Completion Rate and Feasibility of Converting AAN Multiple Sclerosis Quality Measures to Digital Measures
Samantha Tidd1, Sarita Walvekar2, Saswat Sahoo3, Kelly Bowen4, Mengke Du5, Adam Webb6, Marisa McGinley5
1University of Rochester Medical Center, 2Yale Medical Center, 3Johns Hopkins, 4Cleveland Clinic Lerner College of Medicine, 5Cleveland Clinic, 6Emory University School of Medicine
Objective:
To determine the completion rate of the 2020 AAN multiple sclerosis (MS) quality measures (QMs) at a single MS Center and evaluate if they could be collected digitally with discrete data elements in the electronic health record.
Background:
There have been attempts to standardize care metrics in MS through QMs but their practical application have yet to be characterized.
Design/Methods:
In this retrospective cohort study, adult people with MS (PwMS) who had at least 2 visits 24 months ±6 months apart were included. Digital quality measures (dQMs) were created for the 2020 AAN MSQMs and completion rate calculated for the entire sample. In a 20% random sample, manually abstracted quality measures (mQMs) were also collected through chart review. The McNemar test was used to determine if there were differences in mQM and dQMs completion rates in this subgroup.
Results:
dQMs were calculated for 1,101 PwMS (average age 48 years, 73% female, 65% relapsing course) and a subgroup of 218 PwMS had mQMs collected. Several measures could not be converted to dQMs due to the absence of discrete data elements that could be abstracted from the EHR. The completion rates collected were significantly lower for dQMs compared to mQMs for MRI screening (82% vs 93%, p<0.001), symptom management (8.7% vs 24%, p<0.001), cognitive screening (77% vs 88%, p<0.001), and fatigue management (2.1% vs 32%, p=0.046) while completion rates were similar for fatigue screening (91% vs 89%, p=0.43) and cognitive management (0% vs 5.6%).
Conclusions:
The significant differences between manually extracted completion rates and those using digital proxies for multiple care metrics raise concern about the ability to convert many existing AAN QMs into dQMs. Future efforts are needed to develop QMs that leverage EHR data elements allowing for consistent and widespread use in clinical practice that accurately reflect the neurological care that is being delivered.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.