Table 3 Experts' ratings: Choice of antidepressant

Mirtazapine

Round 1: Median 9.0 (7-10); n = 11

Round 2: Median 9.0 (7-10); n = 11

-Some data, good profile regarding side effects, at this time most used

-Seems well tolerated

-When side effects on sleep and appetite are handy

-Too soft

-Has useful sedative properties and well tolerated

-My first-line - evidence is that it is more effective than SSRIs and appetite stimulant and sedative qualities are useful side effects

Citalopram

Round 1: Median 9.0 (5-10); n = 11

Round 2: Median 9.0 (5-10); n = 13

-Side effects due to its antIHJJistaminic action

Venlafaxine

Round 1: Median 7.5 (0-10); n = 8

Round 2: Median 8.0 (0-9); n = 13

-Less side effects than tricyclics

-Useful if patient needs help for sedation

-If depression is associated with chronic high level of anxiety

-Not straightforward - discontinuation symptoms common. Cardiac toxicity

Sertraline

Round 1: Median 7.0 (2-10); n = 11

Round 2: Median 7.5 (0-10); n = 12

-Favourable side effect profile

-Interaction with dopamine system could give different and more side effects than other SSRIs

-Least adverse effects, presently best available

Escitalopram

Round 1: Median 8.0 (0-10); n = 10

Round 2: Median 7.0 (0-10); n = 13

-Not cost effective alternative to citalopram

-Better tolerated than R-citalopram and less interaction with cytocrome 450 2D6

Amitriptyline

Round 1: Median 5.0 (1-9); n = 10

Round 2: Median 4.0 (0-8); n = 12

-Side-effects

-Too sedative, effect too delayed, frequent compliance obstacle due to side effects

-If indicated for neuropathic pain

-For individual TCAs I don't think there is sufficient evidence to differentially rate, although clinically I would favour those drugs that are used most often in palliative care anyway e.g. amitriptyline

-An excellent antidepressant

Paroxetine

Round 1: Median 5.5 (0-9); n = 10

Round 2: Median 3.0 (0-7); n = 13

-Too many pharmacological interactions

-Interactions and withdrawal problems

-No place - withdrawal syndrome

Fluoxetine

Round 1: Median 3.0 (0-6); n = 10

Round 2: Median 3.0 (0-6); n = 13

-Too long half time, not useful for physically ill people

-Long t1/2 and interactions

-Too disinhibitory, but not better antidepressant than other SSRIs

Imipramine

Round 1: Median 5.0 (0-6); n = 7

Round 2: Median 2.0 (0-8); n = 11

-Just in patients with depressive history who have already utilized before

Nortriptyline

Round 1: Median 2.0 (0-9); n = 9

Round 2: Median 2.0 (0-7); n = 11

-If indicated for neuropathic pain

-Less than amitriptyline

Mianserin

Round 1: Median 5.0 (0-9); n = 7

Round 2: Median 0.0 (0-7); n = 11

-Little used in practice compared with other drugs and little evidence available

-No point really - it is pharmacologically similar to mirtazapine, which doesn't hit the white cells