👋 Hi, I'm Khyati Busa

Healthcare Data Analyst | Business Analyst

🦷 Dentist Turned Data Analyst
💻 Excel, Tableau, SQL, Python, R
📍 Boston, MA USA


Skills

  • Tableau

  • SQL

  • Excel

  • Python

  • R

  • EHR Data Management

  • HEDIS, STARS, HIPAA, HL7

  • Medical Coding: ICD-11, CPT

  • Healthcare Process Improvement

  • Project Management


Projects

🏥 Improving hospital efficiency through data analysis

SQL | Descriptive + Exploratory Data Analysis
◾ Queried and analyzed patient demographics, diagnosis patterns, appointment trends, lab test utilization, and readmission rates using complex SQL queries.
◾ Utilized JOINs, CASE, COUNT(), AVG(), EXTRACT(), and DATE_DIFF() to clean, filter, and categorize healthcare data for actionable insights.
◾ Identified high-risk patient groups, common diagnoses, peak appointment times, and readmission trends to support hospital resource allocation and patient care improvements.

.

Patient satisfaction dashboard with Tableau

Tableau + SQL | Data Visualization
◾ Developed an interactive Tableau dashboard using HCAPS survey data from CMS to analyze hospital performance and patient satisfaction.
◾ Created filters, bar charts, and comparison visualizations to assess hospital ratings by state, size, and individual selection.
◾ Integrated SQL for data cleaning and transformation
◾ Designed a survey response rate tracker to evaluate hospital engagement and patient feedback trends.
◾ Identified top-performing and underperforming hospitals using visual analytics

.

Diagnosis and predictors of heart disease

Python | Exploratory Data Analysis & Risk Prediction
◾ Analyzed UCI Heart Disease Dataset to identify key risk factors using Python (Pandas, NumPy, Seaborn, Matplotlib).
◾ Discovered higher heart disease prevalence in males and seniors (60+), with asymptomatic chest pain being most common.
◾ Identified correlations between fasting blood sugar, ST segment slope, and heart disease risk.
◾ Handled missing data, detected outliers, and generated insights to aid early diagnosis and prevention.

.

🏥 Analyzing Hospital Data with EXCEL

EXCEL| Exploratory Data Analysis
◾ Created interactive dashboards using Excel Pivot Tables, Charts, and Conditional Formatting.
◾ Evaluated hospital performance metrics like avoidable conditions, revenue trends, and cost per patient.
◾ Structured and cleaned large datasets from data.gov for meaningful analysis.

.

Emergency Visit Dashboard with Tableau

Tableau | Data Visualization & Dashboard Development
◾ Developed an interactive Tableau dashboard to analyze ER visits, patient demographics, wait times, satisfaction scores, and referral trends.
◾ Visualized age, race, and gender distribution, created heatmaps for peak visit times, and tracked key KPIs like average wait time (35.26 min) and satisfaction (4.99/10).
◾ Utilized calculated fields, filters, and parameters for dynamic insights, aiding hospital resource allocation and performance monitoring.

.

Clinical Quality Measure Mapping

HEDIS Guidelines (Diabetes Care Focus)
◾ Mapped HEDIS diabetes care measures using ICD-10-CM, CPT, CPT II, and LOINC codes.
◾ Simulated EHR data extraction for performance tracking and reporting.
◾ Aligned clinical data with value-based care and quality improvement goals.

.


About Me

👋 Hi, I'm Khyati, and welcome to my portfolio! My passion lies in uncovering meaningful insights from data to drive informed decision-making. With a background in healthcare and health informatics, I’ve combined my analytical skills with industry knowledge to transform raw data into actionable insights.I'm proficient in:SQL
Python
Tableau
Excel
Minitab
PostgreSQL
I’m looking to help organizations leverage data for improved healthcare outcomes and operational efficiency. Feel free to reach out to me at [email protected]!

Previous Experience

Clinical Operations Analyst
Little Smile Dental – Worcester, MA, U.S.A.
◾ Lead a patient experience analytics initiative by collecting, structuring, and analyzing patient feedback datasets to identify key service drivers, workflow gaps, and operational risks; deliver data-driven insights that inform leadership decision-making and service improvements.
◾ Perform theme-based analysis of structured patient review data (Friendly Staff, Pain-Free Care, Professional Care, Family Friendly) to quantify satisfaction trends and highlight improvement opportunities.
◾ Develop data reports and visual summaries to communicate performance insights and track key patient experience metrics for clinic leadership.
◾ Facilitate cross-functional collaboration between clinical providers, administrative staff, and leadership to evaluate operatory and hygiene workflows and implement data-driven process improvements.
◾ Maintain project communication, documentation, and KPI reporting related to workflow improvements, ensuring transparency and alignment across stakeholders.
Dental Assistant Analyst
Asarkof Dental, Lexington, MA, U.S.A.
◾ Assisted in the development and management of electronic health records (EHRs), ensuring accurate and secure data handling that complied with HIPAA regulations.
◾ Participated in a pilot program utilizing SQL to query and analyze large datasets, generating insights that informed supply chain optimization strategies, reducing waste and achieving a cost reduction of 15%.
◾ Conducted regular audits of patient data entries and system backups to maintain high standards of data integrity and reliability.
General Dentist
Sahaj Dental Clinic, Gujarat, India
◾ Performed a wide range of dental procedures, including examinations, cleanings, fillings, and extractions, ensuring high standards of patient care and satisfaction.
◾ Implemented a new patient data management system, increasing the efficiency of patient information retrieval by 30% and reducing administrative errors by 25%.
◾ Analyzed patient treatment data to identify trends and inefficiencies, leading to a redesigned appointment scheduling system that improved patient flow by 20%.

🏫 Evaluating School Success with Tableau Project

In the state of Massachusetts, there are approximately 953,748 high school students, which is almost a million kids! This means that decisions regarding high school education affect one-seventh of the state's population, which is 6.985 million.

Background

The state of Massachusetts asked to analyze their education data. The Secretary of Education is exploring ways to increase college attendance, including investing in building more schools. One idea is to lower the average class size in hopes of boosting the percentage of students attending college. Using Tableau we have taken a deep dive into the data.All the data is available to look at yourself on this Tableau DashboardThe original dataset can be found here 👉 Dataset

Analysis

I was stunned to learn that last year, 162,137 students did not graduate while 791,610 students successfully graduated.Another 205,818 did not attend college, while 585,791 students proceeded to pursue
higher education after graduation.
As per the state's request, I looked at the correlation of class size.The average class size for the entire state is 18, however, the top 5 schools have the largest class sizes. While class size may not be a significant contributor to student grades. There was more data to look at.

Next, we looked at who could help other schools.The top five schools in fourth-grade math proficiency are Hingham, Winchester, Lynnfield, Manchester, and Sherborn. Reaching out to these schools to discover how they are aiding their student's success will be important.

Then we identified schools needing the most urgent help.The schools that are currently struggling the most with their grade point average. These schools include Springfield Public, Boston Charter, Whaling City Junior, Springfield High, and Phoenix Academy. They could benefit greatly from some assistance and that would be a good place to start.Lastly, we found the most helpful piece of information!When it comes to struggling schools, an average of 77% of their students are economically disadvantaged. In contrast, only 13% of students in the top 5 schools face the same situation.According to the state, children who use one or more assistance programs are at a disadvantage.

Conclusion

I recommend reaching out to the high-performing schools and inquiring if they would be willing to share their successful strategies with the lower-performing schools.Instead of investing in building more schools, we should focus on improving or expanding the programs that help economically disadvantaged children. This would be a more effective use of resources.Lastly, It would be beneficial to contact colleges and share this information with them to explore if they're willing to assist high schools in enhancing the transition to college.If you enjoyed reading this. Please connect with me on LinkedIn. I'll be posting more like it. This project was done as part of the DAA Boot Camp Projects. Educational purpose.

Improving Hospital Efficiency Through Data analysis

Introduction

In this hypothetical scenario, I assume the role of a healthcare data analyst at a local hospital where data-driven decision-making is essential for optimizing resource allocation and improving patient care. The hospital's management team has expressed concerns about ensuring efficient service delivery and enhancing patient outcomes. They recognize the potential of leveraging patient data to gain actionable insights but require support in transforming this data into meaningful information. To address these concerns, I conducted a comprehensive analysis of the hospital's patient population, examining demographics, diagnosis patterns, appointment trends, lab test utilization, risk stratification for cardiovascular disease, and readmission rates. This analysis provides the hospital with a detailed understanding of their patient demographics and care needs, enabling informed strategic decisions that aim to enhance operational efficiency and patient care outcomes.

Analysis and Insight

👥Demographic Profile Analysis

• Objective: To understand the age and gender distribution of the patient population, providing insights into the demographic composition.• Methodology: Using SQL queries, I extracted data on patient ages and genders, grouped the results by gender, and categorized patients into different age groups.• Findings: The data revealed a balanced gender distribution with slight variations across age groups. The most populous age groups fell within 19-35 and 36-60 age brackets, suggesting a middle-aged demographic focus.• Implications: This demographic profile can inform the hospital’s resource allocation, allowing for age-appropriate services and staff training.

🏥Diagnosis and Demographic

• Objective: To identify the most prevalent diagnoses and analyze their distribution across different demographic groups (age and gender).• Methodology: Diagnoses data was linked to demographic data using patient IDs. Prevalence rates were calculated for each diagnosis, with further breakdowns by gender and age.• Findings: The most common diagnoses included hypertension, diabetes, and respiratory infections, with higher prevalence rates among older patients. Gender differences were evident in diagnoses like respiratory infections, which were more prevalent among males, while hypertension was more common among females.• Implications: This analysis helps inform targeted healthcare initiatives, enabling the hospital to tailor programs based on demographic characteristics and improve service efficiency.

📅Appointment Patterns

• Objective: To analyze common appointment times and their distribution throughout the day to improve operational planning.• Methodology: Appointment data was grouped by time intervals throughout the day, and the frequency of appointments was calculated for each interval.• Findings: Peak appointment times occurred between 9 AM and 11 AM, with a noticeable decline in the afternoon. Evening appointments were less common, indicating an opportunity to better utilize these slots or adjust staffing levels accordingly.• Implications: This information can help the hospital optimize staffing and resource allocation during peak hours and explore ways to increase appointment availability during less busy times.

🧪Common Lab Test

• Objective: To determine the most frequently ordered lab tests for resource allocation and clinical decision support.• Methodology: Lab test data was analyzed to identify the top lab tests ordered.• Findings: The most commonly ordered tests included Complete Blood Count (CBC), Lipid Panel, and Fasting Blood Sugar tests.• Implications: Understanding the demand for these tests helps the hospital allocate resources efficiently and ensures that essential supplies and personnel are readily available.

🔬Lab Test Insight

• Objective: To identify patients with abnormal fasting blood sugar levels for early intervention.• Methodology: Fasting blood sugar test results were filtered to identify cases with abnormal values, allowing for targeted follow-ups.• Findings: A significant portion of patients showed elevated fasting blood sugar levels, highlighting potential undiagnosed diabetes cases.• Implications: Early identification allows for timely interventions, potentially preventing more severe health issues. This can guide the hospital in developing patient education and lifestyle intervention programs.

❤️⚕️Risk Stratification for Cardiovascular Disease

• Objective: To categorize patients into high, medium, and low risk groups for cardiovascular disease based on smoking status and diagnoses.• Methodology: Patients were classified as high, medium, or low risk based on their smoking status and whether they had diagnoses of hypertension or diabetes.• Findings: A significant number of patients fell into the medium and high-risk categories, particularly among smokers with existing diagnoses of hypertension or diabetes.• Implications: These insights allow the hospital to prioritize care for high-risk patients, inform preventive healthcare initiatives, and allocate resources to cardiovascular health programs.

🔄Readmission Analysis

• Objective: To identify patients readmitted within 30 days, including the details of both the initial and readmission visits.• Methodology: Data on patient visits was analyzed to identify readmissions occurring within 30 days of the initial visit, detailing the reason for both visits and the time between them.• Findings: Readmission rates within 30 days were notable, with common readmission reasons including complications related to diabetes and respiratory issues.• Implications: Addressing readmission trends allows the hospital to enhance discharge planning, improve follow-up care, and implement targeted interventions for conditions with high readmission rates.

Conclusion and Recommendation

The analyses presented above provide a foundation for strategic decision-making aimed at improving patient care efficiency and optimizing resource allocation. Based on these insights, I recommend the following actions:• Develop age-specific programs and services to cater to the hospital’s middle-aged demographic.
• Enhance staffing and resources during peak appointment times and explore strategies to increase the utilization of off-peak hours.
• Prioritize high-risk cardiovascular patients for preventive care initiatives, such as smoking cessation programs and dietary counseling.
• Implement targeted discharge and follow-up protocols for conditions with high readmission rates, aiming to reduce overall readmissions and improve patient outcomes.
By leveraging these insights, the hospital can enhance service delivery, support patient-centered care, and achieve better clinical outcomes, ultimately strengthening its position as a data-driven healthcare provider.

Clinical Quality Measure Mapping – HEDIS Guidelines (Diabetes Care Focus)

Description

  • Mapped HEDIS clinical quality measures for diabetes management using ICD-10-CM, CPT, CPT II, and LOINC codes to define eligible populations, numerators, and exclusions.

  • Developed structured reference tables to support the identification of patients requiring HbA1c testing, nephropathy monitoring, retinal exams, and blood pressure control interventions.

  • Applied clinical informatics principles to simulate extraction of standardized data from EHRs, aiding in performance measurement and quality reporting.

  • Enhanced understanding of population health metrics and value-based care by aligning data elements with quality improvement and compliance initiatives.

HEDIS Criteria for eligible populationCode ClassificationAlpha/numerical CodeCode Description
Diabetes Type I or IIICD-10-CME10.XXXX, E11.XXXXUsed for diagnosis of Type I or II diabetes mellitus.
HbA1c TestCPT83036, 83037HbA1c test for laboratories.
HbA1c test level less than 7.0% (for specific population)CPT II3044FMost recent HbA1C level of less than 7 percent
HbA1c test level < 8.0CPT II3051FHbA1C level of greater than 7 and less than 8 percent
HbA1c test level > 9.0%CPT II3046FHbA1C level of greater than 9
Eye exam (retinal) performedCPT, CPT II, and ICDCPT II- 2022F, 2023F, 2024F CPT- 92227, 92229,92250 ICD- E10.31-E10.35Diabetic Retinal Screening with Eye Care Professional, Eye Exam with Evidence of Retinopathy, Retinal Screening, Type 1 diabetes mellitus with diabetic retinopathy
BP control (<140/90 mm Hg)CPT II3074F, 3075F, 3077F (Systolic), 3078F–3080F (Diastolic)Specific CPT codes are used to identify the patient with specific blood pressure. It measures the blood pressure control in patient with diabetes during the measurement year. (<140/90)
Medical attention for nephropathyCPT, CPT II, ICD-10-CMCPT- - 81000, 81001, 81002, 81003, CPT II- 3066F, 4010F, ICD- E08.21, E08.22, E08.29, E09.21, E09.22, E09.29, E10.21, E10.22Urinalysis for nephropathy screening or monitoring test or evidence of nephropathy, Documentation of treatment of nephropathy, ACE inhibitor/ ARB prescribed or currently being taken

Hospital Analysis Using EXCEL

Project Description

I developed an interactive Hospital Performance Dashboard in Microsoft Excel to analyze key performance indicators (KPIs) related to hospital operations, financials, and patient outcomes. This project was part of a proof-of-concept submission for a healthcare organization’s Business Intelligence (BI) system. The dashboard provides data-driven insights to help stakeholders improve patient care, optimize costs, and enhance operational efficiency.

Key Features & Insights

Readmission Rate Analysis
Identified top 10 surgical categories with the highest readmission rates, emphasizing the need for quality improvement in specific areas such as Alcohol/Chemical Dependency and Oncology.

Avoidable Conditions
Highlighted Low Back Pain as the leading cause of avoidable encounters, signaling potential inefficiencies in patient management.

Financial & Performance Comparison
Assessed Value-Based Incentive Payments across three hospitals, revealing that Coventry Hospital did not meet its goal while others performed better. Additionally, Sandy Coast Hospital maintained a lower per-patient cost compared to others.

Revenue Distribution & Trends
Visualized total paid amounts from 2012 to 2014, demonstrating a significant upward trend. A treemap chart illustrated how different income sources (Medicare, Medi-Cal, Private Coverage, etc.) contributed to overall revenue.

KPI Selection & Justification

I selected Readmission Rate as a key performance indicator (KPI) due to its direct impact on quality of care, hospital costs, and patient satisfaction. Tracking readmission rates helps hospitals identify care gaps, improve treatment protocols, and reduce unnecessary healthcare expenditures.

Technical Skills Utilized

  • Data Cleaning & Manipulation: Processed raw datasets from data.gov, ensuring proper integration for meaningful analysis.

  • Advanced Excel Functions: Used Pivot Tables, VLOOKUP, INDEX-MATCH, and Conditional Formatting to structure and analyze data effectively.

  • Data Visualization: Created bar charts, frequency distributions, treemaps, and trend analyses for intuitive data interpretation.

  • Business Intelligence & Healthcare Analytics: Provided actionable recommendations based on hospital financial performance, readmission trends, and patient care metrics.

Impact & Takeaways

This project reinforced my ability to leverage data analytics in healthcare to derive actionable insights. By combining financial, operational, and clinical data, I demonstrated how hospitals can optimize resources, enhance patient outcomes, and make informed decisions.