The evaluation of the Henry J. Kaiser family foundation's community health promotion grant program: Design

Selected Statistical Issues in Group Randomized Trials

“…Should the groups be individually matched into pairs based on well-chosen, group-level variables, with randomization performed by pair? Some large-scale, community-based program evaluations have been conducted using a matched design [COMMIT (23), Working Well Trials (47)], whereas others have elected not to use matched design [Kaiser Study (49), Stanford Five-City Project (15a)]. The investigation of the impact of matching communities into more homogenous pairs to improve statistical power has been an active area of research in the design of GRTs.…”

“…Note that the definition of "group" is broad, allowing the target population to differ in nature and size, usually depending on how the intervention can be delivered most practically. Examples of GRTs include the National Cancer Institute's Working Well Trial (group = work site) (47), their 5-A-Day Program (group = work site) (27), the Hutchinson Smoking Prevention Project (group = school district) (40), the National Cancer Institute's COMMIT project (group = community) (23), and the Kaiser Family Foundation's Community Health Promotion Grants Program (group = community) (49).…”

Implications of the Results of Community Intervention Trials

“…This program was designed to foster community health promotion efforts targeting cardiovascular disease, cancer, substance abuse, adolescent pregnancy, and injuries (198,220). Comparisons between 11 intervention communities and 11 control communities, however, indicated little evidence of positive changes in the outcomes targeted by the intervention communities (220).…”

Selected Methodological Issues In Evaluating Community-Based Health Promotion And Disease Prevention Programs

“…For example, Figure 1 shows the "small theory" of treatment that guides the evaluation of the Henry J. Kaiser Family Foundation's Community Health Promotion Grants Program (89). The conceptual basis for the model (27,39,40), based in social leaming theory (a "large theory") (2), emphasizes mod ifying community norms and inducing changes in the physical, regulatory, and socioeconomic environments to make them more supportive of healthful behaviors and behavior change.…”

“…These programs are aimed at entire populations , which are usually geographi cally defined , and they attempt to change health behavior and disease risk through mass media campaigns, activation of existing community organiza tions, or changes in the physical or sociocultural environment. Several large programs of this kind have been mounted for cardiovascular disease preven tion (30,33,44,56,71), as reviewed by Shea & Basch (79,80), and the approach is increasingly being applied to other disease areas and populations (3, 34,67,89,92). As investment in community-based programs has grown, so has the importance of evaluating their effectiveness, as evidenced in part by the recent publications of Green & Lewis (38) and Bracht (6).…”