RESUMO
Most estimation algorithms for adaptive treatment strategies assume that treatment rules at each decision point are independent from one another in the sense that they do not possess any common parameters. This is often unrealistic, as the same decisions may be made repeatedly over time. Sharing treatment-decision parameters across decision points offers several advantages, including estimation of fewer parameters and the clinical ease of a single, time-invariant decision to implement. We propose a new computational approach to estimation of shared-parameter G-estimation, which is efficient and shares the double robustness of the "unshared" sequential G-estimation. We use this approach to analyze data from the Scottish Early Rheumatoid Arthritis (SERA) Inception Cohort.
RESUMO
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.
Assuntos
Projetos de Pesquisa , HumanosRESUMO
BACKGROUND: A range of anti-modified protein antibodies (AMPAs) are associated with rheumatoid arthritis. We aimed to assess the relationship between AMPA profiles and radiographic progression in patients with new-onset rheumatoid arthritis. METHODS: In this cohort study, we obtained samples and data from the Scottish Early Rheumatoid Arthritis (SERA) inception cohort and biobank, which recruited patients with new-onset rheumatoid arthritis or undifferentiated arthritis who had at least one swollen joint from 20 hospitals across Scotland. AMPAs in plasma samples were measured by ELISAs at baseline. Paired radiographs of the hands and feet were taken at baseline and at 1 year and were scored with the Sharp-van der Heijde (SvH) method. We calculated differences in radiographic progression using estimated marginal mean changes between baseline and 1 year, with the baseline values of radiographic variables, rheumatoid factor, sex, age at recruitment, symptom duration, and Disease Activity Score 28 with C-reactive protein included as covariates. FINDINGS: Between March 1, 2011, and April, 30, 2015, 1073 patients were recruited to the SERA study. 362 patients with rheumatoid arthritis were included in our study and had their AMPA profiles determined. Patients were grouped into four main autoantibody profiles by reactivities to post-translational modifications: single positivity for anti-citrullinated peptide antibodies (ACPAs; 73 [20%]); double positivity for ACPAs and anti-acetylated peptide antibodies (AAPAs; 45 [12%]); triple positivity for ACPAs, AAPAs, and anti-carbamylated peptide antibodies (151 [42%]); and AMPA negativity (74 [20%]). 19 (5%) patients were in one of the minor autoantibody groups. Of the 233 patients with both antibody data and radiographs of sufficient quality, triple-positive patients had more radiographic progression between baseline and 12 months (estimated mean change in total SvH score 1·8, 95% CI 0·9-2·6, SE 0·4) than did single-positive patients (0·5, 0·1-1·0, 0·2; estimated mean difference in the total change in SvH score 1·2, 95% CI 0·1-2·4, SE 0·5). There was no difference in radiographic progression between single positive patients and AMPA negative patients (estimated mean change in total SvH score 0·7, 95% CI 0·1-1·4, SE 0·3; estimated mean difference in the total change in SvH score -0·2, 95% CI -1·1 to 0·7, SE 0·4). INTERPRETATION: This study suggests that the optimal prediction of future rates of radiographic progression in patients with rheumatoid arthritis will require an assessment of autoantibodies against multiple post-translationally modified proteins or peptides. FUNDING: The EU FP7 HEALTH programme, the Scottish Translational Medicine Research Collaboration, and the Chief Scientist Office Scotland.