Propensity score methods in drug safety studies: practice, strengths and limitations

Wang, Jixian; Donnan, Peter T.

doi:10.1002/pds.656

Cited by 40 publications

(35 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A propensity score estimating the probability of a patient receiving tamoxifen was calculated using a logistic regression model for age, social class, pathological tumour characteristics, Charlson's index and co-prescribing (Wang and Donnan, 2001). This was added to the model to adjust for propensity to receive tamoxifen and so reduce bias in the survival analysis.…”

Section: Discussionmentioning

confidence: 99%

“…The use of the propensity score within the final model helps reduce potential selection bias (Wang and Donnan, 2001) because of the use of observational data, where clinicians may choose patients with a lower risk of mortality to receive tamoxifen. However, our analysis holds after allowing for the predisposition of a patient to receive tamoxifen.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Cohort study examining tamoxifen adherence and its relationship to mortality in women with breast cancer

et al. 2008

View full text Add to dashboard Cite

Increasing duration of tamoxifen therapy improves survival in women with breast cancer but the impact of adherence to tamoxifen on mortality is unclear. This study investigated whether women prescribed tamoxifen after surgery for breast cancer adhered to their prescription and whether adherence influenced survival. A retrospective cohort study of all women with incident breast cancer in the Tayside region of Scotland between 1993 and 2002 was linked to encashed prescription records to calculate adherence to tamoxifen. Survival analysis was used to determine the effect of adherence on all-cause mortality. In all 2080 patients formed the study cohort with 1633 (79%) prescribed tamoxifen. The median duration of use was 2.42 years (IQR ¼ 1.04 -4.89 years). Longer duration was associated with better survival but this varied over time. The hazard ratio for mortality in relation to duration at 2.4 years was 0.85, 95% CI ¼ 0.83 -0.87. Median adherence to tamoxifen was 93% (interquartile range ¼ 84 -100%). Adherence o80% was associated with poorer survival, hazard ratio 1.10, 95% CI ¼ 1.001 -1.21. Persistence with tamoxifen was modest with only 49% continuing therapy for 5 years of those followed up for 5 years or more. Increased duration of tamoxifen reduces the risk of death, although one in two women do not complete the recommended 5-year course of treatment. A significant proportion of women have low adherence to tamoxifen and are at increased risk of death.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Cohort study examining tamoxifen adherence and its relationship to mortality in women with breast cancer

et al. 2008

View full text Add to dashboard Cite

show abstract

“…The ORs did not differ significantly across the two survey samples (P \ .81), verifying our decision to combine the two surveys. To control for possible residual confounding from demographic and health factors and to incorporate a number of covariates in our model without consuming excessive degrees of freedom, we used member characteristics to generate propensity scores (Braitman and Rosenbaum 2002;Wang and Donnan 2001). Using logistic regression, we generated a propensity score for each member, modeling antidepressant treatment (adequate versus inadequate) from the following covariates: age, gender, race, marital status, education level, receipt of SSI/SSDI, self-reported disabling condition and health status (C-statistic = 0.64).…”

Section: Methodsmentioning

confidence: 99%

Antidepressant Adequacy and Work Status Among Medicaid Enrollees with Disabilities: A Restriction-based, Propensity Score-adjusted Analysis

Smith

Henry

Zhang

et al. 2009

Community Ment Health J

View full text Add to dashboard Cite

This cross-sectional study of adult survey respondents with disability and depression (n = 199) enrolled in Massachusetts' Medicaid program examined the association of adequately or inadequately prescribed antidepressant treatment and self-reported work status using conditional logistic regression, controlling for age, gender, race, marital status, education, receipt of SSI/SSDI, self-reported disabling condition, and health status. Confounding by severity was addressed by two methods: restriction of our sample and subsequent stratification by propensity score. Individuals receiving adequate antidepressant treatment had an increased odds of working compared to individuals receiving inadequate treatment, both in analyses in which restriction was used to limit confounding (OR = 3.45, 95% CI = 1.15-10.32, P < .03), and in analyses which combined restriction with adjustment by propensity score stratification (OR = 3.04, 95% CI = 1.01-9.62, P < .05). Among this sample of Medicaid enrollees with disability and depression, those receiving adequate antidepressant treatment were significantly more likely to report working.

show abstract

“…Confounding adjustment by baseline covariates (L(0)) using standard modeling approaches or causal methods for point treatment problems (e.g., propensity score [14][15][16][17] ) can then provide unbiased estimates of the effect of the early therapy exposure (A(0)). Such results are especially useful in studies (e.g., randomized trials) in which changes in therapies are uncommon during the follow-up time, that is, an intention-to-treat (ITT) interpretation of the results is informative.…”

Section: Motivation For Marginal Structural Model Approaches In Compamentioning

confidence: 99%

Dynamic marginal structural modeling to evaluate the comparative effectiveness of more or less aggressive treatment intensification strategies in adults with type 2 diabetes

Neugebauer

Fireman

Roy

et al. 2012

Pharmacoepidemiology and Drug

View full text Add to dashboard Cite

Purpose Chronic disease care typically involves treatment decisions that are frequently adjusted to the patient's evolving clinical course (e.g., hemoglobin A1c monitoring and treatment intensification in diabetes patients). Thus, in comparative effectiveness and safety research (CER), it often is less clinically relevant to contrast the health effects of static treatment decisions than to compare the effectiveness of competing medical guidelines, that is, adaptive treatment strategies that map the patient's unfolding clinical course to subsequent treatment decisions. With longitudinal observational studies, treatment decisions at any point in time may be influenced by clinical factors that also are risk factors for the outcome of interest. Such time-dependent confounders cannot be properly handled with standard statistical approaches, because such confounders may be influenced by previous treatment decisions and may thus lie on causal pathways between the very outcomes and early treatment decisions whose effects are under study. Under explicit assumptions, we motivate the application of inverse probability weighting estimation to fit dynamic marginal structural models (MSMs) in observational studies to address pragmatic CER questions and properly adjust for time-dependent confounding and informative loss to follow-up. Methods We review the principles behind this modeling approach and describe its application in an observational study of type 2 diabetes patients to investigate the comparative effectiveness of four adaptive treatment intensification strategies for glucose control on subsequent development or progression of urinary albumin excretion. Results Results indicate a protective effect of more aggressive treatment intensification strategies in patients already on two or more oral agents or basal insulin. These conclusions are concordant with recent randomized trials. Conclusions Inverse probability weighting estimation to fit dynamic MSM is a viable and appealing alternative to inadequate standard modeling approaches in many CER problems where time-dependent confounding and informative loss to follow-up are expected. Copyright

show abstract

Propensity score methods in drug safety studies: practice, strengths and limitations

Cited by 40 publications

References 11 publications

Cohort study examining tamoxifen adherence and its relationship to mortality in women with breast cancer

Cohort study examining tamoxifen adherence and its relationship to mortality in women with breast cancer

Antidepressant Adequacy and Work Status Among Medicaid Enrollees with Disabilities: A Restriction-based, Propensity Score-adjusted Analysis

Dynamic marginal structural modeling to evaluate the comparative effectiveness of more or less aggressive treatment intensification strategies in adults with type 2 diabetes

Contact Info

Product

Resources

About