Regional Consumer Guide Methodology
The consumer guides use domains to represent the quality of care for a group of people or condition. The domains often included several measures of care relevant to the group or condition. Plan scoring on a domain is determined using one of two methodologies depending on the number of measures in the domain; A) domains with multiple measures, B) domains with single measures.
For domains with multiple measures plan scoring is determined by using a standard score or z-score. The z-score is a calculation that describes the relationship of the plan's rate to the statewide average taking into account the standard deviation of the rate.For the maternal health measures a plan's rate is compared to either a New York City regional rate or a Rest of State regional rate based on the region that the plan covers. If a z-score exceeds a threshold, it is capped to prevent a single measure from impacting overall rates excessively. Then a domain's composite score is calculated by averaging the z-scores for all of the measures.
Star ratings for the domains are assigned based on whether the composite z-score is above or below the 95 percent confidence limit. Composite z-scores that are less than -1.96 are assigned one star, those greater than 1.96 are assigned three stars and those between -1.96 and 1.96 are assigned two stars.
The formula used to create the z-score for the guides is as follows:
For those domains in which there is only one measure a z-score is not calculated. Star ratings instead are based on if the measure is significantly above or below the 95 percent confidence limit for the statewide average for that measure.
If the rate is significantly below the statewide average one star is assigned, if the rate is significantly above the statewide average than three stars are assigned, and if the rate is not significantly different than statewide average two stars are assigned.
Once a star rating is calculated for each domain, an overall rating is calculated by taking the total number of stars received across all ranked domains and dividing that sum by the total number of possible stars. This rate is multiplied by 100 to create a percentage between 0 and 100.