Loading
Population sweep agent
The sweep runs the cascade engine on every patient and writes each finding back into FalkorDB. Everything on this page is a Cypher aggregate over those findings and the prescriptions themselves.
The synthetic cohort is deliberately enriched with cascades so there is something to find. These numbers are not a prevalence estimate.
Red: documented pattern. Amber: inferred from the graph alone.
Discovery
The method pharmacoepidemiologists use to find cascades in national data: among people who start both drug classes within a year, does B start after A more often than before it? Domino adjusts for prescribing trends (the null-effect ratio), then asks the graph for a mechanism. A signal with a path is a candidate cascade. A signal without one is usually a care pathway, like anticoagulation before rate control.
| First class → second class | After / before | Adjusted sequence ratio, log scale | aSR (95% CI) | Mechanism in the graph |
|---|---|---|---|---|
| Anticoagulant → Beta blocker | 29 / 0 | 50.16 (6.42–>64) | none: likely a care pathway | |
Doherty et al., Ann Fam Med 2025: 533,464 adults aged 65+, PSSA with a 365-day window. Stored on each Pattern node in the graph.
FalkorDB betweenness centrality over drugs and effects: the nodes most cascade routes pass through. Good places for a prescriber prompt.
| Anticoagulant → Proton pump inhibitor Usually deliberate: NSAID, antiplatelet or anticoagulant → bleeding risk → stomach protection |
| 16 / 0 |
| 31.24 (3.94–>64) |
| via gi bleeding risk |
| Loop diuretic → Potassium-sparing diuretic | 18 / 0 | 29.56 (3.74–>64) | via hypokalaemia |
| Loop diuretic → Potassium supplement Usually deliberate: Diuretic → low potassium → potassium supplement | 24 / 1 | 14.83 (3.74–>64) | via hypokalaemia |
| Loop diuretic → SGLT2 inhibitor | 11 / 0 | 28.13 (3.5–>64) | none: likely a care pathway |
| Loop diuretic → Beta blocker | 21 / 1 | 13.75 (3.44–>64) | none: likely a care pathway |
| Vestibular sedative (antihistamine) → Cholinesterase inhibitor | 13 / 0 | 25 (3.13–>64) | via cognitive impairment |
| ACE inhibitor → Cough suppressant Documented: ACE inhibitor → dry cough → cough suppressant | 14 / 0 | 23.16 (2.91–>64) | via dry cough |
| Loop diuretic → ACE inhibitor | 24 / 2 | 9.03 (2.9–42.24) | none: likely a care pathway |
| NSAID → Proton pump inhibitor Usually deliberate: NSAID, antiplatelet or anticoagulant → bleeding risk → stomach protection | 34 / 4 | 6.91 (2.83–20.71) | via gi bleeding risk |
| Potassium supplement → Potassium-sparing diuretic | 11 / 0 | 18.72 (2.33–>64) | none: likely a care pathway |
| Antiplatelet → Proton pump inhibitor Usually deliberate: NSAID, antiplatelet or anticoagulant → bleeding risk → stomach protection | 14 / 2 | 6.86 (2.1–32.65) | via gi bleeding risk |
| Anticoagulant → Calcium channel blocker (non-dihydropyridine) | 10 / 0 | 16.38 (2.03–>64) | none: likely a care pathway |
| Alpha-1 blocker → Vestibular sedative (antihistamine) Documented: Alpha-1 blocker → dizziness → vestibular sedative | 14 / 1 | 8.14 (1.98–>64) | via dizziness / vertigo |
| Biguanide → Antidiarrhoeal Documented: Metformin → diarrhoea → loperamide | 9 / 0 | 15.29 (1.88–>64) | via diarrhoea |
| NSAID → Potassium-sparing diuretic | 8 / 0 | 14.13 (1.73–>64) | via hypertension, peripheral edema |
| Alpha-1 blocker → Oral vasopressor | 10 / 1 | 7.25 (1.71–62.73) | via orthostatic hypotension |
| Proton pump inhibitor → Calcium channel blocker (dihydropyridine) | 16 / 3 | 4.8 (1.69–17.42) | none: likely a care pathway |
CALL algo.betweenness({nodeLabels: ['Drug', 'Effect'], relationshipTypes: ['CAUSES', 'TREATED_BY'], samplingSize: 2000, samplingSeed: 20261015})
YIELD node, score
RETURN labels(node)[0] AS label, node.id AS id, node.name AS name, score
ORDER BY score DESC
LIMIT 40