Combining randomized and non-randomized data to predict heterogeneous effects of competing treatments.

Some patients benefit from a treatment while others may do so less or do not benefit at all. We have previously developed a two-stage network meta-regression prediction model that synthesized randomized trials and evaluates how treatment effects vary across patient characteristics. In this article, we extended this model to combine different sources of types in different formats: aggregate data (AD) and individual participant data (IPD) from randomized and non-randomized evidence. In the first stage, a prognostic model is developed to predict the baseline risk of the outcome using a large cohort study. In the second stage, we recalibrated this prognostic model to improve our predictions for patients enrolled in randomized trials. In the third stage, we used the baseline risk as effect modifier in a network meta-regression model combining AD, IPD randomized clinical trial to estimate heterogeneous treatment effects. We illustrated the approach in the re-analysis of a network of studies comparing three drugs for relapsing-remitting multiple sclerosis. Several patient characteristics influence the baseline risk of relapse, which in turn modifies the effect of the drugs. The proposed model makes personalized predictions for health outcomes under several treatment options and encompasses all relevant randomized and non-randomized evidence.

Comparative Effectiveness of Natalizumab, Fingolimod, and Injectable Therapies in Pediatric-Onset Multiple Sclerosis: A Registry-Based Study.

BACKGROUND AND OBJECTIVES: Patients with pediatric-onset multiple sclerosis (POMS) typically experience higher levels of inflammation with more frequent relapses, and though patients with POMS usually recover from relapses better than adults, patients with POMS reach irreversible disability at a younger age than adult-onset patients. There have been few randomized, placebo-controlled clinical trials of multiple sclerosis (MS) disease-modifying therapies (DMTs) in patients with POMS, and most available data are based on observational studies of off-label use of DMTs approved for adults. We assessed the effectiveness of natalizumab compared with fingolimod using injectable platform therapies as a reference in pediatric patients in the global MSBase registry. METHODS: This retrospective study included patients with POMS who initiated treatment with an injectable DMT, natalizumab, or fingolimod between January 1, 2006, and May 3, 2021. Patients were matched using inverse probability treatment weighting. The primary outcome was time to first relapse from index therapy initiation. Secondary study outcomes included annualized relapse rate; proportions of relapse-free patients at 1, 2, and 5 years; time to treatment discontinuation; and times to 24-week confirmed disability worsening and confirmed disability improvement. RESULTS: A total of 1,218 patients with POMS were included in this analysis. Patients treated with fingolimod had a significantly lower risk of relapse than patients treated with injectable DMTs (hazard ratio [HR], 0.49; 95% confidence interval [CI], 0.29-0.83; p = 0.008). After adjustment for prior DMT experience in the unmatched sample, patients treated with natalizumab had a significantly lower risk of relapse than patients treated either with injectable DMTs (HR, 0.15; 95% CI 0.07-0.31; p < 0.001) or fingolimod (HR, 0.37; 95% CI 0.14-1.00; p = 0.049). The adjusted secondary study outcomes were generally consistent with the primary outcome or with previous observations. The findings in the inverse probability treatment weighting-adjusted patient populations were confirmed in multiple sensitivity analyses. DISCUSSION: Our analyses of relapse risk suggest that natalizumab is more effective than fingolimod in the control of relapses in this population with high rates of new inflammatory activity, consistent with previous studies of natalizumab and fingolimod in adult-onset patients and POMS. In addition, both fingolimod and natalizumab were more effective than first-line injectable therapies. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that patients with POMS treated with natalizumab had a lower risk of relapse than those with fingolimod.

Visualizing the target estimand in comparative effectiveness studies with multiple treatments.

Aim: Comparative effectiveness research using real-world data often involves pairwise propensity score matching to adjust for confounding bias. We show that corresponding treatment effect estimates may have limited external validity, and propose two visualization tools to clarify the target estimand. Materials & methods: We conduct a simulation study to demonstrate, with bivariate ellipses and joy plots, that differences in covariate distributions across treatment groups may affect the external validity of treatment effect estimates. We showcase how these visualization tools can facilitate the interpretation of target estimands in a case study comparing the effectiveness of teriflunomide (TERI), dimethyl fumarate (DMF) and natalizumab (NAT) on manual dexterity in patients with multiple sclerosis. Results: In the simulation study, estimates of the treatment effect greatly differed depending on the target population. For example, when comparing treatment B with C, the estimated treatment effect (and respective standard error) varied from -0.27 (0.03) to -0.37 (0.04) in the type of patients initially receiving treatment B and C, respectively. Visualization of the matched samples revealed that covariate distributions vary for each comparison and cannot be used to target one common treatment effect for the three treatment comparisons. In the case study, the bivariate distribution of age and disease duration varied across the population of patients receiving TERI, DMF or NAT. Although results suggest that DMF and NAT improve manual dexterity at 1 year compared with TERI, the effectiveness of DMF versus NAT differs depending on which target estimand is used. Conclusion: Visualization tools may help to clarify the target population in comparative effectiveness studies and resolve ambiguity about the interpretation of estimated treatment effects.

Confounder Adjustment Using the Disease Risk Score: A Proposal for Weighting Methods.

Propensity score analysis is a common approach to addressing confounding in nonrandomized studies. Its implementation, however, requires important assumptions (e.g., positivity). The disease risk score (DRS) is an alternative confounding score that can relax some of these assumptions. Like the propensity score, the DRS summarizes multiple confounders into a single score, on which conditioning by matching allows the estimation of causal effects. However, matching relies on arbitrary choices for pruning out data (e.g., matching ratio, algorithm, and caliper width) and may be computationally demanding. Alternatively, weighting methods, common in propensity score analysis, are easy to implement and may entail fewer choices, yet none have been developed for the DRS. Here we present 2 weighting approaches: One derives directly from inverse probability weighting; the other, named target distribution weighting, relates to importance sampling. We empirically show that inverse probability weighting and target distribution weighting display performance comparable to matching techniques in terms of bias but outperform them in terms of efficiency (mean squared error) and computational speed (up to >870 times faster in an illustrative study). We illustrate implementation of the methods in 2 case studies where we investigate placebo treatments for multiple sclerosis and administration of aspirin in stroke patients.

Early Lifestyle Determinants of Adiposity Trajectories from Childhood into Late Adolescence.

Objective: The aim of this study was to examine associations of childhood physical activity, sedentary behavior, and dietary intake with adiposity trajectories from childhood to adolescence. Methods: Quebec Adipose and Lifestyle Investigation in Youth (QUALITY) cohort (n = 630) data from 3 time points (8-10, 10-12, and 15-17 years) for 377 Caucasian children with parental obesity were analyzed. Height and weight, physical activity and sedentary behavior (7-day accelerometry), screen time (self-reported), and dietary intake (three 24-hour diet recalls) were measured. Group-based trajectory modeling identified longitudinal trajectories of body-mass index z-scores (zBMIs). Inverse probability of exposure-weighted multinomial logistic regressions examined associations between baseline lifestyles and zBMI trajectory groups. Results: Six trajectory groups were identified: Stable-Low-Normal-Weight (two groups, 5.7% and 33.0%, which were combined), Stable-High-Normal-Weight (24.8%), Stable-Overweight (19.8%), Stable-Obesity (8.8%), and Overweight-Decreasers (7.9%). For every additional portion of fruits and vegetables, the likelihood of being in the group of Overweight-Decreasers increased by 29% (odds ratio [OR]: 1.29, 95% confidence interval [CI]: 1.09-1.55) compared with the reference group (Stable-Low-Normal-Weight). For every additional hour of sedentary behavior, the likelihood of belonging to the group of Overweight-Decreasers increased 2-fold (OR: 1.99, 95% CI: 1.28-3.21) and Stable-Obesity increased 1.5-fold (OR: 1.56, 95% CI: 1.08-2.23), compared with the reference. Every additional 10 minutes of moderate-to-vigorous physical activity was associated with a lower likelihood of belonging to the Stable-Obesity group (OR: 0.75, 95% CI: 0.61-0.89) and to the group of Overweight-Decreasers (OR: 0.79, 95% CI: 0.64-0.95) compared with the reference. Finally, children were more likely to belong to the Stable-Obesity group with each additional hour/day of screen time (OR: 1.23, 95% CI: 1.01-1.58). Conclusions: Trajectories of zBMIs from childhood to late adolescence were stable, except for one group which decreased from overweight in childhood to normal weight in adolescence. The latter had more favorable baseline dietary intake of fruits and vegetables. ClinicalTrials.org no. NCT03356262.

Overall and patient-specific comparative effectiveness of dimethyl fumarate versus teriflunomide: A novel approach to precision medicine applied to the German NeuroTrans Data Multiple Sclerosis Registry.

Background: Multiple sclerosis (MS) comparative effectiveness research needs to go beyond average treatment effects (ATEs) and post-host subgroup analyses. Objective: This retrospective study assessed overall and patient-specific effects of dimethyl fumarate (DMF) versus teriflunomide (TERI) in patients with relapsing-remitting MS. Methods: A novel precision medicine (PM) scoring approach leverages advanced machine learning methods and adjusts for imbalances in baseline characteristics between patients receiving different treatments. Using the German NeuroTransData registry, we implemented and internally validated different scoring systems to distinguish patient-specific effects of DMF relative to TERI based on annualized relapse rates, time to first relapse, and time to confirmed disease progression. Results: Among 2791 patients, there was superior ATE of DMF versus TERI for the two relapse-related endpoints (p = 0.037 and 0.018). Low to moderate signals of treatment effect heterogeneity were detected according to individualized scores. A MS patient subgroup was identified for whom DMF was more effective than TERI (p = 0.013): older (45 versus 38 years), longer MS duration (110 versus 50 months), not newly diagnosed (74% versus 40%), and no prior glatiramer acetate usage (35% versus 5%). Conclusion: The implemented approach can disentangle prognostic differences from treatment effect heterogeneity and provide unbiased patient-specific profiling of comparative effectiveness based on real-world data.

Methods for comparative effectiveness based on time to confirmed disability progression with irregular observations in multiple sclerosis.

Real-world data sources offer opportunities to compare the effectiveness of treatments in practical clinical settings. However, relevant outcomes are often recorded selectively and collected at irregular measurement times. It is therefore common to convert the available visits to a standardized schedule with equally spaced visits. Although more advanced imputation methods exist, they are not designed to recover longitudinal outcome trajectories and typically assume that missingness is non-informative. We, therefore, propose an extension of multilevel multiple imputation methods to facilitate the analysis of real-world outcome data that is collected at irregular observation times. We illustrate multilevel multiple imputation in a case study evaluating two disease-modifying therapies for multiple sclerosis in terms of time to confirmed disability progression. This survival outcome is derived from repeated measurements of the Expanded Disability Status Scale, which is collected when patients come to the healthcare center for a clinical visit and for which longitudinal trajectories can be estimated. Subsequently, we perform a simulation study to compare the performance of multilevel multiple imputation to commonly used single imputation methods. Results indicate that multilevel multiple imputation leads to less biased treatment effect estimates and improves the coverage of confidence intervals, even when outcomes are missing not at random.

Overall and patient-level comparative effectiveness of dimethyl fumarate and fingolimod: A precision medicine application to the Observatoire Français de la Sclérose en Plaques registry.

Background: Comparing real-world effectiveness and tolerability of therapies for relapsing-remitting multiple sclerosis is increasingly important, though average treatment effects fail to capture possible treatment effect heterogeneity. With the clinical course of the disease being highly heterogeneous across patients, precision medicine methods enable treatment response heterogeneity investigations. Objective: To compare real-world effectiveness and discontinuation profiles between dimethyl fumarate and fingolimod while investigating treatment effect heterogeneity with precision medicine methods. Methods: Adults initiating dimethyl fumarate or fingolimod as a second-line therapy were selected from a French registry. The primary outcome was annualized relapse rate at 12 months. Seven secondary outcomes relative to discontinuation and disease progression were considered. A precision medicine framework was used to characterize treatment effect heterogeneity. Results: Annualized relapse rates at 12 months were similar for dimethyl fumarate and fingolimod. The odd of treatment persistence was 47% lower for patients treated with dimethyl fumarate relative to those treated with fingolimod (odds ratio: 0.53, 95% confidence interval: 0.39, 0.70). None of the five precision medicine scoring approaches identified treatment heterogeneity. Conclusion: These findings substantiated the similar effectiveness and different discontinuation profiles for dimethyl fumarate and fingolimod as a second-line therapy for relapsing-remitting multiple sclerosis, with no significant effect heterogeneity observed.

Recommendations for the use of propensity score methods in multiple sclerosis research.

BACKGROUND: With many disease-modifying therapies currently approved for the management of multiple sclerosis, there is a growing need to evaluate the comparative effectiveness and safety of those therapies from real-world data sources. Propensity score methods have recently gained popularity in multiple sclerosis research to generate real-world evidence. Recent evidence suggests, however, that the conduct and reporting of propensity score analyses are often suboptimal in multiple sclerosis studies. OBJECTIVES: To provide practical guidance to clinicians and researchers on the use of propensity score methods within the context of multiple sclerosis research. METHODS: We summarize recommendations on the use of propensity score matching and weighting based on the current methodological literature, and provide examples of good practice. RESULTS: Step-by-step recommendations are presented, starting with covariate selection and propensity score estimation, followed by guidance on the assessment of covariate balance and implementation of propensity score matching and weighting. Finally, we focus on treatment effect estimation and sensitivity analyses. CONCLUSION: This comprehensive set of recommendations highlights key elements that require careful attention when using propensity score methods.

The use and quality of reporting of propensity score methods in multiple sclerosis literature: A review.

BACKGROUND: Propensity score (PS) analyses are increasingly used in multiple sclerosis (MS) research, largely owing to the greater availability of large observational cohorts and registry databases. OBJECTIVE: To evaluate the use and quality of reporting of PS methods in the recent MS literature. METHODS: We searched the PubMed database for articles published between January 2013 and July 2019. We restricted the search to comparative effectiveness studies of two disease-modifying therapies. RESULTS: Thirty-nine studies were included in the review, with most studies (62%) published within the past 3 years. All studies reported the list of covariates used for the PS model, but only 21% of studies mentioned how those covariates were selected. Most studies used PS matching (72%), followed by PS adjustment (18%), weighting (15%), and stratification (3%), with some overlap. Most studies using matching or weighting reported checking post-PS covariate imbalance (91%), although about 45% of these studies relied on p values from various statistical tests. Only 25% of studies using matching reported calculating robust standard errors for the PS analyses. CONCLUSIONS: The quality of reporting of PS methods in the MS literature is sub-optimal in general, and in some cases, inappropriate methods are used.

Evaluation of automated microvascular flow analysis software AVA 4: a validation study.

BACKGROUND: Real-time automated analysis of videos of the microvasculature is an essential step in the development of research protocols and clinical algorithms that incorporate point-of-care microvascular analysis. In response to the call for validation studies of available automated analysis software by the European Society of Intensive Care Medicine, and building on a previous validation study in sheep, we report the first human validation study of AVA 4. METHODS: Two retrospective perioperative datasets of human microcirculation videos (P1 and P2) and one prospective healthy volunteer dataset (V1) were used in this validation study. Video quality was assessed using the modified Microcirculation Image Quality Selection (MIQS) score. Videos were initially analyzed with (1) AVA software 3.2 by two experienced investigators using the gold standard semi-automated method, followed by an analysis with (2) AVA automated software 4.1. Microvascular variables measured were perfused vessel density (PVD), total vessel density (TVD), and proportion of perfused vessels (PPV). Bland-Altman analysis and intraclass correlation coefficients (ICC) were used to measure agreement between the two methods. Each method's ability to discriminate between microcirculatory states before and after induction of general anesthesia was assessed using paired t-tests. RESULTS: Fifty-two videos from P1, 128 videos from P2 and 26 videos from V1 met inclusion criteria for analysis. Correlational analysis and Bland-Altman analysis revealed poor agreement and no correlation between AVA 4.1 and AVA 3.2. Following the induction of general anesthesia, TVD and PVD measured using AVA 3.2 increased significantly for P1 (p < 0.05) and P2 (p < 0.05). However, these changes could not be replicated with the data generated by AVA 4.1. CONCLUSIONS: AVA 4.1 is not a suitable tool for research or clinical purposes at this time. Future validation studies of automated microvascular flow analysis software should aim to measure the new software's agreement with the gold standard, its ability to discriminate between clinical states and the quality thresholds at which its performance becomes unacceptable.

Finite sample variance estimation for optimal dynamic treatment regimes of survival outcomes.

Deriving valid confidence intervals for complex estimators is a challenging task in practice. Estimators of dynamic weighted survival modeling (DWSurv), a method to estimate an optimal dynamic treatment regime of censored outcomes, are asymptotically normal and consistent for their target parameters when at least a subset of the nuisance models is correctly specified. However, their behavior in finite samples and the impact of model misspecification on inferences remain unclear. In addition, the estimators' nonregularity may negatively affect the inferences under some specific data generating mechanisms. Our objective was to compare five methods, two asymptotic variance formulas (adjusting or not for the estimation of nuisance parameters) to three bootstrap approaches, to construct confidence intervals for the DWSurv parameters in finite samples. Via simulations, we considered practical scenarios, for example, when some nuisance models are misspecified or when nonregularity is problematic. We also compared the five methods in an application about the treatment of rheumatoid arthritis. We found that the bootstrap approaches performed consistently well at the cost of longer computational times. The asymptotic variance with adjustments generally yielded conservative confidence intervals. The asymptotic variance without adjustments yielded nominal coverages for large sample sizes. We recommend using the asymptotic variance with adjustments in small samples and the bootstrap if computationally feasible. Caution should be taken when nonregularity may be an issue.

Milrinone for refractory cerebral vasospasm with delayed cerebral ischemia.

OBJECTIVE: Intravenous (IV) milrinone is a promising option for the treatment of cerebral vasospasm with delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH). However, data remain limited on the efficacy of treating cases that are refractory to standard therapy with IV milrinone. The aim of this study was to determine predictors of refractory vasospasm/DCI despite treatment with IV milrinone, and to analyze the outcome of rescue therapy with intraarterial (IA) milrinone and/or mechanical angioplasty. METHODS: The authors conducted a retrospective cohort study of all patients with aSAH admitted between 2010 and 2016 to the Montreal Neurological Institute and Hospital. Patients were stratified into 3 groups: no DCI, standard therapy, and rescue therapy. The primary outcome was frequency of DCI-related cerebral infarction identified on neuroimaging before hospital discharge. Secondary outcomes included functional outcome reported as modified Rankin Scale (mRS) score, and segment reversal of refractory vasospasm. RESULTS: The cohort included 322 patients: 212 in the no DCI group, 89 in the standard therapy group, and 21 in the rescue therapy group. Approximately half (52%, 168/322) were admitted with poor-grade aSAH at treatment decision (World Federation of Neurosurgical Societies grade III-V). Among patients with DCI and imaging assessing severity of vasospasm, 62% (68/109) had moderate/severe radiological vasospasm on DCI presentation. Nineteen percent (21/110) of patients had refractory vasospasm/DCI and were treated with rescue therapy. Targeted rescue therapy with IA milrinone reversed 32% (29/91) of the refractory vasospastic vessels, and 76% (16/21) of those patients experienced significant improvement in their neurological status within 24 hours of initiating therapy. Moderate/severe radiological vasospasm independently predicted the need for rescue therapy (OR 27, 95% CI 8.01-112). Of patients with neuroimaging before discharge, 40% (112/277) had developed new cerebral infarcts, and only 21% (23/112) of these were vasospasm-related. Overall, 65% (204/314) of patients had a favorable functional outcome (mRS score 0-2) assessed at a median of 4 months (interquartile range 2-8 months) after aSAH, and there was no difference in functional outcome between the 3 groups (p = 0.512). CONCLUSIONS: The aggressive use of milrinone was safe and effective based on this retrospective study cohort and is a promising therapy for the treatment of vasospasm/DCI after aSAH.

Tune out and turn in: the influence of television viewing and sleep on lipid profiles in children.

BACKGROUND/OBJECTIVES: Physical activity is beneficial to lipid profiles; however, the association between sedentary behavior and sleep and pediatric dyslipidemia remains unclear. We aimed to investigate whether sedentary behavior or sleep predicted lipid profiles in children over a 2-year period. SUBJECTS/METHODS: Six hundered and thirty children from the QUALITY cohort, with at least one obese parent, were assessed prospectively at ages 8-10 and 10-12 years. Measures of sedentary behavior included self-reported TV viewing and computer/video game use. Seven-day accelerometry was used to derive sedentary behavior and sleep duration. Adiposity was assessed using DEXA scans. Twenty-four-hour dietary recalls yielded estimates of carbohydrate and fat intake. Outcomes included fasting total cholesterol, triglycerides, HDL and LDL-cholesterol. Multivariable models were adjusted for adiposity and diet. RESULTS: At both Visit 1 (median age 9.6 year) and Visit 2 (median age 11.6 year), children were of normal weight (55%), overweight (22%), or obese (22%). Every additional hour of TV viewing at Visit 1 was associated with a 7.0% triglyceride increase (95% CI: 3.5, 10.6; P < 0.01) and 2.6% HDL decrease (95% CI: -4.2, -0.9; P < 0.01) at Visit 2; findings remained significant after adjusting for adiposity and diet. Every additional hour of sleep at Visit 1 predicted a 4.8% LDL decrease (95% CI: -9.0, -0.5; P = 0.03) at Visit 2, after adjusting for fat intake; this association became nonsignificant once controlling for adiposity. CONCLUSIONS: Longer screen time during childhood appears to deteriorate lipid profiles in early adolescence, even after accounting for other major lifestyle habits. There is preliminary evidence of a deleterious effect of shorter sleep duration, which should be considered in further studies.

Adaptive Treatment Strategies With Survival Outcomes: An Application to the Treatment of Type 2 Diabetes Using a Large Observational Database.

Sequences of treatments that adapt to a patient's changing condition over time are often needed for the management of chronic diseases. An adaptive treatment strategy (ATS) consists of personalized treatment rules to be applied through the course of a disease that input the patient's characteristics at the time of decision-making and output a recommended treatment. An optimal ATS is the sequence of tailored treatments that yields the best clinical outcome for patients sharing similar characteristics. Methods for estimating optimal adaptive treatment strategies, which must disentangle short- and long-term treatment effects, can be theoretically involved and hard to explain to clinicians, especially when the outcome to be optimized is a survival time subject to right-censoring. In this paper, we describe dynamic weighted survival modeling, a method for estimating an optimal ATS with survival outcomes. Using data from the Clinical Practice Research Datalink, a large primary-care database, we illustrate how it can answer an important clinical question about the treatment of type 2 diabetes. We identify an ATS pertaining to which drug add-ons to recommend when metformin in monotherapy does not achieve the therapeutic goals.

Head Motion Predicts Transient Loss of Consciousness in Human Head Trauma: A Case-Control Study of Mixed Martial Artists.

OBJECTIVE: Concussion with transient loss of consciousness is a commonly observed but poorly understood phenomenon with mounting clinical significance. This study aimed to examine the relationship between head motion in varying planes and transient loss of consciousness in athletes with brain injuries. STUDY DESIGN: A case-control design was used. The Ultimate Fighting Championship database was screened for events ending with knockouts from 2013 to 2016. Time of strike, striking implement, strike location, and head motion were recorded for all knockout strikes (cases) and for a subset of nonknockout strikes (controls). Characteristics of winners and losers were compared using two-tailed t tests. Multivariate logistic regression was used to determine odds ratios for strike characteristics associated with transient loss of consciousness. The Kaplan-Meier estimate was used to describe the temporal distribution of knockouts. RESULTS: One hundred thirty-six fights were identified and 110 videos were included. Head motion in the axial plane was strongly associated with transient loss of consciousness (odds ratio, 45.3; 95% confidence interval, 20.8-98.6). Other predictors of transient loss of consciousness were head motion in sagittal and coronal planes, nonfist striking implements, and strikes to the mandible or maxilla. The Kaplan-Meier survival curve demonstrated a decreasing rate of knockouts through time. CONCLUSIONS: Rotational head acceleration, particularly in the axial plane, is strongly associated with transient loss of consciousness.

Mortality after Spontaneous Subarachnoid Hemorrhage: Causality and Validation of a Prediction Model.

OBJECTIVE: To evaluate primary causes of death after spontaneous subarachnoid hemorrhage (SAH) and externally validate the HAIR score, a prognostication tool, in a single academic institution. METHODS: We reviewed all patients with SAH admitted to our neuro-intensive care unit between 2010 and 2016. Univariate and multivariate logistic regressions were performed to identify predictors of in-hospital mortality. The HAIR score predictors were Hunt and Hess grade at treatment decision, age, intraventricular hemorrhage, and rebleeding within 24 hours. Validation of the HAIR score was characterized with the receiver operating curve, the area under the curve, and a calibration plot. RESULTS: Among 434 patients with SAH, in-hospital mortality was 14.1%. Of the 61 mortalities, 54 (88.5%) had a neurologic cause of death or withdrawal of care and 7 (11.5%) had cardiac death. Median time from SAH to death was 6 days. The main causes of death were effect of the initial hemorrhage (26.2%), rebleeding (23%) and refractory cerebral edema (19.7%). Factors significantly associated with in-hospital mortality in the multivariate analysis were age, Hunt and Hess grade, and intracerebral hemorrhage. Maximum lumen size was also a significant risk factor after aneurysmal SAH. The HAIR score had a satisfactory discriminative ability, with an area under the curve of 0.89. CONCLUSIONS: The in-hospital mortality is lower than in previous reports, attesting to the continuing improvement of our institutional SAH care. The major causes are the same as in previous reports. Despite a different therapeutic protocol, the HAIR score showed good discrimination and could be a useful tool for predicting mortality.

Non-regular inference for dynamic weighted ordinary least squares: understanding the impact of solid food intake in infancy on childhood weight.

A dynamic treatment regime (DTR) is a set of decision rules to be applied across multiple stages of treatments. The decisions are tailored to individuals, by inputting an individual's observed characteristics and outputting a treatment decision at each stage for that individual. Dynamic weighted ordinary least squares (dWOLS) is a theoretically robust and easily implementable method for estimating an optimal DTR. As many related DTR methods, the dWOLS treatment effects estimators can be non-regular when true treatment effects are zero or very small, which results in invalid Wald-type or standard bootstrap confidence intervals. Inspired by an analysis of the effect of diet in infancy on measures of weight and body size in later childhood-a setting where the exposure is distant in time and whose effect is likely to be small-we investigate the use of the $m$-out-of-$n$ bootstrap with dWOLS as method of analysis for valid inferences of optimal DTR. We provide an extensive simulation study to compare the performance of different choices of resample size $m$ in situations where the treatment effects are likely to be non-regular. We illustrate the methodology using data from the PROmotion of Breastfeeding Intervention Trial to study the effect of solid food intake in infancy on long-term health outcomes.

A comparison of bivariate, multivariate random-effects, and Poisson correlated gamma-frailty models to meta-analyze individual patient data of ordinal scale diagnostic tests.

Individual patient data (IPD) meta-analyses are increasingly common in the literature. In the context of estimating the diagnostic accuracy of ordinal or semi-continuous scale tests, sensitivity and specificity are often reported for a given threshold or a small set of thresholds, and a meta-analysis is conducted via a bivariate approach to account for their correlation. When IPD are available, sensitivity and specificity can be pooled for every possible threshold. Our objective was to compare the bivariate approach, which can be applied separately at every threshold, to two multivariate methods: the ordinal multivariate random-effects model and the Poisson correlated gamma-frailty model. Our comparison was empirical, using IPD from 13 studies that evaluated the diagnostic accuracy of the 9-item Patient Health Questionnaire depression screening tool, and included simulations. The empirical comparison showed that the implementation of the two multivariate methods is more laborious in terms of computational time and sensitivity to user-supplied values compared to the bivariate approach. Simulations showed that ignoring the within-study correlation of sensitivity and specificity across thresholds did not worsen inferences with the bivariate approach compared to the Poisson model. The ordinal approach was not suitable for simulations because the model was highly sensitive to user-supplied starting values. We tentatively recommend the bivariate approach rather than more complex multivariate methods for IPD diagnostic accuracy meta-analyses of ordinal scale tests, although the limited type of diagnostic data considered in the simulation study restricts the generalization of our findings.

Maternal sleep-disordered breathing and the risk of delivering small for gestational age infants: a prospective cohort study.

OBJECTIVE: It is unclear whether objectively measured maternal sleep-disordered breathing (SDB) leads to poor fetal outcomes. In this study, we prospectively assessed whether polysomnography-based diagnosis of SDB in the third trimester is associated with the delivery of small for gestational age (SGA) infants. STUDY DESIGN: Participants were recruited from a multicentre pregnancy cohort study. Eligible participants were evaluated for SDB based on symptoms (snoring and/or witnessed apnoeas assessed using the Pittsburgh Sleep Quality Index questionnaire) and in-home complete polysomnography in the third trimester. SGA was defined as <10th centile using customised birthweight centiles adjusted for maternal parity, prepregnancy body mass index (BMI), ethnicity, gestational age and infant sex. RESULTS: Of the 234 pregnant participants who completed a sleep study, 82% were Caucasian, with mean (SD) age of 31 (4.3) years and a prepregnancy BMI of 23 (4) kg/m(2). The delivery of SGA infants occurred in 27 (12%) of the study participants. The symptoms of SDB had poor overall sensitivity and specificity for diagnosing SDB identified by polysomnography. Symptoms of SDB in the third trimester demonstrated a potential association with delivering an SGA infant, however this did not reach statistical significance (OR 2.36 (95% CI 0.85 to 6.54, p=0.10)). However, the odds of delivering an SGA infant were significantly increased with polysomnography-based diagnosis of maternal SDB (using apnoea-hypopnoea index cut-off of 10, OR 2.65 (95% CI 1.15 to 6.10, p=0.02)). CONCLUSIONS: Objectively measured SDB in the third trimester is significantly associated with the delivery of SGA infants.