Table 2.
Multivariable logistic regression models identifying variables independently associated with each of ≥1 year GME research, faculty appointment and federal mentored-K and/or RPG award receipt among board-certified surgeons who had graduated from LCME-accredited US medical schools in 1997–2004 (N = 5328).a
| ≥1 year GME research (yes vs. no) AOR (95% CI) | P-value | Faculty appointment (yes vs. no) AOR (95% CI)b | P-value | K/RPG award receipt (yes vs. no) AOR (95% CI)b | P-value | |
|---|---|---|---|---|---|---|
| Sex | ||||||
| Men | 1.000 (reference) | 1.000 (reference) | 1.000 (reference) | |||
| Women | 1.011 (0.883 – 1.157) | .878 | 1.147 (1.002 – 1.314) | .047 | 1.271 (0.697 – 2.317) | .434 |
| Race/ethnicity | ||||||
| White | 1.000 (reference) | 1.000 (reference) | 1.000 (reference) | |||
| Other/unknown | 1.006 (0.616 – 1.643) | .981 | 0.952 (0.578 – 1.570) | .848 | 1.898 (0.423 – 8.510) | .402 |
| URM | 1.106 (0.917 – 1.334) | .292 | 1.069 (0.886 – 1.289) | .486 | 0.579 (0.173 – 1.932) | .374 |
| Asian/Pacific Islander | 1.317 (1.130 – 1.535) | <.001 | 1.072 (0.917 – 1.253) | .381 | 1.168 (0.634 – 2.154) | .618 |
| Medical-school graduation yearc | 0.980 (0.954 – 1.006) | .126 | 0.964 (0.939 – 0.990) | .008 | 0.765 (0.670 – 0.874) | <.001 |
| Degree program | ||||||
| MD | 1.000 (reference) | 1.000 (reference) | 1.000 (reference) | |||
| MD/PhD | 1.017 (0.544 – 1.902) | .957 | 1.385 (0.746 – 2.573) | .302 | 8.956 (2.445 – 32.807) | .001 |
| Career intention at graduation | ||||||
| Not research-related | 1.000 (reference) | 1.000 (reference) | 1.000 (reference) | |||
| Research-related | 3.483 (3.043 – 3.987) | <.001 | 2.529 (2.197 – 2.910) | <.001 | 2.679 (1.206 – 5.952) | .016 |
| Missing | 2.245 (1.475 – 3.417) | <.001 | 1.926 (1.256 – 2.953) | .003 | 1.233 (0.242 – 6.286) | .801 |
| Total debt at graduation | ||||||
| None | 1.000 (reference) | 1.000 (reference) | 1.000 (reference) | |||
| $1 – $99,999 | 0.960 (0.798 – 1.156) | .669 | 0.842 (0.700 – 1.013) | .068 | 0.651 (0.327 – 1.297) | .223 |
| ≥ $100,000 | 0.948 (0.787 – 1.142) | .575 | 0.817 (0.678 – 0.984) | .033 | 0.518 (0.239 – 1.120) | .094 |
| Missing | 0.746 (0.484 – 1.149) | .184 | 0.629 (0.406 – 0.975) | .038 | 1.136 (0.250 – 5.155) | .869 |
| ≥1 year GME research | ||||||
| No | ---- | 1.000 (reference) | 1.000 (reference) | |||
| Yes | ---- | 1.790 (1.573 – 2.037) | <.001 | 4.596 (2.355 – 8.969) | <.001 | |
| F32 research fellowship award | ||||||
| No | ---- | 1.000 (reference) | 1.000 (reference) | |||
| Yes | ---- | 1.413 (0.906 – 2.203) | .128 | 1.990 (0.744 – 5.318) | .170 | |
| USMLE Step l scored | 1.019 (1.016 – 1.023) | <.001 | 1.004 (1.001 – 1.008) | .011 | 1.027 (1.010 – 1.044) | .002 |
Abbreviations: GME = graduate medical education; RPG = research program grant; AOR = adjusted odds ratio; CI = confidence interval; URM = underrepresented race/ethnicity in medicine; USMLE = United States Medical Licensing Examination.
Each model fit the data (each Hosmer and Lemeshow Test P >.05)
AOR < 1.000 indicates a lower likelihood of outcome with each more-recent graduation year.
AOR >1.000 indicates a greater likelihood of outcome with each point increase in Step l score.