Is there life as an investigator after becoming a pediatric department chair?
Article Outline
Leadership in academic medical centers has become increasingly challenging, with the complex issues of budgetary management, clinical responsibilities, regulatory compliance, and faculty affairs. This has been associated with high turnover rates for both medical school deans and department chairs, with the average tenure being approximately 4 to 5 years.1, 2 Although positive reasons (such as being chosen to become a dean) account for some turnover of chairs, there are also negative reasons, including dissatisfaction with the job or poor job performance, perhaps because of inadequate preparation.3, 4
Another possible explanation for rapid turnover is a concern about decreased academic productivity during chairmanship. Academic chairmen are often promoted to their positions because of prior productivity as an independent investigator funded by the National Institutes of Health (NIH). After becoming chairmen, administrative duties predominate, competing for time that could be devoted to research. We were interested in determining whether chairs in pediatrics were funded NIH investigators at their time of appointment and, if so, whether they were able to continue to be independent investigators in the face of their major administrative responsibilities.
This issue is important for several reasons. If chairmen feel frustrated by an inability to continue performing independent research, a role they found important in their own academic lives, this could affect the duration and success of their tenure as a chair. If chairmen believed that they needed to continue to be productive from a research perspective, they might focus on this activity rather than on the clinical/administrative issues that need to be addressed, resulting in failure in their administrative role. The inability to continue to be productive in a research-intensive institution might be taken by their faculty as a marker of failure on the part of their chair.
To address this issue we compared the NIH grant support among current pediatric department chairmen before versus after their appointment. We hypothesized that accepting an appointment as chairman of pediatrics would result in a significant decline in NIH research grant awards on which they serve as the principal investigator.
Methods
Using the list of chairmen of pediatrics published by the Association of Medical School Pediatric Department Chairs (AMSPDC) in September 2004, we searched the Computer Retrieval of Information on Scientific Projects database of the NIH to identify pre- and post-appointment NIH research grants in which the chair served as principal investigator (PI). Excluded from consideration were the 12 interim and acting chairs. We counted only NIH awards, new and competitive renewals that represented independent research efforts (eg, R01, R21, P01, U54). We excluded institutional/infrastructure grants such as training, symposium, construction, and core center grants (eg, K12, T32, R13/U13, CO6, M01).
Using AMSPDC data, we obtained information about gender, age, and date of appointment of all 125 pediatric department chairs, and categorized the departments of pediatrics by size and by their degree of emphasis on research as described below. The AMSPDC provides a list of departments divided by number of full-time faculty members: small <60 full-time faculty, medium 60-150 faculty, and large >150 faculty members. The NACH (National Association of Children’s Hospitals) publishes an annual survey entitled “NIH Awards to Children’s Hospitals and Pediatric Departments of Medical Schools” that is derived from the NIH website (http://grants1.nih.gov/grants/award/state/state.htm). We used data from the federal FY 2004 NACH document to divide the 125 department of pediatrics/children hospitals into quartiles according to total NIH grant support; we combined the 3rd and 4th quartiles.
Using Poisson Regression analysis, we compared pre-chair awards in the 10 years before appointment with all post-chair awards and presented the data as award rates per year, with 95% confidence limits. We then evaluated the association between success in attaining post-chair grant awards according to key factors that describe the department or the chairman. The association is expressed in terms of the relative success ratio, which is the annual rate of grant successes at one level of a factor (numerator rate), such as department in the top 25th percentile in grant funding, relative to the rate of grant successes at another reference level (denominator rate), such as department is below the 50th percentile in grant funding. When the factor of interest is measured as a continuous variable such as age, the interpretation of the success ratio is the change in the rate of annual grant awards associated with a 1-year difference in age. We accounted for the following factors in estimating annual post-appointment awards: pre-appointment awards, size of department, department ranking in NIH grant support, age, and sex.
Results
Of 125 chairs of Departments of Pediatrics/Children’s Hospitals, 113 held permanent appointments as of September 2004 and were studied; 101 were male (89.4%). The age range was 31 to 66 years (median 50 years), and duration of appointment ranged from just appointed to 27 years (median 6.3 years) (Table I; available at www.jpeds.com). In terms of size of departments, 29 were small, 71 medium, and 12 large.
Table I. Demographics: age appointed and duration as chair
| Demographics | N (missing) | Mean ± Sd | Median | Range |
|---|---|---|---|---|
| Age | 109 (4) | 50.4±5.4 | 49.9 | 31.0,66.0 |
| Duration as chair | 112(1) | 7.0±5.5 | 6.3 | 0,27.5 |
In the 10 years before appointment, 42 (37%) future chairs were PI of one or more NIH grants, including both new grants and competitive renewals. Post-appointment, only 27 (24%) were PIs on new or competing renewal grants. Whereas 27% of chairs held 3 or more such NIH grants before appointment, only 8% held this number of grants after appointment. When we examined only new grant awards, the number dropped further, from 33 (29%) having new grants before appointment to 19 (17%) post-appointment (Figure 1). Using an adjusted annual rate of total grants and new grants, controlling for pre to post differences in paired observations, we found similar decreases of 42% and 35%, respectively (Figure 2; available at www.jpeds.com).
Figure 1.
Percentage distribution of total grants. Pre- to post-differences in proportions of awards are statistically significant at p = .001.
Figure 2.
(a), Adjusted annual rate of total grants. (b), Adjusted annual rate of new grants.
In examining relative success ratios, we found that having held a preexisting grant was a significant predictor of maintaining grant support after appointment as chair, carrying a 56-fold increased chance of an award (Table II; available at www.jpeds.com). Size of department was also predictive, with chairs of small departments having a 2-fold increase in likelihood of retaining grant support. Higher departmental quartile ranking in NIH grant support also carried an almost 3 fold increased chance of the chair maintaining grant support. There was no age effect below 55 years, but a significant fall off of grant support after that time. There were no sex-related differences in grant support.
Table II. Predictors of post-appointment NIH awards based on Poisson modeling
| Factor | Relative success ratio | 95% Confidence interval | |
|---|---|---|---|
| Having a pre-appointment award | 55.90⁎ | 16.02 | 194.74 |
| Size of department | |||
| Small | 2.28† | 1.16 | 4.47 |
| Medium | 1(ref) | ||
| Large | 0.68 | 0.41 | 1.14 |
| Departmental ranking | |||
| Top quartile | 2.78 | 0.99 | 7.83 |
| 2nd quartile | 1.80 | 0.68 | 4.81 |
| 3rd & 4th quartiles | 1(ref) | ||
| Not ranked | 0.41 | 0.15 | 1.10 |
| Age (years) | |||
| 55+ | 0.27 | 0.04 | 1.87 |
| 45–54 | 1.24 | 0.31 | 4.86 |
| <45 | 1(ref) | ||
| Male sex | 1.02 | 0.34 | 3.05 |
⁎Factors that are statistically significant at P < .01. |
†Factors that are statistically significant at P < .05. |
Discussion
These data indicate that becoming a pediatric department chair is associated with a significant decline in NIH grant awards as a PI. Relative to pre-appointment, the post appointment award rates fell dramatically. This change is not explained by pre-appointment status, sex, age, department size, or NIH ranking of department.
Certain factors tended to mitigate these changes. Chairs from small departments and those from research-intensive programs were more likely to retain grants. One could speculate that small departments are less challenging to manage administratively, so that there is more time for the chair to spend on a personal research agenda. Research-intensive departments may have both the infrastructure and an institutional culture that promotes continued research productivity on the part of the chair. The fall off in grant support after age 55 mirrors general NIH grantee information; nationally only 20% of PIs are in this age group.5 Gender played no role, but as noted by others,2 the number of women chairs remains low.
This research suggests that the answer to the question of whether there is life as an independent investigator after becoming a pediatric chair appears to be: not usually, at least not by the standard of being an NIH PI. Less than one quarter of chairs met this criterion. However, this study did not address other sources of funding and did not evaluate rates or perceived quality of research publications, which are traditionally the yardstick of success in academia. It also did not address whether this fall off in grant support was a factor in the turnover rate of chairs of pediatrics.
Nevertheless, this study would predict that becoming chairman has a negative impact on the appointee’s personal research productivity. Whether this is good or bad can be debated. One could posit that chairs should move from being the sun (a PI investigator) to being the moon (receiving reflected glory from their faculty). Perhaps an equally important point is how the chairs feel about their loss of grant support. Does this contribute to malaise and lead to a shortened tenure as chair? A study is currently being performed to attempt to answer this question within the broader issue of understanding the rapid turnover of chairs.2
We thank Jean Bartholomew of AMSPDC for her help in providing data for this study.
References available at www.jpeds.com.
References
- . Tenure of deans of medicine during the last five decades. Acad Med. 1994;89:1–7
- . Leadership trends in academic pediatric departments. Pediatrics. 2005;116;:342–344
- . Attitudes toward and experience in research among Family Medicine chairs. J Fam Pract. 1992;35;:417–421
- . Burnout in chairs of obstetrics and gynecology (diagnosis, treatment and prevention). Am J Obstet Gynecol. 2002;186;:601–612
- National Institutes of Health Office of Extramural Research. 2005 NIH Research Project Awards by age of Principal Investigator, FY 2001. http://grants.nih.gov/grants/award/trends/prininv.htm.
Supported through the Council of Deans Fellowship Program of the American Association of Medical Colleges.
PII: S0022-3476(06)00060-6
doi:10.1016/j.jpeds.2006.01.032
© 2006 Elsevier Inc. All rights reserved.
