Cancer Watch

Symptom Monitoring Improves Cancer Care

Cancer patients talking to doctors doesn’t work very well: Half of all symptoms go unreported. Computerized systems and email prompts may do a better job.


Peter Wehrwein

If you want to understand cancer treatment from the patient’s point of view, Wit is a good place to start. You can watch the first-rate HBO version of the one-act play by Margaret Edson online for free. Emma Thompson plays Vivian Bearing, an English literature professor and expert on John Donne and metaphysical poetry who has stage IV ovarian cancer; “published and perished,” remarks Thompson. The title has several meanings, but what would otherwise be an utterly dreary and depressing story is elevated by Bearing’s delightful dark humor—her wit. One of her wry observations is about the “How are you feeling today?” greeting she gets from doctors and nurses. “I have been asked ‘How are you feeling today’ when I was throwing up into a plastic washbasin,” says Bearing. “I was asked when I was emerging from a four-hour operation with a tube in every orifice.”

Cancer specialists say they want to replace the empty, reflexive “how are you feeling today?” question with meaningful concern for—and data collection from—patients in the throes of cancer treatment (“My treatment imperils my health—and herein lies the paradox,” says Professor Bearing). One aspect of this intention to be more heedful of patients goes by the name “symptom monitoring.” That doesn’t sound very caring, creative, or important, but it changes how cancer is treated and how the treatment is judged. In health care today, nothing seems to gain traction unless it is a) measured and b) can be tied to payment. Symptom monitoring is well suited for both.

“It’s just getting started, but we’re doing some work in defining measurement tools that more accurately reflect the patient voice, the patient report of symptoms, and patient experience overall,” says Robert S. Miller, MD, the senior director of quality and guidelines for the American Society of Clinical Oncology (ASCO) and medical director of CancerLinQ, ASCO’s big data project.

Robert S. Miller, MD

It may take several years for the full complement of performance measures of the patient experience to be ready to use, says Robert S. Miller, MD, of the American Society of Clinical Oncology.

Miller says it may take several years for a full complement of performance measures of the patient experience, including symptom monitoring, to be researched, analyzed, and made ready for use. He definitely has a foot on the brake: ASCO, says Miller, doesn’t want to lend its imprimatur to questions of patients that later need to be reworked—or junked entirely—or settle on a platform (PCs, say) that turns out to be based on faulty assumptions and not the best choice for collecting this type of information. “We want to get this right,” he says, the implication being, of course, that getting it wrong is a distinct possibility.

Yet Miller also notes that ASCO may put some decisions about patient experience measurement and symptom monitoring on a faster timetable perhaps in the next year, as part of its push to improve the quality of care. And the pressure is on to come up with the metrics that underlie quality of care and value-based payment that rewards it. Miller references the HHS announcement early in 2015 that set a goal for this year of having 30% of traditional Medicare fee-for-service payments paid through alternative payment models such as ACOs and bundled payments and linking 85% of the payments to some kind of measure of quality or value. Incorporation of performance measurements that are supposed to reflect the patient experience “is happening and it’s only going to increase,” says Miller. “ASCO wants to be active in this because we want to say to our members, ‘We want to work with you to create the best metrics that make a difference, that really resonate with patients, that give our members the answers they need to help us direct their care better.’”

Automation helps

That sounds nice—and maybe a bit too aspirational and abstract. Taking a look at Ethan Basch’s research lends some nuts and bolts to what Miller is talking about. Basch, MD, is director of the cancer outcomes research program at the University of North Carolina’s Lineberger Comprehensive Cancer Center and a leading expert on symptom monitoring. Basch says symptom monitoring is the cornerstone of quality cancer care—“symptoms related to cancer. Symptoms related to the treatments we give.”

Ethan Basch, MD

Symptom monitoring is the cornerstone of quality cancer care, says Ethan Basch, MD, of the University of North Carolina’s Lineberger Comprehensive Cancer Center.

It’s not that oncologists are oblivious to their patient’s suffering or cavalier about symptoms (although some might be). But traditionally, symptoms have come up only during the in-person patient visit. Often it falls to the patient to mention problems with symptoms, and the clinician, understandably, may be focused on the details of the patient’s treatment. According to Basch, studies have shown that about half of cancer patient symptoms go unreported.

In the kind of twist that we’ve gotten used to, an impersonal, computerized system for collecting information on symptoms may do a far better job than the in-person encounter. In February, Basch and his colleagues reported the results of a study in the Journal of Clinical Oncology that showed an association between a computerized system for monitoring symptoms and an assortment of desirable outcomes, including some that would attract the attention of payers: better quality of life, fewer visits to the emergency department, fewer hospitalizations, and longer time on chemotherapy. Basch conducted the study among 766 patients at Memorial Sloan Kettering Cancer Center. The intervention used the web-based Symptom Tracking and Reporting (STAR) system, which asks questions based on a National Cancer Institute list of 12 symptoms commonly experienced during chemotherapy, including nausea, vomiting, pain, and constipation. Patients graded their symptoms on a 5-point scale, ranging from 0 (not present) to 4 (disabling).

The intervention had three interesting elements that may shape such programs once they leave the cocoon of the research project. First, and most fundamentally, patients reported the symptoms themselves by tapping on a computer screen rather than asking questions. Second, patients who were experienced with computers could receive email prompts to report their symptoms. Third, nurses were notified if patient-reported symptoms worsened by 2 points or more, or reached the absolute score of 3 or more. Automation is key, says Basch: “Rather than having 1,000 patients call the clinic on a Friday afternoon you develop a system that will not only prompt patients but also sort out things that warrant a nurse’s attention.”

The difference the symptom monitoring intervention made in emergency room visits (34% vs. 41%) and hospitalizations (45% vs. 49%) weren’t stunning. But consider the costs, and they start to look pretty good. Basch didn’t crunch the numbers for this study but he says they are modest. His take: “The differences are greater than the drugs that are getting approved and that we are spending $10,000 to $12,000 per month on.”

Basch has two grant proposals in the works to the Patient-Centered Outcomes Research Institute to test some of his findings in a larger national study. ASCO and several other organizations are working with him.

Window of opportunity

Using computers to have patients self-report their symptoms—it’s pretty basic and not that new. The STAR system has been around awhile. The recruitment of Memorial Sloan Kettering patients into Basch’s study started in 2007. But Miller, at ASCO, says interest in systematically collecting data on the patient experience and symptoms is escalating for a confluence of reasons: the democratization of medical information, the smartphone and the ease with which symptoms can be reported and the data analyzed (in stark contrast to the paper survey), and the expense of today’s oncology drugs. “The question now that’s increasingly being asked: What is their value?” says Miller. “So, I think we’re finally at the point where we recognize that the patient’s perception of value as characterized by their own physical and emotional reactions to treatment or to cancer is probably the most important question, rather than what their doctor thinks is happening to them.”

Vivian Bearing might agree.