NIH's dedicated K23 trial pathway was associated with more K-trial leadership but no detectable gain in new trials
- 1Macalester College
Abstract
In 2018, the US National Institutes of Health (NIH) created a dedicated clinical trial pathway for K23 career development awards, and its planned KC1 award will not fund trial conduct for applications due from late 2027. Whether the pathway changed how many awardees led trials, or which trials, is unknown. We assessed leadership of the trial the award supported (K trial), of distinct new trials, and of any trial. In this cohort study of 1847 K23 awardees (fiscal years 2014 to 2021; 837 under the pathway), we linked NIH award records to ClinicalTrials.gov registrations, including archived versions. A difference-in-differences design, adjusted for fiscal year and NIH funding institute, compared awardees whose funded abstracts proposed their own trial (trial plans; n = 721) with other awardees. A language model measured trial plans and leadership, and author review of 280 sampled records corrected its errors. Before correction, 54.4% of earlier and 73.1% of pathway awardees with trial plans led a K trial within 5 years, vs 4.1% and 5.1% of others. This gap widened by 18.6 percentage points uncorrected and 13.7 (95% CI, 5.7-21.2) corrected, mostly in grant-citing trials. The new-trial gap changed by -3.8 (95% CI, -14.4 to 6.9). Assuming parallel trends, every correction excluded gains above 11.5 points (earlier rate, 37.6%). The any-trial gap widened by 4.2 (95% CI, -5.9 to 14.0) when correction probabilities were estimated separately by period and 9.6 (95% CI, 1.5-17.9) when they were pooled across periods. Had early-cohort trends (2.4 points yearly) continued, the K-trial estimate would be 4.1 (95% CI, -17.7 to 26.8). Limitations include more complete registration after concurrent NIH application changes, which the design cannot separate from the pathway, and selection among funded awardees. The dedicated pathway was associated with more awardees with trial plans leading their award-supported trial, although more complete registration or an earlier trend could not be ruled out. New-trial leadership within 5 years did not detectably increase, and whether the pathway produced more trial leaders is unclear. Evaluations of KC1 should track proposed and new trials separately.
Related
- Attenuated hypothalamic response to fructose via a dedicated gut-brain pathway
- How leadership attention is associated with online–offline healthcare service synergy: differentiated effects of noticing, interpreting, and acting
- Acute transcutaneous auricular vagus nerve stimulation is associated with lower baroreflex sensitivity without detectable Mayer-wave coherence change: an exploratory study
- Culturable Leptospira spp. exhibit no detectable spatial or matrix-associated structuring in Estero de Paco, Manila, Philippines
- Microscopically detectable maternal schistosomiasis at delivery in Eastern Uganda: Prevalence, microscopy-based species distribution and associated household WASH characteristics in a hospital-based cross-sectional study
BibTeX
@misc{wang2026nihs,
title = {{NIH's dedicated K23 trial pathway was associated with more K-trial leadership but no detectable gain in new trials}},
author = {H. Wang and H. Fu and Z. Tian},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.09.758109},
url = {https://www.biorxiv.org/content/10.64898/2026.10.09.758109v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - NIH's dedicated K23 trial pathway was associated with more K-trial leadership but no detectable gain in new trials
AU - H. Wang
AU - H. Fu
AU - Z. Tian
PY - 2026
DA - 2026/10/09
PB - bioRxiv
DO - 10.64898/2026.10.09.758109
UR - https://www.biorxiv.org/content/10.64898/2026.10.09.758109v1
AB - In 2018, the US National Institutes of Health (NIH) created a dedicated clinical trial pathway for K23 career development awards, and its planned KC1 award will not fund trial conduct for applications due from late 2027. Whether the pathway changed how many awardees led trials, or which trials, is unknown. We assessed leadership of the trial the award supported (K trial), of distinct new trials, and of any trial. In this cohort study of 1847 K23 awardees (fiscal years 2014 to 2021; 837 under the pathway), we linked NIH award records to ClinicalTrials.gov registrations, including archived versions. A difference-in-differences design, adjusted for fiscal year and NIH funding institute, compared awardees whose funded abstracts proposed their own trial (trial plans; n = 721) with other awardees. A language model measured trial plans and leadership, and author review of 280 sampled records corrected its errors. Before correction, 54.4% of earlier and 73.1% of pathway awardees with trial plans led a K trial within 5 years, vs 4.1% and 5.1% of others. This gap widened by 18.6 percentage points uncorrected and 13.7 (95% CI, 5.7-21.2) corrected, mostly in grant-citing trials. The new-trial gap changed by -3.8 (95% CI, -14.4 to 6.9). Assuming parallel trends, every correction excluded gains above 11.5 points (earlier rate, 37.6%). The any-trial gap widened by 4.2 (95% CI, -5.9 to 14.0) when correction probabilities were estimated separately by period and 9.6 (95% CI, 1.5-17.9) when they were pooled across periods. Had early-cohort trends (2.4 points yearly) continued, the K-trial estimate would be 4.1 (95% CI, -17.7 to 26.8). Limitations include more complete registration after concurrent NIH application changes, which the design cannot separate from the pathway, and selection among funded awardees. The dedicated pathway was associated with more awardees with trial plans leading their award-supported trial, although more complete registration or an earlier trend could not be ruled out. New-trial leadership within 5 years did not detectably increase, and whether the pathway produced more trial leaders is unclear. Evaluations of KC1 should track proposed and new trials separately.
ER -