Phase 1: Conceptualization
Information Gathering
Measure development begins with information gathering – the measure developer conducts an environmental scan, develops a business case and requests input from a broad group of stakeholders including patients. The developer then narrows down the concepts to specific measures.
Phase 2: Specification
Draft Specifications
After the information gathering phase, we begin to draft the measure specifications. Measure specifications provide the comprehensive details that allow the measure to be collected and implemented consistently, reliably, and effectively. The specifications identify the population, the recommended practice, the expected outcome and determine how it will be measured. They also may include age ranges, performance time period and allowable values for medical conditions or procedures, code systems, descriptions.
Measure technical specifications will address the following questions: How will the measure be named? Does the name mean anything to people when they read it? Do they understand what that measure is about? What would the setting of the measure be (e.g., ambulatory office)? How will the data be collected? These questions have to be answered before testing begins.
Harmonization is all about reducing burden. Look at measures currently in practice and determine if there are places where our measure could be harmonized with the existing measure(s).
Phase 3: Testing
Measure Testing
Measure testing assesses the suitability of the quality measure’s technical specifications and acquires empirical evidence to help assess the strengths and weaknesses of a measure. Measure testing involves testing the components of the quality measure such as the data elements, the scales (and items in the scales if applicable), and the performance score.
There are two parts to measure testing: alpha and the beta testing.
Alpha testing helps identify early issues and often begins as early as the conceptualization step and is repeated during the development of the measure specifications.
Beta testing, which is also referred to as field testing, generally occurs after the initial specifications have been developed, and strives for representative sample sizes – multiple sites/settings. The primary purpose for beta testing is to understand the usability of the measure and to test the scientific acceptability of the measure.
After the testing ends, the results are analyzed with a return to the specification phase, or even the conceptualization phase, to rework the measure before testing again.
The PRIME registry is ABFM’s Qualified Clinical Data Registry (QCDR) and serves as our measure testing bed.
Phase 4: Implementation
Measure Implementation
The measure is then submitted for NQF endorsement (not a requirement for use by CMS) and for use in the 18 CMS quality reporting and payment programs.
What’s the difference between submitting to CMS versus submitting to NQF?
The National Quality Forum (NQF) submission is about the endorsement process where a consensus-based entity reviews the measure using five evaluation criteria to assess the measure on its own merit and independent of a CMS program. It essentially gives it that stamp of approval, and so endorsement/NQF submission is separate from CMS implementation. NQF endorsement is valued for measures in CMS programs, but it is not a requirement.
The CMS implementation process takes the measure from being in development to being actively used in 18 of the CMS quality payment programs (QPP).
Phase 5: Use, Continuing Evaluation, and Maintenance
Re-evaluation
This step ensures that the measure continues to add value to quality reporting measurement programs and that its construction continues to be sound. The regular reevaluation of measures is vital as the science and other factors are always changing (e.g., development of new clinical guidelines, new technologies for data collection, discovering a better way to calculate measure results). Continually reviewing the measure will ensure it remains relevant and meaningful. Measures that stop being useful are retired.
Interaction Among Measure Lifecycle Phases
The Measure development lifecycle is not a linear process. Once the measure is conceptualized, it can move throughout the various phases in the measure development lifecycle.
The following decision criteria throughout the measure development cycle ensures a measure meets the applicable standards before moving to the next phase:
Importance to measure and report—including analysis of opportunities for improvement such as reducing variability in comparison groups or disparities in healthcare related to race, ethnicity, age, or other classifications.
Scientific acceptability—including analysis of reliability, validity, and exclusion appropriateness.
Feasibility—including evaluation of reported costs or perceived burden, frequency of missing data, and description of data availability.
Usability—including planned analyses to demonstrate that the measure is meaningful and useful to the target audience. This may be accomplished by the Technical Expert Panel (TEP) reviewing the measure results such as means and detectable differences, dispersion of comparison groups, etc. More formal testing, if requested by CMS, may require assessment via structured surveys or focus groups to evaluate the usability of the measure (e.g., clinical impact of detectable differences, evaluation of the variability among groups).