Clinical Context
Dropout from diabetes care—defined as loss of regular follow-up with healthcare providers—is a common and underappreciated problem. Studies suggest 20-40% of diabetes patients experience periods of care discontinuity, during which glycemic control deteriorates, complications progress undetected, and preventive care lapses. The consequences extend beyond individual patients: discontinuous care increases healthcare costs through emergency visits, hospitalizations, and late-stage complication treatment.
Understanding who drops out of care—and why—is essential for designing retention interventions. Previous research has identified demographic factors (younger age, male sex, lower socioeconomic status), disease factors (shorter diabetes duration, fewer comorbidities), and healthcare system factors (poor access, high costs, long wait times) as dropout predictors. However, behavioral patterns like prior dropout history have received less attention.
This secondary analysis from the large Japanese J-DOIT2 trial examined whether prior dropout history and current glycemic control predicted future dropout risk. Identifying patients at high risk for disengagement could enable targeted retention interventions before care discontinuity occurs.
PICO Summary
Population: Patients with type 2 diabetes enrolled in the Japan Diabetes Outcome Intervention Trial-2 (J-DOIT2-LT008), a large-scale long-term cohort study.
Intervention/Exposure: Assessment of two key predictive factors: (1) prior dropout history from diabetes care, and (2) elevated HbA1c levels (≥10%, indicating very poor glycemic control).
Comparison: Patients without prior dropout history and patients with HbA1c <10%.
Outcome: Prior dropout history was the strongest predictor of subsequent dropout—patients who had previously dropped out were significantly more likely to drop out again. Among patients without prior dropout history, HbA1c ≥10% was associated with elevated dropout risk, suggesting very poor control predisposes to care disengagement. The findings identify two high-risk groups for targeted retention efforts.
Clinical Pearls
1. Past Behavior Predicts Future Behavior: Patients who have dropped out before are likely to drop out again. This pattern suggests dropout reflects stable patient characteristics (health beliefs, coping style, life circumstances) rather than isolated incidents. These patients require ongoing engagement strategies, not just re-enrollment after each dropout episode.
2. Very Poor Control Signals Disengagement Risk: HbA1c ≥10% identified patients at risk for dropout among those without prior dropout history. Extremely elevated HbA1c may reflect existing disengagement (missed appointments, medication non-adherence, lifestyle non-engagement) even before formal care dropout occurs. It’s a warning sign that the patient-provider relationship may be weakening.
3. Paradox of Those Most Needing Care: The patients most likely to drop out—those with very poor control—are precisely those who need intensive engagement most. This creates a troubling paradox where high-risk patients become invisible to the healthcare system at critical moments.
4. Retention Is Treatable: Unlike unchangeable risk factors (age, diabetes duration), dropout history signals a need for intervention. Patients returning after dropout represent a second chance to implement retention strategies—but only if healthcare systems recognize and act on this risk factor.
Practical Application
Implement systematic flagging of patients with prior dropout history in medical records. When these patients return to care, recognize them as high-risk for repeat dropout and implement enhanced engagement strategies: more frequent follow-up appointments initially, proactive outreach before scheduled appointments, addressing barriers to care (transportation, scheduling conflicts, costs), and exploring reasons for prior dropout to prevent recurrence.
For patients with HbA1c ≥10%, interpret this not just as a glycemic problem requiring treatment intensification, but as a potential engagement problem requiring relationship attention. Schedule more frequent visits, ensure cultural concordance and communication quality, and assess for diabetes distress or burnout that might be driving disengagement.
Healthcare systems should track dropout as a quality metric alongside HbA1c control and complication screening. Regular analysis of which patients are dropping out can identify systemic access barriers amenable to intervention. Automated systems that flag missed appointments for immediate outreach can interrupt the dropout trajectory before it becomes prolonged.
Broader Evidence Context
This study adds to literature showing that care continuity in diabetes is as important as care quality. Patients who remain engaged with care, even if their control is imperfect, have better long-term outcomes than those who cycle in and out of the healthcare system. The chronic care model emphasizes proactive, patient-centered engagement rather than reactive, disease-centered treatment.
Interventions to reduce dropout include patient navigation programs, community health worker support, telehealth options for barrier reduction, and self-management support groups that create accountability and community. The evidence base for specific retention interventions is growing but remains less robust than for glycemic treatments.
Study Limitations
Japanese healthcare system may differ from other settings in access, costs, and cultural factors affecting dropout. The definition of “dropout” varied and may capture different phenomena (intentional disengagement vs. unintentional loss of contact). Reasons for dropout were not systematically assessed. The observational design cannot establish causation—both prior dropout and poor control may be markers of underlying factors (depression, social instability, health beliefs) rather than direct causes.
Bottom Line
Patients with type 2 diabetes who have previously dropped out of care, and those with very poor glycemic control (HbA1c ≥10%), face significantly elevated risk of future care dropout. These identifiable high-risk groups should receive enhanced engagement and retention interventions to prevent care discontinuity and its associated adverse outcomes.
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 Japan Diabetes Outcome Intervention Trial-2 Large-Scale Trial 008 (J-DOIT2-LT008).” Read article
