Overview of the world's medical schools: an update

That few data are available on the characteristics of medical schools or on trends within medical education internationally constitutes a major challenge when developing strategies to address physician workforce shortages. Quality and up-to-date information is needed to improve health and education policy planning.

We used publicly available data from the International Medical Education Directory and Avicenna Directories, and an internal education programme database to gather data on medical education provision worldwide. We sent a semi-structured questionnaire to a selection of 346 medical schools, of which 218 (63%) in 81 different countries or territories replied. We contacted ministries of health, national agencies for accreditation or similar bodies to clarify inconsistencies among sources. We identified key informants to obtain country-level specific information. Descriptive statistics were used to analyse current medical school data by country.

There are about 2600 medical schools worldwide. The countries with the largest numbers of schools are India (n = 304), Brazil (n = 182), the USA (n = 173), China (n = 147) and Pakistan (n = 86). One-third of all medical schools are located in five countries and nearly half are located in 10 countries. Of 207 independent states, 24 have no medical school and 50 have only one. Regionally, numbers of citizens per school differ: the Caribbean region has one school per 0.6 million population; the Americas and Oceania each have one school per 1.2 million population; Europe has one school per 1.8 million population; Asia has one school per 3.5 million population, and Africa has one school per 5.0 million population. In 2012, on average, there were 181 graduates per medical school.

The total number and distribution of medical schools around the world are not well matched with existing physician numbers and distribution. The collection and aggregation of medical school data are complex and would benefit from better information on local recognition processes. Longitudinal comparisons are difficult and subject to several sources of error. The consistency and quality of medical school data need to be improved to accurately document potential supply; one example of such an advancement is the World Directory of Medical Schools.