A new study co-authored by Diabetes Action Canada’s (DAC) Conrad Pow found that hospital stays may provide a critical opportunity to identify people with diabetes at higher risk of future foot complications and connect them to preventive care earlier, potentially reducing amputation rates.

Lower-limb amputations remain one of the most devastating complications of diabetes, yet many are preventable with timely screening and intervention.

A new study published in the Journal of Clinical Epidemiology explored how hospitals could play a larger role in preventing these complications before they become severe.

 📖 Read the full paper here.

Researchers developed and tested an AI prediction model that uses routinely collected hospital data to identify people with diabetes who may be at higher risk of developing serious foot complications after leaving the hospital. The study included data from more than 107,000 hospitalizations across Ontario.

The goal: connect people to preventative foot care earlier, when interventions may be most effective. 

What the researchers wanted to learn

Foot ulcers, infections, gangrene, and amputations are among the most serious complications associated with diabetes. Many are preventable with regular screening, education, and access to expert foot care, but access remains inconsistent. 

The research team wanted to understand whether hospitalization could serve as a key opportunity to identify people at elevated risk and connect them to community-based preventive care after discharge. 

The study focused on adults with diabetes discharged from General Internal Medicine units at 29 Ontario hospitals between 2016 and 2023.

The research team used routinely collected hospital data to build and validate a prediction model estimating an individual’s likelihood of developing a serious foot complication within one year of discharge  

Importantly, the model was designed using information already commonly available in hospital records, making real-world implementation more feasible. 

What the study found

Among 107,836 hospitalizations included in the study, 2.7% of patients developed a serious foot complication within one year of discharge. More than 12% died during the same period, underscoring the complexity of care needs among people recently hospitalized with diabetes.  

The prediction model showed good overall performance in identifying people at higher risk for future complications. The paper found that factors such as peripheral artery disease, congestive heart failure, higher HbA1c levels, kidney disease markers, and homelessness were associated with an increased risk.  

The study also found important differences in how the model performed across hospitals, highlighting the need for local validation before tools like this are widely implemented.  

The research team then piloted the model at one Ontario hospital in early 2025. A care coordinator used the tool to identify patients at elevated risk and offer referrals to a community-based preventative foot care clinic after discharge.  

Of the 107 patients successfully reached by phone, 41% accepted referral to preventative foot care services.  

Why this matters

Foot complications can have a major impact on quality of life, mobility, independence, and overall health. They are also a significant burden on healthcare systems and remain one of the leading causes of diabetes-related hospitalizations and amputations. 

The paper highlights a practical opportunity to intervene earlier. 

The research team found that many people who undergo diabetes-related amputations were hospitalized a year prior. By using hospital stays as a point for risk identification and follow-up planning, healthcare institutions may be able to connect patients to preventative care before infections become severe.  

The findings are particularly relevant for older adults and people facing barriers to consistent access to care, including those experiencing housing instability or multiple chronic conditions. The study also highlights the importance of building stronger links between hospital care and community-based prevention services. 

Beyond foot care specifically, the work demonstrates how routinely collected health data can help support more proactive, connected, and person-centred diabetes care. 

Looking ahead

The next step is not only developing prediction tools, but also ensuring they can be effectively integrated into real-world healthcare settings.

Future work should focus on improving referral pathways and expanding access to preventative foot care programs in communities where services may be limited. The study also reinforces the value of integrated care models that connect hospitals, primary care, and community supports more effectively after discharge. 

For DAC, the work reflects a broader focus on improving outcomes through earlier intervention, practical implementation, and care pathways designed around the real experiences of people living with diabetes.