A new day for the Yale Way

Yale Medicine in New Haven, Conn., grew a lot in the past decade. From 2008 to 2018, the largest academic multispecialty practice in New England:

• Broadened clinical departments from 16 to 21;
• Expanded from 891 physicians to 1,416 (a 31% increase); and
• Increased the number of nonphysician providers (NPPs) fivefold, from 149 to 754.

Not surprisingly, the volume of patient encounters saw a big jump in that same time frame: from just under 916,000 to 3.1 million. That created challenges with billing and collecting for all those encounters.

For Yale Medicine’s Joyce Dupee, MHA, CPC, CPC-P, director, coding and billing, and Sally Thibodeau, BA, director, business office operations, this meant a yearlong effort to break through what Dupee called “a very strong, siloed” set of clinical departments with their own set of processes to transform the organization into a centralized revenue cycle operation.

In 2001, Yale Medical Group (later rebranded Yale Medicine) was “still very decentralized,” Dupee said. The various clinical departments “controlled their own fee schedules, scheduling, credentialing, provider schedules, [insurance] follow-up,” until a big change came in 2011 in the form of an 18-month implementation of a new EHR.

That “big bang” event, as Dupee dubbed it, set off an evaluation of the entire system’s practice management applications. The shift back to centralization had begun.

In 2014 following the new EHR launch, pre-authorization teams were centralized. Two years later, Dupee’s part of the transformation began: a centralized billing and coding team.

Bringing it all together

Long before the centralizing effort began, there was a problem with patient registration. More than 300 people had access to the system before it was automated and eligibility checks were still done by calling a payer, Thibodeau said.

“It took time; nobody did it the same way … we had huge lines at the front desk as people were waiting to register,” Thibodeau said. “And for new patients, the system was: you would come an hour early for an appointment, you would register as a new patient and then move onto your appointment. [Insurance] cards were sometimes accepted; sometimes not everything was paper.”

Those problems resulted in a 47% eligibility rejection rate in 1998, forcing additional resources to be reworked.

In centralizing multiple departments’ patient registration:

• New patient demographic and insurance information was verified
• Existing patient demographics were verified every 180 days and at time of check-in
• Insurance eligibility was checked every 90 days for commercial payers and Medicare and every 30 days for Medicaid and self-pay through a series of four mechanisms:
  • Real-time eligibility (RTE) checks
  • Batched RTE checks
  • Payer websites
  • Direct contact with payers

Between the new EHR and an eligibility verification product, more of these tasks were automated and done “behind the scenes,” Thibodeau said. “The more you have to type something in manually, the greater chance that there’s going to be some kind of error.”

Other self-service options have allowed Yale Medicine to establish pre-time of service (TOS) registration in addition to TOS and post-TOS registrations. Paired with eligibility improvements, Yale Medicine in FY2018 had a registration-related rejection rate of 13% — a far cry from where it was three decades earlier.

The addition of real-time appointment filling then took the patient registration and eligibility improvements to eliminate dead space in providers’ schedules. Same-day and day-after work queues for dedicated registrar calls allowed for primary care clinics to prioritize patient calls for appointments the next morning.

The patient registration centralization was rounded out by charge review work queues that identify patient accounts not verified or updated at the time a charge is entered into the EHR, ensuring all information is correct prior to claim submission.

Authorizations and referrals

Prior to the new EHR, Yale Medicine found that almost 250 staff worked on obtaining authorizations in some capacity, and hardly any of the departments had dedicated staff for authorizations, nor was it a priority for front-end units of the departments, transferring the work to the back end with providers. This also caused a lack of accountability with denials and write-offs from rejected claims in which authorization was not obtained.

Six months after the EHR go-live, practice leaders realized tools within the new system were not being used. A combination of lack of awareness and disparate procedures led the group to consider centralizing functions related to prior authorizations and referrals.

Two pilot sections — plastic surgery and endocrine surgery — began using a centralized authorization and referral process in the EHR similar to the centralized registration process. Cross-training of staff was emphasized, though Thibodeau noted that the work, especially with payers, was highly specialty specific — “one size does not fit all,” she said.

Due to the variance, communication became crucial. In-basket messaging was set up with pools for each department to quickly address whether a peer-to-peer review was needed or there was a problem with a claim. Work queue structures in the EHR also prioritized the order of authorizations and referrals for surgical, diagnostic, non-diagnostic, admissions and psychiatry work.

In the first two years, denials based on authorizations not obtained fell 18% and 23%, respectively. These improvements allowed the departments to set minimum daily productivity expectations for staff:

• 50 office visits
• 26 procedures
• 19 surgeries

Coding and billing

Post-EHR implementation, Yale Medicine launched a steering committee and engaged a consultant in 2014 and 2015 to identify opportunities for centralizing coding and billing work, including:

• A comprehensive reconciliation process for charge capture
• Eliminating paper and manual processes
• Improved coder and provider training
• Standardized policies and procedures
• Co-location of all staff

Workgroups formed with representation from the affected departments, and a service-level agreement (SLA) was developed to delineate tasks for a central coding/billing team from the individual departments.

Recognizing the amount of work required, Dupee said the work was phased between April 2016 and September 2017, which included transitioning and relocating departments to join eight clinical departments already co-located. “We were not going to manage staff that resided in 17 different locations … we want them all in the same building; we want them to learn from each other,” Dupee said.

Through July 2018, this process allowed the practice to: 

• Document processes for all the transitioned departments
• Compile quality assurance data on coders to identify improvement opportunities
• Add a monthly coding education session for all staff
• Implement monthly meetings with other centralized business units and clinical departments
• Pilot a new provider education program

Along the way, a key metric to track the success of the centralization work for coding and billing was pre-A/R to A/R lag days. For the first six departments that transitioned, that metric fell from 5.0 days for the original eight co-located departments in December 2015 to 2.3 days in June 2017 with 14 departments together.

In the final phase from September 2017 to June 2018, Yale Medicine brought two behavioral health departments with very high lag day counts into the centralized team. A mix of bundled billing, varying treatment plan requirements and providers who may not close an encounter for months all contributed to lag days ranging between 23 and 48 days at the beginning of the transition, compared to the other 14 departments ranging from 2.2 to 2.5 days once centralized.

By June 2018, those behavioral health departments had their lag days down to 8.4. When combined with the other 14 departments, the overall lag days was 2.7, down from 4.0 in March 2018.

Lessons learned

Dupee said the Yale Medicine team assumed that staff were already doing much of this work in the clinical departments; instead, variances were found across the practice. “Some departments were split by doctor rather than by sub-specialty,” Dupee said. Finding those variances and standardizing was a major hurdle to clear.

Dupee hoped to transition the big binder of documented processes — still in paper — to a searchable PDF, though she acknowledged her coder-billers still intend to “go onto the PDF and print it.” 

The collaborative meetings between clinical departments and the centralized business units managed by Thibodeau also were vital. “The front end and the back end need to work together to understand if they’re seeing rejection trends,” Dupee said.

What’s next

The next step for Yale Medicine is to create a new provider education program as they come on board, and then track results to assess the value of that education to improve on the current level of centrally managed clinical billing of 87%. “We hope to move them up to 100%,” Dupee said.

Yale Medicine also recognizes that as it improves processes, patient access becomes a bigger challenge. The practice scheduled about 1.7 million appointments in FY2018 with about 170 different phone lines for incoming patient calls. In a single year, the practice received 1.5 million calls with an average abandonment rate of 9% and average terminated call rate (for voicemails or answering services) of 13%.

“We found about one out of four calls was just not being answered,” Thibodeau said, pointing toward work to create a single number for an access center with staff who can register and schedule patients. “This is going to be a big culture change,” she added.

“The centralization of certain services has allowed us to deliver clearly better customer service to the physicians, to the departments and to the patients that we serve,” Thibodeau said. “It all works so intricately together.”

The key to it all? “It’s constant tweaking, depending upon the payer: Rules change, codes change, it’s just constant work,” Thibodeau said.
 

Editor's note: This article appears in the January 2019 issue of MGMA Connection magazine.