Chronic Kidney Disease Prediction Dashboard

Published November 25, 2025

Software Developer Janaka Alwis

Introduction

The Chronic Kidney Disease (CKD) Prediction Dashboard with Dashtera is a comprehensive interactive analytics tool designed to predict the likelihood of CKD in patients based on clinical, physiological, and demographic attributes. Built using Dashtera’s no-code platform, this dashboard transforms a real-world patient dataset into actionable insights through interactive charts, feature distributions, and predictive model outputs.

This dashboard helps healthcare practitioners and administrators to:

Identify key clinical and demographic factors associated with CKD.

Analyze lab test results and physiological measures to detect early warning signs.

Deploy data-driven interventions for early diagnosis and preventive care.

The dataset contains 400 patient records, each capturing 24 features including demographics, lab tests, medical conditions, symptoms, and the CKD diagnosis status (ckd / notckd).

Dataset

The CKD dataset captures a wide range of patient attributes relevant for kidney health. Missing values are present in several features, which were handled during preprocessing.

Feature	Description	Value / Unit
age	Patient age in years	Numeric
bp	Blood Pressure	mmHg
hemo	Hemoglobin	g/dL
pcv	Packed Cell Volume	%
bgr	Blood Glucose Random	mg/dL
bu	Blood Urea	mg/dL
sc	Serum Creatinine	mg/dL
sg	Specific Gravity (urine)	1.005–1.025
al	Albumin (urine)	0–5
su	Sugar (urine)	0–5
pc	Pus Cells	Normal/Abnormal
rbc	Red Blood Cells	Normal/Abnormal
pcc	Pus Cell Clumps	Present/Not Present
ba	Bacteria	Present/Not Present
htn	Hypertension	Yes/No
dm	Diabetes Mellitus	Yes/No
cad	Coronary Artery Disease	Yes/No
ane	Anemia	Yes/No
appet	Appetite	Good/Poor
pe	Pedal Edema	Yes/No
class	CKD status	ckd / notckd

Derived & Processed Fields

Age_Group: Child/Teen (0–18), Young Adult (19–35), Middle Age (36–50), Senior Adult (51–65), Elderly (66+)
Hemoglobin_Group, BP_Group, PCV_Group, BGR_Group, BU_Group, SC_Group: Categorized based on clinical ranges

About Dashtera

What is Dashtera?

Dashtera is a no-code, cloud-based dashboarding platform that allows users to connect data sources, perform transformations, and build interactive visualizations without writing code. It’s ideal for quick insights, KPI tracking, and storytelling with data.

Key Features

Connects to various data sources (CSV, Excel, APIs, etc.)
Wide range of chart types, including advanced statistical visuals
Interactive drill-downs and dynamic filters
Shareable dashboards with flexible layouts
Supports calculated fields and transformations
User-friendly drag-and-drop interface

Advantages Over Similar Tools

Extremely easy to use—minimal technical expertise required
Rapid dashboard creation and deployment
Suitable for both beginners and advanced users
Lightweight yet powerful compared to Tableau or Power BI

Dashboard

Demographics & Physiological Measures

This dashboard examines age, blood pressure, hemoglobin, and packed cell volume distributions and their relationship with CKD.

Dashboard provides an overview of patients’ age, blood pressure, hemoglobin, and packed cell volume distributions. Stacked bar charts highlight that CKD prevalence increases with age and is strongly associated with anemia and abnormal blood pressure levels. Box plots and gauge charts allow clinicians to quickly assess individual patient metrics, offering insights into the physiological conditions most linked to CKD.

Present in this visualization:

Blood Pressure Box Plot
Hemoglobin Gauge Chart (last patient)
BP Gauge Chart (last patient)
PCV Gauge Chart (last patient)
Age Box Plot, Age Histogram & Distribution

Hemoglobin Categories (Stacked Bar Chart) :

hemoglobin_group	class	total_patients
3–7 (Severe Anemia)	ckd	12
7.1–11 (Moderate Anemia)	ckd	99
11.1–12.9 (Mild Anemia)	ckd	68
13–17 (Normal)	ckd	25
13–17 (Normal)	notckd	131
>17 (High)	notckd	13

Blood Pressure Categories (Stacked Bar Chart) :

bp_group	class	total_patients
50–89 (Very Low)	ckd	156
50–89 (Very Low)	notckd	148
90–120 (Normal)	ckd	82
140–159 (High)	ckd	1
160–180 (Very High)	ckd	1

Packed Cell Volume Categories (Stacked Bar Chart) :

pcv_group	class	total_patients
09–20 (Severe Low)	ckd	9
21–30 (Moderate Low)	ckd	55
31–35 (Borderline Low)	ckd	52
36–40 (Near Normal)	ckd	43
36–40 (Near Normal)	notckd	8
41–50 (Normal)	ckd	21
41–50 (Normal)	notckd	108
51–54 (High)	ckd	3
51–54 (High)	notckd	30

Age Categories (Stacked Bar Chart):

age_group	class	total_patients
0–18	ckd	16
0–18	notckd	3
19–35	ckd	16
19–35	notckd	40
36–50	ckd	49
36–50	notckd	46
51–65	ckd	99
51–65	notckd	41
66+	ckd	62
66+	notckd	19

Lab Test Results

This dashboard analyzes blood glucose (BGR), blood urea (BU), and serum creatinine (SC).

Lab test results including blood glucose (BGR), blood urea (BU), and serum creatinine (SC) are visualized to identify patterns indicative of CKD. Gauge charts and histograms show that higher BGR, BU, and SC levels are more common in CKD patients, while distribution and box plots reveal variability and outliers. Categorization into clinical ranges provides actionable thresholds for early diagnosis.

BGR Categories:

bgr_group	class	total_patients
70–99 (Normal)	ckd	36
70–99 (Normal)	notckd	53
100–125 (Impaired)	ckd	49
100–125 (Impaired)	notckd	61
126–140 (High)	ckd	17
126–140 (High)	notckd	30
Unknown	ckd	110

BU Categories:

bu_group	class	total_patients
10–20 (Normal)	ckd	23
10–20 (Normal)	notckd	29
21–30 (Borderline)	ckd	32
21–30 (Borderline)	notckd	36
31–40 (Moderate High)	ckd	31
31–40 (Moderate High)	notckd	33
41–50 (High)	ckd	22
41–50 (High)	notckd	46
Unknown	ckd	129

SC Categories :

sc_group	class	total_patients
0.4–0.9 (Normal)	ckd	20
0.4–0.9 (Normal)	notckd	83
1.0–1.2 (Borderline High)	ckd	25
1.0–1.2 (Borderline High)	notckd	62
1.3–1.4 (High)	ckd	15
Unknown	ckd	178

Urine-Related Tests

Urine analysis features such as specific gravity, albumin, sugar, pus cells, red blood cells, pus cell clumps, and bacteria are displayed to detect kidney function abnormalities. Vertical bar charts and treemaps distinguish CKD from non-CKD patients, highlighting that abnormal albumin levels, pus cells, and RBC irregularities are strong indicators of CKD. These visualizations support both patient-level assessment and population-level trends.

Specific Gravity (SG) :

sg_group	class	total_patients
1.005	ckd	7
1.010	ckd	84
1.015	ckd	75
1.020	ckd	31
1.020	notckd	75
1.025	ckd	11
1.025	notckd	70

Albumin (AL) :

albumin_group	class	total_patients
0 (Normal)	ckd	54
0 (Normal)	notckd	145
1 (Trace)	ckd	44
2 (Mild)	ckd	43
3 (Moderate)	ckd	43
4 (Severe)	ckd	24
5 (Very Severe)	ckd	1

Sugar (SU) :

sugar_group	class	total_patients
0 (Normal)	ckd	145
0 (Normal)	notckd	145
1 (Trace)	ckd	13
2 (Mild)	ckd	18
3 (Moderate)	ckd	14
4 (Severe)	ckd	13
5 (Very Severe)	ckd	3

Medical Conditions & Symptoms

This dashboard examines comorbidities (hypertension, diabetes, coronary artery disease, anemia) and symptoms (appetite, pedal edema) in CKD patients. Spider charts reveal the proportion of CKD patients with each condition, while stacked bar charts show symptom prevalence. The visualizations emphasize that CKD frequently coexists with hypertension and diabetes, and that poor appetite or pedal edema are common clinical indicators.

Conditions (Spider Chart)

feature	pct_ckd_yes	pct_ckd_no	pct_notckd_no
Hypertension	36.93	25.88	37.19
Diabetes Mellitus	34.17	28.39	37.19
Coronary Artery Disease	8.54	54.27	37.19
Anemia	15.03	47.61	37.34

Symptoms (Stacked Bar Charts)

Appetite: Poor appetite: 82 CKD vs 1 notCKD
Pedal Edema: Yes: 76 CKD vs 1 notCKD

Overall Insights & Recommendations

Demographic Focus: Elderly and middle-aged adults are at higher CKD risk.
Lab Indicators: Hemoglobin, PCV, BGR, BU, and SC are strong predictive features.
Urine Analysis: Albumin, SG, and RBC abnormalities strongly correlate with CKD.
Medical Conditions & Symptoms: Hypertension, diabetes, anemia, poor appetite, and pedal edema are key indicators.
Dashboard Utility: Enables doctors to input patient data, visualize distributions, and get real-time CKD predictions with interpretability via feature importance and SHAP values.

Conclusion

The CKD Prediction Dashboard with Dashtera transforms a complex multivariate dataset into a user-friendly, interactive platform for early detection of chronic kidney disease. By integrating demographic, physiological, lab, urine, and symptom data, healthcare professionals can identify high-risk patients, improve early intervention, and make data-driven clinical decisions.

Using Dashtera’s no-code framework, the dashboard combines predictive analytics, visualization, and interpretability to support actionable healthcare insights.

Category: