Inadequate, Missing, or Inaccessible Goals of Care Documentation Is Ethical Concern

By Stacey Kusterbeck

Early goals of care discussions help to ensure that patients’ preferences for medical care are aligned with the treatment they receive. Such discussions also result in cost savings, found one study.¹ However, goals of care discussions may be done inconsistently, inadequately, or are missing from the medical record when needed most. Here are some recent research findings regarding ethical concerns with goals of care documentation.

Patients and families may perceive negative content in goals of care documentation (such as judgmental comments or lack of empathy). Researchers asked six community members and six clinicians to review 65 goals of care notes written by clinicians at 14 hospitals and analyzed participants’ perceptions of positive and negative content.² Clinicians and community members had similar perceptions about positive content, such as clear documentation, kind language, and describing patient values, but differed in how they perceived negative content. Community members perceived concerns about patient or family understanding and lack of empathy as negative content. Clinicians perceived lack of important information and poor writing as negative content. Both groups perceived descriptions of bad patient health as negative content.

Patients with Alzheimer’s disease and related dementias (AD/ADRD) are less likely to have values and preferences documented in a goals of care note. Researchers assessed goals of care documentation for 5,475 hospitalized patients with and without AD/ADRD at 21 hospitals.³ Seriously ill patients with AD/ADRD were more likely than seriously ill patients without AD/ADRD to have documentation of family members present during the conversation, more likely to have surrogate decision-makers, and more likely to have a prognosis, but they were less likely to have documentation of patient presence during the conversation and less likely to have patient values and preferences documented. This suggests that patients with AD/ADRD are at risk of receiving goal-discordant care. “Clinicians and ethics consultants should be proactive in assessing patient values and preferences for care for patients with AD/ADRD from these patients themselves (if they are able to communicate) or their surrogate decision-maker,” asserts Gina Piscitello, MD, MS, the study’s lead author and assistant professor at the University of Pittsburgh School of Medicine.

Goals of care discussions often omit psychosocial concerns (such as social stressors) that many patients with life-limiting chronic illness often struggle with. Researchers analyzed 26 transcripts of goals-of-care conversations.⁴ Participants frequently discussed these major themes regarding psychosocial concerns: acceptance of illness, feeling of being a burden, illness transitions, and suffering from disease. Better recognition of these psychosocial concerns can help clinicians to elicit patient goals and preferences when making treatment decisions, assert the study authors.

Asking questions about serious illness experience can help to elicit goals of care. It often is challenging for clinicians to identify and document patients’ goals of care. A group of researchers conducted a study to determine if asking questions about serious illness experience during advance care planning could help to elicit goals of care.⁵ The researchers analyzed 92 transcribed advance care planning discussions with adults 80 years of age and older with cognitive impairment and their care partners. In describing serious illness experience, it became clear where participants’ goals of care fell between longevity and comfort. For example, when participants described experiences with iatrogenic harm or living with severe disability, it provided an opportunity for providers to learn that participants emphasized comfort over longevity.

There are gaps in the content of pediatric goals of care documentation. Researchers analyzed 429 goals-of-care notes written from 2016-2023.⁶ Most notes included treatment decisions, discussion participants, and goals, fears, and priorities. Only 9.8% of notes included information preferences. Just one-third of notes included end-of-life care planning.

“While the notes included key components of goals of care, the depth and quality of these components varied widely. This variation raises concerns as to how well patient values and preferences are being captured and communicated in the electronic health record,” says Hannah Reidy, MD, the paper’s first author and currently a pediatric resident at Boston Children’s.

A surprising finding was the misclassification of notes as goals-of-care documentation. This mislabeling may hinder clinicians’ ability to quickly locate relevant patient goals of care. “With the variation in the content documented, clinicians should consider how they approach documenting goals-of-care conversations,” says Reidy.

Nearly 40% of deceased patients did not have documented end-of-life preferences. This finding highlights a significant gap in communicating and planning for the end-of-life in the pediatric population. “Given the substantial proportion of patients without end-of-life preferences documented, it is imperative for clinicians to develop methods to proactively identify patients who may benefit from early goals-of-care discussions,” says Jessica Ma, MD, MHSc, another of the study authors. Ma is an assistant professor at Duke University School of Medicine and physician in the Geriatric Research Education and Clinical Center at the Durham VA Health Care System.

Ethicists are uniquely positioned to help develop and strengthen institutional policies around goals of care documentation, according to Ma. Ethicists can help advocate and design workflows and policies for standardized, high-quality documentation.

“To support patient autonomy and goal-concordant care, ethicists can partner with providers in identifying opportunities to engage in, and document, meaningful goals of care conversations that honor the patient’s voice and values,” says Ma.

Machine learning models can measure whether and when goals of care conversations are documented. As a critical care and palliative care nurse, Alaa Albashayreh, PhD, MSHI, RN, witnessed patients and families struggling with crucial end-of-life decisions. “I saw firsthand how advance care planning conversations about goals of care and surrogate designation are essential elements of patient-centered care. Yet, tracking when and with whom these conversations occur has been extremely challenging,” says Albashayreh, an assistant professor at the college of nursing at University of Iowa.

Albashayreh and colleagues tested four natural language processing models to detect goals-of-care and surrogate designation conversations in clinical narratives.⁷ The Bidirectional Encoder Representations from Transformers (BERT) model was most accurate, detecting goals-of-care conversations in 85% in the medical records of patients and in 60% of surrogate designation conversations.

The first documentation of goals-of-care conversations occurred at a median of just 28 days before death. Surrogate-designation conversations were documented even later — only eight days before death. “This is concerning. It suggests these conversations are happening too late in the disease trajectory for patients to meaningfully benefit from them,” says Albashayreh.

Conversations that are documented within weeks or days of death become crisis-driven rather than thoughtful planning discussions. “Clinicians should initiate these conversations earlier in the disease trajectory — ideally, when patients can still meaningfully participate and before acute deterioration occurs,” says Albashayreh.

The researchers also found significant disparities in who receives goals-of-care conversations based on race, marital status, and religious affiliation. “This highlights how social determinants of health affect access to these essential conversations,” says Albashayreh. Clinicians should examine their own practices by asking these questions, says Albashayreh:

• Are certain patient populations being overlooked for goals-of-care discussions?
• Are we making assumptions about who wants or needs advance care planning conversations based on race, marital status, or other demographic factors?

“The good news is that our study demonstrates natural language processing can accurately detect advance care planning conversations in clinical notes,” says Albashayreh. Ethicists can help to address this issue at their institutions with these practices, says Albashayreh:

• Ethicists can work with informatics teams to develop systems that track advance care planning conversations. It is not enough to track whether the conversations are happening. Systems also need to track when the conversations occur relative to clinical milestones and patient outcomes, says Albashayreh.
• Ethicists can address structural barriers. Patients referred to specialty palliative care were significantly more likely to have documented goals-of-care conversations. This suggests that goals of care conversations are not happening consistently in primary care settings. “Ethicists can advocate for training and support that enables all clinicians, not just palliative care specialists, to have these conversations,” suggests Albashayreh.
• Ethicists can examine equity issues. Ethicists can start the discussion by asking this question: Why are certain patient groups less likely to have advance care planning conversations documented?
• Ethicists can develop artificial intelligence (AI)-driven clinical decision support. “This is where the future lies,” says Albashayreh. Ethicists can collaborate with other healthcare and informatics teams to develop systems that identify patients who have not had goals-of-care conversations yet but likely would benefit from them. “Our research establishes the foundation for such predictive interventions, using AI to prompt clinicians to initiate these conversations at appropriate times with the right patients, before crisis moments occur,” says Albashayreh.
• Ethicists can include advance care planning timing (not just occurrence) as a quality metric. Ethicists can use the median time from conversation to death as an institutional indicator of whether goals-of-care discussions are happening early enough to benefit patients. “We now have the computational tools to measure these important care processes at scale and, potentially, to identify patients who need these conversations before it’s too late. Ethicists can leverage these capabilities to move from anecdotal concerns to data-driven quality improvement,” offers Albashayreh.

Stacey Kusterbeck is an award-winning contributing author for Relias. She has more than 20 years of medical journalism experience and greatly enjoys keeping on top of constant changes in the healthcare field.

References

1. Piscitello GM, Arnold RM, Schell J. Cost-related outcomes associated with documented goals of care conversations. J Pain Symptom Manage. 2025; Sep 23. doi: 10.1016/j.jpainsymman.2025.09.017. [Online ahead of print].

2. Piscitello GM, Garcia-Fuentes E, Arnold RM, et al. Assessing perceptions of positive and negative content within goals of care notes. J Pain Symptom Manage. 2025;70(5):503-513.e1.

3. Piscitello G, Lingler J, Ramos K, et al. Assessing the content of goals of care documentation for hospitalized patients with Alzheimer’s disease and related dementias. J Am Geriatr Soc. 2025;73(8):2531-2538.

4. Gambino J, Schlichte LM, Haverfield MC, et al. Uncovering psychosocial contexts in goals of care conversations: A qualitative study in a multi-centered randomized controlled trial. Clin Gerontol. 2025;48(5):1320-1327.

5. Thai GH, Hanna V, Zhang P, et al. Asking older adults with impaired cognition and care partners about serious illness experiences can elicit goals of care during advance care planning. Am J Hosp Palliat Care. 2025; Aug 13. doi: 10.1177/10499091251367337. [Online ahead of print].

6. Riedy H, Lyons K, Casarett DJ, et al. Content and comprehensiveness of goals-of-care documentation in pediatric patients. J Pain Symptom Manage. 2025;70(2):e144-e158.

7. Albashayreh A, Herr K, Fan W, et al. Harnessing natural language processing and high-dimensional clinical notes to detect goals-of-care and surrogate-designation conversations. Clin Nurs Res. 2025;34(7):321-331.