Lay Health Workers for End-of-Life Talks in Advanced Cancer

Roxanne Nelson, BSN, RN

July 31, 2018

Patients with advanced cancer may be more comfortable discussing their treatment goals and care preferences with a trained nonclinical worker, authors of a new study conclude.

They found that compared with patients who received usual care, those who spoke with a lay healthcare worker about their personal goals for care were more likely to talk with their oncologist about their preferences, reported higher satisfaction with their care, had higher rates of hospice use, had lower acute care use, and had lower total healthcare costs in the last month of life.

There was no negative effect on survival.

The study was published July 26 in JAMA Oncology.

Lead author Manali I. Patel, MD, MPH, MS, assistant professor at the Division of Oncology at Stanford University School of Medicine, California, commented that medical staff was less than enthusiastic about the study.

"Initially, prior to conducting the study, we found significant resistance among our nurses, social workers, and our oncologists," she told Medscape Medical News.

Her colleagues (oncology providers) and most of the staff were very hesitant about the proposal that a lay health worker could successfully assist in addressing goals of care and advocating for patients to communicate their wishes. "Given this feedback, we designed the study such that the lay health worker would have no direct communication with the oncologists, as many of them were opposed to this approach, but rather the intervention would primarily be conducted in a patient-focused intervention," Patel said.

Oncology providers agreed to this approach and agreed to testing it in a rigorous study. "They would be blinded and  we could obtain evidence to prove us wrong — that lay health workers could indeed conduct these services," she noted.

However, several months into the study, some of the oncology providers requested that their patients be enrolled in the intervention and specifically requested that the lay health worker talk to certain patients outside of the trial. "Similarly, our clinic social workers, behavioral medicine practitioners, nurses, and other staff members proactively contacted our lay health worker in hopes that the lay health worker would conduct the intervention on their patients who were not enrolled in the study," Patel added.  "Our facility is considering expanding this approach to other serious illnesses as well."

Satisfaction Better

In cancer care, lay health workers already assist in cancer screening and treatment adherence but are not commonly integrated into end-of-life care.

The current study was based on prior research that Patel had conducted when she was a fellow at Stanford's Clinical Excellence Research Center, when patients had expressed a preference for having this type of discussion with nonclinical personnel.

The  study involved  213 patients with stage 3 or 4 solid tumors. The mean age was 69.3 years; most (99.1%) were male, and most (77.5%) were of non-Hispanic white race/ethnicity. Half of the patients in the study  were randomly assigned to the intervention group, which was a 6-month program with a lay health worker.

 All the patients in the study received usual care, which was provided by the oncology team and included a social worker who arranged housing and transportation, and mental health counseling by a behavioral medicine practitioner.

The objective was to achieve goals-of-care documentation for at least 75% of patients, and secondary outcomes included advance directive documentation, patient satisfaction, healthcare use, and costs.

Patient satisfaction was measured at baseline and at 6 months after randomization by using the "satisfaction with provider" item of the validated Consumer Assessment of Healthcare Providers and Systems General Survey (scores on a scale of 0 to 10, where 0 is the worst and 10 is the best possible satisfaction). Satisfaction with decision making was assessed within 4 to 6 months of randomization by using the validated 6-item Satisfaction With Decision Scale (scale ranging from 0 to 5).

The authors note that a "clinically meaningful difference in the Satisfaction With Decision scale is 0.5 and in the satisfaction with provider scale is 0.9."

Most patients (n = 97 [92.4%]) in the intervention group had documented their goals of care in the electronic health record within 6 months after randomization, exceeding the primary objective of 75%.

Patients in the intervention group also scored significantly higher than those in the usual care group in terms of having documented goals of care (at 6 months, 97 of 105 [92.4%] vs 19 of 108 [17.6%]; P < .001; at 15 months, 98 of 105 [93.3%] vs 26 of 108 [24.1%]; P < .001), as well as having advance directives (6 months, 71 of 105 [67.6%] vs 28 of 108 [25.9%]; P < .001; 15 months, 78 of 105 [74.3%] vs 36 of 108 [33.3%]; P < .001).

Satisfaction with decision making was also higher in the intervention group compared with the controls, and the intervention groups had higher scores across all domains of satisfaction with decision and reported greater satisfaction with their oncology provider (mean [standard deviation] score, 9.16 [1.44] vs 7.83 [2.36]; P < .001). The change in mean satisfaction from baseline to 6 months between the groups was significantly different (difference-in-difference, 1.53; 95% confidence interval, 0.67 - 2.41; P < .001).

Cost Significantly Reduced at End of Life

Finally, the authors looked at healthcare cost and use. Rates of intensive cancer therapies did not differ between groups at 6 months, but patients in the intervention group were more likely to have used hospice within the 6- and 15-month timeframes after randomization (at 6 months, 37 of 105 [35.2%] vs 20 of 108 [18.5%]; P = .006; at 15 months, 47 of 105 [44.8%] vs 30 of 108 [27.8%]; P = .009).

Patients in the control group had higher rates of hospitalization and emergency department (ED) visits at 15 months, but the differences were not statistically significant. In addition, total health costs in the intervention group were lower during this 15-month period, but the difference between groups was not statistically significant (median, $86,025 vs $111,958; P = .08).

However, the differences in cost and use of healthcare services were far more pronounced at the end of life.

Although mortality rates were similar between groups, patients in the intervention group who died had fewer ED visits (mean ED visits per patient, 0.05 vs 0.60; P < .001) and fewer hospitalizations (mean hospitalizations per patient, 0.05 vs 0.5; P < .001). They were also more likely to receive hospice care (46 [76.7%] vs 29 [48.3%]; P = .002) and accrued dramatically lower total healthcare expenditures within 30 days of death (median, $1048 vs $23,482; P < .001) compared with the control group.

This Intervention "Fills a Gap"

Approached by Medscape Medical News for an independent comment, Todd Barrett, MD, a palliative medicine physician at the Ohio State University Wexner Medical Center, Columbus, noted that some of the resistance to use of lay health workers may have been related to the fear that it would somehow alter the doctor-patient relationship.

"The cultural situation in the US at present is that we avoid discussing end of life," he said, "Or denying end of life. But in this situation we see that a nonclinical person can push us along that path."

The study highlights the fact that discussion of end-of-life care doesn't have to be initiated by a medical professional, and Barrett pointed out that clergy have also had these discussions with patients.

"Rather than interfering with the doctor-patient relationship, I see this as complementary," Barrett said. "The lay workers have helped move the conversation forward, and then the oncologist and palliative care physicians are now able to interpret those goals and values and provide a medical recommendation."

Barrett believes that this is a "huge step forward" and certainly fills a gap — there is an acute shortage of palliative care physicians, and time is often at a premium for oncologists and primary care providers. "There often isn't enough time to initiate these conversations, or to give these discussions the amount of time that they need," he explained.

A main limitation of the study was the population, which comprised mostly older white men. "So we don't know what how generalizable this will be with other, more diverse populations," Barrett said, "as there are cultural and ethnic variations when it comes to end-of-life issues."

This work was supported by a grant from the Department of Veterans Affairs Office of Patient-Centered Care and Cultural Transformation. Patel was supported by a National Institutes of Health provided by the Stanford Clinical and Translational Science Award and by the California Healthcare Foundation. The authors and Barrett have disclosed no relevant financial relationships.

JAMA Oncol. Published July 26, 2018. Abstract

For more news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: