From 3M Health Information Systems
What Can Go Wrong with PCP Attribution and How It Can Be Prevented
Attributing a person to a primary care physician (PCP) is an essential feature of population health management because it enables an accurate and fair assessment of the quality of care a provider delivers. Attribution is based on the concept that a PCP is responsible to a person across time and the entire continuum of care. It establishes this responsibility, creating a relationship between a person and his or her PCP. When members have a designated PCP, plans are able to consider the overall health of a PCP’s unique panel of patients, enabling them to measure and reward provider performance on an apples-to-apples basis.
Outside of health plans that require up-front PCP choice, Treo Solutions, part of 3M Health Information Systems, has a PCP attribution process designed to help determine and assign the relationship between a health plan member and the particular PCP who is primarily responsible for the individual’s overall health care.
When health plan data does not specify an existing PCP-member relationship, we identify the connection through the preponderance of claims data in a multi-step PCP attribution process.
1. Define the PCP: We start by defining a PCP as a physician with a specialty of Family Medicine, Internal Medicine, Pediatrics or General Practice. Some clients choose to include OB/GYNs and/or Nurse Practitioners (NPs) and Physician Assistants (PAs) as PCPs.
2. Identify the physician group: Once the definition of a PCP has been established, we identify the physician group that provided the patient the majority of evaluation & management (E&M) services (99201-99499) from typical PCP place-of-service codes (office, walk-in retail health clinic, independent clinic, federally qualified health center, public health clinic and rural health clinic). In case of a tie, we choose the group with the most non-E&M services for that individual. We use “most dates of services” and “most recent service” for further tie-breaking.
3. Identify the PCP: Using the same process in step 2, we identify the PCP within the group who has the majority of E&M services (99201-99499), with the same tie-breaking process.
Issues that can arise in claims-based attribution
Despite a fine-tuned attribution process, the task of determining the PCP for each health plan member can still encounter issues. There are several that can arise at the early stages and they’re often related to data integrity. Below are a few of the most common and what can be done to solve them.
- A member has not had a recent PCP visit. Individuals with no PCP visits in the 12-month reporting period will be unassigned. At Treo, we counter this issue by extending the look-back period to 24 months to capture more visits.
- A PCP’s information (name, specialty, physician group) is wrong. An inaccurate PCP listing or inaccurate link between a PCP and physician group leads to inaccurate attribution. Health plans and provider systems must review the accuracy of PCP listings as well as the links between PCPs and groups. As the accuracy of these files improves, so will the attribution process.
- A Nurse Practitioner or Physician Assistant is inaccurately assigned. If an NP or PA is assigned to primary care versus specialty care, it can skew the attribution process at Step 1 by misrepresenting which providers qualify as a PCP. Health plans and provider systems must review the list of NPs and PAs to improve the accuracy of their listing as PCP. The more accurate the listing, the more accurate the attribution process will be.
Achieving a high member attribution rate starts with quality data and calls for a fine-tuned PCP attribution process. It is a collaborative effort among payers, provider systems and the analytics vendor that leads to better coordinated patient care and effective pay-for-performance models.
L. Gordon Moore, MD, is Senior Medical Director for Populations and Payment Solutions at 3M Health Information Systems.