Updates to the Depression Module Needed

[tbhallett]

• The Years Lived With Disability due to "Depression / Self-Harm" is very high and this is not well founded, and not in agreement with the GBD data (see below). This should be re-calibrated

  

• The only symptom for Depression is 'Self-Harm' - but it seems likely that there should be a symposium of depression, and which incurs some disability.

• In the calibration process (#844), we have reduced the probability of screening very severely, to reflect the very few MentOPD appointments that happen. But, it may be that this screening actually does happen, as part of other appointments, and so should occur but not incur the MentOPD appointment; or perhaps should only incur the MentOPD if there is a referral to a specialist, for those that do have the (non-severe) symptom.

[BinglingICL]

Hi @tbhallett, thanks a lot for this. Would like to add some notes if helpful.

The only symptom for Depression is 'Self-Harm' - but it seems likely that there should be a symposium of depression, and which incurs some disability.

For example, we could define a symptom of "Depression" in addition to "Self-Harm", which might be caused/implemented in several modules including Depression, Labour and Care of women during pregnancy, etc.

In the calibration process (https://github.com/UCL/TLOmodel/issues/844), we have reduced the probability of screening very severely, to reflect the very few MentOPD appointments that happen. But, it may be that this screening actually does happen, as part of other appointments, and so should occur but not incur the MentOPD appointment; or perhaps should only incur the MentOPD if there is a referral to a specialist, for those that do have the (non-severe) symptom.

In this branch PR #847, we have specifically reduced the pr_assessed_for_depression_in_generic_appt_level1 from 0.5 to 0.01 and the pr_assessed_for_depression_for_perinatal_female from 1.0 to 0.01. This is based on MentOPD usage comparison with DHIS2 data, but not on published literature. If references on these probabilities are found, we will incorporate them in our resource files.

Yes, since MentOPD in tlo model is requiring times from mental health specialists, we might need to consider two types of depression screening, like general depression consultancy with nursing and midwives and specialist depression diagnosis with mental health officer and nurses. The latter might be implemented when the patient has the symptom of "Depression".