Rating scales and measurement-based care — structured clinical viva

Opening frame. Scales quantify constructs; MSE and diagnosis remain clinical. MBC uses serial validated scores to change care, not decorate the chart.^[4]

Core battery. PHQ-9 (0–27; bands 5/10/15/20; item 9 risk); GAD-7 (0–21; screen ~10); HAM-D/MADRS for clinician depression severity/change; PANSS for psychosis trial-grade severity; YMRS for mania severity; CGI-S/I 1–7 global; MoCA/MMSE cognitive screens with education/language caveats.^[1][2][7]

Psychometrics. Reliability = consistency (internal, test–retest, inter-rater). Validity = measuring the intended construct (content, construct, criterion). Responsiveness = sensitivity to true clinical change.^[1][2]

PANSS pearl. Items 1–7 → minimum total 30; percent change uses (total − 30).^[2][5]

Evidence. Guo 2015 MBC vs standard care RCT; STAR*D algorithmic MBC; Fortney tipping point; Lewis implementation barriers/facilitators.^[3][4][6]

Rebuttal to anti-scale consultant. Clinical impression is necessary but prone to drift and handover loss; measured residual symptoms drive earlier step-up; scores never replace empathy or MSE — they discipline follow-through.^[4][6]