Search

Purpose To compare mycophenolate mofetil (MMF) to methotrexate (MTX) as corticosteroid-sparing therapy for ocular inflammatory diseases. Design Retrospective analysis of cohort study data. Methods Participants were identified from the Systemic Immunosuppressive Therapy for Eye Diseases Cohort Study. Demographic and clinical characteristics were obtained via medical record review. The study included 352 patients who were taking single-agent immunosuppression with MTX or MMF at 4 tertiary uveitis clinics. Marginal structural models (MSM)-derived statistical weighting created a virtual population with covariates and censoring patterns balanced across alternative treatments. With this methodological approach, the results estimate what would have happened had none of the patients stopped their treatment. Survival analysis with stabilized MSM-derived weights simulated a clinical trial comparing MMF vs MTX for noninfectious inflammatory eye disorders. The primary outcome was complete control of inflammation on prednisone ≤10 mg/day, sustained for ≥30 days. Results The time to success was shorter (more favorable) for MMF than MTX (hazard ratio = 0.68, 95% confidence interval: 0.46-0.99). Adjusting for covariates, the proportion achieving success was higher at every point in time for MMF than MTX from 2 to 8 months, then converges at 9 months. The onset of corticosteroid-sparing success took more than 3 months for most patients in both groups. Outcomes of treatment (MMF vs MTX) were similar across all anatomic sites of inflammation. The incidence of stopping therapy for toxicity was similar in both groups. Conclusions Our results suggest that, on average, MMF may be faster than MTX in achieving corticosteroid-sparing success in ocular inflammatory diseases.

We discuss some causal estimands that are used to study racial discrimination in policing. A central challenge is that not all police–civilian encounters are recorded in administrative datasets and available to researchers. One possible solution is to consider the average causal effect of race conditional on the civilian already being detained by the police. We find that such an estimand can be quite different from the more familiar ones in causal inference and needs to be interpreted with caution. We propose using an estimand that is new for this context—the causal risk ratio, which has more transparent interpretation and requires weaker identification assumptions. We demonstrate this through a reanalysis of the NYPD Stop-and-Frisk dataset. Our reanalysis shows that the naive estimator that ignores the posttreatment selection in administrative records may severely underestimate the disparity in police violence between minorities and whites in these and similar data.

INTRODUCTION: We evaluated visual acuity (VA) over 5 years in a subspecialty non-infectious uveitis population. METHODS: Retrospective data from 5,530 non-infectious uveitis patients with anterior, intermediate, posterior or panuveitis were abstracted by expert reviewers. Mean VA was calculated using inverse probability of censoring weighting to account for losses to follow-up. RESULTS: Patients were a median of 41 years old, 65% female, and 73% white. Initial mean VA was worse among panuveitis (20/84) than posterior (20/64), intermediate (20/47), and anterior (20/37) uveitides. On average, mean VA improved by 0.62, 0.51, 0.37, and 0.26 logMAR-equivalent lines over 2 years, respectively (each P

In patients with chronic kidney disease (CKD), clinical interest often centers on determining treatments and exposures that are causally related to renal progression. Analyses of longitudinal clinical data in this population are often complicated by clinical competing events, such as end-stage renal disease (ESRD) and death, and time-dependent confounding, where patient factors that are predictive of later exposures and outcomes are affected by past exposures. We developed multistate marginal structural models (MS-MSMs) to assess the effect of time-varying systolic blood pressure on disease progression in subjects with CKD. The multistate nature of the model allows us to jointly model disease progression characterized by changes in the estimated glomerular filtration rate (eGFR), the onset of ESRD, and death, and thereby avoid unnatural assumptions of death and ESRD as noninformative censoring events for subsequent changes in eGFR. We model the causal effect of systolic blood pressure on the probability of transitioning into 1 of 6 disease states given the current state. We use inverse probability weights with stabilization to account for potential time-varying confounders, including past eGFR, total protein, serum creatinine, and hemoglobin. We apply the model to data from the Chronic Renal Insufficiency Cohort Study, a multisite observational study of patients with CKD.

In randomized controlled trials, the evaluation of an overall treatment effect is often followed by effect modification or subgroup analyses, where the possibility of a different magnitude or direction of effect for varying values of a covariate is explored. While studies of effect modification are typically restricted to pretreatment covariates, longitudinal experimental designs permit the examination of treatment effect modification by intermediate outcomes, where intermediates are measured after treatment but before the final outcome. We present a novel application of generalized structural mean models (GSMMs) for simultaneously assessing effect modification by post-treatment covariates and accounting for noncompliance to assigned treatment status. The proposed approach may also be used to identify post-treatment effect modifiers in the absence of noncompliance. The methods are evaluated using a simulation study that demonstrates that our approach retains consistent estimation of effect modification by intermediate variables that are affected by treatment and also predict outcomes. We illustrate the method using a randomized trial designed to promote re-employment through teaching skills to enhance self-esteem and inoculate job seekers against setbacks in the job search process. Our analysis provides some evidence that the intervention was much less successful among subjects that displayed higher levels of depression at intermediate post-treatment waves of the study. We also compare the assumptions of our approach and principal stratification as alternatives to account for differences in effects by intermediate variables.

CKD is a global public health problem with significant mortality and morbidity.We examined the multivariable association of plasma levels of IL-1, IL-1 receptor antagonist, IL-6, TNF-α, TGF-β, high-sensitivity C-reactive protein, fibrinogen, and serum albumin with progression of CKD in 3430 Chronic Renal Insufficiency Cohort study participants.Over a median follow-up time of 6.3 years, 899 participants reached the composite end point of ≥50% decline in eGFR from baseline or onset of ESRD. Elevated plasma levels of fibrinogen, IL-6, and TNF-α and lower serum albumin were associated with a greater decline in eGFR over time. After adjusting for demographics, BP, laboratory variables, medication use, and baseline eGFR, hazard ratios for the composite outcome were greater for the patients in the highest quartile of fibrinogen (hazard ratio, 2.05; 95% confidence interval, 1.64 to 2.55; P0.001), IL-6 (hazard ratio, 1.44; 95% confidence interval, 1.17 to 1.77; P0.01), and TNF-α (hazard ratio, 1.94; 95% confidence interval, 1.52 to 2.47; P0.001) compared with those in the respective lowest quartiles. The hazard ratio was 3.48 (95% confidence interval, 2.88 to 4.21; P0.001) for patients in the lowest serum albumin quartile relative to those in the highest quartile. When also adjusted for albuminuria, the associations of fibrinogen (hazard ratio, 1.49; 95% confidence interval, 1.20 to 1.86; P0.001), serum albumin (hazard ratio, 1.52; 95% confidence interval, 1.24 to 1.87; P0.001), and TNF-α (hazard ratio, 1.42; 95% confidence interval, 1.11 to 1.81; P0.001) with outcome were attenuated but remained significant.Elevated plasma levels of fibrinogen and TNF-α and decreased serum albumin are associated with rapid loss of kidney function in patients with CKD.

Marginal structural models are a class of causal models useful for characterizing the effect of treatment in the presence of time-varying confounding. They are more widely used than structural nested models, partly because these models are easier to understand and to implement. We extend marginal structural models to situations with clustered observations with unit- and cluster-level treatment and introduce an appropriate inferential method. We consider how to formulate models with cluster-level and unit-level treatments. For unit-level treatments, we consider cases with and without interference. We also consider the use of unit-specific inverse probability weights and certain working correlation structures to improve the efficiency of estimators in some situations. We apply our method to different scenarios including 2 or 3 units per cluster and a mixture of larger clusters. Simulation examples and data from the treatment arm of a glaucoma clinical trial were used to illustrate our method.

We address several questions relating to the use of standard regression and Structural Nested Mean Model (SNMM) approach (e. g., Ten Have et al. 2007) to analyze post-randomization effect modifiers of the intent-to-treat effect of a randomized intervention on a subsequent outcome, which has not been well examined. We show through simulations that the SNMM performs better with respect to bias of estimates of the intervention and interaction effects than does the corresponding standard interaction approach when the baseline intervention is randomized and the post-randomization factors are subject to confounding, and even when there is no association between the intervention and effect modifier. However, causal inference under the SNMM makes untestable assumptions that the causal contrasts do not vary across observed levels of the intervention and post-randomization factor. In addition, the precision of the SNMM-based estimators depends on the effect of the randomized intervention on the post-randomization factor varying across baseline covariate combinations. These issues and methods are illustrated with the application of the standard and causal methods to a randomized cognitive therapy (CT) trial, for which there is a conceptual model of negative cognitive styles or distortions impacted by CT but then in turn modifying the effect of CT on subsequent suicide ideation and social problem solving outcomes.

Propensity scores are widely used to control for confounding when estimating the effect of a binary treatment in observational studies. They have been generalized to ordinal and continuous treatments in the recent literature. Following the definition of propensity function and its parameterizations (called the propensity parameter in this paper) proposed by Imai and van Dyk, we explore sufficient conditions for selecting propensity parameters to control for confounding for continuous treatments in the context of regression-based adjustment in linear models. Typically, investigators make parametric assumptions about the form of the dose-response function for a continuous treatment. Such assumptions often allow the analyst to use only a subset of the propensity parameters to control confounding. When the treatment is the only predictor in the structural, that is, causal model, it is sufficient to adjust only for the propensity parameters that characterize the expectation of the treatment variable or its functional form. When the structural model includes selected baseline covariates other than the treatment variable, those baseline covariates, in addition to the propensity parameters, must also be adjusted in the model. We demonstrate these points with an example estimating the dose-response relationship for the effect of erythropoietin on hematocrit level in patients with end-stage renal disease.

OBJECTIVES: ImproveCareNow (ICN) is the largest pediatric learning health system in the nation and started as a quality improvement collaborative. To test the feasibility and validity of using ICN data for clinical research, we evaluated the effectiveness of anti-tumor necrosis factor-α (anti-TNFα) agents in the management of pediatric Crohn disease (CD). METHODS: Data were collected in 35 pediatric gastroenterology practices (April 2007 to March 2012) and analyzed as a sequence of nonrandomized trials. Patients who had moderate to severe CD were classified as initiators or non-initiators of anti-TNFα therapy. Among 4130 patients who had pediatric CD, 603 were new users and 1211 were receiving anti-TNFα therapy on entry into ICN. RESULTS: During a 26-week follow-up period, rate ratios obtained from Cox proportional hazards models, adjusting for patient and disease characteristics and concurrent medications, were 1.53 (95% confidence interval [CI], 1.20–1.96) for clinical remission and 1.74 (95% CI, 1.33–2.29) for corticosteroid-free remission. The rate ratio for corticosteroid-free remission was comparable to the estimate produced by the adult SONIC study, which was a randomized controlled trial on the efficacy of anti-TNFα therapy. The number needed to treat was 5.2 (95% CI, 3.4–11.1) for clinical remission and 5.0 (95% CI, 3.4–10.0) for corticosteroid-free remission. CONCLUSIONS: In routine pediatric gastroenterology practice settings, anti-TNFα therapy was effective at achieving clinical and corticosteroid-free remission for patients who had Crohn disease. Using data from the ICN learning health system for the purpose of observational research is feasible and produces valuable new knowledge.

Objective The race-specific association of inflammation with adiposity and muscle mass in subjects with chronic kidney disease (CKD) was examined. Methods Plasma concentration of interleukin (IL)−1β, IL-1 receptor antagonist (IL-1RA), IL-6, IL-10, tumor necrosis factor (TNF)-α, TGF-β, high-sensitivity C-reactive protein (hs-CRP), fibrinogen, and serum albumin was measured in 3,939 Chronic Renal Insufficiency Cohort study participants. Bioelectric impedance analysis was used to determine body fat mass (BFM) and fat-free mass (FFM). Results Plasma levels of hs-CRP, fibrinogen, IL-1RA, IL-6, and TNF-α increased and serum albumin decreased across the quartiles of body mass index. In multivariable analysis, BFM and FFM were positively associated with hs-CRP, fibrinogen, IL-1β, IL-1RA, and IL-6. One standard deviation (SD) increase in BFM and FFM was associated with 0.36 (95% confidence interval [CI] = 0.33, 0.39) and 0.26 (95% CI = 0.22, 0.30) SD increase in log-transformed hs-CRP, respectively (P

Background Various indicators of progression of chronic kidney disease (CKD) have been used as outcomes in clinical research studies. The effect of using varying measures on the association of risk factors with CKD progression has not been well characterized. Study Design Prospective cohort study. Setting & Participants The Chronic Renal Insufficiency Cohort (CRIC) Study (N=3,939) enrolled men and women with mild to moderate CKD, 48% of whom had diabetes and 42% were self-reported black race. Predictors Age, race, sex, diabetes, baseline estimated glomerular filtration rate (eGFR), proteinuria, and other established CKD risk factors. Outcomes Death, end-stage renal disease (ESRD), and eGFR events, including: (1) eGFR halving, (2) eGFR 2 , (3) eGFR halving and 2 , (4) eGFR decrease of 20mL/min/1.73m 2 , (5) eGFR halving or decrease of 20mL/min/1.73m 2 , and (6) eGFR decrease of 25% and change in CKD stage. Results Mean entry eGFR was 44.9mL/min/1.73m 2 . Annual rates of death, ESRD, and eGFR halving were 2.5%, 4.0%, and 6.1%, respectively, during an average follow-up of 5.4 years. Associations between risk factors and ESRD and eGFR events were similar across different definitions. However, these associations were substantially different from those with death. HRs for ESRD, eGFR halving, and death in the highest compared to the lowest proteinuria category were 11.83 (95% CI, 8.40-16.65), 11.19 (95% CI, 8.53-14.68), and 1.47 (95% CI, 1.10-1.96), respectively. Limitations Participants may not be representative of the entire CKD population. Conclusions Using ESRD or eGFR events, but not death, in the definition of kidney disease outcomes is appropriate in follow-up studies to identify risk factors for CKD progression.

Treatment of many conditions in ophthalmology involves an extended sequence of decisions over a period of time. For example, in managing glaucoma, a target intraocular pressure (IOP) may be selected, and a topical eye drop prescribed, after which the IOP is reassessed. If the IOP goal is not met using the one eye drop, another may be prescribed, and so on1. Failure to adequately control the IOP initially may lead to worsening of glaucoma as manifested by increased optic disk cupping and/or visual field loss, after which the target pressure may be revised further downward, and additional interventions considered. At each visit, a decision is made about whether or not to start a new treatment, or whether to stop an existing treatment. The goal of a clinician contemplating a series of decisions should be to choose the best available decision at each time or the best available strategy over time. Evaluation and comparison of such strategies using observational or non-experimental data is more difficult than the evaluation of simple one-time decisions or treatments, because of the complexities of characterizing the joint effects of a sequence of treatment decisions over time and adjusting for time-varying confounders in these settings.

Purpose Erythropoietin (EPO) improves measures of quality of life and reduces transfusions. Clinical trials have reported higher mortality associated with higher hemoglobin targets in varied clinical settings, making difficult the selection of erythropoiesis stimulation strategies in end-stage kidney disease. Observational studies distinguishing an effect of EPO from underlying conditions are challenging, but promise insights relevant to real-world settings. Methods Using data from the United States Renal Data System, we performed a retrospective cohort study of hemodialysis patients treated between 2000 and 2004. 409 364 Medicare insured patients receiving hemodialysis therapy as of January 2000 or who began dialysis after January 2000 and survived >6 months were studied. We examined the association of EPO dose in any given month with death over subsequent follow-up. Results Within each hematocrit group ( 39%), the hazard ratios comparing the 80th percentile to the median EPO dose were 0.88 (95% CI: [0.87–0.90]), 0.94 ([0.93–0.94]), 0.98 ([0.98–0.99]), 1.06 ([1.05–1.06]) and 1.08 ([1.07–1.09]), respectively. Within the highest hematocrit group, the association of a high EPO dose with elevated mortality was attenuated over time. Among patients with malignancy or indications of EPO resistance, the association of higher EPO dose with lower mortality was attenuated when hematocrit was low, while its association with higher mortality was stronger when hematocrit was high. Conclusions These analyses demonstrate a complex relationship between EPO dosing and mortality, suggesting a possible beneficial effect among severely anemic hemodialysis patients, but possible harm when administered to individuals with higher hematocrit levels. Copyright © 2013 John Wiley & Sons, Ltd.

Tom Ten Have made many contributions to causal inference and biostatistics before his untimely death. This paper reviews Tom's contributions and discusses potential related future research directions. We focus on Tom's contributions to longitudinal/repeated measures categorical data analysis and particularly his contributions to causal inference. Tom's work on causal inference was primarily in the areas of estimating the effect of receiving treatment in randomized trials with nonadherence and mediation analysis. A related area to mediation analysis he was working on at the time of his death was posttreatment effect modification with applications to designing adaptive treatment strategies.

To evaluate the incidence of remission among patients with intermediate uveitis; to identify factors potentially predictive of remission.Retrospective cohort study.Involved eyes of patients with primary noninfectious intermediate uveitis at 4 academic ocular inflammation subspecialty practices, followed sufficiently long to meet the remission outcome definition, were studied retrospectively by standardized chart review data. Remission of intermediate uveitis was defined as a lack of inflammatory activity at ≥2 visits spanning ≥90 days in the absence of any corticosteroid or immunosuppressant medications. Factors potentially predictive of intermediate uveitis remission were evaluated using survival analysis.Among 849 eyes (of 510 patients) with intermediate uveitis followed over 1934 eye-years, the incidence of intermediate uveitis remission was 8.6/100 eye-years (95% confidence interval [CI], 7.4-10.1). Factors predictive of disease remission included prior pars plana vitrectomy (PPV) (hazard ratio [HR] [vs no PPV] = 2.39; 95% CI, 1.42-4.00), diagnosis of intermediate uveitis within the last year (HR [vs diagnosis5 years ago] =3.82; 95% CI, 1.91-7.63), age ≥45 years (HR [vs age45 years] = 1.79; 95% CI, 1.03-3.11), female sex (HR = 1.61; 95% CI, 1.04-2.49), and Hispanic race/ethnicity (HR [vs white race] = 2.81; 95% CI, 1.23-6.41). Presence/absence of a systemic inflammatory disease, laterality of uveitis, and smoking status were not associated with differential incidence.Our results suggest that intermediate uveitis is a chronic disease with an overall low rate of remission. Recently diagnosed patients and older, female, and Hispanic patients were more likely to remit. With regard to management, pars plana vitrectomy was associated with increased probability of remission.

Background Glomerular filtration rate (GFR) is considered the best measure of kidney function, but repeated assessment is not feasible in most research studies. Study Design Cross-sectional study of 1,433 participants in the Chronic Renal Insufficiency Cohort (CRIC) Study (ie, the GFR subcohort) to derive an internal GFR estimating equation using a split-sample approach. Setting & Participants Adults from 7 US metropolitan areas with mild to moderate chronic kidney disease; 48% had diabetes and 37% were black. Index Test CRIC GFR estimating equation. Reference Test or Outcome Urinary 125 I-iothalamate clearance testing (measured GFR [mGFR]). Other Measurements Laboratory measures, including serum creatinine and cystatin C, and anthropometrics. Results In the validation data set, the model that included serum creatinine level, serum cystatin C level, age, sex, and race was the most parsimonious and similarly predictive of mGFR compared with a model additionally including bioelectrical impedance analysis phase angle, CRIC clinical center, and 24-hour urinary creatinine excretion. Specifically, root mean square errors for the separate models were 0.207 versus 0.202, respectively. Performance of the CRIC GFR estimating equation was most accurate for the subgroups of younger participants, men, nonblacks, non-Hispanics, those without diabetes, those with body mass index 2 , those with higher 24-hour urine creatinine excretion, those with lower high-sensitivity C-reactive protein levels, and those with higher mGFRs. Limitations Urinary clearance of 125 I-iothalamate is an imperfect measure of true GFR; cystatin C level is not standardized to certified reference material; lack of external validation; small sample sizes limit analyses of subgroup-specific predictors. Conclusions The CRIC GFR estimating equation predicts mGFR accurately in the CRIC cohort using serum creatinine and cystatin C levels, age, sex, and race. Its performance was best in younger and healthier participants.

We review the concept of time-dependent confounding by using the example in paper “Comparative effectiveness of individual angiotensin receptor blockers on risk of mortality in patients with chronic heart failure” by Desai et al. and illustrate how to adjust for it by using inverse probability of treatment weighting through a simulated example. We discuss a few subtle issues that arise in specification of the model for treatment required to fit marginal structural models (MSMs) and in specification of the structural model for the outcome. We discuss the differences between the effects estimated in MSMs and intention-to-treat effects estimated in randomized trials, followed by an outline of some limitations of MSMs. Copyright © 2012 John Wiley & Sons, Ltd.

Recent theoretical studies have shown that conditioning on an instrumental variable (IV), a variable that is associated with exposure but not associated with outcome except through exposure, can increase both bias and variance of exposure effect estimates. Although these findings have obvious implications in cases of known IVs, their meaning remains unclear in the more common scenario where investigators are uncertain whether a measured covariate meets the criteria for an IV or rather a confounder. The authors present results from two simulation studies designed to provide insight into the problem of conditioning on potential IVs in routine epidemiologic practice. The simulations explored the effects of conditioning on IVs, near-IVs (predictors of exposure that are weakly associated with outcome), and confounders on the bias and variance of a binary exposure effect estimate. The results indicate that effect estimates which are conditional on a perfect IV or near-IV may have larger bias and variance than the unconditional estimate. However, in most scenarios considered, the increases in error due to conditioning were small compared with the total estimation error. In these cases, minimizing unmeasured confounding should be the priority when selecting variables for adjustment, even at the risk of conditioning on IVs.

Brachial artery measures of BP are associated with increasing degrees of proteinuria. Whether central measures of BP or vascular stiffness are associated with increased risk of proteinuria in patients with chronic kidney disease (CKD) is unknown.Measurements of central and brachial artery BP, and aortic pulse wave velocity (PWV) were performed in a cross-sectional cohort of patients with CKD (n = 2144) from the Chronic Renal Insufficiency Cohort (CRIC) study to determine factors which predict increased risk of proteinuria. Multivariate analysis stratified by diabetes included age, ethnicity, gender, estimated glomerular filtration rate (GFR), waistline, smoking, heart rate, and medications to evaluate the relationship of hemodynamic factors and proteinuria.Brachial artery systolic BP (SBP) was important as an explanatory factor for variations in proteinuria among both diabetics (R(2) = 0.40, P0.0001) and non diabetics (R(2) = 0.38, P0.001). Measures of peripheral pulse pressure (PP), central SBP, and central pulse pressure added little to the explained variation in proteinuria beyond brachial artery SBP, whereas PWV as a measure of vascular stiffness incrementally accounted for a significant portion of variation in proteinuria beyond that explained by brachial artery SBP in diabetics (R(2) = 0.42, P0.001) but not non diabetics.Brachial artery SBP and PWV are both associated with variations in proteinuria in patients with CKD.

Background/Aims: Metabolic syndrome may increase the risk for incident cardiovascular disease (CVD) and all-cause mortality in the general population. It is unclear whether, and to what degree, metabolic syndrome is associated with CVD in chronic kidney disease (CKD). We determined metabolic syndrome prevalence among individuals with a broad spectrum of kidney dysfunction, examining the role of the individual elements of metabolic syndrome and their relationship to prevalent CVD. Methods: We evaluated four models to compare metabolic syndrome or its components to predict prevalent CVD using prevalence ratios in the Chronic Renal Insufficiency Cohort (CRIC) Study. Results: Among 3,939 CKD participants, the prevalence of metabolic syndrome was 65% and there was a significant association with prevalent CVD. Metabolic syndrome was more common in diabetics (87.5%) compared with non-diabetics (44.3%). Hypertension was the most prevalent component, and increased triglycerides the least prevalent. Using the bayesian information criterion, we found that the factors defining metabolic syndrome, considered as a single interval-scaled variable, was the best of four models of metabolic syndrome, both for CKD participants overall and for diabetics and non-diabetics separately. Conclusion: The predictive value of this model for future CVD outcomes will subsequently be validated in longitudinal analyses.

We describe causal mediation methods for analysing the mechanistic factors through which interventions act on outcomes. A number of different mediation approaches have been presented in the biomedical, social science and statistical literature with an emphasis on different aspects of mediation. We review the different sets of assumptions that allow identification and estimation of effects in the simple case of a single intervention, a temporally subsequent mediator and outcome. These assumptions include various no confounding assumptions including sequential ignorability assumptions and also interaction assumptions involving the treatment and mediator. The understanding of such assumptions is crucial since some can be assessed under certain conditions (e.g. treatment–mediator interactions), whereas others cannot (sequential ignorability). These issues become more complex with multiple mediators and longitudinal outcomes. In addressing these assumptions, we review several causal approaches to mediation analyses.

Background Aortic pulse wave velocity (PWV) is a measure of arterial stiffness and has proved useful in predicting cardiovascular morbidity and mortality in several populations of patients, including the healthy elderly, hypertensives and those with end-stage renal disease receiving hemodialysis. Little data exist characterizing aortic stiffness in patients with chronic kidney disease (CKD) who are not receiving dialysis, and in particular the effect of reduced kidney function on aortic PWV. Methods We performed measurements of aortic PWV in a cross-sectional cohort of participants enrolled in the Chronic Renal Insufficiency Cohort (CRIC) study to determine factors which predict increased aortic PWV in CKD. Results PWV measurements were obtained in 2,564 participants. The tertiles of aortic PWV (adjusted for waist circumference) were 10.2 m/s with an overall mean (+/- s.d.) value of 9.48 +/- 3.03 m/s (95% confidence interval = 9.35-9.61 m/s). Multivariable regression identified significant independent positive associations of age, blood glucose concentrations, race, waist circumference, mean arterial blood pressure, gender, and presence of diabetes with aortic PWV and a significant negative association with the level of kidney function. Conclusions The large size of this unique cohort, and the targeted enrollment of CKD participants provides an ideal situation to study the role of reduced kidney function as a determinant of arterial stiffness. Arterial stiffness may be a significant component of the enhanced cardiovascular risk associated with kidney failure.

The barriers to live donor transplantation are poorly understood. We performed a prospective cohort study of individuals undergoing renal transplant evaluation. Participants completed a questionnaire that assessed clinical characteristics as well as knowledge and beliefs about transplantation. A participant satisfied the primary outcome if anyone contacted the transplant center to be considered as a live donor for that participant. The final cohort comprised 203 transplant candidates, among whom 80 (39.4%) had a potential donor contact the center and 19 (9.4%) underwent live donor transplantation. In multivariable logistic regression, younger candidates (OR 1.65 per 10 fewer years, p0.01) and those with annual incomeor=US$ 15 000 (OR 4.22, p = 0.03) were more likely to attract a potential live donor. Greater self-efficacy, a measure of the participant's belief in his or her ability to attract a donor, was a predictor of having a potential live donor contact the center (OR 2.73 per point, p0.01), while knowledge was not (p = 0.56). The lack of association between knowledge and having a potential donor suggests that more intensive education of transplant candidates will not increase live donor transplantation. On the other hand, self-efficacy may be an important target in designing interventions to help candidates find live donors.

Background and objectives: Anemia is a common and early complication of nondialysis chronic kidney disease (CKD). One contributing factor is iron deficiency, which may be particularly problematic during erythropoietin replacement therapy. The aim of this study was to examine the prevalence of iron deficiency in nondialysis CKD. Design, setting, participants, & measurements: The National Health and Nutritional Examination Survey (NHANES) data for NHANES III (1988 to 1994) and subsequent NHANES 2-yr datasets, 1999 to 2000, 2001 to 2002, and 2003 to 2004 were analyzed for individuals >18 yr old. Results: It was found that low levels of iron tests [either serum ferritin < 100 ng/ml or transferrin saturation (TSAT) < 20%] were present in most patients with reduced creatinine clearance (CrCl). The percentage of low iron tests was higher among women than men, present in 57.8 to 58.8% of men and 69.9 to 72.8% of women (P < 0.001). With declining levels of CrCl, in women, TSAT levels decreased, whereas, surprisingly, serum ferritin tended to progressively increase. The percentage of anemic subjects increased progressively with declining quartiles of TSAT but was unrelated to serum ferritin quartiles. Conclusions: It was found that low levels of iron tests, following National Kidney Foundation/Kidney Disease Outcomes Quality Initiative guidelines (either serum ferritin < 100 ng/ml or TSAT < 20%) were present in most patients with reduced CrCl.

Delayed allograft function (DGF) is a common adverse event in postrenal transplantation. The etiology of DGF is thought to include both nonimmunologic (donor age, cold ischemia time, and recipient race) and immunologic factors. We examined the association of DGF with amino acid mismatches at 66 variable sites of the HLA-A molecule in a prospective cohort study of 697 renal transplant recipients of deceased donors. Using a multivariate logistic regression model adjusted for nonimmunologic risk factors, we show that combinations of a few amino acid mismatches at crucial sites of HLA-A molecules were associated with DGF. In Caucasian recipients, a mismatch at position 62, 95, or 163, all known to be functionally important within the antigen recognition site, was associated with an increased risk for DGF. Furthermore, a decreased risk for DGF was associated with a mismatch at HLA-A family-specific sites (149, 184, 193, or 246), indicating that evolutionary features of HLA-A polymorphism separating HLA-A families and lineages among donor-recipient pairs may correlate with the magnitude of alloreactivity influencing the development of DGF. These findings suggest that amino acid polymorphisms at functionally important positions at the antigen recognition site of the HLA-A molecule have a significant influence on DGF.

Background and objectives: Hemoglobin variability is common among dialysis patients, and has been associated with increased mortality. The causal nature of this association has been difficult to ascertain because of potential time-dependent confounding, for which traditional statistical methods do not control. Design, settings, participants, & measurements: A retrospective cohort of 34,963 Fresenius Medical care dialysis patients from 1996 was assembled. Hemoglobin variability, absolute hemoglobin level, and temporal hemoglobin trend were measured over rolling 6-mo exposure windows. Their association with mortality was estimated using history-adjusted marginal structural analysis that adjusts for time-dependent confounding by applying weights to observations inversely related to the predictability of observed levels of hemoglobin. Results: In the primary analysis, each g/dl increase in hemoglobin variability was associated with an adjusted hazard ratio (HR) [95% confidence interval (CI)] for all-cause mortality of 1.93 (1.20 to 3.10). Neither higher absolute hemoglobin level nor increasing hemoglobin trend were significantly associated with mortality; adjusted HR (95% CI) 0.85 (0.64 to 1.11) and 0.60 (0.25 to 1.45), respectively. Conclusions: Marginal structural analysis demonstrates that hemoglobin variability is associated with increased mortality among chronic hemodialysis patients, and that this effect is more pronounced than appreciated using standard statistical techniques that do not take time-dependent confounding into account.

Hemoglobin levels vary substantially over time in hemodialysis patients, and this variability may portend poor outcomes. For a given patient, hemoglobin concentration over time can be described by absolute levels, rate of change, or by the difference between observed level and expected level based on the preceding trend (i.e., seemingly random variability). We investigated the independent associations of these different methods of describing hemoglobin over time with mortality in a retrospective cohort of 34,963 hemodialysis patients. Hemoglobin concentration over time was modeled with linear regression for each subject, and the model was then used to define the subject's absolute level of hemoglobin (intercept), temporal trend in hemoglobin (slope), and hemoglobin variability (residual standard deviation). Survival analyses indicated that each 1g/dl increase in the residual standard deviation was associated with a 33% increase in rate of death, even after adjusting for multiple covariates. Patient characteristics accounted for very little of the variation in our hemoglobin variability metric (R2 = 0.019). We conclude that greater hemoglobin variability is independently associated with higher mortality.

The use of propensity scores to adjust for measured confounding factors has become increasingly popular in cohort studies. However, their use in case-control and case-cohort studies has received little attention. The authors present some theory on the estimation and use of propensity scores in case-control and case-cohort studies and present the results of simulation studies that examine whether large-sample expectations are realized in studies of typical size. The application of propensity scores is less straightforward in case-control and case-cohort studies than in cohort studies. The authors' simulations revealed two potentially important issues. First, when using several potential approaches, there is artifactual effect modification of the odds ratio by level of propensity score. The magnitude of this phenomenon decreases as the sample size increases. Second, several potential approaches produce estimated propensity scores that do not converge to the true value as sample size increases and, thus, can fail to adjust fully for measured confounding factors. However, the magnitude of residual confounding appeared modest in our simulations. Researchers considering using propensity scores in case-control or case-cohort studies should consider the potential for artifactual effect modification and their reduced ability to control for potential confounding factors.

We present a linear rank preserving model (RPM) approach for analyzing mediation of a randomized baseline intervention's effect on a univariate follow-up outcome. Unlike standard mediation analyses, our approach does not assume that the mediating factor is also randomly assigned to individuals in addition to the randomized baseline intervention (i.e., sequential ignorability), but does make several structural interaction assumptions that currently are untestable. The G-estimation procedure for the proposed RPM represents an extension of the work on direct effects of randomized intervention effects for survival outcomes by Robins and Greenland (1994, Journal of the American Statistical Association 89, 737-749) and on intervention non-adherence by Ten Have et al. (2004, Journal of the American Statistical Association 99, 8-16). Simulations show good estimation and confidence interval performance by the proposed RPM approach under unmeasured confounding relative to the standard mediation approach, but poor performance under departures from the structural interaction assumptions. The trade-off between these assumptions is evaluated in the context of two suicide/depression intervention studies.

Background: A surrogate marker is a variable commonly used in clinical trials to guide treatment decisions when the outcome of ultimate interest is not available. A good surrogate marker is one where the treatment effect on the surrogate is a strong predictor of the effect of treatment on the outcome. We review the situation when there is one treatment delivered at baseline, one surrogate measured at one later time point, and one ultimate outcome of interest and discuss new issues arising when variables are time-varying. Methods: Most of the literature on surrogate markers has only considered simple settings with one treatment, one surrogate, and one outcome of interest at a fixed time point. However, more complicated time-varying settings are common in practice. In this article, we describe the unique challenges in two settings, time-varying treatments and time-varying surrogates, while relating the ideas back to the causal-effects and causal-association paradigms. Conclusion: In addition to discussing and extending popular notions of surrogacy to time-varying settings, we give examples illustrating that one can be misled by not taking into account time-varying information about the surrogate or treatment. We hope this article has provided some motivation for future work on estimation and inference in such settings.

Structural nested models (SNMs) and the associated method of G-estimation were first proposed by James Robins over two decades ago as approaches to modeling and estimating the joint effects of a sequence of treatments or exposures. The models and estimation methods have since been extended to dealing with a broader series of problems, and have considerable advantages over the other methods developed for estimating such joint effects. Despite these advantages, the application of these methods in applied research has been relatively infrequent; we view this as unfortunate. To remedy this, we provide an overview of the models and estimation methods as developed, primarily by Robins, over the years. We provide insight into their advantages over other methods, and consider some possible reasons for failure of the methods to be more broadly adopted, as well as possible remedies. Finally, we consider several extensions of the standard models and estimation methods., Published in at http://dx.doi.org/10.1214/14-STS493 the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

Efficient semiparametric estimation of longitudinal causal effects is often analytically or computationally intractable. We propose a novel restricted estimation approach for increasing efficiency, which can be used with other techniques, is straightforward to implement, and requires no additional modeling assumptions.

In assessing the efficacy of a time-varying treatment structural nested models (SNMs) are useful in dealing with confounding by variables affected by earlier treatments. These models often consider treatment allocation and repeated measures at the individual level. We extend SNMMs to clustered observations with time-varying confounding and treatments. We demonstrate how to formulate models with both cluster- and unit-level treatments and show how to derive semiparametric estimators of parameters in such models. For unit-level treatments, we consider interference, namely the effect of treatment on outcomes in other units of the same cluster. The properties of estimators are evaluated through simulations and compared with the conventional GEE regression method for clustered outcomes. To illustrate our method, we use data from the treatment arm of a glaucoma clinical trial to compare the effectiveness of two commonly used ocular hypertension medications.

Background Left ventricular hypertrophy (LVH) and myocardial contractile dysfunction are independent predictors of mortality in patients with chronic kidney disease (CKD). The association between inflammatory biomarkers and cardiac geometry has not yet been studied in a large cohort of CKD patients with a wide range of kidney function. Methods Plasma levels of interleukin (IL)-1β, IL-1 receptor antagonist (IL-1RA), IL-6, tumor necrosis factor (TNF)-α, transforming growth factor (TGF)-β, high-sensitivity C-Reactive protein (hs-CRP), fibrinogen and serum albumin were measured in 3,939 Chronic Renal Insufficiency Cohort study participants. Echocardiography was performed according to the recommendations of the American Society of Echocardiography and interpreted at a centralized core laboratory. Results LVH, systolic dysfunction and diastolic dysfunction were present in 52.3%, 11.8% and 76.3% of the study subjects, respectively. In logistic regression analysis adjusted for age, sex, race/ethnicity, diabetic status, current smoking status, systolic blood pressure, urinary albumin- creatinine ratio and estimated glomerular filtration rate, hs-CRP (OR 1.26 [95% CI 1.16, 1.37], p

Anemia and mortality in hemodialysis patients: Accounting for morbidity and treatment variables updated over time.BackgroundThe objective of this study was to gain insight into the associations of anemia with mortality among maintenance hemodialysis (HD) patients and patient subgroups by an analysis that more comprehensively represents hemoglobin (Hb) level, morbidity, and treatment characteristics over time than was possible in prior observational studies.MethodsA cohort study was conducted among 5517 subjects in the American arm of the Dialysis Outcomes and Practice Patterns Study Phase I. We used proportional hazard analysis to model all-cause mortality as a function of Hb level measured 1, 3, and 6 months previously. Forty-five potentially confounding patient-level characteristics were considered, including demographics, comorbidities, and time-updated levels of erythropoietin and parenteral iron dosing, medical events, and laboratory and dialysis measures.ResultsCompared to Hb 11 to

Nonadherence to immunosuppressive medications may partly explain the worse allograft outcomes among black recipients of renal transplants. In a prospective cohort study of recipients of deceased donor renal transplants, microelectronic cap monitors were placed on bottles of one immunosuppressive medication to (1) measure average daily percentage adherence during the first posttransplantation year and (2) determine the factors associated with adherence. A total of 278 transplant recipients who provided sufficient microelectronic adherence data were grouped into four categories of average daily percentage adherence: 95 to 100% adherence (41.0% of patients), 80 to 95% adherence (32.4%), 50 to 80% adherence (12.9%), and 0 to 50% adherence (13.7%). In the unadjusted ordinal logistic regression model, black race was associated with decreased adherence (odds ratio [OR], 0.43; 95% confidence interval [CI], 0.26 to 0.72; P = 0.001). Cause of renal disease, Powerful Others health locus of control, transplant center, and dosing frequency were also associated with adherence. After adjustment for transplant center and dosing frequency, the association between black race and decreased adherence was substantially attenuated (OR, 0.65; 95% CI, 0.38 to 1.14, P = 0.13). Transplant center (P = 0.003) and increased dosing frequency (OR, 0.43; 95% CI, 0.22 to 0.86, for three or four times per day dosing; OR, 2.35; 95% CI, 1.01 to 5.45, for daily dosing; versus two times per day dosing; P = 0.003) remained independently associated with adherence. Other baseline demographic, socioeconomic, medical, surgical, and psychosocial characteristics were not associated with adherence. The transplant center and dosing frequencies of immunosuppressive medications are associated with adherence and explain a substantial proportion of the race-adherence relationship.

The objective of this study was to evaluate whether the apparent relationship demonstrated in prior studies between iron dosing and mortality in hemodialysis (HD) patients was confounded by incomplete representation of iron dosing and morbidity over time. A cohort study was conducted among 32,566 patients who received at least 1 yr of HD at the Fresenius Medical Corporation dialysis centers during 1996 to 1997. The outcome measure was all-cause mortality through mid-1998. A total of 19 demographic, comorbidity, and laboratory characteristics were available. By proportional hazards analysis, no adverse effect on 2-year survival was found for baseline iron dose over 6 mo ofor = 1000 mg, but statistically significant elevated mortality was demonstrated for iron doses1000 mg to 1800 mg (adjusted hazards ratio [HR] = 1.09; 95% confidence interval [CI], 1.01 to 1.17) and1800 mg (adjusted HR = 1.18; 95% CI, 1.09 to 1.27). However, fitting multivariable models that appropriately account for time-varying measures of iron administration as well as other fixed and time-varying measures of morbidity, the authors found no statistically significant association between any level of iron administration and mortality. This study suggests that previously observed associations between iron administration and higher mortality may have been confounded, and it provides cautious support for the safety of the judicious administration of cumulative iron doses1000 mg over 6 mo if needed to maintain target hemoglobin levels among patients treated with maintenance HD.

This article addresses unique causal issues in the context of a randomized study on improving adherence to best practice guidelines by primary care physicians (PCP's) in treating their depressed patients. The study assessed an encouragement strategy to improve PCP guideline adherence. In this context, we compare two causal approaches: the conditional-compliance (CC) Bayesian latent class and the conditional-observable (CO) structural mean model methods. The CC methods estimate contrasts between randomized encouragement and no-encouragement arms [intent-to-treat (ITT) estimand] given latent PCP guideline complier classes. The CO methods estimate contrasts between PCP guideline adherence and nonadherence conditions (as-treated estimand) given observed PCP adherence status. The CC ITT estimand for patients with PCP compliers equals the CO as-treated estimand depending on assumptions. One such assumption pertains to the absence of physician defiers, who do the opposite of what they are encouraged to do in tre...

Levels of BP have been associated with increasing rates of renal allograft failure from cadaveric donors, independent of renal function. The effect of BP, a modifiable risk factor, on the failure rates of renal allografts from living donors is unknown and maybe obscured by the rate of decline of renal function from this source of organs. A nonconcurrent cohort study collecting data from 392 recipients of a renal allograft from a living donor during 1990 to 2001 was performed. Multivariable Cox regression models were fit by means of time-varying terms for systolic BP (SBP), diastolic BP (DBP), mean arterial BP (MAP), pulse pressure, and renal function during the first year after transplantation to study the association of BP and the time to allograft failure. Potential nonlinear relationships were considered by fractional polynomial terms. Recipient gender, preemptive transplantation, and time-varying terms for the natural logarithm of creatinine clearance and acute rejection were retained in the multivariable model. Including separate multivariable models with nonlinear terms for SBP (P = 0.02), for DBP (P = 0.02), or for MAP (P = 0.05) during the first year significantly improved the fit of the respective models and confirmed that there is an association of BP and allograft failure independent of renal function. Pulse pressure had neither a linear nor nonlinear association with allograft failure. In this nonconcurrent study, the level of BP during the first year affected the survival of renal allografts from living donors, independent of renal function. Further investigation is required to confirm the level of BP that is optimal to prevent foreshortened duration of survival.

In traditional regression modeling, to control for confounding by a variable one must include it in the structural part of the statistical model. Marginal structural models are a flexible new set of causal models. The estimation methods used to estimate model parameters use weighting to control for confounding; this allows more flexibility in choosing covariates for inclusion in the structural model and allows the model to more precisely reflect the scientific questions of interest. An important example of this is in multicenter observational studies where there is confounding by cluster. We illustrate these points with data from a study of surgery to provide vascular access for hemodialysis and a study comparing different timings for coronary angioplasty.

Background: Understanding the factors associated with successful arteriovenous fistula (AVF) maturation may provide an opportunity to improve vascular access outcomes. Methods: This prospective cohort study enrolled 348 hemodialysis patients receiving an AVF and followed them until ascertainment of successful AVF maturation. Multivariable logistic regression was used to develop (1) explanatory models of factors associated with AVF success and (2) models for predicting success based on sociodemographic and comorbidity profiles. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the discriminative ability of our predictive model. Predictive probabilities of AVF maturation under the assumption of optimal surgical technique defined by explanatory models were calculated. Results: A total of 55.5% of AVFs successfully matured. A history of stroke or transient ischemic attack, increasing age, and dependence on dialysis at the time of access placement were associated with lower probabilities of maturation. Of potentially modifiable variables, maturation was associated with greater intraoperative doses of heparin, use of large-diameter veins, and mean arterial pressure of 85 mm Hg or greater. Using optimal surgical technique, the probability of successful AVF maturation would have been as high as 84%. A predictive logistic regression model had moderate ability to predict AVF maturation under optimal surgical therapy with an area under the ROC curve of 0.69. Conclusion: Surgical technique potentially may be modified to yield a greater probability of successful AVF maturation. Predictive probabilities of successful AVF maturation under optimal surgical technique reinforce the view that functioning AVFs can be achieved in a large proportion of hemodialysis patients, consistent with experience in selected regions of the United States and abroad.

Objective To examine the relationship between physical violence, controlling behavior, and spontaneous abortion (SAB). Design Nested case-control study. Setting Emergency department of a university hospital. Patient(s) One thousand one hundred ninety-nine pregnant women. Intervention(s) Main outcome measure(s) Physical violence and controlling behavior. Result(s) Cases experienced a SAB (n = 392) and controls maintained their pregnancy through 22 weeks (n = 807). Fifteen percent of women reported violence during the pregnancy, and 49% had reported one or more past episodes of violence. We found no relationship between any measure of physical violence (past, current, or by perpetrator) and the risk of SAB. Conclusion(s) Although physical violence was very prevalent in the study population, exposure to violence did not influence the risk of SAB.

The role of depressive symptoms and optimism on the risk of spontaneous abortion has not been well examined. Prior studies have documented that emotional stress can affect immune and endocrine function, which may in turn influence fertilization, implantation, embryogenesis, and successful fetal growth.We enrolled pregnant women coming to the emergency department at the Hospital of the University of Pennsylvania from January 1999 through August 2001 to examine the role of depressive symptoms and optimism on spontaneous abortion. Cases were women who experienced a spontaneous abortion during the follow-up period (n = 174), and controls were women who maintained their pregnancy through 22 weeks (n = 798). Baseline interviews collected measures of depressive symptoms, using the Center for Epidemiologic Studies Depression Scale (CES-D) and optimism using the Life Oriented Test-Revised (LOT-R).We did not find a significant, prospective relationship between depressive symptoms or optimism and the risk of spontaneous abortion in our specific study population.The impact of depressive symptoms and optimism on pregnancy outcomes should be examined further in other populations.

SUMMARY The compliance score in randomized trials is a measure of the effect of randomization on treatment received. It is in principle a group-level pretreatment variable and so can be used where individual-level measures of treatment received can produce misleading inferences. The interpretation of models with the compliance score as a regressor of interest depends on the link function. Using the identity link can lead to valid inference about the effects of treatment received even in the presence of nonrandom noncompliance; such inference is more problematic for nonlinear links. We illustrate these points with data from two randomized trials.