Actual vs Documented Advance Care Planning in a Primary Care Clinic

Advance care planning (ACP) is a process that enables patients and physicians to define goals and preferences that direct future medical care, as well as identify surrogate decision makers.¹ Primary care physicians are uniquely suited to perform ACP due to their longitudinal relationships and patient-centered model of care.^2,3 However, ACP is a variable entity that takes many forms and occurs across various settings. There is evidence that many physicians feel unprepared to perform ACP^4,5 and that ACP may be underbilled.^6,7 For organizations that use billing for ACP as a quality metric, this may have significant ramifications and indicate a need for further education and training on ACP performance, documentation, and billing. We performed a retrospective chart review for patients aged 70 and older in our family medicine resident clinic to evaluate the various ways in which ACP is charted and the gap between billed and non-billed ACP.

The project was reviewed and approved by the Care New England Insitutional Review Board. We searched the electronic medical record (EMR) for clinic patients over 70 years old seen between August 25, 2020 and Sepetember 25, 2020. Seventy years was chosen as an arbitrary cutoff age to ensure a population most likely to have undergone ACP. We organized results in order of most recent contact (visit with PCP or associated specialist, telephone call, refill, or any interaction with the hospital system) to exclude patients remote from care. We selected the first 50 patients for chart review. Two patients were excluded due to never being seen in our office. We collected demographic information.

We then reviewed charts for ACP using a standardized template specific to the EMR (Epic). This included reviewing billing, all uploaded documents, all encounters from the past calendar year, a special ACP tab, and searching the chart for use of any of the following terms: “code,” “DNR,” “DNI,” “ACP,” “living,” “life,” “death,” “POA,” “goals,” “end,” “MOLST,” “palliative,” and “hospice.” Billing for ACP was defined as use of CPT codes 99497 or 99498 at any point. All ACP was graded on a binary scale of present or not present. Primary outcomes were the percentage of patients with ACP and incidence of ACP documents. Secondary outcome was the proportion of documented ACP conversations in office visits that had billing for ACP.

The average patient age was 80.9 years, and the majority were female, White, and English-speaking (Table 1). Documents found included power of attorney (POA, four patients, 8.3%), medical orders for life-sustaining treatment (MOLST, 2, 4.2%), and advanced directives (AD, 6, 12.5%). Twenty-five patients (52.1%) had a discussion regarding ACP in a PCP visit. Thirty-eight (79.2%) had a documented code status. Eleven (22.9%) were billed for ACP.

While the vast majority of patients (85.6%) had some form of ACP on chart review, only 25% of patients had ACP documents (eg, POA, MOLST, or AD) found in the EMR. This is consistent with prior studies which have demonstrated a low rate of signed, legal ACP documents scanned in the EMR.⁸ Although many of the ACP conversations discussed ACP documents, the majority did not result in signed documents. This may reflect a gap in the ACP process as physicians may not emphasize the importance of returning such documents to the PCP office.

Eleven patients (22.9%) were billed for ACP. This rate of billing compares favorably to nationwide averages showing a peak rate of 5.1% for seriously ill patients⁹ and is consistent with analysis that shows a higher rate of billing in New England.⁸ However, there is evidence that ACP was underbilled. Fourteen (29%) patients had documented conversations about ACP without billing. Although the criteria for billing for ACP are stringent,^10,11 we suspect that some of these visits met criteria based on the quantity and content of documentation.

That ACP is underbilled is consistent with data showing that ACP billing increases with minor educational interventions⁶ and has increased steadily since its inception in 2016.^12,13 Indeed, one study found that having physicians reflect on a patient’s risk of dying doubled rates of ACP billing.¹⁴ The potential underbilling of ACP could have major implications for organizations using billed ACP as a quality care metric and represents an opportunity for further training. Given that many physicians feel unprepared to perform ACP^4,5 and the significant inter-provider variation in billing rates,¹⁴ training on best ACP practices could significantly improve performance and billing rates.

Limitations of our study include its retrospective, observational nature which limits causal inference. Our population was older, majority White, and English-speaking, limiting generalizability. Moreover, lack of a standardized definition for ACP limits application to clinical practice.

Overall, our review demonstrates the complex, multifaceted nature of ACP. At a time when many are calling into question the utility of ACP,^15-18 future research is needed to better define ACP and help organizations apply these definitions in ways that contribute meaningfully to patient care.