Palliative Care EMR Triggers Project

Background
Many patients living with advanced cancer have unmet needs around illness comprehension and coping, advance care planning and decision making, symptoms and daily functioning, and coordination of care. Consequently, many advanced cancer patients experience distress and suffering which diminishes quality of life. These needs remain unmet because they are challenging to address in the cancer clinic setting where oncology clinicians have limited time, competing priorities, and a lack of space for longer conversations. One solution is timely referral to supportive care services, including palliative care specialists, who are adeptly skilled and suited to address these needs. The problem, however, is delivery of palliative care is inconsistent across Canada, and cancer patients who would benefit from palliative care services are referred late, or not at all. Timely or early palliative care is defined as within eight weeks of advanced cancer diagnosis (the American Society of Clinical Oncology definition). But this introduces a second problem: It is unsustainable for palliative care services to see every advanced cancer patient.
 

From prior work, we know that requiring oncologists to systematically evaluate patients’ palliative care needs and make timely referrals is not feasible. Oncologists value screening cues. Consensus palliative care referral criteria have been established by an international Delphi study. Research is now needed to clinically validate these criteria using real world data (from health care administrative databases and the electronic medical record [EMR]) and clinician input. Then, we will test whether the criteria identify cancer patients with a high burden of unmet palliative care needs. Finally, we will use these validated criteria to drive EMR flags to alert busy oncology clinicians to patients with high palliative care needs.
 

The objectives of this project are:
1. To clinically validate performance of the palliative care referral criteria using health databases (including EMR data) in Alberta;
2. To build and pilot EMR-embedded electronic flags (driven by the algorithms developed in Objective 1) to assess the acceptability to oncology clinicians and impact on palliative care referral.
 

Methods
Objective 1. The SaferDx trigger tools development framework will guide Objective 1. Steps in this framework include identifying data sources, constructing the "high palliative care needs" referral algorithms, testing algorithms with clinician and patient advisors, validating algorithms using real-world (anonymized) patient medical records, and finally, assessing algorithm performance. Iterative refinements will lead to the final algorithms that will identify patients with high palliative care needs.
 

Objective 2. EMR-embedded flags that cue providers to patients with "high palliative care needs" will be piloted with 20 oncologists in Alberta for 4 months. We will gather data on the acceptability of these flags to oncologists. Sekhon’s Theoretical Framework of Acceptability for healthcare interventions will be used to ensure that acceptability is comprehensively assessed. We will also gather data on how well the flags worked. For example, did the gap between who should be referred, and who was referred, narrow for clinicians who participated in the pilot when compared to group of clinicians who did not?
 

Significance
Even though early palliative care is proven beneficial for people living with advanced cancer, it is not sustainable for all advanced cancer patient to see a specialist palliative care provider. EMR use is increasing across Canada. Using EMR tools to alert busy oncology clinicians to patients with high palliative care needs holds promise to improve timely palliative care referrals. Validating the palliative care referral criteria using health databases including EMRs (that can be adopted across Canada) is needed to assess feasibility and measure patient volumes generated by each criterion. This knowledge will allow different jurisdictions to selectively pick palliative care referral criteria that are sustainable for their centre. It will also prioritise palliative care resources to those with high palliative care needs while making sure they are not missed. A future randomized controlled trial will provide evidence on whether EMR-embedded flags lead to timely palliative care access. This approach holds promise to be more systematic and sustainable than relying on oncologists to routinely screen and refer to palliative care.
 

Funding
This research is funded by the Canadian Cancer Society (grant #707151). A parallel study in Ontario has been funded in part by a contribution from Health Canada, Health Care Policy and Strategies Program. The views expressed herein do not necessarily represent the views of Health Canada.