Research Agenda

Predicting and Understanding Unreported Adverse Events (AEs):

• What participant and service characteristics (e.g., demographics, socio-economic status, clinical conditions) are most strongly associated with unreported AEs, and how do these differ by AE type?

• How do historical patterns of AEs (trailing cumulative variables) differ between those that are reported and unreported, and how might this difference inform predictive modeling for identifying unreported AEs?

• What categorical clinical variables (e.g., specific diagnoses such as schizophrenia, medication allergies, polypharmacy) distinctly influence the likelihood that certain types of AEs go unreported?

Data Quality and Unreported AE Types:

• How does incomplete or missing participant data impact the ability of predictive analytics to accurately detect or estimate the frequency of unreported AEs across different AE categories?

• What distinct patterns emerge for rare or less frequently reported AE types, and how can predictive modeling be adapted to better detect or anticipate these unreported events?

Factors Influencing AE Reporting Behaviors:

• What organizational, systemic, or individual factors influence whether AEs are reported or remain unreported, and how does this vary across AE types (e.g., behavioral incidents vs. medical emergencies)?

• How do case managers’  and providers' perceptions and assessments influence decisions around reporting versus not reporting particular AE types, and how might predictive analytics support improved reporting accuracy?

Impact of Predictive Alerts on Unreported AEs:

• How effective are automated predictive alerts in prompting case managers to recognize, document, and intervene in cases that might otherwise remain unreported, and do alert effects vary by AE type?

Organizational Implementation and Unreported AE Management:

• What organizational workflows, training, or data-entry practices contribute to underreporting of specific AE types, and how can predictive analytics inform targeted improvements in these areas?

Long-Term Outcomes and Unreported AE Insights:

• What new insights about participant outcomes or system-wide improvements can be gained by accurately predicting, identifying, and managing previously unreported adverse events, and how does this differ by AE type?