Using Concept Mapping to Form Title V Priorities in Wyoming

BACKGROUND:  Every five years the Title V Maternal and Child Health (MCH) program completes a statewide needs assessment and determines its priorities for the subsequent five year cycle.  This process includes gathering input from stakeholders and communities, and compiling data on existing MCH indicators.  It is often difficult to synthesize the large quantity of information gathered into logical groupings for presentation to decision makers to inform priority selection and strategic planning.  Concept mapping is a process of defined steps that yield a conceptual representation of thoughts in an objective form for planning and evaluation.

METHODS:  To determine potential priorities for the Wyoming MCH program, epidemiologists compiled a list of MCH indicators from the following sources: life course indicators, preconception health indicators, Title V proposed National Performance Measures, previous needs assessment data, and emerging issues from community meetings.  The indicators were divided into three population areas Women and Infants (women 15-44 and infants 0-1), Child (1-11), and Adolescent Health (12-24). A modified version of concept mapping was used to determine how best to group the indicators.  MCH staff grouped similar indicators into ‘concepts’ and then named each ‘concept’.  Each indicator was scored using an internal ranking tool. A summed binary square symmetric similarity matrix was created from the grouped indicators.  Two dimensional scaling used the similarity matrix (SAS version 9.2) to assign coordinates to each indicator and map the proximity and conceptual similarity of indicators to one another. The coordinates for each indicator were used in a hierarchical cluster (SAS) analysis for each population.  The cubic cluster criterion (CCC), R square, pseudo-F, and T-squared statistics were used to determine the appropriate number of clusters for each population.

RESULTS:  Scaling maps confirmed indicators were reasonably grouped and illustrated the results to the program staff.  Six clusters were identified for the women and infant group, six for the child group, and seven for the adolescent group.  Each cluster had a ranked score based on indicators, and indicators were ordered within the clusters.  The clusters have become potential priorities which will be further researched and evaluated  in the needs assessment process.

CONCLUSIONS:  Concept mapping framework provides an objective set of statistical tools for grouping indicators based on stakeholder input.  Use of concept mapping allowed for visualization of the process which increased stakeholder buy in and understanding of the methods.