Skip Navigation
 

To organize objectives within topic areas, Dr. Fielding proposed that each topic area have a logic model that would serve as a "structural scaffolding" for discussion about how objectives relate. Dr. Remington said this could become complicated, but ideally it would provide a simple way to think about how determinants relate to outcomes. Although the exact template of the logic models had not been specified, they should display the logic for why certain objectives have been grouped together.

  • Dr. Fielding asked for a vote on the recommendation that logic models be created for each topic area. All members who were present voted in favor of the recommendation; none were opposed.

IV.  Healthy People 2020 Target Setting Recommendations

Dr. Remington, who also chaired the Subcommittee on Target-Setting, gave an overview of that group's efforts. He explained that the group had been charged with answering four questions: 1) what data should be used in target-setting; 2) what processes should be used; 3) should targets be aspirational or realistic; and 4) should they incorporate knowledge of interventions. While the Subcommittee had addressed most of these questions in its report, some unresolved issues remained for discussion. Remaining questions included whether there should be disparities targets; what tools should be provided to states, regions, and localities to help with target-setting, and how target-setting should relate to prioritization of objectives.

He said the quality of data available for target-setting will not be uniform across all objectives. There might be two categories of objectives: those with high-quality data, and those with lower quality data for tracking purposes. Targets should be rooted in past experience and should incorporate knowledge of effective interventions. Ideally, they should be set using evidence to project what the effects of programs and policies would be, and then modeling a science-based objective. Yet targets can be set with less information by seeking a percentage improvement in the level of the objective (e.g., 10%, 20%).

The process of target-setting should vary depending on the objective. The "Better than the Best" approach is of limited value; often those targets have not been meaningful for the entire population (e.g., in situations where there are high rates of a condition). The Subcommittee proposed setting a single population target for each objective at a level that represents an improvement for most of the population, but might not be better than that of the subpopulation with the "best" health status. This method is not based on the status of a reference population. Disparity reduction would be achieved by having all groups reach the target.

Dr. Remington said the target-setting approach recommended by the Subcommittee would yield three types of objectives: a subset that have science-based or S.M.A.R.T. (Specific, Measurable, Achievable, Realistic and Time-bound) targets; a sub-set that are monitored but do not have targets (because not enough data is available for the targets to be meaningful); and a sub-set of developmental objectives that currently lack a data source. Regarding "aspirational" versus "realistic" targets, the Subcommittee preferred using the term "reach" to describe targets that are achievable, but require more effort than perpetuation of the status quo

There was discussion as to whether separate "disparities" or "equity" targets should be set in addition to the general population targets. In the past, the "Better than Best" approach was criticized for creating a single target that was not realistic for subpopulations with the greatest health disparities. A disparities target could be used to measure the gap in health status among subpopulations, which would highlight the issue. States could then be charged with setting disparities targets. Yet setting two targets (one for the population mean, and a second for variance) could lead to confusion. A variance target would be both difficult to measure and difficult to communicate about. Dr. Remington requested the Committee's feedback.

Dr. Fielding said disparities targets could be viewed two different ways. First, they could be seen as targets for groups that are not at the level of the average, mean, or reference group. Second, a disparities target could be seen as measuring dispersion. Dr. Remington expressed the view that, if the population target is based on the mean, the disparities target should focus on reducing disparities for populations at higher risk. If this approach is used, there will be population subgroups that have already reached the target at the outset of the decade. Thus, one would not reduce variation, but one would seek to reduce disparities by focusing efforts on the populations that are at greatest risk.

When asked how a separate disparities target would be expressed, Dr. Remington said one could look at the standard deviation of subpopulations. Groups that are more than one or two standard deviations away from the mean would be the focus of disparity-reduction efforts, so that ultimately there would be less variation across the subpopulation means. Ideally, there should be very little variation across subpopulation means for a particular objective (indicating low health disparities). Nonetheless, in some areas there would be tremendous variation. Dr. Fielding asked the group to address the questions of: 1) what they think of disparities targets, and 2) how such targets might be expressed.