KUALA LUMPUR, June 17 — The Ministry of Health (MOH) has changed the ePlacement system for medical officers from a first-come, first-served basis to one where placements will be decided entirely by the system’s algorithm.
The 8/2025 batch of 2,248 UD10 medical officers for permanent appointments – whose placements on February 27 were nullified due to “technical problems” in the ePlacement system – will be the first to try the new system that requires candidates to choose three facilities for placements. One of the three options must include a facility in either Sabah or Sarawak.
Crucially, a doctor’s posting will be decided solely by a machine learning algorithm – without any human intervention or bias.
During two virtual ePlacement briefings on Google yesterday for medical officers, MOH officials explained that the new “merit-based” algorithm will look at a combination of factors, including the candidate’s personal health, career aspirations, and family background.
“This method is expected to be fairer, and more transparent and equitable,” a Human Resource Division (BSM) official told the morning briefing that was held together with officials from the Information Management Division (BPN) and Digital Health Division (BKD).
According to an FAQ on the new ePlacement system, medical officers do not need to make new registrations on the system as the system will use doctors’ old data. However, candidates are reminded to update their personal profile before the ePlacement session.
Medical officers can begin updating their personal profile on the ePlacement system until June 23 at 11.59pm.
The new ePlacement session will be held from June 24 at 12pm to June 27 at 11.59pm. A doctor’s failure to select placements during the ePlacement session will be considered a rejection of their permanent appointment offer.
Appeals to change the decision on placements will be opened from July 8 at 10am until July 14 at 5pm. Decisions on appeals will be announced on August 19.
Candidates must report for duty at their respective facilities on October 2.
The two briefings by the MOH yesterday, held about a week ahead of the ePlacement session next Tuesday, were attended by some 800 participants combined, including CodeBlue that joined both video calls.
“This merit system for the placement of permanent medical officers was developed to ensure a fair, transparent, and equitable process. It takes into account a combination of various factors, including one’s personal circumstances, family, health, and career aspirations to ensure that decisions on placements are made comprehensively and focused on the needs and vulnerabilities of the officer’s household,” according to the MOH FAQ.
“Marks on merit are calculated based on a weighted scoring system that takes into account various factors holistically. Each component is given certain marks based on set criteria.”
Personal Profile On ePlacement System: Personal Information, Specialist Training, Family Background, Document Checklist
Based on a demo of the ePlacement system during yesterday’s briefings, a candidate’s personal profile comprises personal information, specialist training, family background, and a document checklist.
Personal information includes data fields like an officer’s marital status, place of residence, health problems (if any), and OKU (person with disability) status (if any).
A drop-down list appears for health problems, including genitourinary diseases (for example, chronic kidney disease), cardiovascular diseases (for example, hypertension, heart disease, stroke), neurological conditions (for example, stroke, epilepsy, Parkinson’s disease), immunocompromised conditions (for example, long-term steroid therapy), mobility impairments (for example, wheelchair users, limb loss), metabolic disorders (for example, diabetes, thyroid disorders), autoimmune diseases, and tuberculosis, among others.
For the specialist training category, the officer must state their main interest (hospital or clinic). Optional data fields include specialist training status (MedEX/ parallel pathway/ gazettement) and field of specialty.
For the family background category, officers can list their dependents or family members. Details of their family members include whether or not they have health problems and OKU status. A similar drop-down list of health problems appears, as for the officer themselves.
Under the checklist category, officers can upload a copy of their marriage certificate, medical reports for themselves or their family members, employer’s acknowledgement for spouse’s place of work, OKU status document for themselves or their family members, scholarship offer letter, and exam result slip – whichever document is applicable.
ePlacement Application: Three Choices With Justifications, Sabah/ Sarawak Mandatory
When applying for placements, medical officers will be able to choose three specific facilities, including a mandatory selection of a facility in either Sabah or Sarawak.
The ePlacement system does not allow a candidate to proceed if none of their three choices includes Sabah or Sarawak.
Officers need to provide a justification for each of their three options for placements. A drop-down list for the justifications show factors like the officer or officer’s spouse being pregnant, officer’s own health problems, siblings’ health problems, threat to safety accompanied by a police or official report, parents’ health problems, spouse has a permanent appointment but choose another state health department (JKN), officer wants to continue their postgraduate study, or officer has previously served in Sarawak, among other reasons listed.
However, the drop-down list does not include justifications like “MOH is forcing me to pick Sabah/ Sarawak” or, in the case of candidates actually preferring to be posted to Borneo, “I want to serve in Sabah/ Sarawak.”
CodeBlue understands that the system’s algorithm assumes that first or second choices indicate genuine preference for a Sabah or Sarawak posting, whereas placing these states as the third choice is assumed to be an unwanted compulsion.
Machine Learning Algorithm Solely Decides Placement, ‘Free Of Human Bias’
BKD deputy director Dr Vivek Jason Jayaraj told yesterday’s briefings that the MOH has developed a “machine learning algorithm” that takes into account constraints like vacancies in a particular facility, including the “many vacancies” in Sabah and Sarawak.
“The algorithm will then sift through the data you put in — your family details, comorbidities, interests – and it will then suggest the best placement for you,” said Dr Jason.
“We score for you. Based on that score, we’ll assign you to a location.”
Dr Jason explained that if a candidate chooses Shah Alam Hospital as their first choice, for example, the algorithm will try to allocate that facility. But if that’s not possible, the algorithm will look for hospitals or clinics – based on the candidate’s main interest – within that region. If necessary, the algorithm will move on to the second and third choice respectively.
“Ninety-nine per cent of the time, you will receive one of those three selections as the chairperson has already pointed out earlier,” he said.
When asked whether a human being would be involved in decisions on postings, the BKD official said the placement process would be “entirely automated.”
“We don’t want this person to have a say; we want the algorithm to fairly adjudicate where you go.”
Algorithm Weightage Can’t Be Made Public To Avoid Bias
Dr Jason said the MOH will not be able to disclose the logic of the algorithm or the weights of different criteria in determining a medical officer’s placement, as this could bias the system.
“But what we can tell you is, please answer all the form questions to the best of your ability and populate the system with your documents. All those things come into play.”
When asked why the MOH believes that the algorithm will be fairer than a first-come, first-served basis, Dr Jason said, “it’s quite clear that the algorithm is free of human bias.”
He noted that a first-come, first-served basis was unfair to people without good internet access, claiming that “malicious attacks” were used to “try and sway the system” during the invalidated February ePlacement session.
“It’s not strictly an AI in a large language model,” Dr Jason said about the new algorithm. “But it is machine learning. It’s using elements of machine learning within the algorithm itself.”
He added that the MOH has been running multiple different simulation exercises over the past two months to look at the “error percentage”; the error rates were “very, very low.”
CodeBlue understands that data entry from candidates will flow into another “external system” to work placement decisions; results will then be pushed back into the ePlacement system.
‘Don’t Want To Split Up Families’ With Sick Members
On factors like a medical officer having a child with disability or being married to another doctor who is also selecting permanent placements, Dr Jason explained that candidates in these cases will likely get their first selection or be posted to the same state.
“Again, if this wasn’t clear, the algorithm is free of human hands. This is just meant to be as transparent as possible, as fair as possible, because we understand that humans can have biases, and we don’t want to introduce those biases.”
When asked whether the algorithm was biased against single people, Dr Jason simply said criteria like comorbidities and disabilities are “very, very important to us.”
“We don’t want to split up families that potentially could have sick parents, sick children.”
He also warned medical officers against submitting false data or documents, as the MOH can cross-reference one’s OKU number against existing government databases, for example.
“The validity of the algorithm itself runs through a governance committee. So rest assured, we are trying as much as possible to squeeze out all elements of bias, of human manipulation, from the algorithm. This will be as fair as possible to as many people as possible.”
Being First Or Last In Applying On ePlacement Doesn’t Matter
Dr Jason stressed that medical officers do not need to rush to be the first to apply when the ePlacement session opens on June 24, since placements are no longer decided on a first-come, first-served basis.
The new system allows doctors to choose whichever three facilities they want – albeit with one Sabah/ Sarawak mandatory selection – throughout the ePlacement session from June 24 to 27.
Regardless of whether candidates are the first or last to make their selections during those three days, the algorithm only kicks in and makes a decision, based on what it thinks is best for the doctor, after the ePlacement session is closed.
Last April, CodeBlue published its own survey among more than 900 medical officers on their preference for placements, finding that 62 per cent would rather quit government service than serve away from their home state.
Some of the placement factors cited in the CodeBlue poll are similar to the new ePlacement system, such as preference for type of facility, personal health, and career development. Family was rated as the most important factor in deciding where to serve.
