Clinical Context
Treatment dropout—when patients discontinue regular diabetes care—is a major barrier to optimal outcomes. In Japan, approximately 10-15% of diabetes patients drop out of care annually, leading to uncontrolled hyperglycemia, accelerated complications, and eventual re-presentation with advanced disease. Understanding factors that predict dropout could enable targeted retention interventions before patients disengage.
Previous dropout history is intuitively a risk factor—patients who have dropped out before may be more likely to do so again, reflecting ongoing barriers to care engagement. High HbA1c levels may indicate complex disease requiring intensive management that overwhelms patients, or may reflect disengagement that predates formal dropout. Identifying these high-risk patients allows proactive intervention.
The J-DOIT2 program (Japan Diabetes Outcome Intervention Trial 2) was a large-scale initiative to reduce diabetes treatment dropout through various interventions. This secondary analysis examined whether dropout history and baseline HbA1c predicted subsequent dropout risk, informing strategies for patient retention.
PICO Summary
Population: Japanese patients with type 2 diabetes participating in the J-DOIT2-LT008 trial, a large-scale pragmatic trial of dropout prevention interventions.
Intervention/Exposure: History of prior treatment dropout and/or HbA1c levels ≥10% at enrollment.
Comparison: Patients without prior dropout history and HbA1c <10%.
Outcome: History of treatment dropout was strongly associated with higher subsequent dropout risk. Among patients without prior dropout history, HbA1c ≥10% significantly predicted future dropout. These factors represent actionable identifiers for targeted retention efforts.
Clinical Pearls
1. Past Behavior Predicts Future Behavior: Patients with prior dropout history are at substantially elevated risk for subsequent dropout. This isn’t surprising but is actionable—when patients return after a gap in care, flag them for enhanced retention efforts rather than simply resuming usual care.
2. Very High HbA1c Signals Disengagement Risk: HbA1c ≥10% (approximately 86 mmol/mol) predicted dropout even in patients without prior dropout history. This extreme level may reflect patients already partially disengaged from self-management or overwhelmed by disease complexity. It’s both a glycemic target and a retention warning sign.
3. Dual Risk Stratification: Combining dropout history and HbA1c creates risk categories: prior dropouts with high HbA1c (highest risk), prior dropouts with controlled HbA1c, no dropout history but high HbA1c, and neither factor (lowest risk). This stratification can guide intensity of retention interventions.
4. Japanese Context but Universal Principles: While conducted in Japan, the principles apply broadly: healthcare systems should track dropout history and use it to identify high-risk patients, and very poor control should trigger assessment of engagement barriers alongside medical intensification.
Practical Application
Implement systematic tracking of care gaps and dropout episodes in diabetes patient records. When patients return after gaps in care, document this as a risk factor and implement enhanced follow-up: more frequent appointments, proactive outreach between visits, exploration of barriers to care engagement.
For patients with HbA1c ≥10%, address not only the glycemic emergency but also the engagement risk. Explore barriers: Is the regimen too complex? Are there financial obstacles? Is the patient overwhelmed or experiencing diabetes distress? Intensive medical management without addressing these factors may lead to dropout.
Consider care coordination, patient navigation, or case management for highest-risk patients. The cost of retention interventions is far less than the cost of complications from years of uncontrolled diabetes during dropout periods.
Broader Evidence Context
Treatment dropout is recognized globally as a major barrier to diabetes care. Studies from multiple countries identify similar risk factors: younger age, male sex, lower socioeconomic status, depression, and early-stage disease (when patients feel well). Interventions including reminder systems, patient navigation, and addressing social determinants have shown promise for improving retention.
The J-DOIT2 program represents one of the largest systematic efforts to address diabetes treatment dropout at a population level.
Study Limitations
Secondary analysis of trial data may not fully capture all dropout predictors. Japanese healthcare system characteristics (universal coverage, high access) may limit generalizability to other settings. Definitions of dropout vary across studies, affecting comparability. Reasons for dropout weren’t systematically characterized.
Bottom Line
Prior dropout history and HbA1c ≥10% are strong predictors of subsequent treatment dropout in type 2 diabetes. Healthcare systems should track these factors and implement enhanced retention interventions for high-risk patients to prevent disengagement and its consequences.
Source: Goto A, et al. “Association of Dropout History and HbA1c Levels with Subsequent Dropout Risk in Patients with Diabetes: A Secondary Analysis of the J-DOIT2-LT008.” Read article
