Euthanasia, whether in a medical setting (hospital, clinic, hospice) or not (at home) is often erroneously described as "mercy killing". Most forms of euthanasia are, indeed, motivated by (some say: misplaced) mercy. Not so others. In Greek, "eu" means both "well" and "easy" and "Thanatos" is death.
Euthanasia is the intentional premature termination of another person's life either by direct intervention (active euthanasia) or by withholding life-prolonging measures and resources (passive euthanasia), either at the express or implied request of that person (voluntary euthanasia), or in the absence of such approval (non-voluntary euthanasia). Involuntary euthanasia - where the individual wishes to go on living - is an euphemism for murder.
To my mind, passive euthanasia is immoral. The abrupt withdrawal of medical treatment, feeding, and hydration results in a slow and (potentially) torturous death. It took Terri Schiavo 13 days to die, when her tubes were withdrawn in the last two weeks of March 2005. Since it is impossible to conclusively prove that patients in PVS (Persistent Vegetative State) do not suffer pain, it is morally wrong to subject them to such potential gratuitous suffering. Even animals should be treated better. Moreover, passive euthanasia allows us to evade personal responsibility for the patient's death. In active euthanasia, the relationship between the act (of administering a lethal medication, for instance) and its consequences is direct and unambiguous.
Still, this sophistry obscures the main issue:
If the sanctity of life is a supreme and overriding value ("basic good"), it ought to surely preclude and proscribe all acts and omissions which may shorten it, even when the shortening of life is a mere deleterious side effect.
Moreover, the apparent dilemma between the two values (reduce suffering or preserve life) is non-existent.
There are four possible situations. Imagine a patient writhing with insufferable pain.
1. The patient's life is not at risk if she is not medicated with painkillers (she risks dying if she is medicated)
2. The patient's life is not at risk either way, medicated or not
3. The patient's life is at risk either way, medicated or not
4. The patient's life is at risk if she is not medicated with painkillers
In all four cases, the decisions our doctor has to make are ethically clear cut. He should administer pain-alleviating drugs, except when the patient risks dying (in 1 above). The (possible) shortening of the patient's life (which is guesswork, at best) is immaterial.