Data Management Plan Nih Example

STANDARD OPERATING PROCEDURE: NIH Data Management Plan (DMP) Architecture, Execution, and Compliance Verification

Document ID: SOP-TR-NIH-DMP-2026-V1
Effective Date: March 30, 2026
Version: 1.4.0
Review Cadence: Annual
Owner: Julian Vance, Chief Architect, Template Registry

---

1. EXECUTIVE SUMMARY & PURPOSE

This Standard Operating Procedure (SOP) defines the institutional engineering standard for drafting, reviewing, validating, and executing Data Management Plans (DMPs) mandated by the National Institutes of Health (NIH) Data Management and Sharing (DMS) Policy. The purpose is to ensure absolute alignment with NIH policy requirements, protect human subjects data through cryptographic and administrative controls, and automate long-term digital preservation across institutional repositories.

---

2. SCOPE & PREREQUISITES

2.1 Scope

This SOP applies to all research personnel, principal investigators (PIs), data stewards, and systems architects operating under Template Registry oversight who submit grant applications or manage active awards requiring an NIH DMS Plan.

2.2 Prerequisites & Tooling

• Access Credentials: Institutional Single Sign-On (SSO) with multi-factor authentication (MFA).
• Authoring Environment: DMPTool (institutional subscription tier) or validated markdown-to-PDF toolchain.
• Data Classification Software: SecurEnds or institutional equivalent for identifier auditing.
• Storage Infrastructure: FedRAMP Moderate/High certified environments (e.g., AWS GovCloud, institutional secure enclaves).

---

3. ROLES & RESPONSIBILITIES (RACI MATRIX)

| Role | Principal Investigator (PI) | Data Steward | Systems Architect | Compliance Officer | | :--- | : मिशन | :--- | :--- | :--- | | Data Element Identification | Responsible | Accountable | Consulted | Informed | | Repository Selection | Accountable | Responsible | Consulted | Informed | | Security & Access Controls | Consulted | Responsible | Accountable | Informed | | Budget & Resource Justification| Accountable | Consulted | Informed | Responsible | | Final Sign-off & Submission | Accountable | Consulted | Consulted | Responsible |

Legend: R = Responsible, A = Accountable, C = Consulted, I = Informed.

---

4. STEP-BY-STEP PROCEDURE

Phase 1: Data Characterization & Element Definition

• 1.1 Catalog all anticipated data types (quantitative, qualitative, audio, imaging, code, derived datasets) generated by the proposed research activities.
• 1.2 Estimate the total storage volume (in Gigabytes/Terabytes) and anticipated growth rates over the life of the award.
• 1.3 Evaluate data sensitivity: identify whether data include Protected Health Information (PHI), Personally Identifiable Information (PII), or controlled-access genomic data.

Phase 2: Metadata, Formats, and Standards Formulation

• 2.1 Select community-accepted metadata standards (e.g., DDI for social sciences, CDISC for clinical trials, Darwin Core for biodiversity) or define custom JSON-LD schemas.
• 2.2 Mandate non-proprietary, long-term preservation file formats (e.g., CSV, TIFF, Plain Text, NetCDF) for all final shareable datasets.
• 2.3 Establish a strict version control taxonomy (e.g., Semantic Versioning v[Major].[Minor].[Patch]) for code and tabular datasets.

Phase 3: Repository Selection & Access Architecture

• 3.1 Select a primary NIH-supported or institutional repository (e.g., NIMH Data Archive, dbGaP, Dryad, institutional Dataverse) that guarantees persistent identifiers (DOIs/Accession Numbers).
• 3.2 Define access restriction protocols: outline timelines for data release (no later than publication or end of award, whichever comes first).
• 3.3 For human subjects data, establish controlled-access mechanisms via Data Use Certifications (DUCs) and institutional review board (IRB) oversight frameworks.

Phase 4: Budgeting, Resource Allocation, & Compliance Review

• 4.1 Itemize allowable costs for data management and sharing (e.g., repository deposit fees, specialized curation personnel, storage provisioning) within the grant budget justification.
• 4.2 Execute a cross-functional peer review of the draft DMP against the NIH 6-element evaluation rubric using the Template Registry compliance validator.
• 4.3 Obtain digital sign-offs from the Lead Data Steward and Institutional Compliance Officer prior to final grant package compilation.

---

5. QUALITY ASSURANCE & PRO-TIPS

5.1 Best Practices

• Early Integration: Draft the DMP concurrently with the specific aims page; repository selection often dictates infrastructure design.
• Machine-Actionable DMPs: Utilize JSON exports from DMPTool to enable automated updates throughout the project lifecycle.
• De-identification Pipelines: Standardize automated scrubbing scripts for PII/PHI before data reaches staging servers.

5.2 Common Pitfalls

• Vague Access Timelines: Stating "data will be shared when appropriate" will trigger immediate peer-review penalties. Use exact milestones (e.g., "upon primary publication or within 30 months of award start").
• Ignoring Unstructured Data: Forgetting to account for qualitative interview transcripts or custom analysis scripts in the total volume calculation.

5.3 Metric Thresholds

• Metadata Completeness Score: $\ge 98%$ compliance against Dublin Core or domain-specific schema validator.
• Plan Validation Time: $\le 48$ hours from initial draft submission to compliance officer review.

---

6. FREQUENTLY ASKED QUESTIONS

Q1: How should we handle proprietary software formats required for specialized instrumentation?
A: Convert data outputs to open, non-proprietary formats (e.g., converting proprietary .raw microscopy files to .OME.TIFF) prior to repository ingestion. If conversion results in critical data loss, archive both the raw proprietary format alongside documented conversion scripts.

Q2: What is the exact budget threshold allowable for data sharing costs?
A: There is no fixed dollar limit, but costs must be reasonable, justified, and incurred during the performance period of the award. Costs associated with routine data collection or infrastructure normally maintained by institutional IT cannot be requested.

Q3: Does the NIH DMS policy apply to undergraduate or pilot studies with human subjects?
A: Yes. Any NIH-funded research generating scientific data—regardless of award mechanism (R03, R21, R01, K-series)—must comply with the DMS policy, provided the data are of sufficient quality to be validated and replicated.