Research

Tutorials

The Express Scripts Research Team employs advanced scientific techniques, like those described here, to gather evidence necessary to uncover new insights into prescription-drug use and the pharmacy benefit.

Good Research Practices for Comparative Effectiveness Research: Approaches to Mitigate Bias and Confounding in the Design of Nonrandomized Studies of Treatment Effects Using Secondary Data Sources: The ISPOR Good Research Practices for Retrospective Database Analysis Task Force Report—Part II. (PDF file)
Cox ER, Martin BC, Van Staa T, Garbe E, Siebert W, et al. Value in Health 2009, 12.

This report discusses the inherent biases when using secondary data sources for comparative effectiveness analysis and provides methodological recommendations to help mitigate these biases. The report provides recommendations and tools for researchers to mitigate threats to validity from bias and confounding in measurement of exposure and outcome. Recommendations on design of study included: the need for data analysis plan with causal diagrams; detailed attention to classification bias in definition of exposure and clinical outcome; careful and appropriate use of restriction; extreme care to identify and control for confounding factors, including time-dependent confounding. Design of nonrandomized studies of comparative effectiveness face several daunting issues, including measurement of exposure and outcome challenged by misclassification and confounding. Use of causal diagrams and restriction are two techniques that can improve the theoretical basis for analyzing treatment effects in study populations of more homogeneity, with reduced loss of generalizability.

A Checklist for Retrospective Database Studies — Report of the ISPOR Task Force on Retrospective Databases (PDF file)
Motheral BR, Brooks J, Clark MA, Crown WH, Davey P, Hutchins D, Martin BC, Stang P. Value In Health 2003;6(2):90-97.

Although health-related retrospective databases, claims databases in particular, continue to be a significant data source for outcomes research, they pose many practical challenges. A 27-question checklist was developed in regards to database research or database studies, to guide decision makers as they consider the database, the study methodology, and the study conclusions. A wide range of issues, from data linkages to data interpretation, are covered by the checklist questions.

Evaluating Medication Adherence: Which Measure is Right For Your Program? (PDF file)
Fairman KA, Motheral BR. Journal of Managed Care Pharmacy 2000; 6(6):499-506.

Measuring medication adherence is becoming increasingly important to inform both clinical programs and cost containment policy debate. None of the methods for measuring medication adherence constitutes a "gold standard." Choosing an appropriate measure for a program depends largely on the program's purpose as well as on available resources. This article reviews strengths and limitations of various measures for different types of programs. For those situations in which using prescription claims data is appropriate, the article reviews techniques for dataset preparation and calculation of various claims-based measures.

Comment on "Variations in medication compliance related to individual drug, drug class, and prescribing physician" (PDF file)
Fairman KA. Journal of Managed Care Pharmacy 2000;6:100-101.

The publication of a recent article about medication compliance raises concerns. The study's lack of eligibility information and failure to develop an adequate provider database are serious methodological flaws. The study's primary outcome measure of compliance showed little variation, most likely because patients terminating therapy prematurely were excluded from the sample, limiting the study population to a small subgroup of medication users. Outcomes researchers must dig a little deeper - identify appropriate databases, redesign a study when the initial design proves inappropriate, and use multivariate techniques when descriptive analyses lead to substantial unanswered questions - in order to give the outcomes research field the best possible chance to inform health care policy.

Going to the Source: A Guide to Using Surveys in Health Care Research (PDF file)
Fairman KA. Journal of Managed Care Pharmacy 1999;5(2):150-159.

The objective of this study was to review techniques for producing useful and objective surveys about health care consumers' experiences and opinions. The following principles should guide project planning and questionnaire construction:1) enlist the respondent's interest and trust; 2) maintain trust by keeping respondent burden to a minimum; 3) provide an attractive product; 4) avoid confusing, threatening, or biased questions; and 5) ensure that the questionnaire is consistent with planned data analyses. Several additional procedures enhance the usefulness of survey results: pretesting, follow-ups with initial non-respondents to increase response rate, and performing tests for reliability and validity.

When survey results are to be used for important organizational purposes, or when a questionnaire deals with potentially sensitive topics, employing a well-trained professional survey researcher is wise. Depending on project scope and complexity, survey researchers may be used either as reviewers, consultants or project administrators.

Research Methodology: Hypotheses, Measurement, Reliability, and Validity (PDF file)
Motheral BR. Journal of Managed Care Pharmacy 1998;4(4):382-390.

Using real-world examples, this article examines hypothesis development, measurement, reliability, and validity. This article provides guidance for critically evaluating the published literature. In the changing health care industry, knowledge of research methodology will benefit pharmacists practicing in a managed care environment.

"The doctor told me to cut the pills in half": Practical considerations in using claims databases for outcomes research.
Fairman KA. Drug Benefit Trends 1997; 9(10): 30-35,39.

Outcomes Management: The Why, What, and How of Data Collection (PDF file)
Motheral BR. Journal of Managed Care Pharmacy 1997;3(3):345-351.

The objective of the study was to describe process of outcomes management (OM) and discuss barriers to its success. Quality endeavors, such as outcomes management are receiving increased attention in the U.S. health care system. OM applies continuous quality improvement (CQI) techniques to medical care, assessing both the effectiveness of products and services and the quality with which they are provided. Linking the structure and process of care to the outcomes of care is critical to OM's success. Sources of data for OM include claims databases, medical records, and patient surveys, each with inherent advantages and disadvantages that must be considered. A number of barriers to true outcomes management must be overcome before its full potential will be realized.

The Use of Claims Databases for Outcomes Research: Rationale, Challenges, and Strategies
Motheral BR, Fairman KA. Clinical Therapeutics 1997; 19(2):346-366

Health care payers and policy makers need information about the cost and effectiveness of medical treatments. While randomized controlled trials historically are the primary source of medical information, they are expensive and labor-intensive, and often have limited utility for answering questions about “real-world” patient populations. These problems have led to an increasing reliance on claims database research in making policy decisions about treatment options. However, both researchers and decision makers should recognize the limitations and unique features of claims databases. Recommendations for avoiding or minimizing threats to internal validity, construct validity, and external validity are: (1) use of a study design that includes comparisons; (2) ensuring that the study design and conclusions are consistent with the database; (3) a priori conceptual modeling of the research question; (4) use of appropriate constructs; (5) explicit examination of alternative explanations for study findings; (6) sensitivity analyses of key assumptions; (7) awareness of the distinction between statistical and practical significance of findings; (8) generalization only when appropriate; and (9) reporting of relevant information. Given that any study design or data source has limitations, we hope that this paper will encourage a philosophy of methodological pluralism in outcomes research. Awareness and accurate reporting of validity issues will strengthen and extend the information resources currently available to decision makers.

PDF files require Adobe Acrobat Reader.