Causal inference in a clinical trial: a comparative example.

D F Heitjan

doi:10.1016/s0197-2456(99)00012-4

Home/Archive/PubMed

PubMed · 10440558

Causal inference in a clinical trial: a comparative example.

D F Heitjan

View original source ↗Export metadata ↓

Abstract

Recently there has been much interest in methods for analyzing clinical trials of treatments that are subject to noncompliance. In this paper I study a small, simple dataset from a clinical trial of immunosuppressive therapy in the treatment of multiple sclerosis. I apply and compare a range of methods: the as-randomized (intention-to-treat) analysis, the as-treated analysis, estimates based on a nonignorable selection model, and Rubin's causal model. The results differ substantially even in this small dataset that exhibits modest noncompliance. For this reason, data analysts should be clear about which parameters are of greatest importance in the analysis of a clinical trial.

Explore related subjects

Azathioprine Bias Dose-Response Relationship, Drug Drug Administration Schedule Drug Therapy, Combination Humans Immunosuppressive Agents Methylprednisolone Models, Statistical Multiple Sclerosis Patient Dropouts Treatment Refusal Clinical Trial Comparative Study Journal Article Randomized Controlled Trial

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

D F Heitjan. 1999. Causal inference in a clinical trial: a comparative example.. https://doi.org/10.1016/s0197-2456(99)00012-4

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2007

An autoregressive linear mixed effects model for the analysis of longitudinal data which show profiles approaching asymptotes.

In longitudinal data, a continuous response sometimes shows a profile approaching an asymptote. For such data, we propose a new class of models, autoregressive linear mixed effects models in which the current response is regressed on the previous response, fixed effects, and random effects. Asymptotes can shift depending on treatment groups, individuals, and so on, and can be modelled by fixed and random effects. We also propose error structures that are useful in practice. The estimation methods of linear mixed effects models can be used as long as there is no intermittent missing.

Azathioprine↗

PubMed2006

Long-term graft outcome with mycophenolate mofetil and azathioprine: A paired kidney analysis.

BACKGROUND: Randomized controlled trials and U.S. Registry data have demonstrated that mycophenolate mofetil (MMF) reduces acute rejection rates and improves graft survival. We undertook the first paired kidney analysis comparing the effects of MMF and azathioprine on graft outcome in the United Kingdom. METHODS: In all, 238 deceased donors from 1999 to 2002 who donated one kidney to a patient treated with MMF and the other kidney to a patient treated with azathioprine were identified from the national transplant database held by U.K. Transplant. Graft function and rates of change of graft function were compared using multiple linear regression analyses adjusting for covariates on an intention-to-treat basis. Incidence of acute rejection and delayed graft function were studied using logistic regression. Patient and graft survival censored for death were evaluated with Cox regression analyses. RESULTS: The MMF-treated patients exhibited a nonsignificant trend towards improved graft function but with increased rejection rates (44% versus 31%, P < 0.01). Treatment with MMF did not reduce delayed graft function rates. Univariate analysis showed that graft survival was inferior in MMF-treated patients (90% versus 95%, log-rank, P = 0.02) but in multivariate Cox regression models, MMF treatment was not a significant factor. Surprisingly, in the first year 32% of patients achieved daily doses of less than 2 g of MMF compared to 18% of patients who received less than 100 mg of azathioprine (P < 0.01). CONCLUSION: In this real-life study, there was no difference in patient or graft outcome between MMF and azathioprine treated groups despite increased rejection rates in patients receiving MMF therapy.

Azathioprine↗

PubMed2006

Long term results of use of azathioprine in patients with ulcerative colitis in India.

AIM: To evaluate the role of azathioprine (AZA) in Indian patients with ulcerative colitis over longer duration of time. METHODS: One hundred fifty six patients with ulcerative colitis who were treated with AZA from January 1995 to December 2003 were reviewed. The indications for its use were as follows: (1) steroid dependent and steroid refractory disease; (2) Azathioprine monotherapy for naive patients with severe disease; and (3) combination therapy (AZA + sulfasalazine or 5-aminosalicylates) for naive patients with severe disease. The data included patient and disease demographics, efficacy and toxicity profile of AZA. Patients with a minimum duration of 6 mo use of AZA were included in this report. RESULTS: Of a total of 156 patients treated with AZA, 45 were excluded from analysis for the following reasons- (follow up less than 6 mo, n = 9; poor follow up, n = 18; adverse affects, n = 18). In steroid refractory/dependent group the mean number of relapses prior to and post initiation of AZA therapy were 3.28 (+/- 0.81) and 0.94 (+/- 0.29) respectively. Discontinuation of steroids could be accomplished in 12 of the 15 steroid dependent patients. The proportion of patients with sustained remission of 1, 2, 3, 4 and 5 years duration were calculated. Eighteen patients experienced adverse effects necessitating withdrawal of AZA (pancreatitis, n = 7; hepatitis, n = 3; gastrointestinal intolerance, n = 2; alopecia, n = 2; and hematological, n = 4) while 13 patients needed dose reduction or temporary withdrawal of the drug. CONCLUSION: Azathioprine is well tolerated and has therapeutic benefits lasting as long as 4 years. Adverse effects such as pancreatitis, hepatitis, cytopenias and gastrointestinal symptoms do occur but are controlled by drug withdrawal only.

Azathioprine↗