It is uncertain whether the use of structured assessment tools (PHQ-9, HADS-D or BDI-III) improves outcomes for depressed patients and the QOF depression severity indicators should be abandoned, a systematic review has concluded.

Eight studies met the inclusion criteria. There was only one (quasi) randomised controlled trial; the rest were observational studies.

The rationale for the assessment of depression severity at the onset of treatment is to improve the targeting of treatment to those who will benefit in accordance with the NICE stepped care guideline. In particular, antidepressants should not be used for the initial treatment of mild depression but should be routinely used for those with moderate to severe depression.

There was low quality evidence that severity assessments facilitate appropriate interventions but none of the studies looked at depression-related outcomes. Patients felt that the assessment tools improved diagnostic accuracy and made them feel they were being taken seriously. However, GPs preferred to rely on their clinical judgement and generally ignored severity scores which did not agree with their assessment.¹

An analysis of medical records from 38 UK practices found that exactly the same proportion of patients assessed with the PHQ-9 and HADS were prescribed antidepressants despite a large difference in the severity scores (83.5% and 55.6% respectively rated moderate to severe).²

GPs are right to question the validity and utility of depression severity scores. A recent study compared severity measures obtained with the four commonly used structured assessment tools with a psychiatrist’s assessment using the Hamilton Rating Scale for Depression (HRSD).

The HADS underestimated severity whereas the PHQ-9 and the BDI overestimated severity; to avoid this the authors recommend revised cut-off scores of 9, 12 and 23 respectively. However, even with the revised

cut-offs none of these tools was sufficiently accurate to be used in clinical settings.³

The rationale for follow-up severity assessments is first to identifynon-responders and adjust treatment accordingly and second to encourage patients to continue treatment for an adequate length of time (at least six months after remission).

There was conflicting evidence on whether follow-up severity scores influenced management. The American COMET study found that they did not whereas a UK retrospective cohort study found that patients with an inadequate response to treatment (fall in PHQ-9 score ≤ 1) were nearly five times as likely to have their treatment adjusted than those with an adequate response (fall in PHQ-9 score ≥ 5).⁵

Surprisingly, the COMET study did find that follow-up severity assessments increased the likelihood of response to treatment and remission. One possible explanation is that patients were encouraged to continue their treatment for longer.

Should we continue to use severity assessment questionnaires? I think a case can still be made for using the PHQ-9, for the following reasons:

• It encourages us to apply the DSM IV diagnostic criteria

• A score ≤ 10 can be used to rule out moderate depression (negative likelihood ratio 0.04)6 in the same way that BNP levels below the cut-off value are used to rule out heart failure

• It enhances patients’ confidence in our diagnosis, thereby potentially increasing adherence to treatment

• The main value of follow-up scores is that they encourage case management which in my opinion is the key to improving outcomes in depression.

Dr Phillip Bland