It sounds like a godsend for America's opioid epidemic: genetic tests that can predict how a patient will respond to narcotic painkillers, as well as an individual's risk of misuse, addiction, and potentially deadly side effects.

Proove Biosciences of Irvine, Calif., claims its "opioid response" and "opioid risk" tests are the only precision medicine tools on the market to do all that, giving doctors information "to guide opioid selection and dosage decisions as well as treat side effects."

But while the concept is captivating, addiction researchers say it is not yet possible to use genetic variation to gauge the risk of drug abuse. And ECRI Institute, a Plymouth Meeting nonprofit center that evaluates medical technology, says Proove has not published independently reviewed studies to support its claims.

"We can't say it doesn't work. All we can say is, there's no evidence it does," said ECRI research analyst Jeff Oristaglio.

In an interview, Proove CEO Brian Meshkin defended his five-year-old products, which he said retail for $1,000 a test and were used by about 400 doctors last year in treating 50,000 patients. He said he expects scientific journals to publish results from studies "within the next six months." Three clinical trials of the opioid response test are ongoing.

Consumers may assume that such high-tech genetic tests have to demonstrate safety and effectiveness to win regulatory approval, but they do not. Even though these complex diagnostics use the latest gene-sequencing and data-crunching techniques, they can come to market under 1988 federal regulations designed to ensure the quality of clinical laboratories.

Two years ago, the U.S. Food and Drug Administration proposed a new framework for overseeing "lab-developed tests" that would take into account their complexity and riskiness, because inaccurate or false results can harm patients. But the agency withdrew the controversial proposal after the November election, saying it needed "to continue to work with stakeholders, our new administration, and Congress."

Proove is one of many companies in the fast-growing genetic-susceptibility testing market, a multi-billion industry built on trying to foresee — and thus, forestall — disease, disability, and death.

Opioid-related deaths have become an urgent public-health crisis. Every day, on average, 3,900 people start abusing prescription painkillers, 580 graduate to cheaper heroin, and 78 die of a narcotic overdose, according to federal data.

In theory, genetics provides an opportunity to reduce this toll. Researchers have linked a predisposition to opioid dependency to gene variants involved in the brain's signaling of reward and pleasure. Addictive behavior, particularly alcohol abuse, is known to run in families.

But addiction experts say risky behavior involves the largely unpredictable interplay of environmental, cultural, and biological factors.

"It is hard to conceive of a genetic test or a genetic score that would be valuable as a predictor of opiate abuse or addiction in general," said Michael Vanyukov,  a University of Pittsburgh professor of pharmaceutical sciences, psychiatry and human genetics.

Vanyukov, who wasn't familiar with Proove's products, said heredity plays a relatively small role in determining variation in addiction risk, while choices and perceptions can play a big role. "If the individual is informed of, say, a 'low' risk score, this very piece of information will change the risk. The error of a genetic score is likely to be great, and reliance on it in practice may be dangerous."

Psychiatrist Charles O'Brien, founding director of the University of Pennsylvania's Center for Studies of Addiction, was also unfamiliar with Proove's test, but echoed that sentiment: "I could not in good conscience recommend that someone spend money on these tests."

O'Brien's own center recently identified gene variants associated with response to naltrexone, a drug that blocks the intoxicating effects of alcohol. But when the center studied alcoholics on naltrexone, strong and weak responders had the same number of heavy drinking days.

"We were very disappointed because we're all looking for precision medicine," O'Brien said.

Proove's tests analyze DNA from a cheek swab. The opioid risk test gives the patient a score — low, moderate, or high risk of opioid abuse — that is based on detecting variants in 12 genes, combined with clinical information such as a history of depression. The company's website claims the algorithm is 93 percent accurate.

But when ECRI scientists looked to validate that claim, all they could find were brief summaries of two studies that the company presented at medical conferences. One study of 290 patients compared the Proove risk test with the "opioid risk tool," a standard, one-minute screening questionnaire that doctors use to ask chronic pain patients about risk factors such as a history of mental illness or substance abuse.

"We cannot determine ... whether the test performs better or worse than the opioid risk tool in predicting opioid misuse," ECRI concluded.

Insurance plans either consider Proove's tests unnecessary or have no specific policies, ECRI found,  although Meshkin said insurers are covering the cost "on a case-by-case basis."

"At some point, you've got to stop and produce the evidence if you want people to pay," said Diane Robertson, director of ECRI's health technology assessment service. "Why would anyone want to use something if there is no evidence that it has benefit?"