3 Things to Know About Digital Quality Measures

July 16, 2021 · Andy Reynolds

A two-hour crash course about digital measures at the Digital Quality Summit (DQS) Preconference on July 12 gave attendees a practical grounding in a topic that is vital to quality’s future.

To understand digital measures is to understand where quality is headed: CMS wants all quality measurement to be digital by 2025.

Three key questions the pre-conference event answered about digital measures are below.

1) What is a dQM?

dQM stands for Digital Quality Measure.

The “digital” in the name means the measure is written as computer code—in what’s called a “machine-readable format.” That’s so computers can understand and use dQMs.

Digital Quality Measures:

1. Are built using a common standard for sharing health care information electronically. The standard NCQA uses is Fast Healthcare Interoperability Resources, or FHIR.
2. Use a common data model* (again, FHIR).
3. Use machine-readable measure logic* (e.g., Clinical Quality Language, or CQL).
4. Include clinical terms, codes and other information needed to calculate reliable, comparable measure results.

(*Unsure of the difference between a “data model” and “measure logic”? We define those terms at the bottom of this post.)

2) So what? Why should I care about dQMs? 

dQMs differ from traditional paper-based measures that require manual coding and other arduous preparations before they can be used to get data on clinical performance.

dQMs’ advantages include:

• Standardized Data Definitions
  • dQMs utilize data collected in the normal course of care and perform calculations that previously required additional processes.
  • dQMs’ use of standardized data can improve accuracy and allow more rigorous data validation to occur at different levels of data collection.
  • Standardization also promotes system-wide adoption. The more “plug-and-play” dQMs are, the faster and easier the industry can meet CMS’ 2025 deadline for digital measurement.
• Consistent measure calculation
  • Digital measures reduce programming burden. That’s because they eliminate the manual—and potentially error-prone—step of a person having to read HEDIS Volume 2 and rekey measure specs into their information system. By removing the middleman, dQMs limit possible interpretation errors.
  • dQMS’ consistent way of calculating measures then ensures consistent results.
• Timely, actionable results
  • By using the rich clinical data that reside in disparate digital sources, dQMs measure more of what matters and provide patient-specific insights at the point of care.
  • Traditional measure calculations pertain to the “average” patient and typically summarize care delivered the previous calendar year. dQMs’ ability to provide the right data to the right person at the right time is a huge advance.

3) What’s the difference between dQMs and eCQMs? 

It comes down to data sources.  

dQMs use data from an array of electronic sources—EHRs, claims data, registries, case management systems, health information exchanges, wearable devices, patient experience surveys, and more.

eCQMs, or Electronic Clinical Quality Measures:

• Are a subset of dQMs
• Use data from EHRs.
• Are used mainly in CMS clinician reporting programs.

Here are unofficial mnemonics to differentiate dQMs from eCQMs:

• The “d” in dQM stands for diverse data sources.
• The “e” in eCQM means data come from EHRs.

Bonus: What’s the difference between a “data model” and “measure logic”? And what are examples of each?

• A data model defines how to represent things—such as diagnoses, encounters, medications and screening tests—in a standardized way.
  NCQA uses the Quality Data Model (QDM). That’s also the model Medicare uses for eCQMs in the Merit-Based Incentive Program (MIPS).
• Measure logic ties measure data elements together to produce scores. Put another way, measure logic is the instructions of what to look for and use in the data model.
  The measure logic NCQA uses is Clinical Quality Language (CQL). It’s everywhere in health care—common not only in digital quality measures but also in clinical decision support.

We’ll end with our favorite “explain it like I’m 5” explanation of data models and measure logic:

• A data model (FHIR) is a set of Legos—the colors, sizes and other attributes of each block.
• Measure logic (CQL) says how to assemble the blocks into this: