Burden of Negative Symptoms in Patients With Schizophrenia: A Cluster Analysis

OBJECTIVES: Negative symptoms (NSS) have traditionally been seen as a fundamental aspect of schizophrenia (SCZ). This study aims to characterize distinct groups of schizophrenic patients with negative symptoms using machine learning-driven cluster analysis and to understand their economic burden.

METHODS: A retrospective claims analysis used the STATinMED RWD Insights database. The index date is the earliest recorded NSS diagnosis claim among patients diagnosed with SCZ between January 1, 2016, and September 30, 2022. Eligible patients were ≥13 years old with ≥1 SCZ inpatient claim or ≥2 SCZ outpatient claims ≥30 days apart, ≥12 months of continuous capture pre- and post-index, and no claim for NSS diagnosis or antipsychotics in the 12-month pre-index period. Patients were segmented using ML cluster analysis, determining optimal clusters via the Elbow method and hierarchical clustering. Annual mental health-related healthcare resource utilization (HCRU) and costs were evaluated across all clusters.

RESULTS: Among 15,516 patients, 62.5% were male and mean age was 47. Cluster analysis based on psychiatric and neurodevelopmental comorbidities revealed three distinct clusters among schizophrenic patients with NSS: 68.1% with low negative symptoms (LNS), 17.8% with moderate negative symptoms (MNS), and 14.1% with severe negative symptoms (SNS). SNS patients had severe bipolar disorder and the highest rates of substance abuse, suicidal ideation, and trauma-related disorders. Individuals in the SNS cluster showed the highest rates of mental-health-related inpatient and emergency department utilization compared to the other two clusters. They also incurred significantly higher total mental-health-related costs ($37,920) compared to the MNS ($30,340) and LNS ($22,405) clusters.

CONCLUSIONS: Schizophrenic patients with SNS had higher healthcare utilization and costs than those with lower or moderate severity. Potential strategies to manage other complex disorders in these patients could lead to improved outcomes.