A study protocol for a predictive model to assess population-based avoidable hospitalization risk: Avoidable...

Area of Science:

• Health Services Research
• Epidemiology
• Biostatistics

Background:

• Avoidable hospitalizations indicate primary care effectiveness and are key health system performance indicators.
• Predicting population-level avoidable hospitalizations enables proactive interventions for ambulatory care-sensitive conditions (ACSCs).

Purpose of the Study:

• To develop and validate the Avoidable Hospitalization Population Risk Tool (AvHPoRT).
• To predict the 5-year risk of first avoidable hospitalization for seven ACSCs using population health survey data.

Main Methods:

• Utilized Canadian Community Health Survey (CCHS) data (2000-2006) for model derivation and a hold-out set for validation.
• Linked CCHS data with the Discharge Abstract Database for 394,600 individuals to track hospitalizations over 5 years.
• Developed sex-specific algorithms using Weibull survival models, validated with cross-validation and temporal validation.

Main Results:

• The study developed and validated the AvHPoRT tool for predicting avoidable hospitalizations.
• Performance was assessed using Nagelkerke R², calibration plots, and Harrell's concordance statistic.
• The model development adhered to the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement.

Conclusions:

• The AvHPoRT tool provides a validated method for predicting population-level risk of avoidable hospitalizations.
• This tool can aid health system decision-makers in implementing targeted primary care interventions.
• Dissemination of findings will occur through scientific meetings and peer-reviewed publications.