In this multicenter, randomized trial involving extremely preterm infants, high-dose erythropoietin administered from 24 hours after birth through 32 weeks of postmenstrual age did not result in a ...lower risk of severe neurodevelopmental impairment or death at 2 years of age.
Stepped wedge design is a popular research design that enables a rigorous evaluation of candidate interventions by using a staggered cluster randomization strategy. While analytical methods were ...developed for designing stepped wedge trials, the prior focus has been solely on testing for the average treatment effect. With a growing interest on formal evaluation of the heterogeneity of treatment effects across patient subpopulations, trial planning efforts need appropriate methods to accurately identify sample sizes or design configurations that can generate evidence for both the average treatment effect and variations in subgroup treatment effects. To fill in that important gap, this article derives novel variance formulas for confirmatory analyses of treatment effect heterogeneity, that are applicable to both cross‐sectional and closed‐cohort stepped wedge designs. We additionally point out that the same framework can be used for more efficient average treatment effect analyses via covariate adjustment, and allows the use of familiar power formulas for average treatment effect analyses to proceed. Our results further sheds light on optimal design allocations of clusters to maximize the weighted precision for assessing both the average and heterogeneous treatment effects. We apply the new methods to the Lumbar Imaging with Reporting of Epidemiology Trial, and carry out a simulation study to validate our new methods.
Multiple-period parallel group randomized trials (GRTs) analyzed with linear mixed models can represent time in mean models as continuous or categorical. If time is continuous, random effects are ...traditionally group- and member-level deviations from condition-specific slopes and intercepts and are referred to as random coefficients (RC) analytic models. If time is categorical, random effects are traditionally group- and member-level deviations from time-specific condition means and are referred to as repeated measures ANOVA (RM-ANOVA) analytic models. Longstanding guidance recommends the use of RC over RM-ANOVA for parallel GRTs with more than two periods because RC exhibited nominal type I error rates for both time parameterizations while RM-ANOVA exhibited inflated type I error rates when applied to data generated using the RC model. However, this recommendation was developed assuming a variance components covariance matrix for the RM-ANOVA, using only cross-sectional data, and explicitly modeling time × group variation. Left unanswered were how well RM-ANOVA with an unstructured covariance would perform on data generated according to the RC mechanism, if similar patterns would be observed in cohort data, and the impact of not modeling time × group variation if such variation was present in the data-generating model.
Continuous outcomes for cohort and cross-sectional parallel GRT data were simulated according to RM-ANOVA and RC mechanisms at five total time periods. All simulations assumed time × group variation. We varied the number of groups, group size, and intra-cluster correlation. Analytic models using RC, RM-ANOVA, RM-ANOVA with unstructured covariance, and a Saturated random effects structure were applied to the data. All analytic models specified time × group random effects. The analytic models were then reapplied without specifying random effects for time × group.
Results indicated the RC and saturated analytic models maintained the nominal type I error rate in all data sets, RM-ANOVA with an unstructured covariance did not avoid type I error rate inflation when applied to cohort RC data, and analytic models omitting time-varying group random effects when such variation exists in the data were prone to substantial type I error inflation unless the residual error variance is high relative to the time × group variance.
The time × group RC and saturated analytic models are recommended as the default for multiple period parallel GRTs.
CONTEXT Diabetes is the leading cause of kidney disease in the developed world. Over time, the prevalence of diabetic kidney disease (DKD) may increase due to the expanding size of the diabetes ...population or decrease due to the implementation of diabetes therapies. OBJECTIVE To define temporal changes in DKD prevalence in the United States. DESIGN, SETTING, AND PARTICIPANTS
Cross-sectional analyses of the Third National Health and Nutrition Examination Survey (NHANES III) from 1988-1994 (N = 15 073), NHANES 1999-2004 (N = 13 045), and NHANES 2005-2008 (N = 9588). Participants with diabetes were defined by levels of hemoglobin A1c of 6.5% or greater, use of glucose-lowering medications, or both (n = 1431 in NHANES III; n = 1443 in NHANES 1999-2004; n = 1280 in NHANES 2005-2008).
MAIN OUTCOME MEASURES
Diabetic kidney disease was defined as diabetes with albuminuria (ratio of urine albumin to creatinine ≥30 mg/g), impaired glomerular filtration rate (<60 mL/min/1.73 m2 estimated using the Chronic Kidney Disease Epidemiology Collaboration formula), or both. Prevalence of albuminuria was adjusted to estimate persistent albuminuria.
RESULTS
The prevalence of DKD in the US population was 2.2% (95% confidence interval CI, 1.8%-2.6%) in NHANES III, 2.8% (95% CI, 2.4%-3.1%) in NHANES 1999-2004, and 3.3% (95% CI, 2.8%-3.7%) in NHANES 2005-2008 (P <.001 for trend). The prevalence of DKD increased in direct proportion to the prevalence of diabetes, without a change in the prevalence of DKD among those with diabetes. Among persons with diabetes, use of glucose-lowering medications increased from 56.2% (95% CI, 52.1%-60.4%) in NHANES III to 74.2% (95% CI, 70.4%-78.0%) in NHANES 2005-2008 (P <.001); use of renin-angiotensin-aldosterone system inhibitors increased from 11.2% (95% CI, 9.0%-13.4%) to 40.6% (95% CI, 37.2%-43.9%), respectively (P <.001); the prevalence of impaired glomerular filtration rate increased from 14.9% (95% CI, 12.1%-17.8%) to 17.7% (95% CI, 15.2%-20.2%), respectively (P = .03); and the prevalence of albuminuria decreased from 27.3% (95% CI, 22.0%-32.7%) to 23.7% (95% CI, 19.3%-28.0%), respectively, but this was not statistically significant (P = .07).
CONCLUSIONS Prevalence of DKD in the United States increased from 1988 to 2008 in proportion to the prevalence of diabetes. Among persons with diabetes, prevalence of DKD was stable despite increased use of glucose-lowering medications and renin-angiotensin-aldosterone system inhibitors.
SUMMARY
Cluster randomized trials often exhibit a three-level structure with participants nested in subclusters such as health care providers, and subclusters nested in clusters such as clinics. ...While the average treatment effect has been the primary focus in planning three-level randomized trials, interest is growing in understanding whether the treatment effect varies among prespecified patient subpopulations, such as those defined by demographics or baseline clinical characteristics. In this article, we derive novel analytical design formulas based on the asymptotic covariance matrix for powering confirmatory analyses of treatment effect heterogeneity in three-level trials, that are broadly applicable to the evaluation of cluster-level, subcluster-level, and participant-level effect modifiers and to designs where randomization can be carried out at any level. We characterize a nested exchangeable correlation structure for both the effect modifier and the outcome conditional on the effect modifier, and generate new insights from a study design perspective for conducting analyses of treatment effect heterogeneity based on a linear mixed analysis of covariance model. A simulation study is conducted to validate our new methods and two real-world trial examples are used for illustrations.
To determine if multiple doses of erythropoietin (Epo) administered with hypothermia improve neuroradiographic and short-term outcomes of newborns with hypoxic-ischemic encephalopathy.
In a phase II ...double-blinded, placebo-controlled trial, we randomized newborns to receive Epo (1000 U/kg intravenously; n = 24) or placebo (n = 26) at 1, 2, 3, 5, and 7 days of age. All infants had moderate/severe encephalopathy; perinatal depression (10 minute Apgar <5, pH <7.00 or base deficit ≥15, or resuscitation at 10 minutes); and received hypothermia. Primary outcome was neurodevelopment at 12 months assessed by the Alberta Infant Motor Scale and Warner Initial Developmental Evaluation. Two independent observers rated MRI brain injury severity by using an established scoring system.
The mean age at first study drug was 16.5 hours (SD, 5.9). Neonatal deaths did not significantly differ between Epo and placebo groups (8% vs 19%, P = .42). Brain MRI at mean 5.1 days (SD, 2.3) showed a lower global brain injury score in Epo-treated infants (median, 2 vs 11, P = .01). Moderate/severe brain injury (4% vs 44%, P = .002), subcortical (30% vs 68%, P = .02), and cerebellar injury (0% vs 20%, P = .05) were less frequent in the Epo than placebo group. At mean age 12.7 months (SD, 0.9), motor performance in Epo-treated (n = 21) versus placebo-treated (n = 20) infants were as follows: Alberta Infant Motor Scale (53.2 vs 42.8, P = .03); Warner Initial Developmental Evaluation (28.6 vs 23.8, P = .05).
High doses of Epo given with hypothermia for hypoxic-ischemic encephalopathy may result in less MRI brain injury and improved 1-year motor function.
The clinical research enterprise is not producing the evidence decision makers arguably need in a timely and cost effective manner; research currently involves the use of labor-intensive parallel ...systems that are separate from clinical care. The emergence of pragmatic clinical trials (PCTs) poses a possible solution: these large-scale trials are embedded within routine clinical care and often involve cluster randomization of hospitals, clinics, primary care providers, etc. Interventions can be implemented by health system personnel through usual communication channels and quality improvement infrastructure, and data collected as part of routine clinical care. However, experience with these trials is nascent and best practices regarding design operational, analytic, and reporting methodologies are undeveloped.
To strengthen the national capacity to implement cost-effective, large-scale PCTs, the Common Fund of the National Institutes of Health created the Health Care Systems Research Collaboratory (Collaboratory) to support the design, execution, and dissemination of a series of demonstration projects using a pragmatic research design.
In this article, we will describe the Collaboratory, highlight some of the challenges encountered and solutions developed thus far, and discuss remaining barriers and opportunities for large-scale evidence generation using PCTs.
A planning phase is critical, and even with careful planning, new challenges arise during execution; comparisons between arms can be complicated by unanticipated changes. Early and ongoing engagement with both health care system leaders and front-line clinicians is critical for success. There is also marked uncertainty when applying existing ethical and regulatory frameworks to PCTS, and using existing electronic health records for data capture adds complexity.
Sequential cross-sectional study.
To quantify patterns of outpatient lumbar spine surgery.
Outpatient lumbar spine surgery patterns are undocumented.
We used CPT-4 and ICD-9-CM diagnosis/procedure ...codes to identify lumbar spine operations in 20+ year olds. We combined sample volume estimates from the National Hospital Discharge Survey (NHDS), the National Survey of Ambulatory Surgery (NSAS), and the Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) with complete case counts from HCUP's State Inpatient Databases (SIDs) and State Ambulatory Surgery Databases (SASDs) for four geographically diverse states. We excluded pregnant patients and those with vertebral fractures, cancer, trauma, or infection. We calculated age- and sex-adjusted rates.
Ambulatory cases comprised 4% to 13% of procedures performed from 1994 to 1996 (NHDS/NSAS data), versus 9% to 17% for 1997 to 2000 (SID/SASD data). Discectomies comprised 70% to 90% of outpatient cases. Conversely, proportions of discectomies performed on outpatients rose from 4% in 1994 to 26% in 2000. Outpatient fusions and laminectomies were uncommon. NIS data indicate that nationwide inpatient surgery rates were stable (159 cases/100,000 in 1994 vs. 162/100,000 in 2000). However, combined data from all sources suggest that inpatient and outpatient rates rose from 164 cases/100,000 in 1994 to 201/100,000 in 2000.
While inpatient lumbar surgery rates remained relatively stable for 1994 to 2000, outpatient surgery increased over time.
This analysis of the lumbar epidural steroid injections for spinal stenosis multicenter randomized controlled trial data identifies the degree of and risk factors for cortisol suppression after ...epidural steroid injections in older adults with spinal stenosis. Four hundred patients aged 50 years and older with back or leg pain and central lumbar spinal stenosis completed baseline demographic and psychosocial measures. Morning serum cortisol levels were measured at baseline and 3 weeks after initial injection. Patients were randomized to receive epidural injections of either local anesthetic with corticosteroid (n = 200) or local anesthetic only (n = 200). The specific corticosteroid was chosen at the treating physician's discretion (methylprednisolone, betamethasone, triamcinolone, or dexamethasone). Thirty-two patients (20.3%) treated with corticosteroid experienced cortisol reduction at 3 weeks of >50% compared with 10 patients (6.7%) treated with lidocaine only (adjusted treatment effect = 3.5, 95% confidence interval: 1.6-7.9, P = 0.002). The effect on 3-week cortisol changes did not differ by demographic or patient-level characteristics. Those treated with methylprednisolone or triamcinolone had an average 3-week cortisol reduction of 41.0% (P = 0.005) and 41.6% (P < 0.001) from baseline, respectively, whereas patients treated with betamethasone or dexamethasone were not significantly different than comparable patients in the lidocaine arm. The higher rates of cortisol suppression at 3 weeks in those receiving epidural corticosteroid injections, particularly with longer-acting insoluble corticosteroid formulations, are consistent with sustained systemic absorption of corticosteroid.
PDF
