Intake, output, and drop out in United Kingdom medical schoolsBMJ 1996; 312 doi: https://doi.org/10.1136/bmj.312.7035.885 (Published 06 April 1996) Cite this as: BMJ 1996;312:885
- James Parkhouse, project consultanta
- a Medical Careers Research Group, Unit of Health Care Epidemiology, University of Oxford, Oxford OX3 7LF
- Accepted 8 March 1996
Precise knowledge about production of doctors is important for planning, but existing data make satisfactory analysis difficult.
Intake, output, and drop out data
Intake—Two sets of figures are available: university returns “at the beginning of the autumn term” and a December census for the University Statistical Record of intake “during the year ended 31 July.” Comparison shows differences of 9-56 in the annual intake numbers for 1990-3 over all UK medical schools.
Output—Figures from the University Statistical Record—the medical return—give numbers gaining a first registerable medical qualification “during the year ended 31 July.” Other tables, compiled by the Higher Education Funding Council for England, show students “leaving due to success,” and give “revised calculation of % drop out” year by year during the undergraduate course. The methods of calculation are difficult to understand, but these tables, made available to individual medical schools, seem to be the only official drop out data. Although both sets of figures appear to be based on the academic year, comparison shows discrepancies of up to 355 in the quoted total numbers of qualifiers for individual years 1989-94 (table 1).
Drop out—Medical students who enrolled in October 1991 will mostly qualify in the summer of 1996. Data for the year ended 31 July 1996 will miss those of the 1991 entry who have to resit examinations later in the year and those who have taken an intercalated degree or whose course has been prolonged for other reasons such as ill health. This undercounting will be offset by the inclusion of resitters and other late qualifiers from the 1990 entry or before. So long as cohort sizes and other factors change only slightly it is reasonable to compare intake with output five years later to derive a drop out rate.1
Table 1 shows that, depending on which output figures are used, the drop out rate for the qualifying years 1989-94, calculated in this way, varied from 9.2% to 17.8%. Apart from two figures, the rates were always above 11%, as were the averages for the six year period. Comparing totals for three year periods of intake and output gives estimates ranging from 11.7% to 14.1%.
In striking contrast, the Higher Education Funding Council's “revised calculation of % drop out” tables give rates for 1988-9 to 1993-4 of “4.68% to 5.94%” for United Kingdom and European Union based students and of “5.28% to 13.37%” for overseas based students.
For our Medical Careers Research Group survey of 1993 qualifiers we compiled a cohort for the calendar year, which yielded 3675 doctors. Excluding nine respondents who had not begun their course in the United Kingdom,2 this gives a drop out rate of 12.4%.
Over many years committees have pondered whether we have enough doctors and enough medical school places.3 The Todd Report of 1968 estimated that by 1994 we should need to train more than 4500 doctors a year for the United Kingdom,1 and we are currently producing a thousand a year less.
Getting the right intake needs knowledge and monitoring of drop out rates. Official tables quoted a drop out rate of about 5%. The Medical Workforce Standing Advisory Committee put the figure at around 10%,4 and my calculations, using the methods of the Todd committee, suggest at least 12%.
Relatively high drop out rates have always been quoted for overseas based students. After a 1992 recommendation5 medical school quotas of overseas based students rose from 5% to 7.5%; this has obvious implications during, and also after, the undergraduate course.
There is much confusion and uncertainty in the available figures. Data collection has now been transferred to the new Higher Education Statistics Agency, which proposes to collect information more rigorously. This opportunity should be taken to improve our medical workforce data for the future.
I thank Dr Michael Goldacre, Mr Trevor Lambert, and Ms Karen Hollick for help in preparing this paper.
Funding The Medical Careers Research Group is funded by the Department of Health. The unit of health care epidemiology is funded by the Anglia and Oxford Regional Health Authority.
Conflict of interest None.