Efficiency Analysis of Higher Education Institutions: Use of Categorical Variables

This paper focuses on the Data Envelopment Analysis (DEA) based efficiency evaluation to find the impact of two-step categorical impact on the enrollment efficiency of colleges in Bihar, one of the largest states of India. The objective of the study is to find the impact of factors, other than college-specific, on the efficiency of the colleges. The proposed research includes colleges funded and managed through seven state public universities. To follow the homogeneity condition of DEA, colleges providing courses of Arts (languages and humanities only), Science, and Commerce only, have been selected. The numbers of students enrolled in undergraduate and postgraduate courses are considered as two outputs. Numbers of teaching and non-teaching staff are considered as inputs. Colleges have been classified into two categories based on their presence in the rural or urban areas. The efficiency of a college due to any categorical value is calculated as the ratio of overall efficiency and efficiency calculated with similar categorical Decision-Making Units (DMUs) only. The impact of both the categorical variables, affiliation to university and geographical presence, has been analyzed through the hypothesis testing with the null hypothesis that there is no impact of category on the efficiency of DMUs due to a categorical variable.

More Than Morningness: The Effect of Circadian Rhythm Amplitude and Stability on Resilience, Coping, and Sleep Duration

Self-report tools that measure circadian rhythms have focused primarily on phase. We add to the sparse literature on assessing amplitude and stability. We randomly recruited 1,163 participants who completed several measures. The correlation between the LV scale (amplitude) and FR scale (stability) was −0.12 (p < 0.01). As expected, amplitude was negatively associated with phase (r = −0.64, p < 0.01) while stability showed a weak link with phase (r = 0.07, p < 0.05). Structural equation modeling suggested a close model-fit of the factor structure in the sample (RMSEA = 0.033). The LV scale explained 22% of the variance, while the FR scale explained 23%. Scale reliability was satisfactory for the LV scale (0.68) and good for the FR scale (0.73). Participants with low amplitude or flexible rhythms reported significantly better resilience, coping, and required less daily sleep. We constructed a composite circadian categorical variable to combine the best attributes from the LV and FR scales; participants with both low amplitude and flexible rhythms, reported significantly better resilience, coping, and less sleep need. We found rhythm amplitude decreased with age, while stability remained constant.

Encoding a Categorical Independent Variable for Input to TerrSet’s Multi-Layer Perceptron

The profession debates how to encode a categorical variable for input to machine learning algorithms, such as neural networks. A conventional approach is to convert a categorical variable into a collection of binary variables, which causes a burdensome number of correlated variables. TerrSet’s Land Change Modeler proposes encoding a categorical variable onto the continuous closed interval from 0 to 1 based on each category’s Population Evidence Likelihood (PEL) for input to the Multi-Layer Perceptron, which is a type of neural network. We designed examples to test the wisdom of these encodings. The results show that encoding a categorical variable based on each category’s Sample Empirical Probability (SEP) produces results similar to binary encoding and superior to PEL encoding. The Multi-Layer Perceptron’s sigmoidal smoothing function can cause PEL encoding to produce nonsensical results, while SEP encoding produces straightforward results. We reveal the encoding methods by illustrating how a dependent variable gains across an independent variable that has four categories. The results show that PEL can differ substantially from SEP in ways that have important implications for practical extrapolations. If users must encode a categorical variable for input to a neural network, then we recommend SEP encoding, because SEP efficiently produces outputs that make sense.

The associations between organophosphate esters and urinary incontinence in the general US population

Organophosphate esters (OPEs) impact health in many ways. Since its relationship with urinary incontinence remains unknown, we aimed to explore their associations in the US general population. We combined the results of urine specimens test and self-reported urinary incontinence conditions from the National Health and Nutrition Examination Survey (NHANES) 2013–2014 among 2666 participants and then conducted linear regression and logistic regression to analyse associations between log2-transformed OPE concentrations and urinary incontinence. We found that 0.92% of men and 15.74% of women complained of mixed urinary incontinence (MUI). The concentrations of diphenyl phosphate (DPHP) were significantly correlated to MUI among women when treated as a continuous variable (adjusted odds ratio (OR) = 1.15; 95% confidence interval (CI), 1.01–1.31; p = 0.0369) and as a categorical variable (adjusted OR = 1.24; 95% CI, 1.03–1.49; p for trend = 0.0245), whereas no positive correlation was found in males. There were no significant associations between the other three OPEs: bis(2-chloroethyl) phosphate (BCEP), bis(1,3-dichloro-2-propyl) phosphate (BDCPP) and dibutyl phosphate (DBUP). The association of DPHP with an increased prevalence OR of MUI in women is a public health concern; future prospective studies are needed to explore its potential mechanism.

Recommendations for defining and reporting adherence measured by Biometric Monitoring Technologies (BioMeTs): A systematic review (Preprint)

BACKGROUND Sub-optimal adherence to data collection procedures and/or a study intervention is often the cause of a failed clinical trial. Data from biometric monitoring technologies (BioMeTs) can measure adherence to both digital therapeutics and digital data collection procedures, thereby providing the opportunity to identify the determinants of adherence and thereafter, methods to maximize adherence. OBJECTIVE Our goal was to describe the methods and definitions by which adherence has been captured and reported using BioMeTs in recent years. Identifying key gaps allowed us to make recommendations regarding minimum reporting requirements and consistency of definitions for BioMeT-based adherence data. METHODS We conducted a systematic review of studies published between 2014 and 2019 that deployed a BioMeT outside of the clinical/lab setting for which a quantitative, non-surrogate, sensor-based measurement of adherence was reported. After systematically screening manuscripts for eligibility, we extracted details regarding study design, participants, the BioMeT/s used, and the definition and units of adherence. Primary definitions of adherence were categorized as a continuous variable based on duration (highest resolution), a continuous variable based on the number of measurements completed, or a categorical variable (lowest resolution). RESULTS Our PubMed search terms identified 940 manuscripts; 100 met our eligibility criteria, which contained descriptions of 110 BioMeTs. We identified 37 unique definitions of adherence reported for 110 BioMeTs, and observed that the uniformity of adherence definitions was associated with the resolution of the data reported. When adherence was reported as a continuous time-based variable, the same definition of adherence was adopted for 92% of the tools; however, when adherence data were simplified to a categorical variable, we observed 25 unique definitions of adherence reported for 37 tools. CONCLUSIONS We recommend that: A) quantitative, non-surrogate, sensor-based, adherence data be reported for all BioMeTs when feasible; B) a clear description of the sensor/s used to capture adherence data, the algorithm/s that convert sample-level measurements to a metric of adherence, and the analytical validation data demonstrating that BioMeT-generated adherence is an accurate and reliable measurement of actual usage, be provided when available; and C) primary adherence data be reported as a continuous variable followed by categorical definitions if needed, and that the categories adopted are supported by clinical validation data and/or consistent with previous reports.

Association of stent diameter and target vessel revascularization in patients undergoing percutaneous coronary intervention: a secondary retrospective analysis based on a Chinese cohort study

Background In the treatment of coronary heart disease, target vessel revascularization (TVR) has attracted increasing attention as an efficient means of percutaneous coronary intervention (PCI). The purpose of this study was to explore the association between stent diameter and TVR in patients undergoing PCI. Methods This was a secondary retrospective analysis involving patients with PCI with at least one stent implanted. Information was obtained from the Dryad Digital Repository. Multivariable logistic regression models, interaction analyses, subgroup analyses and piecewise linear regression models were used to evaluate the association between stent diameter and TVR. Results A total of 2522 patients were eventually enrolled in this study, of which 122 (4.8%) had undergone TVR. Significant positive associations were observed between stent diameter and TVR (continuous: odds ratio [OR] 0.485, 95% confidence interval [CI] 0.305–0.773, P = 0.002; categorical variable: T2 vs. T1, OR 0.541, 95% CI 0.348–0.843; T3 vs. T1, OR 0.520, 95% CI 0.334–0.809; P for trend = 0.005). The association remained stable in the fully adjusted model (continuous: OR 0.526, 95% CI 0.306–0.902, P = 0.020; categorical variable: T2 vs. T1, OR 0.510, 95% CI 0.310–0.839; T3 vs. T1, OR 0.585, 95% CI 0.352–0.973; P for trend = 0.042). Among the subgroups of differing clinical presentations, stent diameter was a powerful protective factor for TVR, especially in the delayed PCI group (P for interaction = 0.002). The association was highly consistent across all the other subgroups studied (all P for interaction > 0.05). In the piecewise linear regression model, the need for TVR decreased with an increase in stent diameter when this ranged between 2.5 and 2.9 mm (OR 0.01, 95% CI: 0.01–0.13, P < 0.001). Conclusions A large stent diameter is a powerful protective factor for TVR in PCI patients, especially in the delayed PCI group. This “bigger-is-better” protective effect is remarkable in stents with diameter 2.5–2.9 mm.

Money does not always buy happiness, but are richer people less happy in their daily lives? It depends on how you analyze income

Do people who have more money feel happier during their daily activities? Some prior research has found no relationship between income and daily happiness when treating income as a continuous variable in OLS regressions, although results differ between studies. We re-analyzed existing data, treating household income as a categorical variable and using lowess and spline regressions to explore non-linearities. Our analyses reveal that these methodological decisions provide new insights into the relationship between income and happiness. We find some evidence that higher income is associated with less happiness and no substantive benefit to higher household income in the US after $35-40K and in Germany after €14-18K. Not all analytic approaches generate the same conclusions, which may explain discrepant results.

Neutrophil-to-Lymphocyte Ratio and Treatment Failure in Peritoneal Dialysis-Associated Peritonitis

Objective: We sought to explore if there is an association between neutrophil-to-lymphocyte ratio (NLR) and treatment failure in patients with peritoneal dialysis-associated peritonitis (PDAP).Methods: Our cohort involved 337 episodes of PDAP experienced by 202 patients who were undergoing continuous ambulatory peritoneal dialysis at a single center from 1 July 2013 to 30 June 2018. The exposures were log-transformed NLR and a categorical variable grouped by the tertiles of NLR levels (T1, &lt;3.75; T2, 3.75–6.53; and T3, &gt;6.53) at baseline. Generalized estimating equation (GEE) and restricted cubic spline (RCS) analyses were done to determine the association between NLR and treatment failure, defined as catheter removal or all-cause mortality during therapy.Results: After adjusting for other potential predictors, the log-transformed NLR exhibited an incremental relationship with the risk of treatment failure (odds ratio, 1.82; 95% confidence interval, 1.05–3.15). RCS analyses showed that the relationship was positively and linearly correlated (P for nonlinearity = 0.104). As a three-level categorical variable, in reference to T1, the T3 of NLR showed a 3.41-fold increased venture of treatment failure in fully adjusted model. Subgroup analyses suggested that the prognostic relevance of NLR in PDAP was particularly significant in gram-negative peritonitis.Conclusions: A greater level of NLR at baseline was remarkably associated with a higher incidence of treatment failure among PDAP episodes regardless of other potential risk factors.

Relationship Between Mean Vancomycin Trough Concentration and Mortality in Critically Ill Patients: A Multicenter Retrospective Study

Background: It remains unclear whether the mean vancomycin trough concentration (VTC) derived from the entire course of therapy is of potential benefit for critically ill patients. This study was conducted to explore the association between mean serum VTC and mortality in intensive care units (ICUs).Methods: 3,603 adult patients with two or more VTC records after receiving vancomycin treatment in the eICU Collaborative Research Database were included in this multicenter retrospective cohort study. Mean VTC was estimated using all measured VTCs and investigated as a continuous and categorical variable. Patients were categorised into four groups according to mean VTC: &lt;10, 10–15, 15–20, and &gt;20 mg/L. Multivariable logistic regression and subgroup analyses were performed to investigate the relationship of mean VTC with mortality.Results: After adjusting for a series of covariates, logistic regression analyses indicated that mean VTC, as a continuous variable, was positively correlated with ICU (odds ratio, 1.038, 95% confidence interval, [1.014–1.063]) and hospital (1.025 [1.005–1.046]) mortalities. As a categorical variable, mean VTC of 10–15 mg/L was not associated with reduced ICU (1.705 [0.975–2.981]) and hospital (1.235 [0.829–1.841]) mortalities. Mean VTC of 15–20 mg/L was not correlated with a lower risk of hospital mortality (1.370 [0.924–2.029]). Moreover, mean VTCs of 15–20 and &gt;20 mg/L were significantly associated with higher ICU mortality (1.924 [1.111–3.332]; 2.428 [1.385–4.258]), and mean VTC of &gt;20 mg/L with higher hospital mortality (1.585 [1.053–2.387]) than mean VTC of &lt;10 mg/L. Similar results were observed in patients with different Acute Physiology and Chronic Health Evaluation IV score, creatinine clearance, age, and body mass index subgroups.Conclusion: Mean VTC was not associated with reduced ICU/hospital related mortality. Our results suggested that VTC monitoring might not guarantee vancomycin efficacy for ICU patients.