Any death that comes at the hands of another person is a homicide, the exception being most motor vehicle deaths. Knowing that Brown was injected by someone else makes the determination easy, Fowler says. But his doctors usually aren't supplied that crucial detail from investigators.
Heroin is the most common cause of overdose death in Baltimore, representing 75 percent of such cases, though the overall number has been on the decline in recent years. In 1999, 283 people died from heroin overdoses; that total declined to 118 in 2009, according to the city health department.
Drug dealers themselves have in some instances been charged with selling drugs that led to an overdose; in Washington State, a drug dealer was convicted of "controlled-substance homicide" for selling black tar heroin to a 24-year-old man who later overdosed, while in January a Virginia man was charged after a two-year investigation with felony homicide for selling Opana, a powerful narcotic, to a woman who later died.
Federal statutes also allow prosecutors to seek sentencing enhancements against convicted drug dealers if it can be proved that a buyer died from drugs the defendant sold to them. Prosecutors in Maryland have secured such enhancements — leading to a mandatory minimum of 20 years in prison without parole — in three recent cases.
But the lack of insight into many deaths leads to a high number of undetermined deaths here, officials say. A 2006 study by the Centers for Disease Control showed that Maryland, followed by Massachusetts and Rhode Island, had the highest number of undetermined deaths in the country between 1999 and 2003. The states share the same methodology, which Fowler calls "intellectually honest."
While that study's findings have been used by critics to raise questions about the number of deaths, the study's author, Matthew Breiding, said in an interview last week that in other states, these undetermined deaths would likely have been called "unintentional." He said his study did not find that Maryland's approach drove suicides down, and it didn't look at the effect on homicides, though it does cite research from California that many "undetermined" infant deaths appeared to be "unrecognized homicides."
In 2009, the most recent year for which statistics from Fowler's office are available, Maryland's medical examiners conducted 3,856 autopsies, listing 19 percent as undetermined, compared with 21 percent accidental and 8 percent suicide.
North Carolina reported 4.5 percent of its autopsied cases as "undetermined," with 34 percent ruled accidental and 8.4 percent suicides. In Virginia, 28 percent of cases were accidental and 13 percent were suicides, with just 5 percent undetermined.
Leah Bush, the chief medical examiner for Virginia, said her office tries to use other available information to fill in the gaps when they lack details about the circumstances of a death. For example, if the person has a long history of drug abuse or "drug seeking behavior but no history of suicide, then we call it an accident," she said in an e-mail.
If the drug levels in the person's body are "very high such that the … the usual recreational user cannot possibly survive, then we call it a suicide because the intent is obvious from the very high drug levels," Bush said. To Fowler, those are assumptions.
City prosecutors won't say where their investigation into Brown's death stands, or whether anyone could be charged. Calling it an "uncommon situation," spokesman Mark Cheshire said it is "very much under investigation."