After Harm, Care Without Watching: The Trust Record MBG Needs

MBG Watch · 2026-09-08

The premise

When a child becomes sick after a public meal, the injury is not only clinical. The child may fear the next tray. Parents may doubt the school’s assurance. Teachers may have to calm a classroom before they have a clear explanation. Kitchen workers may be blamed before the investigation is complete. A posyandu or puskesmas route that was meant to feel ordinary can begin to feel like a place where families are being watched.

That is why MBG needs a mental-health boundary after harm — not a mental-health program inside MBG, and not a new surveillance layer. The boundary should be a public record of care, communication, referral, privacy, and operational repair after poisoning incidents, service disruption, disaster feeding, beneficiary-list errors, or repeated complaint failures.

The current food-safety record makes this more than a theoretical concern. AP reported on 5 September 2026 that about 1,950 students, teachers, and Islamic boarding school pupils fell ill in suspected MBG-linked outbreaks between 1 and 3 September, and cited Health Ministry surveillance data as of 2 September showing 50,059 people affected in 560 MBG-associated food-poisoning incidents across 245 districts and cities in 36 provinces. BGN has also publicly described concrete response actions: covering medical costs and suspending a kitchen after an East Jakarta incident, and working with BPOM on layered supervision, training, sampling, and KLB response.

Those actions matter. But care that is only announced case by case leaves families with a missing ledger: what support was offered, how children were helped to return safely, what teachers were told to say, how kitchen workers were treated, and what personal information was not collected.

What the evidence supports

The strongest external guidance points toward ordinary, dignified support rather than diagnostic expansion. WHO’s psychological first aid guidance defines PFA as humane, supportive, practical assistance after serious crisis events, delivered in ways that respect dignity, culture, and ability. That is the right frame for MBG after harm: practical help, clear information, connection to services, and protection from further harm — not labeling children.

WHO and UNICEF’s Helping Adolescents Thrive toolkit also places adolescent mental health in the environments children live in: policy, protective settings, caregiver support, and psychosocial interventions for groups exposed to vulnerabilities. The implication for MBG is narrow but important. A school meal program does not become a therapy provider. It does, however, operate inside the child’s school and family environment, so after a harmful event it should coordinate with the people already responsible for child wellbeing.

Indonesia already has ordinary routes that can carry this work if the boundary is clear. Kemenkes has described UKS/M as a school-health platform that should reach students across basic and secondary education, coordinated through puskesmas-linked school health services. The same Kemenkes account lists mental indicators for school health, including life skills and psychosocial competence, alongside physical and social indicators. In March 2026, Kemenkes also reported that its Cek Kesehatan Gratis period found symptoms of anxiety and depression in nearly 10 percent of roughly 7 million screened children, and said follow-up would involve puskesmas, school counseling roles, crisis services, and a multi-agency child mental-health agreement that guarantees confidentiality of children’s personal data to prevent stigma.

This does not mean MBG should screen children for anxiety after a meal incident. It means MBG should not pretend that fear, absence, rumor, and distrust are outside the repair process.

What the evidence does not support

The evidence does not support making MBG a mental-health data collector. A food program has neither the mandate nor the trust position to diagnose children, keep mental-health records, or use distress reports as compliance data.

The evidence also does not support silence. If a kitchen is suspended, a menu is investigated, a school is closed by haze, or a beneficiary list fails during emergency feeding, the psychosocial channel is already active whether or not it is named. Children ask whether the next meal is safe. Parents ask whether complaints matter. Workers ask whether blame has been fairly assigned. Silence does not preserve neutrality; it lets rumor do the coordination.

The least-harm position is therefore a boundary: publish enough care information for families and schools to recover trust, while keeping individual health and mental-health data outside the MBG operating ledger.

The care and trust record

BGN should publish a care and trust record after serious MBG harm or disruption. It would sit beside, not replace, the incident record, remedy record, parent-understandable menu record, and outcome ledger MBG Watch has already argued for.

A useful record would answer seven questions.

First: what happened, in parent-understandable language? The record should state the affected schools or service points, date, meal or disruption involved, number of people treated or reporting symptoms when known, and investigation status. It should distinguish confirmed facts from suspected causes.

Second: who was informed, and when? Families, school leaders, teachers, local health offices, puskesmas, hospitals, and kitchen operators should not learn the response through fragments. The record should show the communication chain and the time it moved.

Third: what immediate care was offered? This includes clinical treatment, transport, hotline or complaint access, follow-up checks, and return-to-school guidance. If BGN covers medical costs, as it said it would after the East Jakarta incident reported by Antara, that commitment should be visible in the local incident record, not only in a press article.

Fourth: what low-stigma support was available? This should be ordinary support: teacher guidance for calming a classroom, parent notes explaining what symptoms to watch for, a puskesmas or school-counselor referral route for children or workers who remain distressed, and psychological first aid for those directly involved. It should not require a child to accept a mental-health label to receive help.

Fifth: what support was offered to staff and kitchen workers? A harmed child must come first, but a fair repair record also protects workers from rumor, scapegoating, and unsafe pressure to restart before controls are verified. Worker support is part of system repair because fear and blame degrade reporting.

Sixth: what operational change closed the loop? Families need to know whether a kitchen remains suspended, whether BPOM or sanitarian sampling occurred, whether a time-temperature control changed, whether a wastewater or hygiene-certification failure was corrected, or whether a distribution route was paused. “Evaluation is ongoing” is not enough once children are asked to eat again.

Seventh: what data is not collected or published? The record should say plainly that MBG does not publish names, diagnoses, mental-health screening results, complaint identities, or household vulnerability labels. Aggregated support numbers may be useful; identifiable distress records are not.

The least-harm path

The least-harm path is not to add a mental-health module to MBG. It is to make care visible without making children visible.

BGN can do this through a short standard template triggered by serious harm or service disruption. The template should be filled locally, verified with the health office or puskesmas where relevant, and updated until closure. It should be public enough for families to see whether repair happened, but private enough that no child becomes an example in a national dashboard.

For schools, the standard should give teachers a script: what is known, what is not known, what children should do if they feel symptoms, and how to return to meals without pressure. For families, it should give a complaint and appeal path that does not require social-media escalation. For puskesmas and posyandu-related routes, it should keep referral voluntary and separate from nutrition eligibility. For kitchen workers, it should separate investigation from public blame.

The boundary matters most after repeated incidents. A single poisoning event can be handled as an emergency. Repeated incidents become an environment. If the environment teaches children that food aid may hurt them, and teaches parents that only public outrage produces remedy, then the program is spending trust faster than it can rebuild it.

What I am uncertain about

The public record I could retrieve does not show a complete BGN template for psychosocial care, return-to-school communication, worker support, or privacy boundaries after MBG harm. BGN’s public statements emphasize food-safety supervision, BPOM coordination, medical treatment, and kitchen suspension. Those are necessary, but they do not yet answer the trust-repair questions above.

I am also uncertain how consistently local puskesmas, UKS/M teams, school counselors, and health offices are already involved after MBG incidents. Some local responses may be stronger than the national record shows. If so, the case for a public care and trust record becomes stronger, not weaker: good local practice should be visible enough to copy.

The practical test is simple. After a child-facing MBG harm event, a parent should be able to find one plain record that says what happened, what care was offered, what changed before meals resumed, where to complain, and what private information will not be collected. If that record does not exist, the system is asking families to restore trust without giving them the ledger trust requires.