values should be performed.

P–669 “Follitropin”: A retrospective, observational study comparing the efficacy of follitropin alpha biosimilar therapy in different ovarian stimulation protocols: real-world data

Study question To investigate the therapeutic efficacy of follitropin alpha biosimilar therapy in nonselected patients undergoing IVF. Summary answer This large retrospective study demonstrated similar therapeutic efficacy for follitropin alpha biosimilar therapy in women who underwent ovarian stimulation (OS) using different protocols. What is known already Based on data from the last meta-analyses (Budani et al., 2020), follitropin alpha biosimilars showed similar efficacy and safety in randomized controlled trials aimed at proving the therapeutic equivalence in terms of oocytes retrieved in women undergoing OS. In most cases, normogonadotrophic patients were enrolled in such studies without any endocrine or ovarian disturbances. The absence of real-world data can be compensated by additional post-marketing studies aimed at investigating the efficacy of biosimilars in different OS protocols using antagonists and agonists of GnRH and OS with mixed gonadotropins. Study design, size, duration A retrospective, observational, anonymized cohort study conducted at 35 IVF clinics in Russia, named “FOLLITROPIN”, compared the efficacy of OS in 2020. The OS protocols analysed where follitropin alpha biosimilar (Primapur®) was applied for at least 5 days. All of the analysed subjects underwent OS using GnRH antagonist/agonist protocols, with no restrictions on the OS protocol or food supplements/vitamins. No inclusion or exclusion criteria were applied. Overall, 5484 OS protocols were analysed. Participants/materials, setting, methods The efficacy of 5484 OS protocols was calculated, and two subgroups were extracted: (1) mixed gonadotropin OS protocols (N = 2625) vs monotherapy with Primapur® (N = 2859); (2) GnRH antagonist OS (N = 2183) vs GnRH agonist (N = 676) using only Primapur®. Demographic and clinical characteristics: (1) Age 34.9±4.8 vs 32.9±4.7 (p &lt; 0.001), BMI 23.9±4.7 vs 23.6±4.5 (p &lt; 0.001), IVF attempt 1.4±0.7 vs 1.3±0.6 (p &lt; 0.001); (2) Age 32.9±4.6 vs 33.1±4.9 (p = 0.449), BMI 23.7±4.6 vs 23.1±4.5 (p = 0.019), and IVF attempt 1.2±0.5 vs 1.4±0.9 (p &lt; 0.001). Main results and the role of chance The total efficacy of OS with Primapur®: oocytes retrieved: 9.5±7.2, MII: 6.8±6.6, 2PN: 6.1±5.8, clinical pregnancy per ET (PR) 6 weeks after ET: 38.4%. Subgroup 1 analysis: oocytes retrieved: 8.6±6.8 vs 10.3±7.4 (p &lt; 0.001), MII: 6.7±6.2 vs 7.7±6.9 (p &lt; 0.001), 2PN: 5.8±5.2 vs 7.2±6.2 (p &lt; 0.001). There were statistically significant differences between oocyte yields in mixed vs monotherapy protocols due to subgroup differences, including age, BMI and IVF attempts. No statistically significant differences were found for PR in subgroup 1: 39.3% [95% CI: 36.9–41.7%] vs 37.6% [95% CI: 35.3–39.8%] (p = 0.314). The major accompanying gonadotropin in the mixed protocol was menotropin (75% for mixed OS protocols). Subgroup 2 analysis: oocytes retrieved: 10.5±7.5 vs 9.6±7.0 (p = 0.032), MII: 7.6±6.9 vs 6.7±5.7 (p &lt; 0.001), 2PN: 7.3±6.3 vs 5.7±5.0 (p &lt; 0.001). There were statistically significant differences between oocyte yields in GnRH antagonist vs GnRH agonist monotherapy due to subgroup differences, including BMI and IVF attempts. No statistically significant differences were found for PR in subgroup 2: 37.9% [95% CI: 35.5–40.5%] vs 35.9% [95% CI: 30.8–41.1%] (p = 0.482). All medicines were well tolerated at the injection site, and no serious adverse reactions were reported. This large retrospective cohort study in a nonselected population demonstrated similar clinical efficacy for follitropin alpha biosimilar therapy. Limitations, reasons for caution The real-world patient data analysed in this study were representative, showing the ability of follitropin biosimilars to develop both folliculogenesis and clinical pregnancy in a nonselected population. Additional comparative studies are needed to confirm the efficacy of the biosimilars in patients with classified types of infertility causes, including unexplained infertility. Wider implications of the findings: In this study, we demonstrated the therapeutic efficacy of biosimilars in terms of oocyte yield and clinical pregnancy development in women undergoing different OS protocols. Further large-scale studies with known hormonal levels before and during OS, as well as the micro- and macronutrient status of both parents, are needed. Trial registration number None

COVID-19 and the academy: opinions and experiences of university-based scientists in the U.S.

Much of the available evidence regarding COVID-19 effects on the scientific community in the U.S. is anecdotal and non-representative. We report findings from a based survey of university-based biologists, biochemists, and civil and environmental engineers regarding negative and positive COVID-19 impacts, respondent contributions to addressing the pandemic, and their opinions regarding COVID-19 research policies. The most common negative impact was university closures, cited by 93% of all scientists. Significant subgroup differences emerged, with higher proportions of women, assistant professors, and scientists at institutions located in COVID-19 “hotspot” counties reporting difficulties concentrating on research. Assistant professors additionally reported facing more unanticipated childcare responsibilities. Approximately half of the sample also reported one or more positive COVID-19 impacts, suggesting the importance of developing a better understanding of the complete range of impacts across all fields of science. Regarding COVID-19 relevant public policy, findings suggest divergence of opinion concerning surveillance technologies and the need to alter federal approval processes for new tests and vaccines.

Understanding the evaluation of mHealth app features based on a cross-country Kano analysis

While mobile health (mHealth) apps play an increasingly important role in digitalized health care, little is known regarding the effects of specific mHealth app features on user satisfaction across different healthcare system contexts. Using personal health record (PHR) apps as an example, this study identifies how potential users in Germany and Denmark evaluate a set of 26 app features, and whether evaluation differences can be explained by the differences in four pertinent user characteristics, namely privacy concerns, mHealth literacy, mHealth self-efficacy, and adult playfulness. Based on survey data from both countries, we employed the Kano method to evaluate PHR features and applied a quartile-based sample-split approach to understand the underlying relationships between user characteristics and their perceptions of features. Our results not only reveal significant differences in 14 of the features between Germans and Danes, they also demonstrate which of the user characteristics best explain each of these differences. Our two key contributions are, first, to explain the evaluation of specific PHR app features on user satisfaction in two different healthcare contexts and, second, to demonstrate how to extend the Kano method in terms of explaining subgroup differences through user characteristic antecedents. The implications for app providers and policymakers are discussed.

Population subgroup differences in the use of a COVID-19 chatbot

COVID-19 chatbots are widely used to screen for symptoms and disseminate information about the virus, yet little is known about the population subgroups that interact with this technology and the specific features that are used. An analysis of 1,000,740 patients invited to use a COVID-19 chatbot, 69,451 (6.94%) of which agreed to participate, shows differences in chatbot feature use by gender, race, and age. These results can inform future public health COVID-19 symptom screening and information dissemination strategies.