Dataset Information

Estimating causal log-odds ratio using the case-control sample and its application in the pharmaco-epidemiology study.

ABSTRACT: One important goal in pharmaco-epidemiology studies is to understand the causal relationship between drug exposures and their clinical outcomes, including adverse drug events. In order to achieve this goal, however, we need to resolve several challenges. Most of pharmaco-epidemiology data are observational and confounding is largely present due to many co-medications. The pharmaco-epidemiology study data set is often sampled from large medical record databases using a matched case-control design, and it may not be representative of the original patient population in the medical record databases. Data analysis method needs to handle a large sample size that cannot be handled using existing statistical analysis packages. In this paper, we tackle these challenges both methodologically and computationally. We propose a conditional causal log-odds ratio (OR) definition to characterize causal effects of drug exposures on a binary adverse drug event adjusting for individual level confounders. Using a case-control design, we present a propensity score estimation using only case samples and we provide sufficient conditions for the consistency of the estimation of the causal log-odds ratio using case-based propensity scores. Computationally, we implement a principle component analysis to reduce high-dimensional confounders. Extensive simulation studies are performed to demonstrate superior performance of our method to existing methods. Finally, we apply the proposed method to analyze drug-induced myopathy data sampled from a de-identified subset of medical record database (close to 5 million patient records), The Indiana Network for Patient Care. Our method identified 70 drug-induced myopathy (p?

SUBMITTER: Zhu A

PROVIDER: S-EPMC6416084 | biostudies-literature | 2019 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Estimating causal log-odds ratio using the case-control sample and its application in the pharmaco-epidemiology study.

Zhu Anqi A Zeng Donglin D Zhang Pengyue P Li Lang L

Statistical methods in medical research 20180122 7

One important goal in pharmaco-epidemiology studies is to understand the causal relationship between drug exposures and their clinical outcomes, including adverse drug events. In order to achieve this goal, however, we need to resolve several challenges. Most of pharmaco-epidemiology data are observational and confounding is largely present due to many co-medications. The pharmaco-epidemiology study data set is often sampled from large medical record databases using a matched case-control design ...[more]

PMID: 29355073

Similar Datasets

Project description:BACKGROUND:Globally, colorectal cancer (CRC) is the third and second leading cancer in men and women respectively with 600,000 deaths per year. Traditionally, clinicians have relied solely on nodal disease involvement, and measurements such as lymph node ratio (LNR; the ratio of metastatic/positive lymph nodes to total number of lymph nodes examined), when determining patient prognosis in CRC. The log odds of positive lymph nodes (LODDS) is a logistic transformation formula that uses pathologic lymph node data to stratify survival differences among patients within a single stage of disease. This formula allows clinicians to identify whether patients with clinically aggressive tumours fall into higher-risk groups regardless of nodal positivity and can potentially guide adjuvant treatment modalities. The aim of this study was to investigate whether LODDS in colon cancer provides better prognostication compared to LNR. METHODS:A retrospective study of patients on the prospectively maintained Cabrini Monash University Department of Surgery colorectal neoplasia database, incorporating data from hospitals in Melbourne Australia, identified patients entered between January 2010 and March 2016. Association of LODDS and LNR with clinical variables were analysed. Disease-free (DFS) and overall (OS) survival were investigated with Cox regression and Kaplan-Meier survival analyses. RESULTS:There were 862 treatment episodes identified in the database (402 male, 47%). The median patient age was 73 (range 22-100?years). There were 799 colonic cancers and 63 rectosigmoid cancers. The lymph node yield (LNY) was suboptimal (<?12) in 168 patients (19.5%) (p?=?0.05). The 5-year OS for the different LNR groups were 86, 91 and 61% (p?<?0.001) for LNR0 (655 episodes), LNR1 (128 episodes) and LNR2 (78 episodes), respectively. For LODDS, they were 85, 91 and 61% (p?<?0.001) in LODDS0 (569 episodes), LODDS1 (217 episodes) and LODDS2 (75 episodes) groups (p?<?0.001). Overall survival rates were comparable between the LNR and LODDS group and for LNY?<?12 and stage III patients when each were sub-grouped by LODDS and LNR. CONCLUSION:This study has shown for that the prognostic impact of LODDS is comparable to LNR for colon cancer patients. Accordingly, LNR is recommended for prognostication given its ease of calculation.

Dataset Information

Estimating causal log-odds ratio using the case-control sample and its application in the pharmaco-epidemiology study.

Publications

Estimating causal log-odds ratio using the case-control sample and its application in the pharmaco-epidemiology study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets