Antidepressants are a safe and effective treatment for depression in patients with a chronic physical illness, a NICE systematic review and meta-analysis has concluded.

A total of 63 studies involving 5,794 participants met the inclusion criteria. The studies covered: cancer (8 trials); heart disease (9); musculoskeletal disorders (1); respiratory disorders (4); neurological disorders (8); diabetes (7); stroke (11); CKD (1); HIV (5) and multiple/unspecified conditions (9).

There were 48 placebo-controlled studies, of which 35 compared SSRIs with placebo, and nine compared tricyclics with placebo. Remission was defined as a rating scale score below the cut-off for depression and response was defined as a reduction of 50% or more from the baseline score. For SSRIs, the relative risk (RR) of non-remission was 0.81 (95% CI: 0.73-0.91) and RR of non-response 0.83 (95% CI: 0.71-0.97). For tricyclics, the RRs were 0.70 (95% CI: 0.40-1.25: non-significant) and 0.55 (95% CI: 0.43-0.70) respectively.

The calculated effect size varied according to the scales used and the inclusion/exclusion of non-blinded studies but was around 0.3 for SSRIs and 0.6 for tricyclics.

Participants receiving antidepressants were twice as likely to leave the study early because of adverse events (RR 1.80 for SSRIs and 2.00 for tricyclics). There was a small improvement in quality of life for patients treated with SSRIs but there was insufficient data to assess the effect of tricyclics.

Fourteen studies compared SSRIs with tricyclics. No significant differences were found between the two groups for remission rate, response rate, effect size or tolerability.

Three studies were eligible for inclusion in the safety review. These were all trials of depression treatment following a myocardial infarction. Only one of the three reported a significant reduction in cardiovascular events, and this was a post hoc analysis conducted after the original study had failed to show any benefit from the psychotherapeutic intervention.¹ Nevertheless, sertraline, mirtazapine and citalopram do appear to be safe for patients with heart disease.

The effect sizes were small to moderate and comparable with those seen in depression without associated chronic physical illness. An effect size ≥0.5 or RR of non-remission ≤0.8 have been suggested as markers of clinical efficacy.² According to these criteria, the efficacy of SSRIs is of borderline significance, although the adoption of such a cut-off is arbitrary and controversial.³ For SSRIs, assuming a placebo response rate of 30%,⁴ I calculate the NNT to be eight.

There are several reasons for not prescribing antidepressants to patients with chronic conditions who present with mild to moderate depression of recent onset:

• Antidepressants are little better than placebo
• The physical illness may increase susceptibility to side effects
• There is an increased risk of drug interactions
• Patients may be reluctant to take them

NICE recommends that they should only be prescribed if low-intensity psychosocial interventions are ineffective, or if there is a previous history of moderate or severe depression, or if the depression complicates the care of the chronic condition.⁵

If an antidepressant is to be prescribed, SSRIs are the drug group of first choice as they may be better tolerated and have been shown to improve quality of life. NICE recommends sertraline and citalopram as they are less likely than other SSRIs to interact with other medication and appear to be safe in heart disease.⁵

Dr Phillip Bland

• Taylor D, Meader N, Bird V et al. Pharmacological interventions for people with depression and chronic physical health problems: systematic review and meta-analyses of safety and efficacy. Br J Psychiatry 2011;198:179-88