Design, Methods, Biostatistics, and Economic Analysis Core

Columbia University Health Sciences, New York NY

Investigators

Linked publications & trials

Abstract

B. Background and Significance The rationale for inclusion of this Core in CEBP is the number of significant methodological issues in the body of self-management literature suggesting the need for application of additional designs and analytic techniques toward the goal of advancing the knowledge base on this important topic. In this section, we focus primarily on design and analytic issues including those related to economic analyses. We discuss other methodological issues specifically related to theoretical frameworks, cultural-appropriateness, and measurement instruments in the Self-Management, Biobehavioral, and Informatics Core. B1. Methodological Issues Related to Self-Management Research In recent years, a number of systematic reviews and meta-analyses focused on self-management/self-care interventions have been published. These syntheses highlight a number of methodological issues. Chodosh et al. screened 780 studies and included 53 in their random-effects meta-analysis (diabetes=26, osteoarthritis=14, hypertension=13). [1] Their analysis concluded that self-management interventions led to statistical and clinical significance for hemoglobinA1c and blood pressure, but only statistical significance for pain and function in osteoarthritis. Of primary relevance to the rationale for inclusion of the Design, Methods, Biostatistics, and Economic Analysis Core in CEBP, are their conclusions that studies had variable quality and that "the elements of programs most responsible for benefits cannot be determined from existing data, and this inhibits specification of optimally effective or cost-effective programs." (p.427) Warsi et al. critically analyzed the outcomes and methods for 71 trials of self-management education. [2] They identified methodological issues related to comparison groups including convenience samples of concurrent controls (15%) and randomization at the level of the clinic or physician without assessment of a possible center effect in one third of the studies. Only 8 studies were conducted using intent-to-treat analyses and only two of the 71 interventions were conducted by investigators independent of the self-management education program developer. Similar to Chodosh, the authors also identified the issue of inadequate description of the intervention. This issue has been widely acknowledged in medication adherence literature as well. [3, 4] In a 2005 publication, Bakken (CEBP PI) described an approach based upon standardized nursing terminology to address the issue. [5] Specifically, as part of the delivery of the Client Adherence Profiling and Intervention Tailoring Intervention (CAP-IT), [6] the nurse interventionist used the Home Health Care Classification to document the aspects of the tailored adherence intervention (nursing diagnoses and interventions) that were delivered to the patient and also documented the associated time. This strategy enabled calculation of an "intervention dose" and provided evidence that a tailored intervention was delivered. B2. Economic Evaluations in Self Management Research Demographic trends, the explosion in health care technology, and a changing health care system have focused attention on both the cost and the effectiveness of health care delivery. Although total U.S. health care costs are rising at a decreased rate than in the past, national health care spending is projected to reach 20% of the gross domestic product by 2016 ($4.1 trillion). [9] Because of this level of spending, it is necessary, but no longer sufficient, to demonstrate the efficacy of an intervention on clinical outcomes. Increasingly, health policy decision-makers request economic analyses before new interventions or models of health care delivery are accepted into practice. While some may suggest that cost-effectiveness analyses are only applicable for high technology high cost interventions, this is not the case. More recently, the need for economic evaluations related to patient self-management interventions has become apparent and some researchers have begun to address this topic. [10-12] A first step in economic analysis related to self-management is assessing health services utilization (e.g., [13, 14]). However, as noted in a systematic review of cost-effectiveness of interventions to support self-management, the overall quality of economic analyses in this field is poor. [15] These investigators reviewed 39 published economic evaluations analyzing self-management studies using a quality checklist adapted from well-developed guidelines for economic methods. They identified a set of issues that make economic analyses of self-management interventions more challenging than those of more conventional interventions: ill-defined comparator (e.g., "usual" care), placebo and Hawthorne effects, higher risk of control group contamination, and difficulty of ascertaining a societal perspective on costs. The investigators concluded that although the majority of studies claimed that interventions were cost-effective or cost-saving, rigorous economic methods were lacking, which limits the ability of these analyses to affect policy. In a review focused on psycho-educational interventions for children with chronic disease, Barlow and Ellard specifically identified the lack of evidence regarding cost-effectiveness as a knowledge gap in the body of literature. [8]