Science & validation
Validated instruments, standard scoring, clinical judgment first.
Spectra is built on validated, published screening instruments — administered as published, with standard scoring and cutoffs. The brief Tier-1 questions that decide which instrument to open are original, sensitivity-tuned wording: a “yes” opens a validated instrument; it is not itself a diagnosis or finding.
PHQ-9
Depression9 items (0–27); clinically significant at ≥10
Kroenke, Spitzer & Williams, J Gen Intern Med, 2001 — ~88% sensitivity / ~88% specificity for major depression at ≥10.
Public domain.
GAD-7
Anxiety7 items (0–21); cutoff ≥10
Spitzer, Kroenke, Williams & Löwe, Arch Intern Med, 2006 — ~89% sensitivity / ~82% specificity for GAD at ≥10.
Free for clinical use.
AUDIT-C
Alcohol use3 items, sex-specific cutoffs (≥4 men, ≥3 women)
Bush et al., Arch Intern Med, 1998 — validated to identify hazardous drinking and active alcohol-use disorders in primary care.
Public domain (WHO).
NIDA single-question screen
Drug useOne item on past-year illicit/non-medical prescription use
Smith, Schmidt, Allensworth-Davies & Saitz, Arch Intern Med, 2010 — ~100% sensitivity / ~73.5% specificity for a drug-use disorder in primary care.
Public domain.
SSS-8
Somatic symptom burden8 items
Gierk et al., JAMA Internal Medicine, 2014 — a validated brief measure of somatic symptom burden.
Suicidality (PHQ-9 item 9)
Self-harm ideation flagAn endorsement of thoughts of being better off dead or self-harm
Raises a clinical-review flag. This is a risk indicator that prompts clinician assessment, not a standalone suicide-risk instrument.
Reported performance figures are from the cited validation studies and vary by patient population and setting. Spectra is decision-support tooling for licensed professionals; it does not diagnose. All results require provider review.