Summary

  • WHO’s proposed framework gives ministries a voluntary, adaptable starting point across policy, workforce training and incident management; it is capacity-building guidance, not a binding rule or certification.
  • Participation and formal endorsement are not operational accountability. Each joint operation needs a clear, safe and survivor-centred route from disclosure to support, investigation and an authorized remedy.

One operation, several lines of responsibility

A vaccination campaign, emergency response or development programme can put ministry personnel, WHO staff and implementing partners in the same field setting. People receiving care may experience the intervention as one service. Yet the organizations involved do not automatically share the same employer, policies, investigative authority or disciplinary powers. That gap is where a written framework can help—and where a slogan about “joint accountability” can mislead.

WHO’s proposed Prevention of and Response to Sexual Exploitation, Abuse and Harassment (PRSEAH) accountability framework addresses Member States in the context of joint government–WHO operations. Its institutional route matters. WHO’s Executive Board discussed the proposal in 2025; WHO later said the governing bodies noted it. The framework is described as non-binding, voluntary, adaptable to country circumstances and a minimum starting point for ministries of health. It is not a treaty, a uniform national law or a WHO certificate that an operation is safe.

The framework names three capacities: policy and standards; training for personnel and implementing partners; and incident management. The last is the decisive test. The leaflet calls for safe, accessible reporting, communication to populations and workforces at risk, survivor-centred support, and investigation followed by disciplinary or legal measures when allegations are substantiated. A code of conduct without a usable reporting route is not an incident system. Training without an accountable decision-maker cannot guarantee a response. And an investigation that cannot lead to a remedy leaves a broken chain.

Support is not substitution

The framework explicitly treats ministries of health as ultimate duty-bearers for protecting populations, while allowing WHO and health-sector partners to support capacity-building. That allocation should not be blurred. WHO can help develop standards, train staff, convene partners and provide technical assistance. It cannot, by publishing guidance, assume a government’s responsibilities under its own law or decide every case involving another organization’s workforce.

Nor does the word “joint” create one joint employer or investigator. The applicable authority may differ according to who employed the person, where the conduct occurred, which policies govern the operation and what national law requires. The framework’s short leaflet does not settle each jurisdictional question. Those limits are a reason to make responsibilities legible before an incident, not to imply that WHO has supplied a universal case-handling mechanism.

The difference between support and substitution is central to the agency problem. The organization that finances or coordinates a programme may not control the personnel implicated in an allegation. The ministry responsible for public protection may not hold the employment file. A local partner may receive the first disclosure but lack authority to investigate. If these roles are left implicit, each actor can plausibly expect another to act. A named lead, an escalation path and a written statement of decision rights make that hand-off testable.

A meeting is not an implementation record

WHO’s December 2025 report records a regional conference in Pretoria intended to advance operationalization and says the South-East Asia Region was responding to requests for technical support. Those are meaningful signs of engagement. They are not, by themselves, evidence that every participating country adopted the framework, equipped every operation with safe channels or delivered survivor support in practice. The same report asks the Executive Board what support would help accelerate adoption and operationalization. That question is evidence that the work remained in progress at the time of reporting—not proof that no country had acted.

The available materials also use different regional counts: a WHO news account refers to 42 African Member States, while the Board report describes a conference intended to operationalize the framework across 47 countries. The documents do not explain the relationship between those figures. Neither should be turned into a claim that a given number of governments implemented the framework. Conference participation, political support, adoption of a policy, operational capacity and improved outcomes are distinct stages.

An honest progress account would keep them separate. It could state whether a ministry has issued a policy; whether staff and partners have received role-specific training; whether reporting is safe and accessible; whether referral and survivor-support capacity is available; whether investigators and disciplinary authorities are designated; and whether recommendations are followed through. Each measure answers a different question. None should be presented as a proxy for the number of incidents or the quality of survivor experience.

Accountability must protect the person who reports

Transparency here cannot mean publishing case files. A useful public record would be de-identified and limited to governance: which policy applies, which state and WHO contacts receive reports, how a person can reach confidential support, which authority investigates, who can impose discipline or refer a matter under national law, and whether remedial actions were completed. The record should not reveal a survivor’s identity, location, health details or facts that could make a person identifiable in a small community.

The reporting route must also be designed around the people who may need it. A channel inaccessible to patients, local-language speakers, people with disabilities or workers dependent on the programme is formally present but practically weak. Communication should make clear what happens after a report, who can see it, how urgent protection is arranged and what confidentiality cannot be promised. Survivor-centred support should not be conditional on a finding against the accused; investigation and support are related but separate responsibilities.

These are not additional clauses that the WHO framework already mandates. They are tests an editor, ministry, partner or funder can apply when assessing whether a voluntary framework has moved from paper to practice. The framework supplies a shared vocabulary. Local law and responsible authorities still govern the case.

The measure of a voluntary standard is not whether it sounds comprehensive at launch. It is whether a person at risk can identify a safe first door, whether that door leads to the right decision-maker, and whether the process closes with support and an authorized remedy. Until evidence shows those links working, a conference is evidence of convening, a policy is evidence of adoption, and neither is proof that people are protected.

Sources