Peers can work well, but only with training

Problem-solving skills make a big difference

Peer outreach workers have long been valued for their street-smarts and empathy. But not all peers are created equal, suggests a pair of studies that used peers to try to coax patients through six months of isoniazid prophylactic therapy (IPT).

"It’s easy to say peers will work, but you have to give some thought to who the peers are and what it is you’re trying to train them for," says Wafaa El-Sadr, MD, director of the infectious disease department at Harlem Hospital Center in New York City.

Peers drawn from former TB patients

In a two-year-old study still under way at Harlem Hospital, the peers are selected from a pool of former TB patients. Following the patients during their six- to 12-month course of TB therapy gives El-Sadr’s staff a good chance to see which patients are well-suited for the role of peer advisers, she says. Traits she looks for include "good communication skills, flexibility, a total devotion to the mission of TB control, and a willingness to share their experiences with others," she says.

Although the original plan was to provide the peers with training about TB, El-Sadr says she soon discovered that the selected peers already knew plenty and were equipped with a first-hand knowledge of symptoms, medications, and side effects — along with a thorough acquaintance with the workings of the local health care infrastructure.

El-Sadr provided the peers with training to equip them to handle the "tricky situations" patients provide and showed them how to fill out the requisite paperwork for the study.

To hold down the expense of providing twice-weekly observed therapy, the peers are asked instead simply to keep in touch with patients and to make sure they take their medication. In practice, that means they contact their patients at least once a week, El-Sadr says.

"It could be a phone call, or it could be getting together," she says. In some cases, it’s much more: One peer monitor, for example, had to intercede with a landlord who was about to evict his tenant. Then, because the tenant was mentally ill, the peer arranged appointments with a community clinic and with a social worker.

The patients — El-Sadr estimates there are between 70 and 80 enrolled in the study — are randomly assigned to one of two groups, with one group getting peer-monitoring and the other receiving "standard care," meaning they were simply told to show up at a clinic and collect their medicines.

So far, the results have been quite satisfactory, El-Sadr says — only one patient out of 35 has been lost to follow-up. He also says that using peers has been "very rewarding," she says.

The San Francisco story

In San Francisco, the story on peers is more dismal.

One arm of a pilot study conducted three years ago by researchers at the University of California, San Francisco (UCSF), used peer monitors to deliver directly observed preventive therapy to homeless patients. This time, the peers — all described as "sweet people" endowed with plenty of empathy and devoted to their mission — were simply taught to observe for side effects of medication and received no special training in problem-solving.

The results were not encouraging. Peer- monitored patients actually did worse than patients on standard care. Only 21% of the peer-monitored group completed IPT, compared with 25% of patients who finished on standard care, says Jacqueline Tulsky, MD, assistant clinical professor of medicine at UCSF.

Granted, Tulsky’s peers were called to work with some tough customers — the homeless. Yet in the down-and-out environment of the streets, peers are often "thought to be a good thing because they can move freely through [rough] neighborhoods," Tulsky adds.

What gets overlooked, Tulsky says, is how complicated and unpredictable life on the mean streets can be. "If you look at the statistics on drug use and rates of mental illness and alcoholism among the homeless, the numbers are very high. In some ways, they’re very complicated people."

It can be daunting for the peers even to regularly find the subjects and watch them take their pills, adds Jennifer Lorvick, TB project director of Urban Health Studies at UCSF. Typically, the peer "goes to the street corner where the person usually hangs out, but they’re not there," Lorvick says. "So he goes to the person’s house, and they’re not there, either. So he goes to where the person recycles aluminum cans, and they’re not there, either . . . then it’s back to the streets again."

But what really hampered the San Francisco peer monitors was their lack of problem-solving skills, Tulsky says. When patients failed to show up or complained that their medication was making them sick, the peers were baffled, she says. "They couldn’t think of what to say when the patient missed his appointment because he decided he had to go see his caseworker." When a patient stopped taking his meds because of perceived side effects, "they couldn’t figure out how to persuade the person to come into the clinic for evaluation."

In another study that’s a bigger sequel to this one, Tulsky didn’t even bother to include a peer-monitored element. "I wouldn’t like to say that peers would never work; I’d just say they just didn’t work for us," she concludes.