NIHデータ管理・共有ポリシーにおける構造化データ管理・共有計画（DMSP）テンプレートは非構造化テンプレートよりも優れた実績を上げた

The National Institutes of Health Data Management and Sharing (DMS) policy (NOT-OD-21-013) mandates the submission of a Data Management and Sharing Plan (DMSP) for all NIH-funded research that generates scientific data. However, little information is available about how academic medical centers have implemented the policy.

The study aimed to characterize our institution's implementation of the DMS policy and compare structured versus unstructured approaches to producing policy-conformant DMSPs.

We monitored all NIH grant submissions from our institution for 18 months, evaluating policy implementation through DMSP completeness and reviewer comments during the Just-in-Time period. A rubric was developed to assess whether each required DMSP element and sub-element was addressed. Eight DMSP templates (three NIH-provided, five institutionally developed) and two categories of investigator-created DMSPs were scored. Researchers' feedback was collected through surveys and interviews.

79.3% of submitted DMSPs addressed all NIH-required DMSP elements. Element-level compliance ranged from 98.9% (data type) to 82.7% (tools and software). Sub-element scores showed greater variability, with 98.9% completion for data description and 49.3% for data generation. Unstructured DMSPs consistently underperformed compared to structured DMSPs. Survey and interview feedback, along with reviewer comments, reinforced these findings.

A notable 20.7% of DMSPs omitted one or more required elements, indicating a need for improved DMS policy conformance. Structured DMSP templates demonstrated greater alignment with NIH policy. We recommend using structured templates to enhance the quality and consistency of data management and sharing plans.