Skip to content

Systems Approach to Alerting Hospitalists of Need for Serious Illness Conversations Falls Short

Key findings

  • Researchers at Massachusetts General Hospital devised an automated system for identifying recently admitted patients who were at risk of having unmet palliative care needs
  • Over a three-month study period, 38 clinicians in intervention units were informed of the screening results and 23 in control units were not; all clinicians had previously been trained in conducting serious illness conversations
  • Clinicians on the intervention units experienced more growth in confidence in individual communications skills over time, but the differences were not significantly different from the growth noted among clinicians on the control units
  • Both groups reported small but significant declines in their perception of the importance of holding serious illness conversations
  • Clinicians in the intervention units reported a small decrease in their sense that having serious illness conversations were meaningful to their work

In light of the national shortage of specialty palliative care providers, it would be optimal for hospitalists to conduct goals and values conversations with seriously ill patients. These conversations potentially improve quality of life, reduce symptom burden and health care utilization, and even lengthen survival.

However, communication with seriously ill patients can be challenging for hospitalists because of lack of time, uncertainty about prognosis and complicated patient/family dynamics. Some providers find the conversations emotionally draining or uncomfortable.

Juliet Jacobsen, MD, of the Division of Palliative Care and Geriatric Medicine at Massachusetts General Hospital, Jeffrey L. Greenwald, MD, a hospitalist in the Department of Medicine, and colleagues devised a systems approach to improving hospitalists' communication skills and confidence while increasing their sense of the importance of serious illness conversations. However, in the Journal of Palliative Medicine, they report limited success.

Study Details

The researchers studied 61 hospital medicine clinicians who had been trained in the Serious Illness Care Program. 72% were physicians (the others were nurse practitioners or physician assistants); the mean number of years in practice was four.

Before and about three months after enrollment, clinicians completed a survey assessing their confidence in communicating with patients about serious illness and perceptions of the impact such conversations have on patients and themselves.

The general medical units the participants practiced in were designated as:

  • Intervention (38 clinicians): A research assistant used an electronic health record algorithm to screen recently admitted patients and identify those at risk of unmet palliative care needs. During morning interprofessional rounds, the assistant recommended consideration of a serious illness conversation if appropriate
  • Control (23 clinicians): The research assistant screened admissions similarly but did not share the information with clinicians

Clinician Confidence, Communication & Perceptions

Clinicians on intervention units demonstrated a significant increase in confidence in nine of 10 communication measures (e.g., responding to emotions, exploring hopes, exploring worries). Clinicians on control units reported a significant increase in confidence in four measures. However, the mean between-group change in confidence did not differ significantly between the two study groups on any communication measure.

Clinicians in both groups noted improvement in certain communication measures (e.g., asking about fears and worries) but not others (e.g., evaluating prognostic awareness). None of the improvements in clinician communication differed significantly between those in the intervention versus control units.

Neither group perceived a negative effect of serious illness conversations on patient mood. Both groups noted a significant decline from the start of the study to the follow-up survey in their perceived sense of the importance of having serious illness conversations (control, 3.50–3.17 out of 4; intervention, 3.50–3.23; P < .05 for both). Over time, clinicians on the intervention units reported a relative decrease in the perception that having serious illness conversations was meaningful to their work (4.35–4.08 out of 5; P < .05).

Interpreting the Findings

Many of the positive findings are probably more attributable to the training the clinicians received before the study than to the automated process for identifying patients in need of palliative care.

The decrease in the perception that serious illness conversations are meaningful to one's work is critical to explore further because it may be relevant to enhancing job satisfaction. Lack of a sense of meaning is a significant contributor to burnout, and burnout rates are high in hospital medicine.

Learn more about palliative care at Mass General

Refer a patient to the Division of Palliative Care and Geriatric Medicine


According to Massachusetts General Hospital Cancer Center researchers, offering palliative care early in the course of cancer treatment empowers patients and families to cope more successfully.


In a randomized trial, patients with advanced cancer who were assigned to view an educational video about hospice were more likely than controls to use hospice, and their hospice stay was longer.