Riddle me this: I am a system that is active and essential in healthcare, but a complete mystery to most healthcare professionals. Answer: a patient attribution system! Hopefully you did not peek at the title of my post.

Did you get the right answer? Don’t worry if you did not guess a patient attribution system. Most of us professionals working in healthcare know little about how patients are assigned a primary care provider. To illustrate further, I will describe a perfect situation on how this enigmatic system could work in perfect harmony with healthcare service delivery.

A Perfect Situation

Imagine you are a primary care provider (PCP) with integrated behavioral healthcare (IBH) at your clinic. All the patients that walk through the clinic doors are assigned to you by their respective payers. The payers make these assignments to you based on a standard algorithm that considers patient preference, patient health needs, and the quality of your service (including IBH).

As the PCP, you clearly understand these patient assignments because the payers have a great communication strategy that helps you immensely with your performance goals. The payer notifies you of patients who are at risk for poor outcomes, assists you in tracking patient progress over time, and supports you in increasing the level of integrated care.

The extra reimbursement from the alternative payment model (APM) you are participating in allows you to hire an integrated care manager. This individual, who was cross-trained in medical and behavioral health services, works closely with the entire care team to facilitate all payer-PCP communication and engage patients in the IBH service.

This perfect situation works because you clearly know your patient panel, you have a clear communication line with the payer, and you have the resources to connect patients to your IBH service. What a dream!

Patient Attribution Methods and Problems

Most healthcare professionals know little about how payers identify a patient-provider relationship (i.e., patient attribution). Patient attribution is the process of assigning patients to providers for the purpose of measuring and improving the quality and cost of care. IBH cannot realize its potential without an accurate and reliable patient attribution system.

There are numerous methods for assigning patients to providers. The common method is to attribute a patient to the PCP with the most documented primary care charges (e.g., E&M codes for wellness and preventive services).

From there, each method can pull from various elements, resulting in hundreds of algorithm permutations that ascribe a provider-patient relationship. If the patient has not established care with any PCP, then the payer will assign a relationship. This is where attribution gets complicated and inefficient.

I recommend reading this open access article by Riley et al. in JAMA Network to learn more about the limitations of current patient attribution methodology. In short, these limitations will keep IBH from reaching its full potential. Primary care teams participating in APMs need accurate information about the patients they are responsible for caring.

For example, one limitation is that most attribution models depend on retrospective assignment. This means that a PCP will learn about a patient on their attribution roster within months or years after the close of a performance period. That’s a long delay! This interruption can negatively impact both patients who have or have not established care.

Let’s say a patient has already established care and received excellent primary care services. If the payer attribution method has assigned that patient to another clinic, then the original PCP will not receive credit for treating that patient. On the flip side, if a patient has not established care and the PCP has no idea the patient is on their attribution roster, then the patient will miss out on critical primary care and behavioral health services and the PCP will miss out on higher compensation.

The accuracy of payer-derived attribution ranges from 20% to 69%.² One research team observed that each of the seven health plans participating in a project used a different patient attribution model.¹ The wide swings in accuracy and variability of patient attribution will make the transformation toward APMs more difficult.  

Toward a Reliable and Accurate Attribution System

To be clear, some attribution methods have strengths and are working as designed (e.g., Medicare). However, clinicians may not share that view and may see the current methods as unreliable, inaccurate, and untimely.

Riley et al. propose six recommendations for improving patient attribution in healthcare, including shared standards across all payers, new prospective-focused methods, accurate provider performance reports, a standard roster reporting system, a focus on vulnerable populations, and the inclusion of patient preferences.¹

IBH models were designed to expand patient access to behavioral health services and contribute to the quadruple aim. When a healthcare organization decides to add IBH, they need a business model that works. Under a fee-for-service model, providers tend to know how rate structures affect their solvency and their ability to keep the doors open with appropriate staffing models.

Under an APM, providers manage additional risk; to survive financially, they need a clearly defined patient-payer mix and the infusion of revenue to cover expenses (like IBH!) for the attributed patients they serve.³ In short, IBH is not solvent under APMs without that clear delineation of the provider-payer mix. IBH, like any other part of the new primary care model, needs a timely, reliable, and accurate patient attribution system.

1. Riley W, Love K, Wilson C. Patient Attribution—A Call for a System Redesign. JAMA Health Forum. 2023;4(3):e225527. doi:10.1001/jamahealthforum.2022.5527
2. McCoy  RG, Bunkers  KS, Ramar  P,  et al.  Patient attribution: Why the method matters.   Am J Manag Care. 2018;24(12):596-603.
3. Harker, M. Patient Attribution: The Basis for All Value-Based Care. Society of Actuaries. August 2018. Retrieved from https://www.soa.org/globalassets/assets/Files/resources/research-report/2018/patient-attribution.pdf