HIPAA Safe Harbor Compliant

Turn Patient Data Into Published Evidence.

Build cohorts from your vestibular rehab caseload, de-identify with one click, and export publication-ready datasets with auto-generated methods paragraphs and codebooks. From clinical data to journal submission in minutes.

Start Free — 3 Months on Us See how it works
Cohort Preview
47
Patients matched
18.3
Avg sessions
62%
Mean improvement
34
Avg days
Post-concussion VOR-x1 DHI < 40
Generate Research Bundle

Clinicians collect detailed vestibular rehabilitation data every day — session metrics, symptom scores, assessment outcomes. But extracting it for a conference abstract, a grant application, or a journal submission means hours of manual de-identification, Excel formatting, and writing methods sections from scratch. Most outcomes data never gets published.

Research Tools Built for Vestibular Clinicians

Everything you need to go from clinical data to publication-ready datasets.

Dynamic Cohort Builder

Filter across 15 dimensions — diagnosis, exercise type, date range, session count, assessment scores, symptom improvement, demographics, and patient status. Real-time preview shows matched patients and aggregate statistics before you export.

HIPAA De-identification

Two levels: Safe Harbor removes all 18 identifiers per 45 CFR §164.514(b)(2). Basic de-identification preserves temporal intervals with per-patient date shifting for IRB-approved studies. Patient identifiers are shuffled to prevent re-identification by row order.

Publication-Ready Exports

Research Bundles include sessions CSV, assessments CSV, Excel workbook with data dictionary, codebook, auto-generated methods paragraph, summary statistics PDF, and study metadata. Quick Reports provide PDF + CSV for conference presentations.

Built for Researchers and Clinic Leaders

Researchers & Academics

  • Case series and cohort studies — export de-identified datasets for retrospective vestibular rehabilitation outcome analyses
  • Grant preliminary data — generate aggregate outcome statistics and summary PDFs for NIH, PCORI, or foundation applications
  • IRB-ready documentation — codebook, methods paragraph, and de-identification methodology included in every Research Bundle
  • Reusable filter sets — save cohort definitions and re-run analyses across time periods as your caseload grows

Clinic Directors

  • Outcomes reporting — demonstrate aggregate patient improvement rates across diagnoses and exercise types for accreditation bodies
  • Marketing with proof — back clinic marketing claims with real de-identified outcome statistics instead of anecdotes
  • Payer negotiations — present aggregate treatment duration, session counts, and symptom reduction data to support reimbursement discussions
  • Quality improvement — identify which protocols produce the strongest outcomes and adjust clinical practice accordingly

From Caseload to Dataset in Three Steps

1

Filter

Define your cohort using any combination of 15 filter dimensions. Diagnosis, exercise type, date range, assessment scores, demographics — combine them to match exactly the patient group you need.

2

Preview

See real-time cohort statistics before exporting — patient count, average sessions, mean symptom improvement, age and gender distributions, assessment score summaries, and weekly session trends.

3

Export

Choose your de-identification level and export type. Quick Reports for presentations, Research Bundles for publications. Every export is audit-logged and includes provenance metadata for IRB compliance.

What You Get in a Research Bundle

Every Research Bundle is a self-contained ZIP file ready for statistical analysis or journal submission.

sessions.csv

Exercise session data — type, duration, metrics, pre/post symptom scores

assessments.csv

DHI, VOMS, and mBESS scores — intake, discharge, and deltas

dataset-with-dictionary.xlsx

Excel workbook with data sheet and embedded variable definitions

codebook.txt

Field definitions, value ranges, and de-identification methodology

methods-paragraph.txt

Auto-generated methods section describing data collection and de-identification for direct use in manuscripts

summary-statistics.pdf

Cohort demographics, outcome distributions, and aggregate metrics formatted for appendices

Two De-identification Methods

Aspect Safe Harbor Basic
HIPAA standard 45 CFR §164.514(b)(2) Direct identifier removal
Names & emails Removed Removed
Dates Generalized to YYYY-MM Shifted by random offset (preserves intervals)
Ages 89+ bucketed as "89+" Exact ages preserved
Patient IDs Shuffled sequential (Patient-001, -002...) Shuffled sequential
Consent required No (data is no longer PHI) Yes (attestation required)

15 Cohort Filter Dimensions

Combine any filters to define exactly the patient population you need. All filters apply with AND logic.

Diagnosis

Filter by clinical diagnosis

Exercise Types

VOR-x1, saccades, convergence, and 7 more

Date Range

Start and end date boundaries

Session Count

Minimum and maximum sessions completed

DHI Scores

Intake, discharge, or improvement delta

VOMS Scores

Intake, discharge, or improvement delta

mBESS Scores

Intake, discharge, or improvement delta

Symptom Improvement

Minimum percentage improvement threshold

Custom Program

Filter by enrollment in specific programs

Patient Status

Active, discharged, or graduated

Gender

Male, female, other, or undisclosed

Age Range

Minimum and maximum patient age

Exercise Frequency

Sessions per week threshold

Treatment Duration

Days from first to last session

Adherence Rate

Percentage of prescribed sessions completed

Frequently Asked Questions

Is the exported data HIPAA compliant?
Yes. Eye Rehab offers two levels of de-identification. HIPAA Safe Harbor removes all 18 identifiers specified in 45 CFR §164.514(b)(2) — including names, emails, dates of birth, and geographic data below the state level. Ages 89 and older are bucketed as '89+' and dates are generalized to month-year granularity. Basic de-identification removes direct identifiers and shifts dates by a random per-patient offset while preserving relative time intervals for temporal analysis. Both methods include automatic audit logging of every export.
What export formats are available for research?
Eye Rehab provides two export types. Quick Reports generate a PDF summary with a companion CSV file — useful for conference presentations and grant applications. Research Bundles include a comprehensive ZIP with sessions CSV, assessments CSV, an Excel workbook with embedded data dictionary, a codebook documenting de-identification methodology, an auto-generated methods paragraph for publications, summary statistics PDF, and study metadata. Both formats include audit-logged provenance for IRB review.
How many patients do I need to start building cohorts?
The cohort builder activates with any number of linked patients, but Eye Rehab recommends at least 5 patients for meaningful aggregate statistics. Preview statistics update in real time as you adjust filters, so you can see sample sizes before exporting. For publishable research, typical vestibular rehabilitation case series include 10–30 patients, though the tool supports cohorts of any size.
What filters can I use to build a research cohort?
The cohort builder supports 15 filter dimensions: diagnosis, exercise types (VOR-x1, VOR-x2, saccades, convergence, pursuits, and 5 more), date range, session count range, assessment scores (DHI, VOMS, mBESS — including intake, discharge, and delta values), symptom improvement percentage, custom program enrollment, patient status (active, discharged, graduated), gender, and age range. Filters combine with AND logic, and you can save filter configurations for reuse across multiple time periods.
Can I use this data for grant applications?
Yes. The Research Bundle export is designed specifically for grant applications and IRB submissions. It includes study metadata fields (study title, principal investigator, IRB number, institution), a methods paragraph describing the data collection and de-identification process, and a codebook suitable for inclusion in appendices. Quick Reports provide summary statistics for preliminary data sections of grant proposals.
What clinical assessments are tracked?
Eye Rehab tracks three validated clinical assessments: the Dizziness Handicap Inventory (DHI), the Vestibular/Ocular Motor Screening (VOMS), and the Modified Balance Error Scoring System (mBESS). Each assessment captures intake and discharge scores, and the cohort builder lets you filter by score ranges and improvement deltas across any combination of these instruments.
How long are exports retained?
All exports are retained for 30 days from the date of generation. During this window, you can download the export files from the Export History tab. After 30 days, exports are automatically deleted in compliance with data minimization principles. You can regenerate any export at any time by re-running the same filters.
Can I track symptom outcomes across my caseload?
Yes. Eye Rehab tracks five symptom dimensions — dizziness, brain fog, headache, eye strain, and nausea — on a 0-to-10 scale before and after every exercise session. The cohort builder's preview panel shows mean symptom improvement percentage, an improvement distribution histogram, and weekly session trends. You can filter by minimum improvement threshold to isolate high-responders or identify patients who may need protocol adjustments.
Do I need to enable research separately?
Yes. Research & Grants is an opt-in feature that you enable from your Pro Portal settings page. Once enabled, the Research tab appears in your navigation. This gating ensures that data export capabilities are only available to professionals who explicitly intend to use them, supporting HIPAA minimum necessary access principles.
What is the difference between Safe Harbor and Basic de-identification?
HIPAA Safe Harbor de-identification follows the standard defined in 45 CFR §164.514(b)(2). It removes all 18 types of identifiers, generalizes dates to month-year, and buckets ages 89+ — producing a dataset that is no longer considered protected health information under HIPAA. Basic de-identification removes direct identifiers (names, emails) but preserves more temporal detail by shifting dates with a random per-patient offset. Basic requires explicit consent attestation and is suitable for research conducted under an IRB protocol or with patient authorization.

Start publishing your vestibular outcomes

Create your Pro Portal account today. Your first 3 months are free — no credit card required. Build your first cohort in minutes.

Start Free — 3 Months on Us

Full access to cohort builder, de-identification, and exports. Cancel anytime.