Equity, Merit and Need of the Country as the Basis for Selection of Students for Medical Education in Sri Lanka

Equity, Merit and Need of the Country as the Basis for Selection of Students for Medical Education in Sri Lanka

Most of the concerns seem to be based on the perception that the opportunities for medical education in Sri Lanka are severely restricted and the GCE A/L is not a valid, reliable and fair system for student selection. This is an attempt to analyze these issues through available data and evidence.

Does the current system restrict opportunities for biology students?

Let’s look at some data.

In 2013, 40253 students sat for the GCE A/L in Biology stream; 20215 students got three S or more (50 % pass rate). Out of those 20215, over 7000 students got selected to the state universities (30% or one out of every three students who have passed) ¹.

Now, lets compare this with the Arts stream,

In 2013, 70, 000 students passed the examination and only 7000 got selected to the state universities (Only 10 %) ¹.

Now where is the real restriction?

·       
The opportunities to study in Biology stream are restricted. In the whole country, there are less than 800 schools  with advanced level biology classes. Only 50% of biology students pass GCE A/Ls – The main problem is within the secondary education system, where the standard of science education has remained at a lower level. 

·       There is a severe lack of higher education opportunities in Arts stream and a serious issue of employability of Arts graduates. Why there is no enthusiasm to open up private universities in the arts stream? The reason is obvious.

What is the situation regarding the opportunities for state-sector medical education in Sri Lanka?

In 2013, 1300 students out of 20215 students got three S or more were selected for medicine (6.5%). If we look at higher performers, 2388 students got 3 B s and above¹. Therefore approximately 50%  of those who got 3 Bs and above were selected for medicine in state universities. Admission percentage for 3 As is almost always 100% (with very few exceptions).

Is this overly competitive?

Let’s compare these figures with UK system, which has many similarities with the Sri Lankan system due to historical reasons. In the UK, the minimum entry criteria in the standard medical school entry pathway is 3 As. Out of those students with 3 As, only one out of every 5 students get a chance to study medicine². For some high ranking universities like Oxford, this could be 1 out of 17 applicants².

Entry into high standard medical education is competitive, anywhere in the world. The acceptance rate at Harvard Medical School was 165/ 7000 (2%) in 2016³. The All India Institute of Medical Sciences in New Delhi, accepts only 72 students for its undergraduate courses out of the 80,000 to 90,000 students who apply⁴.

How are the University admission opportunities distributed within the Biology Stream?

In 2014, 1255  students out of 7000 Biology stream university admissions  were selected for Medicine¹. That means 1 out of every 5^th Biology stream university student got opportunity to do Medicine.  Only a total of 700 students are admitted for all other health related state university courses⁵.

Is this annual intake of 1300 medical students adequate for the country?

Yes. This is approximately the number that can be afforded by the country’s health system. According to the government’s own admission, there are 6000 vacancies for doctors within the state sector. This means our health system will be saturated with doctors within next 5 years unless there is a rapid expansion of the government and the private sector.  Projections made by sophisticated HRH modeling systems predict the same. The serious HR issue in Sri Lanka is maldistribution and mismanagement than adequacy of numbers.

Disproportionate expansion for seats in Medicine without a concomitant expansion in other health related fields would lead to a top-heavy staff structure resulting in doctors being forced to undertake tasks that a nurse or a paramedical worker could do equally competently and at lower cost.  The proportions are already bad with 25600  doctors, 50000 nurses and just 800 physiotherapists. Even now some allied health courses struggle to fill the full quota of seats.

Therefore the real need for the expansion of opportunities  is not in medicine, but in other fields such as nursing, allied health and science-based disciplines.  There is also a need to improve the employability, remuneration and professional status of these disciplines.

There are some concerns are expressed regarding the validity and reliability of GCE A/L as a selection test.

Is GCE A/L valid admission system?

There are three broad reasons why A/levels should have predictive validity for medical school performance;

1.     
      Cognitive ability—difficulty index of over 90% of the MCQs in 2015 Biology paper are within the ideal range of 0.3 to 0.8.

2.     Substantive content—A levels provide students with a broad array of facts, ideas, and theories  relevant future medical students. Analysis of the questions  (2015 A/L) shows a substantial coverage of human biology, molecular biology, cell biology and genetics

3.     Motivation and personality—achieving high grades at A level requires appropriate motivation, commitment, personality, and attitudes, traits that are also beneficial at medical school and for lifelong learning.

Validity of a selection test is measured through predictive validity. Worldwide evidence show that the academic merit demonstrated during selection exams as the best predictor of medical school performance compared to other methods such as interviews, aptitude testing and letters of recommendation.  Evidence from a large scale meta-analysis conducted in the UK suggest that A/L performance as an excellent predictor of performance along the continuum of medical education starting from undergraduate to entry in to specialist register of the GMC⁶. According to another meta-analysis on US medical school entrance, the biological sciences subset of the MCAT was the best predictor of medical school performance⁷

In Sri Lanka, a study conducted at the Faculty of Medicine, Colombo showed 0.4 correlation between z-score and the medical school performance (In statistical terms, 0.4 is a moderately high correlation).The A/L z-score, A/L attempt and English placement test marks were all significant predictors of outcome at the First Examination⁸.

There may be concerns to why the correlations shown in Sri Lankan studies have been either low or moderately high. The reasons are contextual and statistical.

One main reason is that most state medical faculties admit students based on merit as well district quota system. This gives a scattered distribution of z-scores. The moderately high correlation shown by Colombo, which admits students only based on merit further supports that.

Sri Lankan studies have used direct correlation or regression methods. The more recent UK meta analysis is based on the studies that have used more sophisticated statistical methods to remove the impact of compounding factors and only academic merit is correlated with the performance⁶. 

In the light of above evidence, the UK studies recommend that measures to improve reliability and validity of selection have to be made within the existing framework of A/L system.

Do we have a fair an equitable  admission system?

According to the article 26.1 of the universal declaration of human rights, admission to higher education should be based on the merit and equity.” Many countries have found it challenging to ensure equity in access to higher education. Countries such as UK, USA and Canada have historically experienced low representation of students from lower socio-economic strata in medical education programmes.

The district quota system in Sri Lanka is an attempt to ensure equity in access to higher education. This system presupposes different opportunity levels exist for education at economically diverse districts. This system allocates 40% of selection as all island merit and out of the remaining 60%, 55% of the opportunities are distributed to all the administrative districts of the country and each district is given a quota according to the estimated midyear population of the district. This quota is filled by the students who achieve highest marks next to those who were included in the all island merit category. The remaining 5% of the opportunities are distributed across 16 identified ‘educationally disadvantaged’ districts. District quota system ensures that students from all geographical areas of the country are given an opportunity to enter into a state run medical school.

In spite of many serious limitations of, there is some evidence to suggest that the district quota system has served some purpose in achieving equity

1.     
      Mannar, Mulaitiv and Monaragala (3 M districts) are the districts that are widely considered as poor in educational resources would not have a single student entering the medical faculties if 100% merit based selection was applied.

2.     Relatively low correlation between advanced level results (Z score) and undergraduate performance in state medical faculties.

3.     One analysis has shown that the z-scores within a batch of students show a left-skewed distribution whereas the first year GPA follows Gaussian curve. It is possible that despite variations in cutoff z scores in different districts, those students perform as part of a cohort within a normal distribution when they are given the similar opportunities at the medical school.

Why BBB or equivalent Z Score a better minimum criterion than a lower grade?

Academic merit, need of the country and equity in access to medical education could be considered three principles based on which a higher grade is justified.

1.     
      Academic merit

Based on available evidence, academic merit qualifies to be the sole criterion predictive of later performance in medical education. Thus, in a country like UK, AAA has become the minimum entry criterion. Malaysia allows 3 Bs in Malaysian A/L examination, but has faced with the problem of incompetent doctors⁹. Despite lack of evidence on comparability of these examinations, the importance of high grades in A/L s as minimum entry criteria to ensure a quality product cannot be disputed.     

The decision regarding selection cut-off of students should also be based on the baseline performance applicable to that population.  Over time, the performance of students at the A/L has improved. This grade inflation is a well-recognised phenomenon that can be explained by many factors. The baselines of 3 S or 3 C are not applicable anymore since the current baseline performance of those who are selected to follow medicine are far higher than that.  As grade inflation occurs, that rational and logical measure should be to make the bar higher. How UK has responded to this was by increase the bar as 3 As minimum, as mentioned above.

The current cohort of Sri Lankan medical students, 3 B could be argued as a reasonable baseline due to following reasons.

1. Even within the underprivileged districts, those who a getting these low grades are a very small minority of outliers. For example in 2014, 30 students from Nuwaraeliya had 3 B and above and 32 students were selected for Medicine.

2. Even the grades of those who have been selected based on the lowest Z-scores have been on the rise.

2010      C          C           S             Mulaitive

2011      C          S            B            Nuwara Eliya

2012      C         C            C             Mulaitive

2013      C         C            B             Mannar

2014      C         B            B             Nuwara Eliya

2015      B        B             B

The need to raise the bar is evident by ERPM results, in 2010, pass rates were very law and has increased by 200% in 2016.

It is only due to competition or does it represent the minimum level to complete the medical course?

The extremely low dropout rates of Sri Lankan medical students (less than 1%) also suggests that the selection method has been effective in selecting those who are capable of completing the medical course.

Therefore, the baseline for current cohort for students can be assumed as 3 B and above.

Therefore, minimum entry criteria imposed based on A/level results is not overly restrictive, but allows selection of best students of a cohort to study medicine.

2
      Need of the country

In 2013, 2014 and 2015 the number of students who obtained BBB or above were 2388, 2605 and 2888 respectively. With 1310 students admitted to state medical faculties in 2015, 1578 more students would have been eligible for private or foreign medical education in that year, if BBB criterion was in place. Therefore, a BBB minimum entry criterion can potentially double the annual production of doctors for the country. Recent studies have suggested that with the current numbers in the pipeline, country is expected to saturate with doctors in next 5 years and any further expansion in numbers need to be bases on a sound HRH policy. Further, contrary to common beliefs, a surplus of doctors can lead to problems in health system. Therefore, currently there is no requirement to lower the entry criteria further to expand opportunities in medical education.         

3  
      Equity

Students with BBB or above are obviously more deserving than students with lower grades. Therefore, these students must be provided with financial support in terms of scholarships or educational loans to pursue their education. This will ensure equity to greater degree than considering a lower grade to support financial interests of institutions that provide private medical education abroad or locally.

1
      Department of examination of Sri Lanka (2017), Official statistics provided to SLMC

2.     Medical Schools council of UK (2017)

3.     Harvard Medical School, selection factors and admission statistics (2017) https://hms.harvard.edu/departments/admissions/applying/selection-factors-admissions-statistics

4.     AAIM http://www.aiims.edu/en.html

5.     University Grants Commission (2016) Sri Lanka University Statistics <http://ugc.ac.lk/en/component/content/article/1709-sri-lanka-university-statistics-2015.html>

6.     McManus et al. (2013) Construct-level predictive validity of educational attainment and intellectual aptitude tests in medical student selection: meta-regression of six UK longitudinal studies. BMC Medicine, 11:243 http://www.biomedcentral.com/1741-7015/11/243

7.     Donnon et al. (2007) The Predictive Validity of the MCAT for Medical School Performance and Medical Board Licensing Examinations: A MetaAnalysis of the Published Research, Academic Medicine, 82: 1

8.     Mettananda et al (2006) The Ceylon Journal of Medical Science 2006; 49: 1-12

9.     WHO (2014) HRH Country profile, Malaysia.