BSC.

Computer Science

Student Name: Qandeel Maqsood

Student ID No: PN1141131

Subject Name: Applied Data Science

Subject Code:
Date: 16-11-24
 Table of contents
Introduction to Project Plan and Schedule
 Objective of the Project Plan
 Phases of the Project Plan
 Benefits of a Clear Project Plan
Research Background About the Problem
 Key Challenges in Healthcare Data
 The Role of Business Intelligence Dashboards
 Why XYZ Healthcare Needs a BI Dashboard
 Technological Advancements in BI Tools
Outline Specification of the Problem
 Introduction
 Key Issues in the Current System
 Need for a Business Intelligence Dashboard
Design Documentation for the Software System
 Overview of the Dashboard Design
 System Architecture
 Key Features of the Dashboard
Implementation Overview Reporting the Use of Tools
 Overview of the Tools and Technologies
 Power BI Selection and Justification
 Data Preparation and ETL Process
 Dashboard Development
 Testing and Deployment
Evidence
 Data cleaning
 Dashboard
Conclusion
References

G-drive Link:
https://drive.google.com/drive/folders/1zubbgIFj91locK4AxQPI4pDV24CnCoGL
Introduction to Project Plan and Schedule

In today’s healthcare industry, organizations are inundated with vast amounts of data from
patient records, operational metrics, and financial systems. Without an effective means to
process and analyze this data, its potential to improve decision-making and operational
efficiency remains unrealized. Business Intelligence (BI) dashboards have emerged as a powerful
solution to these challenges. This project focuses on developing an interactive BI dashboard for
XYZ Healthcare to enhance data visualization and facilitate strategic decision-making.

Objective of the Project Plan

The primary objective of this project is to create a centralized, interactive BI dashboard that
provides real-time insights into key performance indicators (KPIs). The goal is to empower XYZ
Healthcare to identify trends, monitor progress, and address inefficiencies in a timely manner. A
structured plan is essential for ensuring successful implementation, adherence to deadlines, and
alignment with organizational goals.

Phases of the Project Plan

To streamline the process and deliver a robust solution, the project is divided into the following
key phases:

1. Research and Tool Selection The initial phase involves identifying the most suitable BI
tool for the project. Tableau and Power BI are both evaluated for their ability to handle
large datasets, provide customizable dashboards, and integrate with healthcare data
systems.
2. Data Collection and Preparation Data is collected from XYZ Healthcare’s internal
systems, including patient records, operational logs, and financial reports. This data
undergoes cleaning, formatting, and normalization to ensure it is ready for integration
into the dashboard.
3. Dashboard Design In this phase, the user interface is designed to ensure it is intuitive
and visually appealing. The dashboard will include a mix of visualizations, such as bar
charts, line graphs, and heatmaps, to present data effectively.
4. Testing and Iteration Once the dashboard is developed, it undergoes rigorous testing to
identify and resolve any technical or functional issues. Feedback from key stakeholders is
incorporated to refine the design and functionality.
5. Deployment and Maintenance The final phase involves deploying the dashboard for use
by healthcare professionals and the board of directors. A maintenance plan is outlined to
ensure the dashboard remains updated with the latest data and functionalities.

Benefits of a Clear Project Plan

A well-structured plan provides several advantages:

 Efficient Resource Management: Allocates time, tools, and expertise effectively.

  Risk Mitigation: Identifies potential challenges, such as data integration issues, and
provides mitigation strategies.
 Stakeholder Engagement: Regular updates during key phases keep stakeholders
informed and aligned with project goals.
 Quality Assurance: Ensures the final product meets expectations and provides accurate
insights.

Research Background About the Problem

The healthcare sector is one of the most data-intensive industries, with organizations managing
vast amounts of information daily. From patient records and clinical metrics to financial
performance and resource allocation, healthcare providers face the challenge of deriving
actionable insights from this data. Without proper tools to analyze and visualize these datasets,
the potential for improving patient care and operational efficiency remains untapped.

Key Challenges in Healthcare Data

1. Fragmented Data Systems: Healthcare organizations often use multiple platforms to

store data, leading to silos and difficulties in data integration.
2. Operational Inefficiencies: The absence of real-time insights hampers effective decision-
making, causing delays in resource allocation and patient care.
3. Escalating Costs: Rising healthcare costs necessitate tools that help identify inefficiencies
and optimize resource usage.
4. Regulatory Compliance: Healthcare providers must adhere to stringent regulations,
requiring accurate and transparent reporting.

The Role of Business Intelligence Dashboards

Business Intelligence dashboards provide a solution to these challenges by consolidating data

into an accessible, interactive format. These dashboards are instrumental in:

 Real-Time Decision-Making: Offering stakeholders immediate insights into operational

and clinical performance.
 Trend Analysis: Highlighting patterns in-patient admissions, resource utilization, and
financial metrics.
 Enhanced Reporting: Automating data aggregation and visualization for compliance
and performance reporting.

Why XYZ Healthcare Needs a BI Dashboard?

XYZ Healthcare faces similar challenges in managing its data efficiently. By implementing a BI
dashboard, the organization can:
  Improve Patient Outcomes: Monitor KPIs like readmission rates, patient satisfaction,
and average treatment time.
 Optimize Resources: Visualize data on staff allocation, equipment usage, and facility
occupancy to identify inefficiencies.
 Support Strategic Decisions: Provide the board with a comprehensive view of
operations to guide long-term planning.

Technological Advancements in BI Tools

The adoption of BI dashboards in healthcare has been driven by advancements in data

visualization technologies. Key trends include:

1. AI Integration: Advanced analytics and predictive models powered by AI help forecast

patient volumes and resource needs.
2. Cloud Solutions: Cloud-based BI platforms offer enhanced scalability and accessibility.
3. Interactive Dashboards: Tools like Tableau and Power BI provide customizable
dashboards that allow users to drill down into specific data points.

Outline Specification of the Problem

Introduction

XYZ Healthcare, like many organizations in the healthcare industry, relies on data to make
informed decisions. However, managing and extracting actionable insights from vast amounts of
data can be a significant challenge. The data is spread across multiple systems, ranging from
electronic health records (EHR) to financial databases, and it is often difficult to present this data
in a comprehensive, easily interpretable manner. Currently, XYZ Healthcare lacks an integrated
platform to visualize key performance indicators (KPIs) and track trends in real time, which is a
critical need for efficient decision-making.

Key Issues in the Current System

1. Data Fragmentation
The healthcare organization’s data is stored in different silos, such as patient databases,
staff performance data, and hospital logistics data. This fragmentation makes it time-
consuming and inefficient to access, aggregate, and analyze data across the various
departments.
2. Lack of Real-Time Insights
Stakeholders do not have access to real-time data, which hampers their ability to make
timely decisions based on current performance or trends. Currently, data reports are static
and outdated by the time they reach decision-makers.
3. Manual Reporting Process
The healthcare system still relies on manually generated reports, which are prone to
human error and take up valuable time that could be spent on analysis and strategic
 initiatives. This manual approach limits the organization's ability to quickly adapt to
changing circumstances.
4. Ineffective KPI Monitoring
KPIs such as patient wait times, hospital bed occupancy rates, and patient satisfaction
scores are critical in evaluating healthcare performance. However, without a dedicated
tool for visualizing these metrics, stakeholders lack a clear understanding of the hospital's
performance, limiting their ability to make informed decisions.

Need for a Business Intelligence Dashboard

A Power BI dashboard can address these issues by providing an interactive, real-time, and
consolidated platform for data visualization. It will help XYZ Healthcare:

 Centralize Data: Integrate various data sources into a single, easy-to-access platform.
 Provide Real-Time Insights: Display up-to-date data through real-time connections to
various data sources.
 Improve Decision-Making: Empower stakeholders to make informed decisions with
dynamic and easily interpretable visualizations of the organization’s key metrics.
 Enhance Operational Efficiency: Automate data reporting and analysis, freeing up
resources for strategic tasks.

The goal is to transform the healthcare data landscape at XYZ Healthcare by creating a
centralized, real-time BI dashboard that provides all stakeholders with easy access to key data,
enabling them to make better, more informed decisions.

Design Documentation for the Software System

Overview of the Dashboard Design

The Power BI dashboard will serve as a centralized platform for aggregating, visualizing, and
analyzing key healthcare metrics in real-time. This section outlines the design decisions made for
the dashboard, focusing on system architecture, the flow of data, and the dashboard’s key
features.

System Architecture

1. Data Integration Layer

o Data Sources: The dashboard will integrate data from various sources, including
patient records, hospital occupancy data, staff performance, financial records, and
treatment outcomes. Data will be pulled from these diverse sources using Power
BI’s Power Query tool, which supports easy connections to SQL databases, Excel
files, cloud services (such as Azure), and web-based APIs.
 o ETL Processes: Power Query will be used to clean, transform, and prepare data
for analysis. This step will involve removing duplicates, normalizing data
formats, and dealing with missing or inconsistent data.
2. Data Storage
o Power BI Service: The data will be stored in the Power BI cloud service,
allowing for scalability and secure, easy access from multiple devices. The cloud
service will also allow real-time data updates, ensuring that stakeholders always
have access to the latest information.
3. Data Model
o Data Modeling: Power BI’s data modeling feature will be used to link and
structure data from different sources, creating relationships between tables to
allow for cross-source analysis. Data will be transformed into a data model that
supports aggregation, filtering, and detailed analysis.

Key Features of the Dashboard

1. Interactive Visualizations
o Charts: The dashboard will feature a range of visualizations such as bar charts,
line graphs, pie charts, and heat maps to display data such as patient wait times,
bed occupancy rates, and department performance.
o Interactive Filters: Users will be able to filter data by parameters such as date,
department, patient demographics, and geographic location. This allows
stakeholders to drill down into specific metrics or view the data at different
granular levels.
2. Real-Time Data Updates
o The dashboard will connect to live data feeds to ensure real-time data
visualization. For instance, data on hospital occupancy will be updated
continuously, providing the most up-to-date picture of available resources.
3. User Accessibility
o The dashboard will be designed to be intuitive and accessible to users with
varying levels of technical expertise. The layout will be simple and user-friendly,
enabling stakeholders to easily navigate between visualizations and derive
insights without advanced training.
4. Role-Based Access Control
o User permissions will be implemented to ensure that sensitive data is only
accessible by authorized personnel. Access to certain data sets, such as financial
information, will be restricted to key personnel based on their role within the
organization. (Biswal, 2024)

Implementation Overview Reporting the Use of Tools

Overview of the Tools and Technologies

For the development of the Business Intelligence (BI) dashboard at XYZ Healthcare, Power BI
was chosen as the primary tool due to its user-friendly interface, robust data integration
capabilities, and flexibility. Below is an overview of the implementation process, including the
use of Power BI and other tools in the system.

Power BI Selection and Justification

Power BI was selected for this project for several reasons:

 Ease of Use: Power BI offers a highly intuitive drag-and-drop interface for creating data
visualizations, making it accessible even to users without a technical background.
 Data Connectivity: Power BI supports integration with a wide range of data sources,
including SQL databases, Excel files, web services, and cloud platforms. This allows
XYZ Healthcare to consolidate data from various sources into a single, unified platform.
 Real-Time Data: Power BI supports live data feeds, which is essential for tracking
healthcare metrics such as hospital bed occupancy and patient wait times in real-time.
 Scalability: Power BI’s cloud-based service can scale with the growing needs of XYZ
Healthcare, supporting the addition of new data sources and users.

Data Preparation and ETL Process

This section outlines the implementation process of the Business Intelligence (BI) solution for
XYZ Healthcare, specifically focusing on data collection, cleaning, transformation, and
visualization using Power BI. The primary goal was to create an interactive dashboard that
enables stakeholders to make informed decisions by providing insights into key healthcare
metrics. (Franklin, 2022)

Step 1: Data Collection

Data collection involved extracting multiple datasets from different sources within XYZ
Healthcare. The data used in the project primarily came from CSV files containing various types
of records:

 Patient Records: Information on patient demographics, treatment details, and health

outcomes.
 Financial Data: Data relating to hospital revenues, treatment costs, and operational
expenses.
 Operational Metrics: Information on hospital performance, such as bed occupancy rates,
patient wait times, and staffing levels.

These CSV files were provided by XYZ Healthcare’s internal systems, including the CRM
(Customer Relationship Management) system, ERP (Enterprise Resource Planning) system,
and other databases. These data files served as the foundation for analysis and were initially
imported into Jupyter Notebooks for cleaning and transformation before being loaded into Power
BI for visualization.
Step 2: Data Cleaning

Data cleaning was conducted using Jupyter Notebooks, an interactive environment that allowed
for efficient manipulation and preparation of data using Python and relevant libraries such as
pandas and numpy (Agarwal).

Missing Value Handling:

 Missing values in the datasets were identified and handled by either imputing or
removing them based on the context. For example, missing numerical values (e.g., patient
wait times) were imputed using the mean or median, while rows with critical missing
data (e.g., patient IDs) were removed.

Outlier Detection and Treatment:

 Outliers were detected using statistical methods such as Z-scores and the Interquartile
Range (IQR) method. Outliers, such as unusually high hospital occupancy rates or
abnormal wait times, were either corrected or removed to ensure they did not distort the
analysis.

Data Normalization:

 Data normalization was performed to ensure that all datasets were consistent and
comparable. For example, time-based data (such as patient wait times) was standardized
into minutes, and categorical data (such as department names) was encoded to follow a
consistent format across all datasets.

Data Type Conversion:

 Data types were converted to appropriate formats to ensure compatibility during

analysis. Numerical fields such as treatment costs were cast to float types, categorical
data (e.g., department names, patient conditions) were converted into string or categorical
types, and date fields were converted into datetime objects for easier handling.

The cleaning process in Jupyter Notebook helped ensure that the data was accurate and ready for
transformation, making it suitable for analysis in Power BI.

Step 3: Data Transformation and Loading

Once the data was cleaned, it was ready for transformation and loading into Power BI. The
transformation process involved several key steps:

Data Aggregation:

 The data was aggregated to summarize information at different levels. For example,
financial data was aggregated by treatment type, and operational metrics were aggregated
 by time period or department. This helped simplify the data and provided a high-level
overview of the hospital's performance.

Data Filtering:

 Data filtering was applied to focus on specific metrics of interest. For example,
operational data was filtered to focus on key metrics like bed occupancy rates or average
patient wait times by department.

Data Enrichment:

 To enhance the analysis, additional data sources, such as regional health trends and
public health data, were integrated into the dataset. This provided a more comprehensive
view of the hospital's performance in relation to external factors.

Loading Data into Power BI:

 After cleaning and transforming the data, it was imported into Power BI using the
platform’s built-in connectors for CSV files. The datasets were loaded into the Power BI
data model, where relationships between different tables were defined, allowing for
cross-analysis between patient demographics, financial metrics, and operational data.

Data Modeling in Power BI:

 A data model was created in Power BI, establishing relationships between different
tables, such as linking patient records to treatment data and financial information.
The model was designed to facilitate easy querying and analysis, allowing users to
examine various aspects of the data in detail.

Step 4: Dashboard Creation in Power BI

Once the data was successfully loaded into Power BI, the next step was to create an interactive
and visually appealing dashboard to display the key insights (Carl, 2024).

Visualizations:

 The Power BI dashboard was designed to showcase the most important metrics for XYZ
Healthcare, such as patient satisfaction rates, hospital occupancy rates, and treatment
costs. Various types of visualizations were used, including bar charts, line graphs, and
pie charts, to represent trends and distributions across different dimensions.

Interactive Filters:

 Power BI’s interactive filtering capabilities allowed users to drill down into specific data
points. Users could filter data by time period, department, or treatment type, enabling
them to gain deeper insights into specific aspects of hospital operations.
Real-Time Data:

 The dashboard was set up to update in real-time, ensuring that the most current data was
always displayed. For example, patient wait times and hospital bed occupancy rates
were continuously refreshed, giving stakeholders up-to-date information for timely
decision-making.

Dashboard Development

1. Visualizations
o The dashboard will include a variety of visual elements, such as bar charts, line
graphs, and geographic heatmaps, to visualize KPIs like bed occupancy, treatment
efficiency, and patient satisfaction.
o DAX (Data Analysis Expressions) will be used to create custom calculations and
aggregations to support more advanced metrics, such as patient recovery rates or
department performance comparisons.
2. Interactivity
o The dashboard will feature slicers and drill-down capabilities, allowing users to
interact with the data and filter visualizations by parameters such as department,
date range, and patient category.
3. User Interface
o The layout of the dashboard will be simple and intuitive, with a clear focus on key
metrics. Users will be able to easily navigate through different sections of the
dashboard, accessing detailed insights with just a few clicks.

Testing and Deployment

Once the dashboard is developed, it will undergo thorough testing to ensure the accuracy and
reliability of the data visualizations. Testing will include:

 Data Accuracy Testing: Ensuring that all data points displayed in the visualizations are
accurate and up-to-date.
 Usability Testing: Ensuring that the dashboard is easy to use for both technical and non-
technical users.
 Performance Testing: Ensuring that the dashboard performs efficiently, even with large
datasets.

Evidence

Data Cleaning
Dashboard

Conclusion
In conclusion, the successful implementation of a Business Intelligence (BI) dashboard for XYZ
Healthcare will serve as a transformative tool for improving decision-making, operational
efficiency, and overall performance. The structured approach outlined in this project plan, which
includes clear phases such as research, data collection, dashboard design, testing, and
deployment, ensures that all critical aspects of the project are addressed with precision and care.
A clear project plan is essential for guiding the development process, managing resources
effectively, and mitigating risks. By leveraging the power of data visualization and real-time
insights, XYZ Healthcare will be empowered to make informed decisions that drive both short-
term improvements and long-term sustainability. Ultimately, this BI dashboard will play a vital
role in advancing the organization’s ability to navigate challenges, enhance patient care, and
optimize operational workflows in a dynamic healthcare environment.
References
Agarwal, M. (n.d.). Retrieved from https://realpython.com/python-data-cleaning-numpy-pandas/

Biswal, A. (2024, June 5). Retrieved from https://www.simplilearn.com/tutorials/power-bi-

tutorial/power-bi-dashboard

Carl, T. (2024, March 29). Retrieved from https://www.newhorizons.com/resources/blog/power-bi-

benefits

Franklin, R. (2022, July 12). Retrieved from https://www.precisely.com/blog/big-data/etl-data-

preparation-differences