2027 A Data Dictionary for a Synthetic Registry: A Healthcare Administration Student Portfolio Project
Co-Founder and Chief Data Scientist
A registry dashboard can look convincing while hiding a basic problem: nobody knows exactly what its fields mean. For a healthcare administration student, a synthetic registry and its data dictionary offer a way to demonstrate careful measurement without publishing patient records.
This guide shows what to build, how to define and check the data, where privacy boundaries lie, and how to explain the project to employers. It also helps you decide whether this focused portfolio project meets your goal or whether you need broader healthcare management or technical training.
Key Things You Should Know
- Build the dictionary around one administrative question, such as a simulated 30-day readmission rate; define the eligible population before calculating the measure.
- For each field, document its meaning, record level, type, allowed values, missing-value rule, and synthetic origin - not just its column name.
- Use wholly invented records where possible. "Synthetic" is not a privacy exemption for information derived from real patients; HIPAA de-identification has separate requirements under 45 CFR § 164.514.
What is a synthetic registry healthcare administration portfolio project?
A registry is an organized collection of records about a defined population, event, or service. In this project, the records are invented to model a healthcare operation - for example, hospital discharges and subsequent utilization. A data dictionary explains every field and the rules for using it. The portfolio artifact combines the records, dictionary, validation checks, and a short account of an administrative decision the data could support.
Choose a question you can answer without pretending to evaluate real patients or a real hospital. A simulated discharge registry might help a manager see how eligibility rules change a readmission measure. Students comparing coursework with a larger credential can examine online Master of Health Administration programs; this project is evidence of a specific skill, not a substitute for an MHA or professional experience.
How does a data dictionary support a synthetic healthcare registry?
The dictionary is the agreement between whoever creates, analyzes, and interprets the registry. It prevents a field called readmitted from meaning "any later visit" to one analyst and "an eligible inpatient admission within 30 days" to another. That distinction changes both the numerator and the denominator of a reported rate.
It also separates administrative definitions from clinical judgments. A student pursuing health information management may need deeper training in coding, information governance, and records systems; the distinction between those responsibilities and management work is useful when considering HIM and healthcare administration educational pathways. For this portfolio, the goal is to make a management measure reproducible, not to claim clinical or coding expertise.
What should a healthcare administration student include in the project?
A small, complete submission is more persuasive than a large spreadsheet with unexplained columns. Include artifacts that let another person understand the decision, reproduce the result, and identify the limits of the simulation.
- Write a one-page brief naming the hypothetical organization, administrative question, intended user, and decisions the analysis cannot support.
- Provide an invented, consistently structured dataset and a separate data dictionary with field-level definitions.
- Document eligibility, exclusions, measure calculations, and validation checks in a readable methods note or reproducible workbook.
- Present one chart or compact dashboard, then explain a plausible operational response and why real-world validation would be necessary.
Keep software spending proportionate to the assignment; a spreadsheet and a clearly written document can be enough. If the project is part of a degree decision, compare its modest tool costs separately from healthcare administration master's tuition planning.
Start with the question, then add only fields needed to define the population, calculate the measure, or interpret it. For a simulated discharge registry, the following groups provide a manageable scope. They are design choices for a class project, not a universal registry standard.
The table distinguishes fields that support the calculation from fields that merely add context. That distinction helps prevent a dashboard from implying more than the synthetic records can show.
| Field group | Example fields | Administrative purpose |
| Record linkage | invented_person_id; discharge_id | Identify distinct people and discharge events without names. |
| Event timing | admission_date; discharge_date; next_admission_date | Establish event order and a follow-up window. |
| Eligibility | discharge_disposition; index_eligible | Specify which discharges enter the denominator. |
| Service context | service_line; facility_code | Allow cautious comparison across invented operating units. |
| Derived outcome | readmission_30d | Mark whether an eligible discharge meets the documented outcome rule. |
Do not add detailed diagnoses, addresses, or demographic fields merely to make the file look realistic. A narrow project can fit alongside coursework when comparing a healthcare management graduate study timeline comparison; the appropriate scope depends on the course's actual requirements.
Write definitions that someone who did not create the dataset could apply. State the record level first: if each row represents a discharge, one person may legitimately appear on several rows. Then specify type, valid values, missingness, source or generation method, and any transformation used to create a derived field.
This excerpt shows how a concise dictionary can document both raw and derived fields. Its outcome rule is illustrative; a real quality measure would require the applicable measure specification and validated source data.
| Field | Definition and type | Allowed values or rule | Missingness and origin |
| discharge_id | Unique identifier for one simulated discharge; text | Unique and nonblank | Never missing; invented for this project |
| discharge_date | Date of the simulated discharge; date | On or after admission_date | Never missing for an eligible row; generated, not extracted |
| index_eligible | Whether the discharge enters the example denominator; Boolean | True or false under the documented exclusion rules | Never missing; calculated from simulated fields |
| readmission_30d | Whether an eligible discharge is followed by a qualifying simulated inpatient admission within 30 days; Boolean | True or false only when follow-up can be evaluated | Null when not assessable; calculated from simulated events |
A null outcome should not silently become "no readmission." If follow-up is incomplete, mark the record as not assessable and explain its treatment in the measure. Students working within 12-month health administration graduate degree options can prioritize these essential definitions over a wider but poorly documented dataset.
Wholly invented data is the safest starting point for a public student portfolio. If no record is based on an actual person, the file is not a route for exposing that person's protected health information. By contrast, replacing names in real patient records - or generating records from them - does not automatically make a public file safe.
HIPAA's Privacy Rule governs covered entities and their business associates in relevant circumstances; its de-identification provisions at 45 CFR § 164.514 describe recognized approaches for information derived from protected health information. Re-identification risk, contracts, institutional policy, and research oversight may also matter. Students should not access, export, or publish real records without the necessary authorization and institutional guidance, and should not label a dataset "HIPAA compliant" merely because it is called synthetic.
A privacy-conscious project can help applicants explain their judgment, including those considering MHA applications without a prior healthcare job. It does not replace a school's admissions requirements or supervised experience.
A well-explained dictionary demonstrates requirements gathering, data governance, measure design, quality checking, and communication with nontechnical decision-makers. The strongest evidence is not the chart alone: it is your explanation of why a record qualifies, how you handled missing follow-up, and what a manager should not conclude from the result.
Those are management-facing skills, not proof that you can conduct clinical research, administer a production electronic health record, or lead an organization. If your longer-term goal involves doctoral study, consider healthcare leadership doctorate credit transfer and duration separately; a portfolio artifact should not be mistaken for transferable academic credit or doctoral research.
The same deliverables work in either format. An online student can share a version-controlled workbook and short recorded explanation; a campus student can present the same materials in class. Neither format requires access to a hospital system. Set checkpoints around definitions and testing rather than leaving the dictionary until after the dashboard.
- Agree on one administrative question and define the row-level unit before generating records.
- Draft the dictionary and eligibility rules, then create a small set of wholly invented records that includes missing and edge cases.
- Test uniqueness, allowed values, date order, eligibility, and the outcome calculation; revise the dictionary whenever a rule changes.
- Prepare a brief presentation that distinguishes simulated findings from claims about real healthcare performance.
For compressed coursework, reserve time for review rather than generating more rows. That trade-off matters when planning an accelerated online health administration degree, where course schedules and assessment deadlines should be checked rather than assumed flexible.
Lead with the problem and your decisions: "I designed a simulated discharge registry, defined eligibility and a 30-day outcome, tested edge cases, and documented why the resulting rate cannot describe a real facility." Show a sample dictionary page, the validation results, and one visual. State your individual contribution if the work was collaborative.
Label the dataset wholly synthetic and the findings illustrative. A hiring manager should be able to tell which rules you designed, which tools you used, and what you would verify before applying the method to authorized operational data. For students weighing further education, the portfolio can inform that decision without settling it: review doctoral tuition in healthcare administration only if a doctorate serves a specific career goal, not because this project needs a more advanced credential to be credible.
Which healthcare administration careers value registry and data skills?
The project is most directly relevant to entry-level or developing roles in quality improvement, healthcare operations, performance reporting, and registry coordination. Responsibilities differ: an operations analyst may define and monitor service measures, while a registry coordinator may focus more on field definitions, completeness, and reporting procedures. Employers may still require healthcare experience, specific software skills, or specialized training.
For a student changing careers, the project can make transferable skills visible, but it does not establish eligibility for clinical, licensed, or senior leadership positions. The U.S. Bureau of Labor Statistics reports a May 2025 median annual wage of $126,520 for all management occupations; that broad figure is not an entry-level healthcare registry salary or a forecast of what this project will earn you. Compare actual job descriptions and local pay for the roles you seek.
Build experience first if employers consistently ask for operational exposure that your portfolio cannot provide. If your eventual aim is research or executive education, assess healthcare administration doctoral study schedules and formats against that goal rather than treating further study as the default next step.
Other Things You Should Know About Healthcare Administration
No. A spreadsheet can support a small registry, dictionary, checks, and chart. Use code if it helps you make the work reproducible or matches the jobs you are targeting.
Possibly, if its license permits your use. Identify its actual source and limitations; do not call real public data synthetic. Follow your course and institution's rules.
Only if you can follow and cite the applicable specification accurately. Otherwise, describe your calculation as an illustrative classroom measure, not an official performance score.
No. The project can stand on its own. Consider additional education when job requirements and your knowledge gaps justify its cost and time.
- An overview of synthetic administrative data for research - PMC https://pmc.ncbi.nlm.nih.gov/articles/PMC10464868/
- Data Dictionary | NNLM https://www.nnlm.gov/resources/data/data-glossary/data-dictionary
- Data Dictionaries | U.S. Geological Survey https://www.usgs.gov/data-management/data-dictionaries
- United States Core Data for Interoperability (USCDI) | healthit.gov https://isp.healthit.gov/united-states-core-data-interoperability-uscdi
- Obtaining Data From Electronic Health Records - - NCBI - NIH https://www.ncbi.nlm.nih.gov/books/NBK551878/
- Using synthetic data to tell the right story and improve care | The University of Chicago https://biologicalsciences.uchicago.edu/news/synthetic-data-improve-care
- Data Dictionary | harvard.edu https://datamanagement.hms.harvard.edu/collect-analyze/documentation-metadata/data-dictionary
- What is Considered PHI under HIPAA? Updated for 2026 | The HIPAA Journal https://www.hipaajournal.com/considered-phi-hipaa/
- Introduction to data types and field properties | Microsoft Support https://support.microsoft.com/en-us/access/introduction-to-data-types-and-field-properties
