First-Order Learning Models With the GDINA: Estimation With the EM Algorithm and Applications.

In learning environments, understanding the longitudinal path of learning is one of the main goals. Cognitive diagnostic models (CDMs) for measurement combined with a transition model for mastery may be beneficial for providing fine-grained information about students' knowledge profiles over time. An efficient algorithm to estimate model parameters would augment the practicality of this combination. In this study, the Expectation-Maximization (EM) algorithm is presented for the estimation of student learning trajectories with the GDINA (generalized deterministic inputs, noisy, "and" gate) and some of its submodels for the measurement component, and a first-order Markov model for learning transitions is implemented. A simulation study is conducted to investigate the efficiency of the algorithm in estimation accuracy of student and model parameters under several factors-sample size, number of attributes, number of time points in a test, and complexity of the measurement model. Attribute- and vector-level agreement rates as well as the root mean square error rates of the model parameters are investigated. In addition, the computer run times for converging are recorded. The result shows that for a majority of the conditions, the accuracy rates of the parameters are quite promising in conjunction with relatively short computation times. Only for the conditions with relatively low sample sizes and high numbers of attributes, the computation time increases with a reduction parameter recovery rate. An application using spatial reasoning data is given. Based on the Bayesian information criterion (BIC), the model fit analysis shows that the DINA (deterministic inputs, noisy, "and" gate) model is preferable to the GDINA with these data.

A Hidden Markov Model for Learning Trajectories in Cognitive Diagnosis With Application to Spatial Rotation Skills.

The increasing presence of electronic and online learning resources presents challenges and opportunities for psychometric techniques that can assist in the measurement of abilities and even hasten their mastery. Cognitive diagnosis models (CDMs) are ideal for tracking many fine-grained skills that comprise a domain, and can assist in carefully navigating through the training and assessment of these skills in e-learning applications. A class of CDMs for modeling changes in attributes is proposed, which is referred to as learning trajectories. The authors focus on the development of Bayesian procedures for estimating parameters of a first-order hidden Markov model. An application of the developed model to a spatial rotation experimental intervention is presented.

Bayesian Estimation of the DINA Q matrix.

Cognitive diagnosis models are partially ordered latent class models and are used to classify students into skill mastery profiles. The deterministic inputs, noisy "and" gate model (DINA) is a popular psychometric model for cognitive diagnosis. Application of the DINA model requires content expert knowledge of a Q matrix, which maps the attributes or skills needed to master a collection of items. Misspecification of Q has been shown to yield biased diagnostic classifications. We propose a Bayesian framework for estimating the DINA Q matrix. The developed algorithm builds upon prior research (Chen, Liu, Xu, & Ying, in J Am Stat Assoc 110(510):850-866, 2015) and ensures the estimated Q matrix is identified. Monte Carlo evidence is presented to support the accuracy of parameter recovery. The developed methodology is applied to Tatsuoka's fraction-subtraction dataset.

Case-control study of cervical cancer and gynecologic screening: A SEER-Medicare analysis.

OBJECTIVE: To determine the association between Pap smear and pelvic examination screenings and the development of invasive cervical cancer in a Medicare population using a matched case-control design. METHODS: Matched case-control data sets were constructed from the SEER-Medicare database that links the Surveillance Epidemiology End Results (SEER) cancer registry data and Medicare enrollment and claims data of subjects who received care between the years 1991 and 1999 aged 65years or older. The study identified 1267 cervical cancer cases. Controls (N=10.137) were matched to cases representing up to eight matched controls (on age and registry geographic location) for a single case. The association between gynecologic screenings and the development of invasive cervical cancer was ascertained using conditional logistic regression analysis. RESULTS: Having had a Pap smear during the PIDP (pre-invasive detectable phase - 2 to 7years prior to diagnosis) was significantly negatively associated with the development of invasive cervical cancer (OR=0.64, 95% CI=0.53-0.78) which was reduced after taking into account the estimated prevalence of hysterectomy among controls (OR=0.38, 95% CI=0.32-0.46). The negative association between Pap smear screenings and cervical cancer was strongest for squamous tumors (OR=0.48, 95% CI=0.37-0.61). Restricting the subjects to those 72 and over did not affect risk. CONCLUSIONS: There is a reduction in risk for invasive cervical cancer when women over age 65 are screened. This suggests that cervical cancer screening in the aged population may be beneficial.

A phase I study to determine the maximum tolerated dose and safety of oral LR-103 (1α,24(S)Dihydroxyvitamin D2) in patients with advanced cancer.

BACKGROUND: The objective of this study was to determine the maximum tolerated dose and safety of LR-103, a Vitamin D analogue, in patients with advanced cancer. METHODS: In Step A, patients received oral LR-103 once daily in 14-day cycles with intra-patient dose escalation per accelerated dose escalation design. Dose limiting toxicity for Step A was defined as ≥grade 2 hypercalcemia and/or >grade 2 other toxicities. Starting dose was 5 µg/day. Step B used a 3+3 design starting at Step A maximum tolerated dose with 28-day cycles. Dose limiting toxicity was defined as ≥grade 3 hypercalcemia or any grade 3 or 4 non-hematologic toxicity, except hypercalciuria. RESULTS: Twenty-one patients were enrolled; eight were treated in Step A. At dose level 3 (15 µg/day), two patients had dose limiting toxicity. One had grade 4 hyperuricemia. The other had grade 4 GGT plus grade 3 alkaline phosphatase, fatigue and urinary tract infection (UTI). Dose level 2 (10 µg/day) was the maximum tolerated dose for Step A and was starting dose for Step B. The dose was escalated to dose level 5 (30 µg/day) with a patient experiencing grade 3 dose limiting toxicity of hypercalcemia. The study was discontinued before reaching the maximum tolerated dose due to sponsor decision. Modest increases in serum osteocalcin and calcium and decrease in parathyroid hormone were noted. Best response was stable disease; four patients were on therapy for six months or longer. CONCLUSION: Step A dose limiting toxicities limited accelerated dose escalation. The maximum tolerated dose of LR-103 was not reached prior to study termination and this agent is no longer being developed.

The linear transformation model with frailties for the analysis of item response times.

The item response times (RTs) collected from computerized testing represent an underutilized source of information about items and examinees. In addition to knowing the examinees' responses to each item, we can investigate the amount of time examinees spend on each item. In this paper, we propose a semi-parametric model for RTs, the linear transformation model with a latent speed covariate, which combines the flexibility of non-parametric modelling and the brevity as well as interpretability of parametric modelling. In this new model, the RTs, after some non-parametric monotone transformation, become a linear model with latent speed as covariate plus an error term. The distribution of the error term implicitly defines the relationship between the RT and examinees' latent speeds; whereas the non-parametric transformation is able to describe various shapes of RT distributions. The linear transformation model represents a rich family of models that includes the Cox proportional hazards model, the Box-Cox normal model, and many other models as special cases. This new model is embedded in a hierarchical framework so that both RTs and responses are modelled simultaneously. A two-stage estimation method is proposed. In the first stage, the Markov chain Monte Carlo method is employed to estimate the parametric part of the model. In the second stage, an estimating equation method with a recursive algorithm is adopted to estimate the non-parametric transformation. Applicability of the new model is demonstrated with a simulation study and a real data application. Finally, methods to evaluate the model fit are suggested.

Collection of the BD SurePath Pap Test with a broom device plus endocervical brush improves disease detection when compared to the broom device alone or the spatula plus endocervical brush combination.

OBJECTIVE: Here we examine the diagnostic utility of the US Food And Drug Administration (FDA) approved Spatula + endocervical brush combination for the BD SurePath Pap Test (SPPT) and compare it to SPPT collection with the broom alone or to an off-label combination of broom + EC brush. This question is important due to lingering concerns over the value of EC detection to a satisfactory Pap test. METHODS: 20,125 SPPT vials were examined for the collection devices contained. The SPPT collection device combinations allowed were: Rovers Cervex-Brush (broom, FDA approved), Medscand Pap Perfect Spatula + Medscand CytoBrush Plus GT (spatula + GT brush, FDA approved) or Rovers Cervex-Brush + Surgipath C-E Brush (broom + CE brush, off label). RESULTS: Examination of SPPT vials revealed 11,130 collected with the broom, 4,687 collected with the spatula + GT brush and 2,921 collected with the broom + CE brush. Absence of an endocervical/transformation zone was seen in 22.86% of broom cases, 13.10% of spatula + GT brush cases (p= 0.00005 vs broom) and 10.17% of broom + CE brush cases (p= 0.00005 vs broom, p= 0.00005 vs spatula + GT brush). Importantly, LSIL detection was: broom 2.99%; spatula + GT brush 2.45% (p= 0.053 vs broom); broom + CE brush 4.18% (p= 0.034 vs broom, p= 0.0001 vs spatula + GT brush). CONCLUSION: When broom + brush combination is compared to broom alone or to spatula + GT brush, the broom + CE brush combination better sampled the endocervical/transformation zone and increased LSIL detection.

Phase II study of bevacizumab in patients with platinum-resistant ovarian cancer or peritoneal serous cancer.

PURPOSE: We evaluated the efficacy and safety of bevacizumab in patients with platinum-resistant epithelial ovarian carcinoma (EOC) or peritoneal serous carcinoma (PSC) who had experienced disease progression during, or within 3 months of discontinuing, topotecan or liposomal doxorubicin. PATIENTS AND METHODS: No more than three prior treatment regimens were allowed. Patients received single-agent bevacizumab 15 mg/kg intravenously every 3 weeks. Response was assessed by computed tomography (CT) scan every 6 weeks using Response Evaluation Criteria in Solid Tumors (RECIST). RESULTS: Of 44 patients treated, 83.7% were primarily platinum resistant, 59.1% had received liposomal doxorubicin, 25% topotecan, 15.9% both agents, and 47.7% had received three prior chemotherapy regimens. A median of five (range, two to 16) bevacizumab doses were administered. Partial responses were observed in seven patients (15.9%). Median progression-free survival was 4.4 months (95% CI, 3.1 to 5.5 months), with a median survival duration of 10.7 months at study termination. Bevacizumab-associated grade 3 to 4 events included hypertension (9.1%), proteinuria (15.9%), bleeding (2.3%), and wound-healing complications (2.3%). The incidence of GI perforation (GIP; 11.4%) was higher than reported in bevacizumab trials of other tumor types. GIP occurred in 23.8% of patients receiving three prior chemotherapy regimens, compared with 0% of patients receiving two prior chemotherapy regimens (P < .01). A trend toward higher risk of GIP was observed for patients with bowel wall thickening or bowel obstruction on CT scan. Arterial thromboembolic events occurred in three patients (6.8%). Three deaths were related to bevacizumab treatment. CONCLUSION: Bevacizumab has single-agent activity in patients with platinum-resistant EOC or PSC. A higher than expected incidence of GIP was noted in these heavily pretreated patients.

Study of common functional genetic polymorphisms of FCGR2A, 3A and 3B genes and the risk for cryptococcosis in HIV-uninfected patients.

A pilot candidate gene association study was conducted to investigate the role of three common functional genetic polymorphisms of the low-affinity Fc gamma receptors, FCGR2A (131H/R), FCGR3A (158F/V) and FCGR3B (NA1/NA2) in Cryptococcus neoformans infections in individuals not infected with HIV. The FCGR2A 131RR and FCGR3A 158VV genotypes were over-represented [OR: 1.67 (1.05-2.63) and 2.04 (1.06-4.00), respectively] whereas the FCGR3B NA2NA2 was under-represented in patients with cryptococcosis (28% vs. 40% in controls). An analysis of haplotypes showed a significant difference in distribution between cases and controls overall and in Caucasians.

Quality of life of children with cystic fibrosis.

OBJECTIVE: To review health-related quality of life (QOL) and associated issues and to describe a study investigating "Child Health Questionnaire" (CHQ) scores in relationship to newborn screening (NBS) for cystic fibrosis (CF) and markers of disease severity. METHODS: A total of 36 patients from 10-15.5 years old who were enrolled in the screened or control group of the Wisconsin CF Neonatal Screening Project completed the CHQ. Scale scores comprised the dependent variables. Independent variables included study group and measures of disease severity. Analyses included Fisher's exact, 2-sample Wilcoxon, and t tests. RESULTS: QOL did not differ significantly between the screened and control groups for any of the scales. None of the comparisons of CHQ scale scores across measures of disease severity were significant in this small sample, but the CHQ and power were limiting. CONCLUSIONS: Our results did not demonstrate a benefit of CF NBS on QOL; however, the CHQ may not be adequately sensitive to QOL in children with CF with disease severity comparable to our sample. The Cystic Fibrosis Questionnaire, a recently validated CF-specific QOL measure for pediatric samples, is likely to provide a more informative evaluation of the effects of CF NBS on patients' QOL.

A randomized, controlled trial to improve advance care planning among patients undergoing cardiac surgery.

BACKGROUND: Although many healthcare providers and researchers consider it necessary to assist patients with end-stage chronic illnesses to plan for the end of life, they tend to avoid discussing end-of-life issues with patients before major surgery. Consequently, surgical patients and their families generally have insufficient knowledge to make plans in case of life-threatening complications. OBJECTIVE: The objective of this study was to evaluate short-term effects of Patient-Centered Advance Care Planning (PC-ACP). DESIGN AND SUBJECTS: Thirty-two dyads of patients undergoing cardiac surgery and their surrogates were randomly assigned to receive either the PC-ACP intervention (PC-ACP) or usual care. MEASURES: Measures studied were patient-surrogate congruence regarding goals for future medical care, patient and surrogate knowledge of advance care planning and anxiety, and patient decisional conflict. Congruence and anxiety were measured before and after the intervention. Decisional conflict and knowledge of advance care planning were measured after the intervention. RESULTS: Compared with the control group, PC-ACP significantly improved patient-surrogate congruence (Delta=1.27, P<0.01) and reduced patients' decisional conflict (Delta=-0.77, P<0.05). Anxiety change (pre/post) did not differ between treatment and control groups. No difference in knowledge of advance care planning was found between the 2 groups. CONCLUSIONS: The PC-ACP can be an effective approach to advance care planning. Its specificity and relevance to patients' actual medical conditions (as exemplified by its plans for potential complications related to cardiac surgery) can lead to greater patient-surrogate congruence without increasing decisional conflict and anxiety.

Worklife and satisfaction of general internists.

BACKGROUND: Prior studies have reported relatively low job satisfaction for general internists. We used data from a large US physician survey to assess correlates of satisfaction of general internists. METHODS: The Physician Worklife Survey was mailed to a national random stratified sample of 5704 US physicians. General internists were assessed for their satisfaction, training, patient mix, work hours, the likelihood of recommending their specialty to medical students, and job stability. We then compared them with a specialist sample (internal medicine subspecialists [IMSSs]) and a primary care sample (family physicians [FPs]). Logistic regression was used to model predictors of satisfaction, stress, and medical student recruitment. RESULTS: There were 2326 respondents (adjusted response rate, 52%): 450 (19%) were general internists; 502 (22%), FPs; and 438 (19%), IMSSs. General internists were less satisfied than were IMSSs with their relationships with colleagues and with patient care issues (P<.01 for both) and less satisfied than were FPs with community ties (P =.001). Global job, career, and specialty satisfaction were significantly lower for general internists vs FPs and IMSSs (P<.05). General internists spent proportionately more of their work week in the hospital than did FPs (20% vs 13%; P<.001) and more time providing outpatient care than did IMSSs (56% vs 42%; P<.001). General internists had more patients with complex medical and psychosocial problems than did FPs (P<.01) but fewer patients with complex medical problems than did IMSSs (P<.001). Higher satisfaction for general internists was associated with older physician age, less time pressure during office visits, fewer work hours, and fewer patients with complex psychosocial problems (P<.05 for all). General internists were less likely than were FPs to recommend their specialty to medical students (P<.001). Specialty satisfaction, female gender, and control of hassles predicted medical student recruitment by general internists. CONCLUSIONS: General internists' role of caring for patients with complex problems is associated with lower levels of satisfaction than for IMSSs and FPs. Adjusting caseload for patient complexity, expanding time for office visits, and additional training in the care of patients with psychosocially complex problems may improve the job satisfaction of general internists and medical student recruitment into the specialty.

Item response models for joint analysis of quality of life and survival.

A complication when assessing quality of life data longitudinally is that in many trials a substantial percentage of patients die before completing all of the assessments. Furthermore, a patient's risk of dying might be predicted by his current quality of life. This suggests jointly modelling quality of life and survival, and using this combined information to summarize the outcome. The aim of this paper is to address the complicated issues, such as death, present in analysing multiple-item ordinal quality of life data in clinical trials while recognizing the psychometric properties of the quality of life instrument being used. This is done by combining an item response model and Cox's proportional hazard model, where a latent variable process for quality of life determines the probability of selecting various options on quality of life items, and also serves as a time-dependent covariate in the survival model. We accomplish this by using Markov chain Monte Carlo methods to obtain parameter estimates. Then we compute a summary measure, area-under-QOL curve, to compare the efficacy of the treatments. The methods are illustrated with analysis of data from the Vesnarinone trial of patients with severe heart failure, in which quality of life was assessed with the Minnesota Living with Heart Failure Questionnaire.