Following my post noting the occurrence of differences between men and women K99 awardees in their likelihood of receiving an R01 grant NIH, through Sally Rockey's blog, noted that application rates may play a role:
"Of the 2007 cohort of K99 PIs, 86 percent of the men had applied for R01s by 2013, but only 69 percent of the women had applied."
This point has been taken up over at DrugMonkey.
Although such differences in application rates between genders are common in NIH statistics, I was surprised that the rates were this different since the K99 cohort from a single year is, presumably, relatively uniform in terms of career stage, accomplishment (having successfully competed through the same program), and so on.
In considering factors that could contribute to this difference, I thought of the nature of the institutions at which these individuals get their R00 awards (if they do transition). As one (certainly imperfect) measure of institutional characteristics, I used the FY2013 institutional ranking of NIH funding. For the FY2007 cohort, of the 108 men with R00 awards, the median ranking for their R00 institution is 37 and the mean is 81. In contrast, for the 62 women, the median is 57 and the mean is 113. For the FY2008 K99 cohort, the median for men is 44 and the mean is 71. For women, the median is 45 and the mean is 103.
For the purposes of further analysis, I divided institutions into 5 groups (NIH funding ranking 1-25, 6-50, 51-75, 76-100, and >100. The distributions for men and women for the two cohorts are shown below:
The distributions are relatively similar for the institutions near the top of NIH funding rankings. However, there are differences in the remainder of the distribution, most strikingly for institutions with NIH funding ranking >100. For men, 20-21% of the R00 awardees were at such institutions whereas 31-36% of the women were. This reveals that a larger percentage of women over men with K99 awards are beginning their independent careers at institutions that are relatively less research intensive, by opportunity or choice.
How does this relate to the likelihood of receiving an R01 award? The results for the FY2007 cohort are shown below:
For the investigators at institutions with rank 1-25, the percentages of investigators who have achieved R01 funding is comparable for men and women. However, this is not true for the other sets when a higher percentage of men than women have received R01 funding. For example, more than 20% of all women in this cohort are at institutions with NIH funding rankings >100 and have not received R01 funding compared with 7% of all men.
The corresponding plot for the FY2008 K99 cohort is shown below:
Again, more than 20% of all women are at institutions with NIH funding rankings >100 and have not received R01 funding. In addition, for this cohort, the fraction of women at institutions with NIH funding rankings from 1-25 who have received R01 funding is substantially lower than that for men at the same set of institutions.
These data provide insights into some factors that may contribute to the disparities in R01 funding for women and men in the K99-R00 program. Of course, as one parses the program into smaller groups, the statistical power decreases. Nonetheless, these analyses should provide guidance to allow a better understanding of the role of different factors in NIH funding outcomes.