Table 1. Principal component analyses of influences underlying career choice in medical graduates to five years after graduation

Table 1 shows the results of the principal component analyses, demonstrating well-formed factors at each time point. Factors are reported in the descending order of their variance explained (VE), and with a factor score, being the mean (SD) of the Likert question scores making up the factor. The VE is a measure of the amount of variability in the participants’ responses that can be explained by the factor, hence higher levels of VE indicate agreement by a larger number of graduates. The factor score indicates the degree to which graduates consider the factor to be important or unimportant in their choices.  

For example, at Exit from medical school (Table 1A), Factor 1 accounts for the largest VE (15.8%), comprising the 9 question items that are, on average, the least influential in determining career choice for graduates, with a factor score of 2.3 (out of 5). These least influential items are: Costs of Vocational Training; Costs of Medical School; Insurance Risk; Parents/Relatives; Prestige; Training Years; Financial Prospects; Research/Teaching; and Location. By contrast, Factor 4 at Exit, with the highest factor score of 3.9 and explaining 9.3% of the variance, contains the 4 items rated as the most influential by graduates in determining career choice. These most influential items are: Medical School Experiences; Specialty Experience; Mentors; and Self-Appraisal. These results are consistent with new graduates making the most of their abilities and opportunities to secure their first healthcare role. Other factors at Exit fall within these two extremes. 

It is worth noting that the item Parent/Relatives fails to load over the 0.3 threshold on any factor at PGY1 or PGY3 (hence does not appear). Some change in factor structures over time do occur, reflecting changing priorities for graduates. For example, Factor 2 at Exit has a relatively high factor score of 3.6 (VE=11.1%) indicating that the items Flexible Hours, Working Hours, Domestic Circumstances, and Work Culture are important for new graduates. However, by PGY1 (Table 1B) this factor then picks up the items of Training Years, Vocational Training Availability, Location and Job Security, and becomes important to a greater number of graduates by becoming the factor with the largest variance explained (VE=12.4%). This result suggests that graduates are adjusting to their new working lives and are planning for their futures in terms of further training. 

Factor 1 and Factor 2 consistently demonstrate high levels of variance explained and contain a common core of three influencing items. Factor 1, with a median (range) score of 2.1 (1.8 to 2.3) across all time points, continues to describe influences on career choice rated as the least important for medical graduates, and consistently contains the items Costs of Vocational Training, Costs of Medical School, and Insurance Risk. By contrast, Factor 2 is consistently rated as relatively important, with a median (range) score of 3.4 (3.1 to 3.6) across time points, and consistently contains the items Flexible Hours, Working Hours, and Domestic Circumstances.  

Our results demonstrate the existence of well-formed factors in the MSOD data at each time point. Despite some change in factor structure over time, the scores for each factor remain relatively stable, with a median (range) variation in scores of only 0.5 (0.3 to 1.0) Likert points. Table 1 contains results which allow substantial scope for hypothesis formation and future research, including targeted work to better understand the decision points in the critical first five years of a graduate’s career. 

IV. DISCUSSION

Better understanding the influences underlying career choice in medical graduates is a strategically important and practical concern when aiming to match graduate production with professional and community needs. This study is the among the first to conduct a quantitative analysis of the self-reported influences underlying medical graduate career choice in a prospective, national cohort of the same graduates over the critical first five years of their careers.  

The financial burden of completing medical school and vocational training is one of the most widely studied influences in career choice for medical graduates. It is therefore reassuring that our findings demonstrate that these costs are among the least influential considerations at all time points in the five years after graduation for our cohort.  

Factor 2 in the present study consistently contains the items Flexible Hours, Working Hours and Domestic Circumstances, and is rated as important over the first five years of graduates’ careers. Flexibility around working hours and a desire to practice part-time has traditionally been thought of as largely important for female medical graduates (Heiliger & Hingstman, 2000). However, this is no longer the case, with many male graduates in recent decades also desiring more lifestyle-friendly working arrangements allowing the flexibility to spend more time with family (Heiliger & Hingstman, 2000). Taken together with the evidence that the costs of medical school and further training are the least influential in career choice, our results therefore strongly suggest that the ability to offer greater flexibility in working hours is likely to be useful in recruiting medical graduates to underserved specialties. 

It is a practical and pressing necessity that healthcare workforce planning is guided by the best available evidence. A strength of the current study is the ability to link the same participants longitudinally, thus eliminating an important source of bias. Women in the current study made up 57% of medical graduate respondents, reflecting the fact that in recent years in New Zealand and Australia female graduates have outnumbered male graduates. A further strength is the high response rates, yielding a sampling margin of error of only 3% or less at all time points, which compares favourably with many questionnaire studies of medical graduates. 

V. CONCLUSION

Despite widespread concern over rising debt levels and the cost of medical school, our results are reassuring in that the costs of medical school and vocational training were consistently rated as the least important influences in career choice. Our results also suggest that offering greater flexibility around working hours may assist in attracting medical graduates to underserved specialties. Our description of well-formed factors in the influences underlying career choice in the national MSOD questionnaire data provides a useful basis for further research to better understand key decision points in the critical first five years of graduates’ careers. 

Notes on Contributors

Craig Webster was involved in the conceptualisation of this paper, data analysis, writing and revision.

Jack Forsythe was involved in the conceptualisation of this paper, data analysis, writing and revision.

Antonia Verstappen was involved in accessing data for this paper, writing and revision.

Phillippa Poole was involved in the writing and revision of this paper.

Tim Wilkinson was involved in the writing and revision of this paper.

Marcus Henning was involved in the writing and revision of this paper. 

Ethical Approval

This study was carried out in accordance with all regulations of the host organisations and with the approvals of the Human Participants Ethics Committees of the University of Auckland (approval numbers 022388 and 018456) and the University of Otago (approval number 07-155), New Zealand. All participants gave written informed consent to participate in the study, including for anonymised aggregated data to be published. 

Data Availability

The ethics approval for the longitudinal MSOD project currently does not permit the sharing of non-aggregated data. However, this restriction is under review and so non-aggregated data may be available from the corresponding author in the near future.

Acknowledgement

We thank the Health Career Pathways Project, Faculty of Medical and Health Sciences, University of Auckland, and the Medical Schools Outcomes Database Longitudinal Tracking Project at the University of Auckland and the University of Otago for assistance and data access. 

Funding

JF received a summer studentship stipend from the Faculty of Medical and Health Sciences, University of Auckland, New Zealand in support of this research. The Medical Schools Outcomes Database Longitudinal Tracking Project is supported by a grant from Health Workforce New Zealand. 

Declaration of Interest

All authors have no potential conflicts of interest. 

References

Gorman, D. (2018). Matching the production of doctors with national needs. Medical Education, 52(1), 103-113. https://doi.org/10.1111/medu.13369

Heiliger, P. J., & Hingstman, L. (2000). Career preferences and the work-family balance in medicine: Gender differences among medical specialists. Social Science and Medicine, 50(9), 1235-1246. https://doi.org/10.1016/s0277-9536(99)00363-9

Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 31–36. https://doi.org/10.1007/BF02291575

Poole, P., Wilkinson, T. J., Bagg, W., Freegard, J., Hyland, F., Jo, C. E., Kool, B., Roberts, E., Rudland, J., Smith, B., & Verstappen, A. (2019). Developing New Zealand’s medical workforce: Realising the potential of longitudinal career tracking. New Zealand Medical Journal, 132(1495), 65-73.

Webster, C. S., Ling, C., Barrow, M., Poole, P., & Henning, M. (2017). A cross-disciplinary assessment of student loans debt, financial support for study and career preferences upon graduation. New Zealand Medical Journal, 130(1459), 43-53.

Webster, C. S., McKillop, A., Bennett, W., & Bagg, W. A. (2020). A qualitative and semiquantitative exploration of the experience of a rural and regional clinical placement programme. Medical Science Educator, 30(2), 783-789. https://doi.org/10.1007/s40670-020-00949-6

*Craig Webster
Centre for Medical and Health Sciences Education,
School of Medicine, University of Auckland,
Private Bag 92-019
Auckland 1142, New Zealand
+649 923 6525
Email: c.webster@auckland.ac.nz