Requiring doctors to register with state database programs that track patients' opioid prescriptions reduced the amount of addictive painkillers Medicaid patients receive — and saved money, according to a new study.

Simply mandating that physicians register with the state databases led to a nearly 10 percent reduction in prescriptions for the most potent painkillers, the researchers reported in the April issue of Health Affairs, using data collected from 2011 to 2014.

New Jersey is among the states that require registration, although that change to the five-year-old program did not take effect until last November. Pennsylvania's current Prescription Drug Monitoring Program began in August, too late to be included in the study.

As the epidemic of opioid addiction and overdose has worsened over the last decade, many states have created or strengthened databases that providers and pharmacies can check to see if patients are getting prescriptions from multiple doctors. Doctor-shopping is a clear indicator that people are addicted — they have turned to antisocial or criminal behavior to find the drugs they need to avoid withdrawal symptoms — or are selling excess drugs illegally on the street.

Previous studies have generally shown that monitoring programs reduce opioid prescriptions, but the findings have been mixed. Some addiction experts also fear that if physicians stop patients' opioid prescriptions before they can get into treatment programs, they could end up buying painkillers or, worse, heroin on the street.

People on Medicaid, the joint state-federal program for the poor and disabled, have higher rates of illness, pain, and opioid use than the general population, the paper notes. More than 165 million opioid prescriptions were dispensed to Medicaid enrollees nationwide in 2014, accounting for over 7 percent of all prescription drugs paid for by Medicaid, according to the paper. Enrollees also are at least five times as likely to die of opioid-related overdoses than people with other forms of insurance.

Studying Medicaid also allowed the researchers, from Weill Cornell Medical College in New York and the University of Kentucky in Lexington, to analyze a vast trove of data to examine the impact of changes that different states made to their prescription-monitoring programs over several years.

The updates typically added mandatory registration to what had been voluntary programs, or mandated that doctors check the database when, for instance, prescribing an opioid to a patient for the first time.

Although rates of prescribing vary significantly between states, the researchers did determine that the number of prescriptions per person for schedule II opioids such as Vicodin, OxyContin, and Dilaudid was about 10 percent lower in states with any kind of database mandate. Medicaid spending on those drugs also was about 10 percent less than in states that lacked mandates.

When opioids that are most often used for cancer pain were analyzed separately, the mandates had no effect, suggesting that the required checking was, as intended, mainly influencing prescribing to patients whose need for the drugs was not so clear.

Perhaps the biggest surprise was that mandatory checking of the database — a provision to which physician organizations frequently object — was no more effective than simply requiring providers to register. (New Jersey and Pennsylvania now require both.)

The study authors speculated that this was because the first step, mandatory registration, was the bigger hurdle.

"In addition," they wrote, requiring registration "might have further raised prescribers' awareness of misuse and abuse of controlled substances among their patients, leading to subsequent changes in prescribing practices."