Project Milestone Data Plan: Analyzing Synthetic Patient Data for Clinical Insights

Overview

This milestone assignment allows you to receive feedback on your data plan before proceeding with your midterm paper. While this assignment is optional and not graded, completing it will provide valuable guidance to strengthen your final midterm submission.

Purpose

The purpose of this milestone is to help you:

Refine your research question from Discussion Assignment 2 (ASSIGNMENT ATTACHED)

Identify appropriate variables from the dataset for your analysis

Develop a preliminary plan for data analysis

Receive constructive feedback before investing significant time in your midterm paper

Instructions

Research Question Selection (1 paragraph)

Select one research question from the options you developed in Discussion Assignment 2

Refine your question to ensure it is clear, focused, and answerable using the provided dataset

Briefly explain why you selected this particular question and its relevance to nursing informatics and healthcare quality

Variable Selection (1-2 paragraphs)

Identify the specific variables from the dataset you plan to include in your analysis

For each variable, provide:

The exact name of the variable as it appears in the dataset

The type of variable (categorical, continuous, etc.)

A brief explanation of why this variable is relevant to your research question

Identify your dependent (outcome) and independent (predictor) variables

Proposed Data Analysis Plan (1-2 paragraphs)

Describe the descriptive statistics you plan to calculate

Identify any measures of association or simple statistical tests you plan to perform

Explain how these analyses will help you address your research question

Note any potential data quality issues you anticipate and how you plan to address them

Submission Format

Please submit your Data Plan as a 1-2 page document (double-spaced, 12pt font) with the following sections clearly labeled:

Research Question

Variable Selection

Proposed Data Analysis Plan

Example Structure

Research Question:

Is there an association between polypharmacy (5+ medications) and emergency department utilization among patients in the dataset? Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

Variable Selection:

Dependent variable: EmergencyVisits (continuous) – This represents the number of emergency department visits and will serve as my outcome variable to measure healthcare utilization.

Independent variable: TotalMedications (continuous) – This will be recoded into a binary variable (Polypharmacy: Yes/No) where 5+ medications = Yes.

Covariates:

Age (continuous) – Included because older patients may have more medications and higher ED utilization

ChronicConditionCount (continuous) – Included to account for disease burden

InsuranceType (categorical) – Included because insurance coverage may influence both medication access and ED utilization patterns

Proposed Data Analysis Plan:

I will first calculate descriptive statistics for all variables, including means and standard deviations for continuous variables (EmergencyVisits, Age, ChronicConditionCount) and frequencies for categorical variables (Polypharmacy, InsuranceType). I will then compare the mean number of emergency visits between patients with and without polypharmacy using an independent samples t-test. If the data are not normally distributed, I will use a non-parametric alternative like the Mann-Whitney U test. I will also create a contingency table to examine the relationship between polypharmacy and high ED utilization (defined as 2+ visits) and perform a chi-square test to assess this association.

Feedback Process

Your instructor will provide feedback on:

The clarity and feasibility of your research question

The appropriateness of your selected variables

The suitability of your proposed statistical analyses

Suggestions for strengthening your approach for the midterm paper

Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

Patient Age and Hospital Readmissions

Research Question

Does patient age influence the likelihood of hospital readmission within 30 days?

The research question that will be answered in this study will be to determine any relationship between the age of the patient and 30-day readmission to the hospital. I have chosen this question because readmission to hospitals is one of the most important quality indicators in healthcare and is frequently related to the insufficiency of discharge plans or the absence of proper follow-ups. One of the well-known factors that affect the complexity of recovery and care is age, particularly in older adults. This question is especially applicable in nursing informatics to the development of predictive models and clinical decision support systems that are proactive in identifying the high-risk patients. Knowledge of this relationship can aid in the optimization of the discharge process and the facilitation of specific intervention strategies, which will eventually enhance the quality of healthcare and lower expenses.

Variable Selection

I will apply the following variables of the synthetic patient dataset to analyze this research question:

Age

Name of variable in the data: Age

Type: Continuous

Age is the first independent variable and a possible predictor of readmission. Patients with age-related issues usually experience more comorbidities and functional limitations that can put them at risk of readmission.

30-Day Readmission

Name of variables in dataset: Readmission30Days

Type: Categorical (Binary: Yes/No)

Justification: This is the dependent variable. It specifically deals with the research finding: was the patient readmitted back to a hospital within 30 days of being discharged? There will be a binary classification in order to analyze the probability that can be attributed to various groupings of age.

Chronic Conditions Count

Name of the variable in the dataset: ChronicConditionCount

Type: Continuous

Justification: This was used as a covariate since patients with multiple chronic conditions would have a high risk of developing complications during the post-discharge period, regardless of age. By controlling this, one isolates the effect of age.

Discharge Destination

Name of variable in data: DischargeDestination

Type: Categorical such as home, skilled nursing facility, rehabilitation

Rationale: The place of discharge can be a risk factor for readmission. Such as readmission rates may be lower in patients discharged to skilled nursing facilities as compared to those discharged to home.

Insurance Type

Name of variable in data set: InsuranceType

Type: Categorical such as Medicare, Medicaid, private, etc.

Rationale: Type of insurance might have an impact on access to post-discharge facilities and post-discharge care, which can affect the readmission outcomes. This inclusion will be used to control the socioeconomic factors.

Overall, the dependent variable will be Readmission30Days, whereas the primary independent variable will be Age. To enhance the validity of the analysis, other covariates are ChronicConditionCount, DischargeDestination, and InsuranceType. Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

Data Analysis Plan Proposed

I will start by performing a descriptive statistic of all the variables. I will compute means, standard deviations, and ranges of continuous variables (age, chronic condition count). I will compute frequencies and percentages in the case of categorical variables (Readmission30Days, DischargeDestination, InsuranceType).

To determine the correlation between age and readmission, I will apply the logistic regression analysis because the outcome variable, readmission, is binary (Miswan et al., 2021). Age shall be handled as a continuous predictor. Some of the covariates that will be controlled in the logistic regression model include the number of chronic conditions, discharge destination, and type of insurance.

I will also carry out a bivariate analysis prior to the regression. Specifically: Chi-square test to test the unadjusted relationship of age groups (e.g., <65 vs. >=65) and readmission. A t-test or a Mann-Whitney U test to compare the mean age of readmitted and non-readmitted groups based on the assumptions of normality. The possible data quality problems can be; the lack of values in Readmission30Days (or Age), which I will evaluate through frequency checks. In case of a low level of missing data, I will apply listwise deletion, and in case of a high level of missing data, I will take the case of multiple imputation (Woods et al., 2024). Outliers in age like unrealistic ages such as 120+ will be considered and possibly rejected.

Such analyses will assist in establishing whether age is a factor in the risk of 30-day readmission independent of clinical and demographic characteristics. The findings will justify specific interventions in the discharge planning and transitional care, particularly among the elderly.

References

Miswan, N. H., Chan, C. S., & Ng, C. G. (2021). Hospital readmission prediction based on improved feature selection using grey relational analysis and LASSO. Grey Systems: Theory and Application, 11(4), 796-812. https://www.emerald.com/insight/content/doi/10.1108/gs-12-2020-0168/full/html

Woods, A. D., Gerasimova, D., Van Dusen, B., Nissen, J., Bainter, S., Uzdavines, A., … & Elsherif, M. M. (2024). Best practices for addressing missing data through multiple imputation. Infant and Child Development, 33(1), e2407. https://onlinelibrary.wiley.com/doi/abs/10.1002/icd.2407

Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

Analyzing Synthetic Patient Data for Clinical Insights

Based on the synthetic patient dataset and guided by the Module materials, I will develop three potential research questions aimed at identifying patterns that can support clinical decision-making and population health initiatives.

Research Question

Does patient age influence the likelihood of hospital readmission within 30 days?

The key variables

Independent: Age

Dependent: Readmission within 30 days

Rationale

Age is a known risk factor for readmission. Quantifying this relationship can help target high-risk age groups for follow-up care.

Clinical Relevance

The insights can help support discharge planning and early intervention strategies, particularly for geriatric patients, aligning with quality improvement outcomes such as reduced readmission rates.

Research Question

Is there a difference in the average length of stay based on the discharge destination, such as home or a skilled nursing facility?

The key Variables

Independent: Discharge Destination

Dependent: Length of Stay

Rationale

This relationship can reveal patterns in patient acuity and recovery needs.

Clinical Relevance

The results can guide nursing assessments and resource allocation for discharge planning and care transitions, which is consistent with data-driven quality improvement.

Research Question

Does the type of insurance, such as Medicare, Medicaid, or Private, affect the frequency of emergency department visits? Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data

The key Variables

Independent: Insurance Type

Dependent: Number of ED visits

Rationale

Emergency department utilization patterns may reflect access to preventive care and socioeconomic barriers.

Clinical Relevance

Understanding payer-based disparities can inform policy interventions and resource planning to improve equity, which aligns with healthy community planning.

These questions demonstrate how understanding variable measurement levels, such as distinguishing nominal from ratio data, directly informs data analysis strategy. Following Pruinelli et al.’s (2020) roadmap, such exploratory questions form the basis for evidence-based practice and strategic population health interventions.

References

Pruinelli, L., Johnson, S. G., Fesenmaier, B., Winden, T. J., Coviak, C., & Delaney, C. W. (2020). An applied healthcare data science roadmap for nursing leaders: A workshop development, conceptualization, and application. CIN: Computers, Informatics, Nursing, 38(10), 484-489. Project Milestone: Data Plan – Analyzing Synthetic Patient Data for Clinical Insights – Assessing the Influence of Age on Hospital Readmissions Using Synthetic Patient Data