The goal of this clinical study is to learn whether a physician-nurse collaborative intervention guided by an Intelligent Hypoglycemia Risk Early-Warning System can reduce hypoglycemia and improve glycemic control in hospitalized adults with type 2 diabetes receiving intensive insulin pump therapy. The study also aims to understand how risk-stratified management influences the time needed to reach glycemic targets.

The main questions this study aims to answer are:

1. Does the physician-nurse collaborative intervention reduce the incidence of hypoglycemia, particularly level 2 and symptomatic hypoglycemia?
2. Does this intervention help patients reach their glycemic targets sooner during intensive insulin pump therapy?
3. What differences in outcomes are observed between patients classified as high-risk and low-risk by the Intelligent Hypoglycemia Risk Early-Warning System?

Participants will:

1. Receive either standard insulin pump therapy alone or insulin pump therapy combined with the physician-nurse collaborative intervention
2. Undergo hypoglycemia risk assessment using the Intelligent Early-Warning System
3. Receive individualized insulin dose adjustments, intensified glucose monitoring, and tailored hypoglycemia education based on their risk category
4. Be monitored for hypoglycemic events and time to achieve glycemic targets during hospitalization This study will compare the collaborative intervention with standard care to evaluate its effectiveness in preventing hypoglycemia and accelerating glycemic stabilization among hospitalized patients with type 2 diabetes.

Inclusion Criteria:

• Diagnosed with type 2 diabetes mellitus.
• Age 18 years or older.
• Hospitalized for 3 days or longer.
• Receiving continuous subcutaneous insulin infusion (CSII) therapy during hospitalization.

Exclusion Criteria:

• Acute diabetic complications (e.g., diabetic ketoacidosis, hyperosmolar hyperglycemic state).
• Severe cardiovascular or cerebrovascular disease, hepatic dysfunction, or renal dysfunction.
• Malignant tumors.
• Severe cognitive impairment or psychiatric disorders that prevent cooperation.
• Premature discontinuation of CSII therapy due to non-glycemic reasons.

Individual participant data will not be shared because the dataset contains sensitive personal health information and is restricted by institutional and ethical requirements.