Dashboard
💡 Overview: This page consolidates high-level analytics from patient demographic data and physician consultation logs. It helps hospital administrators analyze geographic patient distribution (to support cross-provincial medical insurance settlement planning) and identify high-workload core clinical physicians (based on Task 6, Queries 2 & 5).
📍 Top 5 Provinces by Patient Origin
Statistics based on patient permanent residence data (Query 2)
👨⚕️ Top 3 Attending Physicians by Patient Volume
Ranked by historical inpatient admission case volume (Query 5)
Current Inpatient Count
1,248
Scheduled Weekly Discharges
312
Active Inter-Department Transfers
15
💡 Clinical Workbench: Dedicated patient management interface for nurses and on-duty physicians. Combines core patient profiles, critical allergy warnings, emergency contact information (Query 1), and real-time admission/discharge status (Query 3). The search bar supports fast filtering by patient full name.
Current & Recently Discharged Patient List
| Full Name | Admission Date | Patient Status | Emergency Contact | Contact Phone | Allergy Warnings |
|---|
💡 Cross-Department Transfer Monitoring: Data extracted from the `TREATMENT_LOG` dataset (Query 4). Tracks complete in-hospital patient transfer histories, including handoffs from emergency departments or general wards to specialized units. Critical for monitoring complex condition progression and clarifying inter-department care handover responsibilities.
Recent In-Hospital Transfer Records (Top 5)
💡 Task 7 Classification Analysis: This query evaluates whether physician experience level can predict patient discharge outcomes. Doctors are grouped into three tiers (Junior/Intermediate/Senior). For each tier, the majority outcome becomes the prediction rule.
📊 Classification Results: Experience Level → Discharge Prediction
| Experience Level | Predicted Outcome | Accuracy (%) |
|---|
Interpretation: R = Recovered | T = Transferred | D = Deceased
📈 Outcome Distribution by Experience Level
Accuracy values shown on tooltip when hovering over each bar group
💡 Task 7 Regression Analysis - Length of Stay (LOS) Prediction: Using linear regression, we model hospital stay duration based on patient characteristics.
Formula: LOS = β₀ + β₁·Age + β₂·BMI + β₃·Severity + β₄·Allergies
Interpretation: Severity has the strongest positive impact on LOS (+2.45 days per severity level). Allergies show a slight negative coefficient (patients with allergies tend to have shorter stays in this model).
🧮 LOS Prediction Calculator
Enter patient characteristics to predict hospital stay duration
Predicted Length of Stay
--.--
days
Formula: LOS = -2.63 + 0.0028·Age + 0.0118·BMI + 2.4469·Severity - 0.0225·Allergies(1=yes)
📋 Regression Model Summary
Total Samples
567
Mean LOS
4.41 days
Age Coefficient (β₁)
+0.0028
BMI Coefficient (β₂)
+0.0118
Severity Coefficient (β₃)
+2.4469
Clinical Insight: Severity is the dominant predictor of LOS. Each 1-point increase in severity adds ~2.45 days to hospital stay.