This article was produced in partnership with the Malheur Enterprise, which is a member of the ProPublica Local Reporting Network.

Editor’s Note: This story updates a Nov. 14 report by ProPublica and the Malheur Enterprise that was based on data provided by Oregon’s Psychiatric Security Review Board that turned out to be incomplete. The board, in response to a series of public records requests, turned over the names of 334 people it said it had released from 2008 to 2017. After publication, the board informed the Enterprise that it had actually freed 526 people found not guilty by reason of insanity for felonies during those years. What follows is a revised version that includes analysis of what happened to all 526 people after their release. The earlier story remains online and can be read here.

Studies of the criminal justice system rarely consider what happens after insanity defendants are freed. In Oregon, that has meant state leaders have had no objective measure of the state’s approach, which places more people into community treatment and frees them from state supervision more quickly than other states.

The Malheur Enterprise and ProPublica calculated outcomes by reviewing public records collected from the Psychiatric Security Review Board, courts and state agencies. The analysis is among a handful of recidivism studies of insanity defendants, experts said.

Reporters searched available electronic court records for all 526 people freed by the review board from 2008-17 who had been found guilty except for insanity in felony cases. Because digital court records in Oregon are incomplete before 2016 and because not all states were searchable online, the analysis likely understates the number of new crimes committed.

People’s identities were confirmed by birthdate or matching the original insanity case to a person later charged with a new crime who shared the same state offender ID. Often, both were available to match.

We then looked at what happened during the three-year period after people were released, the standard approach among scholars. Of the 526 people freed between 2008-17, that period could be examined for 418 of them.

About 28 percent, or 118 people, were criminally charged again within three years. An additional 69 people committed new crimes after that period or had not yet been free three years. Most often, people freed after being found not guilty by reason of insanity were back in criminal court in little more than a year. Many appear to have become homeless shortly after leaving the state’s supervision, even though housing is identified as a primary need when they are discharged.

They were charged with felonies more often than people freed after serving prison terms — 23 percent compared with 16 percent within three years — according to our analysis and the Oregon Department of Corrections. Experts said research studies find little evidence that mentally ill people are more likely to commit violent crimes and other factors tend to drive that behavior.

One researcher studied rearrest rates among Oregon insanity defendants in 1986. The analysis, published in the International Journal of Law and Psychiatry, was done by Dr. Joseph Bloom, a forensic psychiatrist who once served as dean of medicine at Oregon Health & Science University and has been president of the American Academy of Psychiatry and the Law.

His results paralleled those found by the Enterprise-ProPublica analysis.

Bloom searched for criminal records in an Oregon police database that is not available to the public but that is regularly used by the review board to monitor clients. In 1982, he tallied the new arrests of 144 people freed from state supervision from 1978-80. He found that 42 percent had been arrested within 18 months and that 16 percent were arrested for felonies.

A more recent study looked at Connecticut, which is one of three states with a Psychiatric Security Review Board similar to Oregon’s.

Dr. Michael Norko and five other authors tracked the recidivism of 215 people freed by that state’s board between 1985 and 2015. They found that 16 percent were later re-arrested, with about a third of those being for felony crimes, according to the article published in the Behavioral Sciences and the Law.

“Given the significant commitment of resources in the state devoted to the PSRB’s supervision, monitoring, and community support of acquittees, these results have important policy and public safety implications,” the authors wrote.

Jayme Fraser is a reporter for the Malheur Enterprise. She’s spending the year examining the way states manage people with mental illnesses who commit violent crimes — in Oregon and nationally. Are there problems with the way your state handles people who are violent and have mental illnesses? Email her at jayme@malheurenterprise.com and follow her on Twitter @JaymeKFraser.