Documented Human Implementation of NIH SBIR/STTR-Funded Neurosurgical Technologies: A Retrospective Cohort Study
En palabras de los autores
Background: Federal small business awards from the National Institutes of Health fund neurosurgical devices and procedural technologies, but how often the specific funded capability reaches human use is unknown. We determined how often these technologies had documented human implementation and in which settings. Methods: We assembled a retrospective public-source cohort of technologies supported by Small Business Innovation Research and Small Business Technology Transfer awards from fiscal years 2005-2020 by keyword filtering and funded-purpose review. Confirmation required living-human use of the funded capability or a supported direct extension after a qualifying same-technology award, with public evidence by October 8, 2026. Exploratory comparisons examined program, observed phase, award era, and cumulative support. A 500-family random sample assessed screening omissions. Results: The cohort comprised 209 technologies, 221 award families, and 453 applications receiving $206,825,343. Human implementation was documented for 31/209 technologies (14.8%; Wilson 95% confidence interval, 10.7%-20.3%); evidence was not located for 156 and was indeterminate for 22. Confirmed settings were neurosurgical-scope care (26), other clinical care (2), mixed surgical cohorts (1), and nonclinical human evaluation (2). Confirmation was more frequent with observed Phase II support than Phase I only (21/98 vs 10/111; difference, 12.4 percentage points; company-component bootstrap 95% confidence interval, 2.8-22.4). None of four conventional tests met a Holm-adjusted .05 threshold. Screening scenarios estimated 40-76 omitted families. Conclusion: Approximately one in seven included technologies had documented human implementation. Linking individual awards to implemented capabilities provides an empirical benchmark for early neurosurgical translation and a practical foundation for tracking subsequent clinical evaluation.
Apareció: martes, 17 de noviembre. medRxiv. Preprint, todavía sin revisión por pares.