Health care graphs, stethoscope and professional using laptop to review records

In this current day and age of electronic health records (EHRs) it is hard to remember a time when all health records were maintained on paper. Clinical data was collected as a part of a time limited study or a sample, and full population data was almost unheard of. Some of the earliest population-based quality measures were built with administrative claims data. These types of process measures were useful for screening or medication adherence measures based on billable information, however there was limited clinical data for understanding health outcomes. Data collection was and continues to be very labor intensive.

Transition to EHRs Brings New Opportunity to Reduce Burden

A little over a decade ago, MNCM was in its beginning stages of developing new measures and its new direct data submission (DDS) process. At that time, about 40% of Minnesota clinics were still on paper records. It was not uncommon to turn to the progress note for a piece of information that could not be obtained any other way. Fast forward ten years and about 99% of Minnesota clinics are now on EHRs. Clinics have gained proficiency in extraction and query capabilities, and along with that have come more requests for timely clinical data. There has been an increasing emphasis on clinical data and outcome measures that have the most potential to improve population health. It is no longer practical to manually abstract even a few pieces of information from the EHR as a part of ongoing quality measurement.

Guiding Principles for a New Approach

For this reason, we need guiding principles to reduce burden for medical groups, support automation of quality measurement, and enable more timely feedback to inform quality improvement efforts. These principles will be used in implementing MNCM’s new PIPE system of data collection, and will also inform MNCM’s future measure development decisions.

For existing measures, elements will be considered for retention when identified by a standardized set of codes, i.e. codable. Currently, most elements that are not codable are contained in exclusions or exceptions. A few of the exceptions, after guideline or medication evolution, no longer made sense and needed a little bit of tuning up.

Future measures developed and stewarded by MNCM must include:

  • Reliable (and consistently used) codes for all components of the measure construct.
  • A consideration for a more inclusive measure with no exclusions.
  • Exclusions and exceptions need to be present in a significant percentage of the population (i.e., 5%).
  • If an exclusion or exception is rare (and codable), there needs to be a strong justification for including it.
  • PRO-based measures should rely on a validated PRO tool with strong psychometrics, validated cut-points and scoring for desired outcomes, and the tool should be free for clinical use & multiple modes of administration.

With these principles in mind, MNCM has refined the specifications for several existing measures.

Learn More

by Collette Cole, Clinical Measure Developer, MN Community Measurement