Identification of mental health and quality of life outcomes in primary care databases in the UK: a systematic review.

Observed Findings

• Codelists were reported in a minority of eligible studies: 17/42 depression studies, 5/12 anxiety studies, 3/6 fatigue studies, and only 1/2 sleep disorder studies.
• Depression was most commonly defined using diagnostic codes (37/42 studies) and/or antidepressant prescriptions (21/42 studies), with symptom-based definitions rare (6/42 studies).
• Fatigue definitions showed consistency across studies but were infrequently reported: definitions typically included chronic fatigue syndrome, neurasthenia, and postviral fatigue syndrome.
• Only 21/120 studies (18%) performed validation of their outcome definitions.
• Substantial variability existed across studies in how the same outcome was defined and coded.

Inferred Conclusions

• There is urgent need for standardized, validated codelists for mental health and quality of life outcomes in UK primary care databases.
• Current research practices lack transparency: most studies did not publish their codelists, limiting reproducibility and comparability.
• Depression and anxiety definitions rely heavily on diagnostic codes and prescriptions rather than symptom-based criteria, potentially missing undiagnosed or untreated cases.
• Fatigue is underrepresented in research using these databases, with minimal codelist reporting and no systematic framework for case identification.

Remaining Questions