High-Intensity Billing for ‘Treat and Release’ ED Visits Has Shot Up Since 2006

High-intensity billing during “treat and release” emergency department (ED) visits rose dramatically over the last two decades, according to an observational study.

From 2006 to 2019, the share of treat-and-release visits at the ED (those where patients were not admitted) that included high-intensity billing jumped from 4.8% to 19.2%, reported researchers led by Alexander Janke, MD, MHS, of the VA Ann Arbor Healthcare System and University of Michigan.

Also, the share of visits by patients with more comorbidities, older patients, and those with “nonspecific but potentially serious diagnoses” also went up, they stated in Health Affairs.

“Mechanistically, there is a tug of war between physician groups and payers for reimbursement, and one of the places where that tug of war plays out is in high-intensity billing,” Janke, who is a fellow in the VA National Clinician Scholars Program, told MedPage Today in a phone call.

His group found that only 47% of the increase in high-intensity billing was “expected” due to changes in administrative measures related to the patient case-mix and kinds of services available in claims data. “High-intensity billing” was defined as ED visits that included a CPT code signaling “high complexity” (99285) or “critical care” (99291 or 99292).

“Upcoding,” or submitting codes for more expensive procedures and diagnoses than were performed, was one possible reason for the upward trajectory of high-intensity billing, the researchers said. Janke’s group suggested that other reasons could be “a correction for historical downcoding” when coding practices were more simplistic, or “broader changes” in the evolution of care in the ED. For example, Janke recalled a senior colleague telling him that, when she began practicing, any very elderly patient who presented to the ED after passing out would be admitted to the hospital.

But emergency medicine has evolved and continually seeks ways to manage patients safely, while using fewer resources and minimizing hospital admission, he explained.

Now, an older adult with multiple chronic comorbidities who presents with a nonspecific complaint and is evaluated in the ED may receive a careful risk stratification and plan to manage their condition safely at home, without requiring hospital admission, he said.

“That’s sort of the most dramatic way in which emergency care’s underlying complexity has changed over the past two decades, and that’s what plays out in the data in the paper,” Janke stated.

He noted that the “expected” increase in high-intensity billing was based on information gleaned from basic claims data such as sex, age, and diagnosis codes in a patient’s chart. What isn’t included are the patient’s social determinants of health and the clinical complexity involved in their care, he explained.

“As we move in the direction of alternative payment models [APM] for emergency care, we have to be sensitive to how the complexity of emergency care has changed, Janke stated. For that reason, claims data alone are likely insufficient for understanding that evolution of care, he argued.”

“Any health policy work to better calibrate emergency care billing with value must account for what’s missing in these simple measures,” Janke noted in a follow-up email. “That includes things like the comprehensive management of older adults with social or functional barriers to safe discharge, the growth of risk stratification tools to safely avoid expensive hospital admissions, or the expanding toolkit of acute care providers to link patients experiencing homelessness to community resources.”

Janke said his group’s future research includes building out data sets to better characterize the complexity of emergency care. That will be “essential” to informing policy conversations of billing codes and developing APMs that “truly improve how we take care of patients,” he stated.

For the current study, the authors used the Nationwide Emergency Department Sample (NEDS) focusing on the period of 2006 to 2019, looking at variables such as age, sex, insurance status visit disposition and ED site region, among others. The analysis excluded patient visits that led to hospital admissions, to transfers to other short-term facilities, or to death. They cautioned that, because data on observation care are not reliably captured in the NEDS database, some visits in the treat-and-release sample may have included patients receiving observation services.