Evaluation.

(a) Using the outcome measures established in subdivision 3, the commissioner shall conduct a biennial evaluation of the statewide health improvement program new text begin grants new text end funded under this section. new text begin The evaluation must use the most appropriate experimental or quasi-experimental design suitable for the grant activity or project. new text end Grant recipients shall cooperate with the commissioner in the evaluation and provide the commissioner with the information necessary to conduct the evaluation, including information on any impact on the health indicators listed in section 62U.10, subdivision 6 , within the geographic area or among the population targeted.

(b) Grant recipients will collect, monitor, and submit to the Department of Health baseline and annual data and provide information to improve the quality and impact of community health improvement strategies.

(c) For the purposes of carrying out the grant program under this section, including for administrative purposes, the commissioner shall award contracts to appropriate entities to assist in designing and implementing evaluation systems. new text begin The commissioner shall consult with the commissioner of management and budget to ensure that the evaluation process is using experimental or quasi-experimental design. new text end

(d) Contracts awarded under paragraph (c) may be used to:

(1) develop grantee monitoring and reporting systems to track grantee progress, including aggregated and disaggregated data;

(2) manage, analyze, and report program evaluation data results; and

(3) utilize innovative support tools to analyze and predict the impact of prevention strategies on health outcomes and state health care costs over time.

new text begin (e) For purposes of this subdivision, "experimental design" means a method of evaluating the impact of a strategy that uses random assignment to establish statistically similar groups, so that any difference in outcomes found at the end of the evaluation can be attributed to the strategy being evaluated; and "quasi-experimental design" means a method of evaluating the impact of a strategy that uses an approach other than random assignment to establish statistically similar groups, so that any difference in outcomes found at the end of the evaluation can be attributed to the strategy being evaluated. new text end