WASHINGTON — The career of military lab analyst Phillip Mills started unraveling the day a colleague made a discovery that would rattle military justice.
On April 28, 2005, Dr. Timothy Kalafut was reviewing Mills' analysis of DNA samples when he noticed something strange. The sample ID numbers looked odd.
"Literally, in the first 10 seconds, I looked at the case file and knew that there was something fishy going on," Kalafut later testified.
Kalafut and other colleagues began investigating and found that Mills had signed off on work he never completed.
Mills later admitted falsifying the test run report at the U.S. Army Criminal Investigation Laboratory outside Atlanta, but he defended his overall record. He called the incident a "case of bad judgment," one unique in his three decades of law enforcement forensics work.
The discovery, however, prompted a review of hundreds of his cases and revealed serious problems. It also, slowly, forced lab officials to reconsider Phillip Randolph Mills himself.
"The real Phillip Mills, as we now know, was not as he advertised himself to be," Navy Lt. Michael Torrisi subsequently declared in a June 2010 legal brief written to attack Mills' credibility. "The real Mills displayed incompetence, deceit and sloth."
Mills remains a mystery. He couldn't be reached over the course of many weeks, and he didn't respond to two letters sent by certified mail and Federal Express.
Numerous interviews, internal investigations and thousands of pages of court documents reveal a complicated man. Mills was a hard worker who sometimes cut corners. He was a technical expert who didn't always know his science. He swore to tell the truth, but he didn't always uphold his oath.
Mills, who's now 65, resigned shortly after being told in November 2005 that he'd be fired, according to court records. The holder of a master's in forensic science from the John Jay College of Criminal Justice, he's defended his reputation.
"I have never even thought about manipulating data to affect the outcome of evidence," Mills wrote in a June 2, 2005, memo.
His one-time colleagues have agreed.
"I still think he's a person with integrity in that I don't think he would have ever intentionally switched or contaminated a sample to provide a false result in a case," lab official Thomas Overson testified at a court hearing.
The 2005 discovery nonetheless led to an effort to re-examine Mills' work. Delays and previous destruction of evidence often thwarted the study, which was completed Sept. 30, 2008, but the final results were alarming. Of 465 cases reviewed, investigators found problems in 118 of them.
The review found that "in general ... Mills did not follow good scientific practice or lab protocols and did not properly document his work." He was "not thorough." He consumed too much evidence in his testing, making retests impossible. He missed potential clues and sometimes falsely documented his work.
Army officials say that these problems were confined to Mills, adding in a statement that the lab "immediately took corrective action regarding that employee and his work."
Still, the withering review of Mills' work called into question the results of courts-martial and marked a sad professional end for a man who'd built his life around the Army Criminal Investigation Laboratory.
The lab "represents my entire forensic career," Mills reminded his supervisors in the June 2005 memo, "and I will never intentionally do anything illegal to damage the reputation of the DNA Division" or the lab.
From the start of his military career in the mid-1970s, Mills was a hardworking and physically courageous soldier, graduating as a paratrooper from the punishing Army Airborne School at the age of 33. He earned medals and high marks, serving in Japan before being assigned in 1995 to the criminal investigation lab at Fort Gillem.
Mills contributed to innumerable courts-martial. From 1995 to 2005, he handled evidence in 253 Army cases, 117 Navy and Marine Corps cases and 95 Air Force cases.
"I have done hundreds and hundreds of rape DNA-type investigations," Mills testified in one 2002 Navy court-martial. "I've handled a whole bunch of condoms."
Mills could be an effective witness with an impressive technical resume, though sometimes there was confusion about his training.
In a 1999 Navy court-martial, an attorney asked Mills to describe his educational background.
"I have a bachelor's degree in chemistry from Catholic University," Mills replied, the trial transcript shows.
Mills' degree, though, came not from the better-known Catholic University of America but from the lesser-known Pontifical Catholic University of Puerto Rico. A Justice Department attorney twice called Mills, inaccurately, "Dr. Mills" in a legal brief filed in February 2009. This mistake was the government attorney's, not Mills', but it shows the stature that witnesses with scientific expertise are accorded.
Mills' supervisors, at any rate, rated his work ethic highly. He showed up early and stayed late, processing evidence submitted by military investigators.
"Phillip R. Mills is the most dedicated and industrious soldier that I ever have had the privilege to supervise," his supervisor, Clement Smetana, wrote in one job evaluation. "Mills, is without exception, the finest officer I have ever known."
The cases Mills worked on exposed the military's criminal underside.
For instance, he tested evidence in the 2002 murder case of Army Lt. Col. David Bartlett, who pleaded guilty to killing his wife at the Army War College in Carlisle, Pa.
Mills also examined clothing taken as evidence for the 2002 murder trial of Army Privates First Class Jonathan Schroeder and Andrew Humiston, charged with beating a fellow 1st Armored Division soldier to death in Germany.
Mills later was found to have missed some bloodstains in that case. His oversight apparently didn't matter, as the two young soldiers already had pleaded guilty. In hindsight, though, it called into question whether Mills valued speed over quality.
"His better-than-average productivity," now-retired Penn State forensic scientist Robert Shaler said of Mills, in a review the lab commissioned, "was due, probably in part, to ... his desire to be recognized as the most productive analyst in the lab."
After retiring from the Army in 1997 as a chief warrant officer, Mills returned to the Army lab as a civilian in early 1998. Over the next few years, even as he told supervisors he'd pick up the pace, he struggled to keep up with the science.
In November 2002, Mills failed an annual test in hair-analysis proficiency, though he wasn't removed from hair-analysis cases until September 2003. He continued working on other cases.
Mills' problems became apparent again on Dec. 23, 2003, when another lab analyst discovered that Mills had allowed cross-contamination of samples from different cases.
"We're dealing with over 100 (test) tubes, and one may have gotten switched or cross-contaminated ... because there were so many tubes involved," Mills explained in a 2006 court hearing.
Mills' supervisor suspended him. The unusually long suspension finally was lifted in September 2004 even as some colleagues continued to question Mills' performance.
Seven months later, Dr. Timothy Kalafut opened one of Mills' cases and started a review that would turn the lab upside down.
MORE FROM MCCLATCHY
Follow the latest legal affairs news at McClatchy's Suits & Sentences blog