Skip to main content
Psych Congress  

Cluster Analysis of Care Pathways in Major Depressive Disorder With Suicidal Ideation or Probable Suicide Attempt


Maryia Zhdanava, MA-Manager, Analysis Group Inc.; Dominic Pilon, MA-Manager, Analysis Group Inc.; Tom Cornwall, PhD-Associate, Analysis Group Inc.; Laura Morrison, MPH-Senior Research Professional, Analysis Group Inc.; Maude Vermette-Laforme, BA-Research Professional, Analysis Group Inc.; Patrick Lefebvre, MA-Managing Principal, Analysis Group Inc.; Abigail Nash, MD, PhD-Associate Medical Director, Janssen Scientific Affairs, LLC; Kruti Joshi, MPH-Director, Janssen Scientific Affairs, LLC; Cheryl Neslusan, PhD-Director, Janssen Scientific Affairs, LLC

Janssen Scientific Affairs, LLC

BACKGROUND: To understand heterogeneity among the population with major depressive disorder (MDD) and suicidal ideation (SI) or attempt (SA) (MDD+SI/SA), clusters were identified through a data-driven process, and care pathways were described by cluster.

METHODS: Adults with ≥1 MDD diagnosis during the 6-month pre-index period before a claim for SI or probable SA were identified in the MarketScan® Databases (10/01/2014-04/30/2019). The first SI/SA claim defined the index event. Hierarchical clustering divided patients into clusters based on characteristics measured pre-index. Silhouette width was used to optimize number of clusters. Patients' care pathways were described by cluster up to 12 months post-index, including the index event.

RESULTS: Among 38,876 patients with MDD+SI/SA, three clusters were identified. Cluster 1 (N=16,025) was least exposed to the healthcare system pre-index, Cluster 2 (N=5,640) moderately exposed, and Cluster 3 (N=17,211) most exposed. Patients with the first MDD diagnosis at index event comprised 86.0% of Cluster 1 and 72.8% of Cluster 2; in Cluster 3, all patients were diagnosed pre-index. Within 30 days post-index, 79.3%, 85.2%, and 88.2% used mental health services including outpatient visits for MDD in Clusters 1, 2, 3, respectively. Within 12 months post-index, these proportions were 83.8%, 90.4% and 93.6%. Within 12 months post-index, 61.5%, 91.5%, and 84.6% had ≥1 claim for an antidepressant in Clusters 1, 2, 3, respectively.

CONCLUSIONS: Patients with MDD+SI/SA least exposed to healthcare system pre-index received less care post-index. An opportunity exists to optimize care following a suicide-related event through psychotherapy, pharmacotherapy, and referral to mental health services.

Back to Top