Abstract This article introduces a general method for Bayesian computing in richly parameterized models, structured Markov chain Monte Carlo (SMCMC), that is based on a blocked hybrid of the Gibbs sampling and Metropolis—Hastings algorithms. SMCMC speeds algorithm convergence by using the structure that is present in the problem to suggest an appropriate Metropolis—Hastings candidate distribution. Although the approach is easiest to describe for hierarchical normal linear models, we show that its extension to both nonnormal and nonlinear cases is straightforward. After describing the method in detail we compare its performance (in terms of run time and autocorrelation in the samples) to other existing methods, including the single-site updating Gibbs sampler available in the popular BUGS software package. Our results suggest significant improvements in convergence for many problems using SMCMC, as well as broad applicability of the method, including previously intractable hierarchical nonlinear model settings. Key Words: BlockingConvergence accelerationGibbs samplingHierarchical modelMetropolis-Hastings algorithm
Although physical activity ameliorates the metabolic impact of high body mass index (BMI), runners with BMI≥25 kg/m2 are relatively understudied.This study had two goals: 1) Identify differences in body composition, as measured by dual X-ray absorptiometry (DXA), between overweight (BMI≥25 kg/m2) runners (OWR) and normal weight (BMI<25 kg/m2) runners (NWR) and 2) Examine whether an 90 minute run alters total or regional fat mass, as measured by DXA, in OWR and NWR. We hypothesized that OWR would have higher total body fat than NWR and OWR with greater changes in visceral fat after a prolonged run.Body composition analysis before and after a supervised run.We recruited NWR (n=16,F:n=7,28.1±1.4 years, BMI 22.0±0.4 kg/m2, results as mean±SE) and OWR (n=11,F:n=7,32.0±1.6 years, BMI 30.5±1.4kg/m2) participants. DXA based body composition was measured before and after a supervised, 90 minute run at 60% heart rate reserve.OWR had higher body fat than NWR in all measured regions.. Both groups did not significantly reduce fat mass at any measured fat depots after the running exposure.OWR had higher body fat in all measured regions than NWR. DXA could not demonstrate any acute fat mass changes after a prolonged run.
Objective: Compare effectiveness of two differently formatted training programs in educating night-time postural care implementers. Design: Mixed-methods parallel-group double-blind design with random assignment. Setting: United States academic institution. Participants: Thirty-eight adult caregivers/providers of children with cerebral palsy. Interventions: Both 2-hour online programs included content on night-time postural care evidence, risk-factor monitoring, sleep-system types, positioning methods, and assessments. Group A used interactive videos, Group B summary information with web-links. Main Measures: We measured self-perceived competence via questionnaires (baseline, post-training, post-simulation) containing 4-point rating-scales of knowledge, ability, and confidence and measured positioning ability via a simulation observation instrument comprising 16 positioning-task ratings with space for describing performance. We recorded participant actions/statements using fieldnotes. Results: Thirty-eight completed training (19 per group). Group A (vs B) showed significantly greater self-perceived competence changes post-training (0.46 points (SE 0.17), P = 0.008). Thirty-seven positioned a standardized “client,” with groups not differing significantly on total tasks completed correctly ( F(1, 92.32) = 1.91, P = 0.17) averaging 11.85 (SE 0.83) and 12.60 (SE 0.84) of 16 tasks correct. Group A’s post-positioning/simulation self-ratings were significantly associated with actual ability ( r = 0.53, P = 0.019). In both groups ⩾47% of caregivers incorrectly completed the tasks of placing head and neck in neutral and snugging up all [positioning] parts. Conclusion: The sleep care positioning training program (interactive video-based format) is effective in building caregivers’ self-perceived competence for night-time postural care. While the lesson was well-received by caregivers and considered a “match [to their] learning style,” the lesson did not lead to greater improvement in actual ability to position the “client” compared to control training.
Abstract Bayesian analyses with the arm‐based (AB) network meta‐analysis (NMA) model require researchers to specify a prior distribution for the covariance matrix of the treatment‐specific event rates in a transformed scale, for example, the treatment‐specific log‐odds when a logit transformation is used. The commonly used conjugate prior for the covariance matrix, the inverse‐Wishart (IW) distribution, has several limitations. For example, although the IW distribution is often described as noninformative or weakly informative, it may in fact provide strong information when some variance components are small (eg, when the standard deviation of study‐specific log‐odds of a treatment is smaller than 1/2), as is common in NMAs with binary outcomes. In addition, the IW prior generally leads to underestimation of correlations between treatment‐specific log‐odds, which are critical for borrowing strength across treatment arms to estimate treatment effects efficiently and to reduce potential bias. Alternatively, several separation strategies (ie, separate priors on variances and correlations) can be considered. To study the IW prior's impact on NMA results and compare it with separation strategies, we did simulation studies under different missing‐treatment mechanisms. A separation strategy with appropriate priors for the correlation matrix and variances performs better than the IW prior, and should be recommended as the default vague prior in the AB NMA approach. Finally, we reanalyzed three case studies and illustrated the importance, when performing AB‐NMA, of sensitivity analyses with different prior specifications on variances.
BackgroundDuring third molar removal, the mandible is supported by a dental assistant (DA) to counter downward forces during surgery, and with sedation, to maintain airway patency. The Restful Jaw device (PEP Design; Saint Paul) provides this support instead of the DA.PurposeThis study compared the occurrence of postoperative preauricular and masticatory muscle pain symptoms (PMMPS) between the device and DAs providing mandibular support, using two outcome measures. Secondary aims identify predictors of outcome and providers' opinions of the device.Study Design, Setting, SampleIn this multisite, single-blind, two-arm parallel randomized trial, participants without preoperative PMMPS had surgical removal of third molars, with sedation and bite blocks were randomly assigned to manual support or the device.Exposure VariableThe exposed group was randomly assigned to the device and the nonexposed group to manual support.Main Outcome Variable(s)The primary outcome was patient-reported PMMPS. Two secondary outcomes were pain assessed with the temporomandibular disorder Pain Screener and providers' views on the device. Outcomes were assessed at 1-, 3-, and 6-month postsurgery.CovariatesThe covariates are baseline demographics (eg, sex), clinical characteristics (eg, eruption status), and third molar surgeries.AnalysesFor occurrence of pain, generalized estimating equations assessed differences between groups. Logistic regression analysis assessed predictors of pain at 1 month, per the Screener. The level for statistical significance was 5%.ResultsEnrollment was 86 and 83 participants in the device and DA groups, respectively. The average age was 20.8 years; the majority were female (65%) and Caucasian (66%). The retention rate was ≥95.9%. The groups did not differ significantly for occurrence of pain using the primary and secondary outcome measures at any follow-up (P ≥ .46). Fully impacted molars were associated with occurrence of pain (odds ratio = 3.44; 95% confidence interval 1.49-7.92; P = .004).Conclusion and RelevanceOccurrence of pain using the primary and secondary outcome measures did not differ significantly between groups at any follow-up and was associated with removal of fully impacted third molars. Four out of five surgeons reported wanting to use the device on a regular basis when performing this procedure in sedated patients. During third molar removal, the mandible is supported by a dental assistant (DA) to counter downward forces during surgery, and with sedation, to maintain airway patency. The Restful Jaw device (PEP Design; Saint Paul) provides this support instead of the DA. This study compared the occurrence of postoperative preauricular and masticatory muscle pain symptoms (PMMPS) between the device and DAs providing mandibular support, using two outcome measures. Secondary aims identify predictors of outcome and providers' opinions of the device. In this multisite, single-blind, two-arm parallel randomized trial, participants without preoperative PMMPS had surgical removal of third molars, with sedation and bite blocks were randomly assigned to manual support or the device. The exposed group was randomly assigned to the device and the nonexposed group to manual support. The primary outcome was patient-reported PMMPS. Two secondary outcomes were pain assessed with the temporomandibular disorder Pain Screener and providers' views on the device. Outcomes were assessed at 1-, 3-, and 6-month postsurgery. The covariates are baseline demographics (eg, sex), clinical characteristics (eg, eruption status), and third molar surgeries. For occurrence of pain, generalized estimating equations assessed differences between groups. Logistic regression analysis assessed predictors of pain at 1 month, per the Screener. The level for statistical significance was 5%. Enrollment was 86 and 83 participants in the device and DA groups, respectively. The average age was 20.8 years; the majority were female (65%) and Caucasian (66%). The retention rate was ≥95.9%. The groups did not differ significantly for occurrence of pain using the primary and secondary outcome measures at any follow-up (P ≥ .46). Fully impacted molars were associated with occurrence of pain (odds ratio = 3.44; 95% confidence interval 1.49-7.92; P = .004). Occurrence of pain using the primary and secondary outcome measures did not differ significantly between groups at any follow-up and was associated with removal of fully impacted third molars. Four out of five surgeons reported wanting to use the device on a regular basis when performing this procedure in sedated patients.
Insulin independence after total pancreatectomy and islet autotransplant (TPIAT) for chronic pancreatitis is limited by a high rate of postprocedure beta cell apoptosis. Endogenous glucagon-like peptide-1 and glucose-dependent insulinotropic peptide, which are increased by dipeptidyl peptidase 4 inhibitor therapy (sitagliptin) may protect against beta cell apoptosis. To determine the effect of sitagliptin after TPIAT, 83 adult TPIAT recipients were randomized to receive sitagliptin (n = 54) or placebo (n = 29) for 12 months after TPIAT. At 12 and 18 months after TPIAT, participants were assessed for insulin independence; metabolic testing was performed with mixed meal tolerance testing and frequent sample intravenous glucose tolerance testing. Insulin independence did not differ between the sitagliptin and placebo groups at 12 months (42% vs. 45%, p = 0.82) or 18 months (36% vs. 44%, p = 0.48). At 12 months, insulin dose was 9.0 (standard error 1.7) units/day and 7.9 (2.2) units/day in the sitagliptin and placebo groups, respectively (p = 0.67) and at 18 months 10.3 (1.9) and 7.1 (2.6) units/day, respectively (p = 0.32). Hemoglobin A1c levels and insulin secretory measures were similar in the two groups, as were adverse events. In conclusion, sitagliptin could be safely administered but did not improve metabolic outcomes after TPIAT.
