How much can you remember about 1991? The first website was built and put online, and if you were online you would have accessed it with a 14.4 kbps modem and a telephone line. Some of the first laptop computers were being produced, and they cost about $3,500. Healthcare was only 12.8% of GDP compared with 18% in 2019. And the National Committee for Quality Assurance (NCQA) had released version 1 of the now ubiquitous Healthcare Effectiveness Data and Information Set (HEDIS). If you aren’t familiar with HEDIS, it’s a set of 90 healthcare outcome measures that are required under Medicare Advantage and used by most payers to measure performance in ways that can significantly impact their incentives. Hospital systems also use HEDIS measures to be able to report performance back to payers and improve their reimbursement.
The definition of HEDIS measures have been updated and expanded regularly over the past 28 years, but the mechanism for data collection has always remained a 20th century process: manual surveys conducted exclusively by certified companies, manual review of medical charts, and evaluation of insurance claims. It’s a time consuming and costly process to report on HEDIS measures. One regional hospital in TX reports having spent $1M to build out the necessary infrastructure and about $2M per year to maintain ongoing reporting. On the other hand, inaccurate reporting can also prevent healthcare organizations from being fully compensated for their outcomes.
Is it possible to achieve the same measurements while also reducing the administrative burden related to HEDIS and other reporting requirements?
Living in a Dream World?
As a data-person, my dream is a world where all of the data is consistently collected through low-burden electronic processes. Systems effectively track data provenance from its original source (human or digital) through system transfers, sharing across organizations, standardization, aggregation, and finally through to a final outcome that can be reliably used to measure performance. Everyone trusts the outcomes and no humans had to do any non-value-add work in order to produce that information. The humans are able to be in service to patients rather than to systems and data.
To my delight, HEDIS is finally making the transition to truly digital measures! Maybe this utopian vision of mine will come true some day! Beginning in 2019, NCQA is releasing six measures for its Electronic Clinical Data Systems (ECDS) program. The new digital measures program will allow data from EHR, HIE, claims, lab, pharmacy, and care management systems to be used directly in the reporting of outcomes, rather than only through traditional manual processes. Over the next few years, NCQA will make digital versions of all 90 measures available, but it also recognizes that adoption of the new digital measures will require a difficult transformation from the healthcare industry at large. Perhaps my utopian vision is still a long way off.
Making it Work in the Meantime
In the meantime, there are things payers and healthcare systems can do to reduce some of the burden of their manual reporting requirements. Robotic Process Automation (RPA) can streamline some manual data collection activities and fully automate most of the activities required to compile data into required formats and submit results online. Healthcare organizations that spend days each month compiling and submitting reports to registries, CMS, partners and state agencies can all benefit from automation of routine process with a minimal investment in technology. Combined with advanced analytic capabilities, these automated processes can even identify potential errors or negative impacts to reimbursement before data is submitted.
The world changes too fast for us to hope to attain any utopian future. Instead, it’s better to make the best of the challenges we face and the technology available to us today. While you’re waiting for a future with perfectly accessible, perfectly integrated, perfectly clean data to get here, take a look at other healthcare processes that can be improved by RPA and AI today and contact us to see what an RPA Proof of Value might look like for you.