The first-order Markov conditional linear expectation approach for analysis of longitudinal data.

We consider longitudinal discrete data that may be unequally spaced in time and may exhibit overdispersion, so that the variance of the outcome variable is inflated relative to its assumed distribution. We implement an approach that extends generalized linear models for analysis of longitudinal data and is likelihood based, in contrast to generalized estimating equations (GEE) that are semiparametric. The method assumes independence between subjects; first-order antedependence within subjects; exponential family distributions for the first outcome on each subject and for the subsequent conditional distributions; and linearity of the expectations of the conditional distributions. We demonstrate application of the method in an analysis of seizure counts and in a study to evaluate the performance of transplant centers. Simulations for both studies demonstrate the benefits of the proposed likelihood based approach; however, they also demonstrate better than anticipated performance for GEE.

Physician-patient communication at prescription of an additional oral drug for type 2 diabetes and its links to patient outcomes - New findings from the global IntroDia® study.

AIMS: To investigate experiences of people with type 2 diabetes (T2DM) at the clinic visit when an additional oral antidiabetes drug (OAD) is prescribed, and how this affects their quality of life, self-management and key outcomes. METHODS: We surveyed adults with T2DM from a large multinational study of patient-physician communication during early T2DM treatment (IntroDia®). We examined their experiences when an additional OAD is prescribed ("add-on") after initial OAD monotherapy, focusing on 24 key conversational elements, overall patient-perceived communication quality (PPCQ), and associations with current patient-reported outcomes. The links between PPCQ and people's efforts to delay add-on therapy were also assessed. RESULTS: 4235 people with T2DM prescribed an additional OAD, or a combination of two, were analysed. Exploratory factor analyses of the conversational elements during add-on yielded three coherent, meaningful factors: Encouraging (Cronbach's α = 0.62), Collaborative (α = 0.81), and Discouraging (α = 0.81). PPCQ was positively associated with Encouraging (ß = +1.252, p < 0.001) and Collaborative (ß = +1.206, p < 0.001), but negatively associated with Discouraging (ß = -0.895, p < 0.001). Better PPCQ at add-on was associated with less diabetes distress, greater well-being and better self-care at the present time. Approximately 20% of people bargained (two-thirds successfully) with their physician to delay additional medication. Non-bargaining individuals reported significantly better mean PPCQ, diabetes distress, well-being and self-care than those who bargained. CONCLUSIONS: Encouraging and patient-inclusive conversations at add-on moments may improve patient well-being and self-care outcomes. People with T2DM who attempted to delay additional medication reported poorer PPCQ and outcomes.

Physician experiences when discussing the need for additional oral medication with type 2 diabetes patients: Insights from the cross-national IntroDia® study.

AIMS: Physician-patient communication when discussing the need for additional oral medication for type 2 diabetes (add-on) may affect the self-care of people with this condition. We aimed to investigate physicians' recalled experiences of the add-on consultation. METHODS: We conducted a cross-sectional survey of physicians treating people with type 2 diabetes in 26 countries, as part of a large cross-national study of physician-patient communication during early treatment of type 2 diabetes (IntroDia®). The survey battery included novel questions about physician experiences at add-on and the Jefferson Scale of Physician Empathy. RESULTS: Of 9247 eligible physicians, 6753 responded (73.0% response rate). Most (82%) agreed that physician-patient discussions at add-on strongly influence patients' disease acceptance and treatment adherence. Half the physicians reported ≥1 challenge in most or all add-on conversations, with a significant inverse relationship between frequency of challenges and Jefferson Scale of Physician Empathy score (standardised ß coefficient: -0.313; p < 0.001). Physicians estimated that only around half their patients with type 2 diabetes follow their self-care advice. Exploratory factor analysis of physician beliefs about why their patients did not follow recommendations yielded two distinct dimensions: psychosocial barriers (e.g. depressed mood) and personal failings of the patient (e.g. not enough willpower) (r = 0.37, p < 0.001). CONCLUSIONS: Physicians' empathy and beliefs about their patients may play a significant role in their success with the add-on conversation and, consequently, promotion of patient engagement and self-care. Although the study was limited by its retrospective, cross-sectional nature, the findings from IntroDia® may inform efforts to improve diabetes care.

Physician-patient communication at diagnosis of type 2 diabetes and its links to patient outcomes: New results from the global IntroDia® study.

AIMS: To investigate patient experiences during the diagnosis of type 2 diabetes mellitus (T2DM), focusing on how physician-patient communication at diagnosis influences patients' psychosocial stress and subsequent self-management and outcomes. METHODS: We surveyed adults with T2DM in 26 countries in a large cross-national study of physician-patient communication during early T2DM treatment (IntroDia®). The self-report questionnaire assessed retrospectively patient experiences during diagnosis conversations (focusing on 43 possible conversational elements, and communication quality) and potential effects on patient-reported outcomes. RESULTS: Data from 3628 people with T2DM who had been prescribed oral treatment at diagnosis were analysed. Exploratory factor analyses of the conversational elements yielded four coherent, meaningful factors: Encouraging (Cronbach's α=0.86); Collaborative (α=0.88); Recommending Other Resources (α=0.75); and Discouraging (α=0.72). Patient-perceived communication quality (PPCQ) at diagnosis was positively associated with Encouraging (ß=+1.764, p<0.001) and Collaborative (ß=+0.347, p<0.001), negatively associated with Discouraging (ß=-1.181, p<0.001) and not associated with Recommending Other Resources (ß=+0.087, p=0.096), using a stable path model. PPCQ was associated with less current diabetes distress, greater current well-being and better current self-care. Conversation elements comprising factors associated with better PPCQ (Encouraging and Collaborative) were recalled more frequently by patients than elements associated with poor PPCQ (Discouraging). CONCLUSIONS: Better physician-patient communication at T2DM diagnosis may contribute to subsequent greater patient well-being and self-care, and may be enhanced by greater physician use of Collaborative and Encouraging conversation elements.

Maximum likelihood based analysis of equally spaced longitudinal count data with first-order antedependence and overdispersion.

This manuscript implements a maximum likelihood based approach that is appropriate for equally spaced longitudinal count data with over-dispersion, so that the variance of the outcome variable is larger than expected for the assumed Poisson distribution. We implement the proposed method in the analysis of seizure data and a subset of German Socio-Economic Panel data. To demonstrate the importance of correctly modeling the over-dispersion, we make comparisons with the semi-parametric generalized estimating equations approach that incorrectly ignores any over-dispersion in the data. Our simulations demonstrate that accounting for over-dispersion results in improved small-sample efficiency and appropriate coverage probabilities. We also provide code in R so that readers can implement our approach in their own analyses.

Rationale for and design of the Idiopathic Pulmonary Fibrosis-PRospective Outcomes (IPF-PRO) registry.

BACKGROUND: Idiopathic pulmonary fibrosis (IPF) is a chronic interstitial lung disease characterised by progressive loss of lung function. Its clinical course is variable but ultimately fatal. There is a need for a multicentre patient registry incorporating longitudinal clinical data and biological samples to improve understanding of the natural history of IPF and contemporary practice patterns. METHODS/DESIGN: The Idiopathic Pulmonary Fibrosis-PRospective Outcomes (IPF-PRO) registry is a national IPF registry in the USA. This registry will enrol approximately 300 patients with newly diagnosed IPF over 2 years at approximately 14 tertiary pulmonary care sites. Participants will be followed for 3-5 years and will receive usual care, as defined by their physician. Clinical data from the year prior to diagnosis will be collected from medical record review on enrolment. Subsequently, data on diagnostic evaluations, pulmonary function tests, physical examinations, laboratory data and clinical events will be collected at routine clinical visits and via a call centre. Participants will complete patient-reported outcome questionnaires at enrolment and then at approximately 6-month intervals. Blood samples for cellular, genetic and transcriptomic analyses will be collected at the same intervals. RESULTS: The first results from the IPF-PRO registry will be presented in 2015. CONCLUSIONS: The IPF-PRO registry will improve understanding of the natural history of IPF, its impact on patients and current practice in the diagnosis and care of patients with IPF. The registry will establish a repository of biological samples from a well-characterised patient population for future research. CLINICAL TRIAL NUMBER: NCT01915511.

Late consequential surgical bed soft tissue necrosis in advanced oropharyngeal squamous cell carcinomas treated with transoral robotic surgery and postoperative radiation therapy.

PURPOSE: A subset of patients with oropharyngeal squamous cell carcinoma (OP-SCC) managed with transoral robotic surgery (TORS) and postoperative radiation therapy (PORT) developed soft tissue necrosis (STN) in the surgical bed months after completion of PORT. We investigated the frequency and risk factors. MATERIALS AND METHODS: This retrospective analysis included 170 consecutive OP-SCC patients treated with TORS and PORT between 2006 and 2012, with >6 months' of follow-up. STN was defined as ulceration of the surgical bed >6 weeks after completion of PORT, requiring opioids, biopsy, or hyperbaric oxygen therapy. RESULTS: A total of 47 of 170 patients (28%) had a diagnosis of STN. Tonsillar patients were more susceptible than base-of-tongue (BOT) patients, 39% (41 of 104) versus 9% (6 of 66), respectively. For patients with STN, median tumor size was 3.0 cm (range 1.0-5.6 cm), and depth of resection was 2.2 cm (range 1.0-5.1 cm). Median radiation dose and dose of fraction to the surgical bed were 6600 cGy and 220 cGy, respectively. Thirty-one patients (66%) received concurrent chemotherapy. Median time to STN was 2.5 months after PORT. All patients had resolution of STN after a median of 3.7 months. Multivariate analysis identified tonsillar primary (odds ratio [OR] 4.73, P=.01), depth of resection (OR 3.12, P=.001), total radiation dose to the resection bed (OR 1.51 per Gy, P<.01), and grade 3 acute mucositis (OR 3.47, P=.02) as risk factors for STN. Beginning May 2011, after implementing aggressive avoidance of delivering >2 Gy/day to the resection bed mucosa, only 8% (2 of 26 patients) experienced STN (all grade 2). CONCLUSIONS: A subset of OP-SCC patients treated with TORS and PORT are at risk for developing late consequential surgical bed STN. Risk factors include tonsillar location, depth of resection, radiation dose to the surgical bed, and severe mucositis. STN risk is significantly decreased with carefully avoiding a radiation dosage of >2 Gy/day to the surgical bed.

The immunomodulatory effects of pegylated liposomal doxorubicin are amplified in BRCA1--deficient ovarian tumors and can be exploited to improve treatment response in a mouse model.

OBJECTIVE: Women with BRCA-associated ovarian cancer demonstrate excellent responses to Pegylated Liposomal Doxorubicin (PLD). PLD has also been shown to enhance T cell recognition of tumor cells. Here we characterize immunophenotypic changes associated with BRCA1 dysfunction in ovarian cancer cells, and evaluate the T cell contribution to the therapeutic efficacy of PLD in a BRCA1- ovarian cancer model to determine whether enhanced anti-tumor immunity contributes to the improved response to PLD in BRCA1- ovarian cancers. METHODS: The immunophenotype of BRCA1- and wild-type (WT) ovarian cancer cells and their response to PLD were compared in vitro using flow cytometry. T cell recruitment to BRCA1- tumors was evaluated with flow cytometry and immunohistochemistry. The contribution of T cell populations to the therapeutic effect of PLD in a BRCA1- model was evaluated using immunodepleting antibodies with PLD in vivo. RESULTS: The cytotoxic response to PLD was similar in BRCA1- and WT cells in vitro. BRCA1- inactivation resulted in higher expression of Fas and MHC-I at baseline and after PLD exposure. PLD prolonged the survival of BRCA1- tumor bearing mice and increased intratumoral T cell recruitment. CD4+ depletion combined with PLD significantly prolonged overall survival (p=0.0204) in BRCA1- tumor-bearing mice. CONCLUSION: Differences in the immunophenotype of BRCA1- and WT cells are amplified by PLD exposure. The enhanced immunomodulatory effects of PLD in BRCA1- tumors may be exploited therapeutically by eliminating suppressive CD4+ T cells. Our results support further study of combination therapy using PLD and immune agents, particularly in women with BRCA gene mutations.

Phase I/II randomized trial of dendritic cell vaccination with or without cyclophosphamide for consolidation therapy of advanced ovarian cancer in first or second remission.

PURPOSE: In spite of increased rates of complete response to initial chemotherapy, most patients with advanced ovarian cancer relapse and succumb to progressive disease. Immunotherapy may have potential for consolidation therapy. EXPERIMENTAL DESIGN: This randomized open-label phase I/II trial evaluated responses of patients with advanced ovarian cancer in remission for vaccination with monocyte-derived dendritic cells (DC) loaded with Her2/neu, hTERT, and PADRE peptides, with or without low-dose intravenous cyclophosphamide. All patients also received pneumococcal vaccine and were randomized to cyclophosphamide 2 days prior to first vaccination. Blood samples were analyzed by ELISPOT and flow cytometry. RESULTS: Of 11 patients, 2 recurred during vaccination. Nine received all 4 doses: 3 patients recurred at 6, 17, and 26 months, respectively, and 6 have no evidence of disease at 36 months. No grade 3/4 vaccine-related toxicities were noted. The 3-year overall survival was 90%. Patients receiving cyclophosphamide showed a non-significant improvement in survival over controls. Patients receiving cyclophosphamide had a transient reduction in neutrophils, but no change in total lymphocytes or regulatory T cells. Modest T-cell responses to Her2/neu and hTERT were seen post-vaccine by IFN-γ ELISPOT. Patients demonstrated below normal responses to the diphtheria conjugate protein CRM197, a component of the pneumococcal vaccine. CONCLUSIONS: In this setting, peptide-loaded DC vaccination elicits modest immune responses, but survival is promising. Pneumococcal vaccination revealed substantial immune suppression, even in patients in remission. Rational design of consolidative strategies for ovarian cancer will need to overcome tolerance and immunosuppression.

The influence of gender, race, and marital status on survival in lung cancer patients: analysis of Radiation Therapy Oncology Group trials.

OBJECTIVE: A meta-analysis was conducted to determine the influence of gender, race, and marital status on overall survival (OS) in Radiation Therapy Oncology Group nonoperative non-small cell lung cancer trials. MATERIALS AND METHODS: Data from 1365 patients treated on nine prospective Radiation Therapy Oncology Group studies activated during the 1990s were analyzed. Impact of gender, marital status, and race was considered in the Cox proportional hazards models. Age, Karnofsky performance status, weight loss, stage, histology, location of primary tumor, biologic equivalent dose, deviation from protocol dose, and education level were adjusted in the model. A two-sided p value <0.05 was considered statistically significant. RESULTS: Males had significantly higher mortality than females adjusted for other covariates (hazard ratio [HR] 1.22, 95% confidence interval 1.08 -1.38). Race and marital status were not independently predictive for OS. Single females had significantly better OS than single males (HR 0.72), and married males had lower OS than single females (HR 1.36). CONCLUSIONS: These results suggest that although certain subgroups of gender, race, and/or marital status have better outcomes with respect to OS; gender seems to be the most significant factor influencing survival results among nonoperative non-small cell lung cancer patients.