With about 80,000 Americans dying each year, excessive alcohol use remains the third most-preventable cause of death in the United States, topped only by smoking and obesity.

Alcohol remains a stubborn killer of people in their prime. The tragedy is propagated over generations, through poverty, violence, broken families and harm to fetal brains. The consequences of excessive alcohol use are also a disaster in crass economic terms.

At a quarter trillion dollars a year, alcohol-use disorders are some of the most neglected and mismanaged medical conditions.

A comparison with nicotine addiction makes this abundantly clear. Much work remains, but sustained policy efforts have reduced U.S. smoking rates from about 40 percent a few decades ago to about 20 percent.

For remaining smokers, scientifically supported treatments have become widely available. Success rates of behavioral interventions are modest but over the years have been supplemented by increasingly effective medications approved by the Food and Drug Administration (FDA).

The situation is fundamentally different when it comes to alcohol-use disorders. Only about one in 10 people with alcoholism ever receives treatment. For those who do, treatment in the United States is almost synonymous with joining Alcoholics Anonymous (AA).

AA was once critical for advancing a view of alcoholism as a disease rather than a moral defect, and it created an admirable fellowship of people willing to support each other. But AA was formed three-quarters of a century ago. At the time, medicine had little to offer alcoholics beyond treating the shakes of acute withdrawal. Much has happened since.

We now know that the effects of behavioral treatments for alcoholism, including AA attendance, are modest. A rigorous academic analysis by the Cochrane Collaboration states that "available . . . studies did not demonstrate the effectiveness of AA or other 12-step approaches in reducing alcohol use and achieving abstinence compared with other treatments."

AA's own surveys have indicated that for every 100 alcoholics entering their first meeting, only about 30 will be attending and sober a year later. This is very close to the spontaneous relapse rates consistently found by research over the past four decades.

With this in mind, one would think that the arrival of relapse-preventive medications would be greeted with great enthusiasm, but nothing could be further from the truth. The first medication in this class, naltrexone, was approved by the FDA in 1994. It is by no means a panacea, but clinical research has established that it reduces the risk of heavy drinking by about 20 percent.

Naltrexone is in fact twice as effective in patients with a particular genetic makeup, showing that patients with alcoholism differ in important ways and allowing a personalized medicine approach to treatment. It is also safe, well-tolerated and cheap. Another medication, acamprosate, also has FDA approval for alcoholism, while several additional drugs look promising in well-designed academic studies.

However, less than 10 percent of patients with alcoholism receive a prescription for any alcoholism medication. This is largely because medications targeting brain function continue to be viewed unfavorably in many 12-step programs. This way, patients lose out on the benefits of treatment, while the pharmaceutical industry gets a clear message to stay away from investing in alcoholism therapies, despite extensive unmet patient needs.

Twelve-step programs are diverse, and many are adopting the medical evidence. But a broadcast of "Larry King Live" from 2007 shows that something fundamental needs fixing. Thirteen years after the FDA had determined that naltrexone is beneficial in treating alcoholism, with additional studies showing the same, Susan Ford Bales, then-chair of the Betty Ford Center, dismissed anti-craving medications upon being confronted with these data.

"We do not use relapse preventative medications at the Betty Ford Center," she said. "We will look at them once there are some statistics and that sort of information behind them."There is no other area of medicine where disregarding easily available evidence this way would be tolerated.

Almost equally problematic is the uncompromising AA tenet that "once an alcoholic, always an alcoholic," and the implication that abstinence is the only worthwhile treatment outcome. It is true that many alcoholics remain highly susceptible to relapse for life, and abstinence is always the safest bet.

But studies also show that others are able to return to social alcohol use. And many patients are not ready to pursue abstinence. Attempting to impose on them a treatment goal they are not ready for raises unnecessary barriers to treatment. It probably contributes to the low proportion of people with alcohol problems who seek treatment.

Earlier this year, the European Medicines Agency approved nalmefene for doctors to give to patients who were not ready to attempt abstinence. This "as-needed" treatment means taking a pill prior to situations associated with high relapse risk, such as a party.

This approach was clearly able to reduce drinking and improve the health of patients. But it would currently not meet with approval by the FDA because of prevailing U.S. dogma that abstinence is all that counts.

The system is broken. It must be fixed before too many more people die. If AA founders Bill Wilson and Bob Smith were alive today, I am certain they would enthusiastically pursue opportunities for treating alcoholism through research and medication.

So should we.

- - -

Markus Heilig is a physician scientist at the National Institutes of Health. The views expressed here are his own.