Homicide is a metric that links a behavior (a violent act) to an outcome (a death). Modern medicine has drastically increased the survival rates for serious injuries in the past several decades. Since death from a given injury has become less common, especially in urban areas close to high-tech hospitals, metrics attached to death - such as homicide - will drop even if there is no change in the frequency of the kind of violent assaults that would have been murders if only 1931 technology were available. Since those kinds of violent attacks have increased dramatically, it appears that using the homicide rate as a metric allows improvements in medical technology to mask a major increase in violence in recent decades.graphic).
Keep that in mind when you consider that American suicide rates have been mostly flat since 1950 - not decreasing, like homocide rates, even though medical technology and injury survivability has vastly improved. In fact, suicide rates have been increasing since 1999. Suicides in the 35-64 age group increased by 28% between 1999 and 2010.
Suicide, like homicide, is a metric that links a behavior (a self-injury) to an outcome (a death). Like homicide, the reduction in death for a given injury should reduce the suicide rate even if self-injuries are constant. We do not see the suicide rate decreasing; in recent years, even as the homicide rate continues to drop, the suicide rate is increasing. Medical technology may be masking an even greater rise in suicidal behavior than the completed suicide rate would indicate.
Case-fatality rates are three to four times higher for self-inflicted gunshots than for gunshots inflicted intentionally by others; that is, suicide attempts by gunshot are more lethal than assaultive shootings. In 2007, 21% of intentional gunshot wounds inflicted by someone other than the victim were lethal; 80% of self-inflicted gunshot wounds were lethal. Gunshot is the most common method of suicide in the United States, accounting for 54% of suicides.
However, almost half of completed suicides have used methods other than firearms - and methods other than gunshot are only 10-15% likely to be fatal. Since 20% of self-inflicted gunshot wounds are currently nonfatal, the firearm suicide rate would "only" rise by about 25% of its current level if modern medicine were not aggresively saving lives that are unwanted by their possessors. But in the absence of modern medicine, up to ten times as many people who poison, cut, hang, or suffocate themselves might succeed in killing themselves.
Proximity to a hospital is a major factor in trauma survival. This may partly explain why rural suicide rates have leapt ahead of urban rates in recent decades: suicidal behavior may be similar, but rural people die from their injuries more frequently than urbanites.
Note: St. Rev hates this graphic as the graphs for weapon injury are of radically different scales (inter alia) and indeed it is probably seriously flawed but it provides a cheap and tasty way of immediately visualizing the phenomenon.