Identifying multimorbid patients with high care needs - A study based on electronic medical record data

Area of Science:

• Health Services Research
• General Practice
• Digital Health

Background:

• Patients with multiple chronic diseases (multimorbidity) often require significant healthcare resources.
• Identifying 'high need' patients is crucial for proactive integrated care interventions.
• General practitioners (GPs) may lack awareness of which patients have high care needs.

Purpose of the Study:

• To identify predictors of high care needs within general practice electronic medical records (EMRs) for patients with multimorbidity.
• To evaluate the predictive accuracy of these identified predictors.

Main Methods:

• A large dataset of 245,065 patients with multimorbidity was analyzed.
• EMRs were linked with hospital claims data.
• Probit regression models predicted high general practice contact rates, emergency department visits, and unplanned hospitalizations using patient demographics, morbidity, and prior healthcare utilization.

Main Results:

• 11% of multimorbid patients had ≥12 general practice contacts annually, predictable by prior year's contact frequency (PPV 42%).
• Predicting emergency department visits (12% of patients) and unplanned hospitalizations (7% of patients) was less accurate (PPV 27% and 20%, respectively).
• High general practice users showed minimal overlap with emergency department visitors or those with unplanned hospitalizations.

Conclusions:

• Distinct 'high need' patient groups exist within multimorbid populations.
• Past general practice utilization effectively identifies patients with high needs for primary care.
• Additional data sources are necessary to accurately predict emergency department visits and unplanned hospitalizations.