Table 4: Sensitivity analyses exploring different publication cut-offs for definition of a physician-scientist
Publication cut-off; no. (%) of graduates*
Characteristic 4 publications
(n = 52)
 5 publications
(n = 51)
 6 publications
(n = 50)
 7 publications (n = 47) 8 publications (n = 45) 9 publications (n = 44) 10 publications (n = 43)
Total no.†
Male35 (67.3)34 (66.7)33 (66.0)33 (70.2)31 (68.9)30 (68.2)30 (69.8)
Female17 (32.7)17 (33.3)17 (34.0)14 (29.8)14 (31.1)14 (31.8)13 (30.2)
Master's students
Male13 (54.2) 12 (52.2)11 (50.0)11 (57.9)10 (55.6)9 (52.9)9 (56.3)
Female11 (45.8)11 (47.8)11 (50.0)8 (42.1)8 (44.4)8 (47.1)7 (43.8)
PhD students
Male22 (78.6)22 (78.6)22 (78.6)22 (78.6)21 (77.8)21 (77.8)21 (77.8)
Female6 (21.4)6 (21.4)6 (21.4)6 (21.4)6 (22.2)6 (22.2)6 (22.2)
Residency choice‡
Internal medicine22 (42.3)22 (43.1)22 (44.0)19 (40.4)19 (42.2)19 (43.2)19 (44.2)
Neurology5 (9.6)5 (9.8)5 (10.0)5 (10.6)5 (11.1)5 (11.4)5 (11.6)
Pediatrics5 (9.6)5 (9.8)5 (10.0)5 (10.6)4 (8.9)4 (9.1)4 (9.3)
OR (95% CI) for classification as a physician-scientist§
4 publications while in the joint program5.06
 (2.38-10.77)
5.34
 (2.50-11.41)
6.53 
(3.09-13.83)
7.89
 (3.66-16.99)
7.74 
(3.58-16.75)
8.31
 (3.81-18.15)
7.63
 (3.51-16.6)

Note: CI = confidence interval, OR = odds ratio.

*Except where indicated otherwise.

†Total number of individuals classified as physician-scientists, according to the specified publication cut-off, of which one publication had the person as first or senior author and was published in the last 5 years, with graduation from the joint program more than 5 years ago.

‡Top 3 residency choices listed.

§Significant estimates from the stepwise logistic regression model that included sex of student, year of graduation, degree obtained (PhD v. master's), number of publications while in the joint program (stratified as < 4 and ≥ 4) and residency choice (internal medicine v. other).