Claims data stops at the diagnosis code. NeuroBlu goes to the symptom.
46M+ patients. 3,099 clinician-validated mappings. 63 symptoms. 5 disorders.
One clinician-validated map. Three sources claims data can't combine.
Symptom Intelligence is NeuroBlu's symptom-level RWE capability. At its core is symptom_lookup, a clinician-validated reference table that turns the Mental Status Exam (a semi-structured clinical assessment), validated scales, clinical-note NLP, and diagnoses into queryable, symptom-level evidence across five neuropsychiatric disorders: schizophrenia, MDD, dementia, anxiety, and bipolar disorder. v2 shipped in 26R2, built on multi-source triangulation across MSE, validated scales, NLP, and ICD, so a symptom finding is never resting on a single, fragile data point.
Ask four people on a research or commercial team "how many patients have anhedonia?" and you'll get four different answers.
That's not a data discipline problem, it's a claims problem. Claims data stops at the diagnosis code, so it can't see negative symptoms, anhedonia, dementia behavioral symptoms, or whether a patient is currently in a manic or depressed state. Today, the Commercial team answers symptom questions with a slide pointing to five or more data locations, and a study team spends roughly three weeks manually mapping symptoms by hand, and still misses an estimated 40% of the signal. Symptom Intelligence replaces that slide and that three-week effort with a single, queryable layer.
Treatment-resistant depression and suicidality
Claims show an MDD diagnosis and a list of antidepressants prescribed, which isn't enough to build a defensible treatment-resistant depression cohort. NeuroBlu adds two or more failed antidepressant trials, C-SSRS-documented suicidality, and MSE-documented active suicidal ideation, each with timing and trajectory attached. The result is a precise, label-aligned cohort suitable for value dossiers, trial feasibility assessments, and market access conversations. The C-SSRS provides validated severity scoring here; the underlying evidence holds without relying on MSE severity ranking.
Schizophrenia negative and cognitive symptoms
67k+schizophrenia-diagnosed patients have documented negative symptoms, and 65k+ have documented cognitive symptoms, conditions no ICD code captures. The source is always the Mental Status Exam combined with NLP, never a diagnosis code, which is exactly why claims-based research misses it. This matters for therapies targeting negative symptoms, programs addressing cognitive impairment, and long-acting maintenance treatment strategy.
MDD anhedonia
1.54 million anhedonia patients sit within the broader MDD cohort of ~12 million patients. Across all disorders, 2.3 million anhedonia patients are identified. About 98% of this signal comes from structured PHQ-9 and QIDS scale items, making it a reliable, structured-scale finding rather than a notes-derived one. It's a clear example of how a single validated instrument, applied consistently, can surface a symptom population claims data has no way to identify.
Dementia neuropsychiatric symptoms
209,985 dementia-diagnosed patients have documented neuropsychiatric symptoms, and platform-wide that figure reaches 1.48 million. As with the schizophrenia findings, the source is the Mental Status Exam, with no ICD code involved. Neuropsychiatric symptoms drive caregiver burden, institutionalization, and cost, and they open a path to evaluating non-amyloid therapeutic strategies that a diagnosis-only view would never surface.
Ready to put symptom-level evidence to work? Log in to NeuroBlu Analytics, or request a demo to see Symptom Intelligence applied to your own research question.
