Article Text

Download PDFPDF

Advance care planning in Medicare: an early look at the impact of new reimbursement on billing and clinical practice
  1. Gawin Tsai1 and
  2. Donald H Taylor2
  1. 1 Duke University, Duke Clinical Research Institute, Durham, North Carolina, USA
  2. 2 Duke University, Sanford School of Public Policy, Durham, North Carolina, USA
  1. Correspondence to Dr Gawin Tsai, Duke University, Duke Clinical Research Institute, 2400 Pratt Street #6551, Durham 27705, North Carolina, USA; gawint{at}


Objective In this study, we examined the US Medicare programme’s (government-funded social insurance for the elderly or disabled) new reimbursement for advance care planning (ACP) that began on 1 January 2016. This single-centre study addressed whether clinicians who have ACP conversations with patients will use the new reimbursement code and if the new reimbursement is successful at motivating clinicians to have more ACP conversations with patients.

Methods This is a multimethod study. To gain a general sense of ACP practice and code visibility, we first surveyed 493 clinicians in a large academic medical centre (20% response rate). Then, for more in-depth answers and to illuminate the reasons behind survey findings, we conducted semistructured interviews with 28 physicians.

Results We found that while clinicians are open to using the reimbursement codes, organisational barriers such as low visibility and documentation make it difficult for clinicians to bill for ACP. Moreover, structural and professional factors have rendered Medicare’s ACP reimbursement largely ineffective at motivating healthcare providers to perform more ACP conversations during the first 3 months of this policy.

Conclusions It does not appear that Medicare’s reimbursement of ACP has made a significant, direct impact on ACP billing or practice during the policy’s first 90 days. However, there is a symbolic role that this change can serve, and the policy could have more impact as its existence becomes more widely known. Barriers to ACP that we identify should be addressed directly to expand the use of ACP.

  • advance care planning
  • health policy
  • palliative medicine

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


With the ageing of the American population, topics pertaining to the chronically ill and aged are now at the forefront of many initiatives in the Medicare programme. Medicare is a single-payer, tax-financed, fee-for-service health insurance programme that covers approximately 57 million aged (>65) or disabled Americans.1 On 1 January 2016, Medicare started reimbursing clinicians for having advance care planning (ACP) conversations with their patients at the rate of $80.00–$86.00 for the first 30 min and about $75.00 for each additional 30 min.2 3

ACP has been defined as "making decisions about the care you would want to receive if you become unable to speak for yourself."4 Studies have found that patients often do not want aggressive care towards the end of life,5 6 and a 2014 poll found that about 25 million Americans over 50 have experienced "excessive or unwanted medical treatment."7 ACP conversations give patients the opportunity to express and document their end-of-life wishes and have the potential to result in better end-of-life care by better aligning healthcare with patient preferences.

Decreasing the amount of unwanted, aggressive end-of-life care should also result in health savings. According to the Agency for Healthcare Research and Quality, in both 2008 and 2009, about 5% of the American population accounted for nearly 50% of healthcare expenditures. This 5% was more likely to be elderly and in fair to poor health, the persons at the heart of the cost and quality crisis in Medicare.8

While ACP has the potential to be beneficial to patients and the health system, barriers have been found on both the patient and clinician sides that have made ACP discussions difficult.9 10 In this early stage of the Medicare change, we addressed two questions: Will clinicians who have ACP conversations with patients use the new billing code? Does Medicare’s reimbursement motivate clinicians to have more ACP conversations with their patients?


For our study sample, we identified 493 clinicians through a large academic centre’s website. Those included in the study were either clinicians found under the primary category of ‘primary care’ and subcategory of ‘adult care’ or clinicians found under the primary category of ‘specialist’ and subcategory of ‘cancer’. Physicians, nurse practitioners and physician assistants were all included in our study as these professional are all allowed to bill for ACP under the new Medicare reimbursement.

We used a multimethod research approach for breadth and depth. First, to gain a broader viewpoint on the issue, we emailed a five-question survey to the 493 clinicians asking about their practice of ACP, their awareness of the Medicare change, and whether or not the change would affect their practice. Clinicians were given 2 weeks to fill out the survey (6–20 January 2016) with a reminder email sent after the first week. Survey analysis was done using the survey software Qualtrics. A total of 99 (20% response rate) clinicians took our survey, 54 primary care clinicians and 45 specialists.

Next, for a more in-depth perspective, we conducted 28 semi-structured interviews (21 January to 6 April 2016) with clinicians. Individual emails were sent to the 493 clinicians surveyed asking for their participation in the interview portion of the research. Due to a low response rate, snowball sampling was also used to recruit more interviewees within the same research pool.11 Researchers were careful to start with diverse informants and specify that the research could benefit from clinician viewpoints regardless of their ACP practices. Nurse practitioners and physician assistants were also sent follow-up emails requesting their participation. However, despite recruitment efforts, only physicians participated in interviews. Interviews lasted 30–60 min each. Interviews were analysed using applied thematic analysis with intracoder reliability efforts. A total of 28 physicians were interviewed, 10 primary care physicians and 18 specialists.


Will clinicians use the new ACP reimbursement codes?

Survey findings

Our survey assessed ACP reimbursement visibility. Although 83% of responding primary care clinicians reported having ACP conversations with their patients, only 35% reported knowing about Medicare’s change to reimburse for ACP. As for specialists, 59% of respondents reported practicing ACP, and 24% reported knowing about Medicare’s reimbursement change.

Interview findings

Prior to 1 January 2016, if clinicians had extended ACP conversations, they might have billed by time, increased complexity of visit or not billed for ACP at all. Most interviewed physicians were open to use of the codes, however noted that barriers such as uncertainty on how to document, information overload, or technological barriers were significant. None of the physicians interviewed felt confident enough in their knowledge of the details behind Medicare’s change to accurately bill for ACP today. For a few, the time it would take to figure out how to document would  not be worth the reimbursement.

We also found that the structure of the billing code needing 30 min of face-to-face counselling clashed with the organisational structures of most outpatient clinic schedules. Appointments were oftentimes scheduled to be shorter than 30 min. Also, some clinicians who practice ACP practice it in short bursts throughout many different appointments rather than at one long appointment.

Does Medicare’s reimbursement motivate clinicians to have more ACP conversations with their patients?

Survey findings

Sixteen per cent of responding primary care clinicians indicated that the new reimbursement was enough for them to try to incorporate (more) ACP in their practice. For responding specialists, only 4.4% noted that the Medicare change motivates them to practice more ACP.

Interview findings

The majority of interviewed physicians did not think that Medicare’s new reimbursement is enough to motivate them to practice more ACP, because they think they are already doing the best they can. Interviewed physicians settled on two distinct categories of barriers to having (more) ACP conversations: structural barriers and professional barriers (table 1). Structural barriers to having ACP conversations involved issues such as the lack of time, competing demands during a clinic appointment, or difficulties in provider coordination. Professional barriers to having ACP conversations were those such as ACP conflicting with the physician identity of being an eradicator of disease or a clinician’s lack of comfort with having the conversation.

Table 1

Selected illustrative quotes: structural and professional barriers

While Medicare’s reimbursement was not a significant motivating factor, a number of interviewed physicians said that they would make a concerted effort to increase ACP conversations if they were convinced that having ACP conversations is crucial to good medical practice. Some of these physicians have been waiting to hearing more about the ACP reimbursement change from their professional societies and would consider attending workshops that show clinicians how to have productive ACP conversations.

Although most physicians denied a direct impact of the reimbursement on their practices, interviewed physician advocates of ACP in our research setting have been working on some potentially impactful changes. Some are trying to change the electronic health record system to include better documentation of ACP conversations and end-of-life issues. Others are brainstorming ways to work within the existing structure, such as looking at scheduling a patient for back-to-back appointments in order to have enough time for ACP conversations. A few are developing ACP resources such as templates or scripts.


In this study, we have found that, at this time, Medicare’s new reimbursement for ACP has not been able to make a significant impact on the billing or practice of ACP. Our survey’s low response rate, and this being a single-centre study, are limitations of our research. However, this study provides useful information that can inform the development of new ACP reimbursement programs. Low visibility and documentation barriers have made it difficult for clinicians to make use of the billing codes, and the reimbursement has not been enough to combat practice barriers and motivate clinicians to strive for more ACP conversations. However, we acknowledge the new reimbursement’s potential to serve as a legitimiser and catalyst for change as it supports ACP advocates’ efforts in breaking down ACP barriers. Additional ways in which the medical community could support these advocates and more directly target the barriers to ACP are needed.

For instance, ACP benefits could be promoted to clinicians. In interviews, physicians stated that they would be more motivated to do ACP if they were convinced that ACP is a crucial part of good medical care. The most obvious change would be to better incorporate the importance of aligning patient and health system goals during medical training with ACP as a specific example. However, for already established practitioners, including ACP in continuing medical education classes and promoting it through professional medical societies can convince clinicians of the benefits to ACP and incite true buy-in on their part which is vital for any policy goal. Medicare could also consider increasing the ACP reimbursement. With a larger reimbursement, health administrators would take more notice and put more organisational emphasis on ACP, coming up with solutions to organisational barriers.

These suggestions could complement the new Medicare reimbursement and support ACP advocates in order to combat the structural, professional, and organisational barriers to ACP, facilitating the process in making ACP a part of standard American healthcare. This study, while limited to the American context, could also be applicable in other fee-for-service healthcare systems.


View Abstract


  • Contributors Study concept and design: GT and DHT. Acquisition, analysis, or interpretation of data: GT. Drafting of manuscript: GT. Critical revision of manuscript for important intellectual content: GT and DHT. Obtained funding: DHT.

  • Funding The project described was supported by Grant Number 1C1CMS331331 from the department of health and human services, Centers for Medicare & Medicaid Services. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the US Department of Health and Human Services or any of its agencies.

  • Competing interests None declared.

  • Ethics approval Duke Medicine Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.