A public bill should not disappear inside a verification queue

A hospital waiting for the state to settle a legitimate bill should not have to depend on a political intervention to learn where its paperwork has gone. Nor should the taxpayer be asked to pay a claim merely because it has been outstanding for a long time. A credible public payment system has to protect both sides. It must establish what is owed and explain what happens next, with enough information for a provider to challenge an error and for the public to follow the money.

The Social Health Authority’s notice of 25 October provides a useful starting point. It sets out the resumption of verification for pending NHIF claims of at least Sh10 million at specified facility levels. A separate schedule names providers and visitation dates. That is an administrative step worth recognising. It gives affected facilities something more concrete than a general assurance that the issue is receiving attention. But an inspection calendar is the beginning of an accountability chain, not its completion.

My argument is that the government should now publish the rest of that chain. What happens after documents are inspected? Who receives a written finding? What is the route for correcting a missing invoice or disputing a rejection? When does an accepted claim become an approved payment, and what determines its place in the queue? These are proposals for a better process. The current notice does not answer all of them, and it should not be made to carry promises it does not contain.

The distinction between a claim and an established obligation matters. Someone may demand money that is not owed, submit the same service twice or produce an incomplete record of a genuine service. Those possibilities require investigation. Equally, an incomplete administrative file does not prove that the underlying treatment never happened. A fair system must be able to reject a false claim, resolve a disputed one and complete a genuine one without pretending that all three are the same problem.

That is why broad language about clearing pending bills often creates more confusion than confidence. A single announced total can combine amounts received, amounts reviewed, amounts disputed and amounts approved. Unless those categories are separated, a reduction in the headline number can mean several different things. The public cannot tell whether a provider has been paid, an error corrected or a claim removed from the system. Each may be legitimate; each requires a different explanation.

The first reform should therefore be a clear status for every claim, visible securely to the provider. A claim should have a reference number that survives changes of office, software and personnel. Its record should show when documents arrived, when additional information was requested, when that information was supplied and when a decision was made. Such a record would not decide the merits by itself. It would make administrative delay identifiable and give a supervisor a specific problem to resolve.

The public version should be less detailed but equally disciplined. It could show the value and number of claims in each stage, the age of unresolved claims and the amount actually paid during a defined period. Personal medical information should remain private. An accountability system does not need to expose a patient’s diagnosis to explain whether an institution has processed an invoice. Protecting confidentiality and publishing useful financial information are compatible objectives if the information is designed carefully.

A second reform is to give a reason for every material decision. Rejection should not be a code that only an official can interpret. A provider should be told whether the problem concerns eligibility, duplication, evidence of treatment, the amount charged or another defined issue. Where correction is possible, the instructions should say what is missing. Where the authority considers a claim fraudulent, that allegation should be handled through an appropriate process rather than slipped into a general administrative label.

The same discipline should apply to claims that pass verification. Approval should come with a statement of what has been accepted, what remains disputed and what further financial authorisation is required. This would prevent a common misunderstanding: a technical finding that a service was delivered is not necessarily the same document as a payment instruction. Providers deserve clarity about those stages, while officials need to avoid giving assurances that the budget or the law does not permit them to fulfil.

Third, the payment order should be explained in advance. There may be defensible reasons to prioritise particular obligations, including the age of an accepted bill or the effect of delay on essential services. But a priority rule should be stated and applied consistently. It should not emerge only when a well-connected claimant reaches the front. If exceptions are necessary, they should be recorded with reasons that an auditor can examine and that do not reveal confidential patient details.

This is also an argument for honest budgeting. A government can acknowledge a valid obligation while lacking enough immediately available cash to settle every accepted claim at once. That is an uncomfortable position, but concealment makes it worse. The proper response is to state the funding gap, identify the proposed source of payment and provide a realistic timetable. An announcement that everybody will be paid soon offers no durable relief if the cash arrangements have not been made.

For a small provider, uncertainty itself can be damaging. Consider a hypothetical clinic deciding whether it can replace equipment or retain a supplier. An acknowledged payment due on a credible date gives it a basis for planning. A claim that remains unexplained for months does not. This is not evidence that any named facility in the SHA schedule faces that situation. It illustrates why the quality of the administrative answer matters independently of the eventual amount paid.

The taxpayer has a parallel interest. Money paid to settle genuine services should be distinguishable from money lost through duplicate claims, inflated charges or poor controls. It is possible to demand prompt payment and rigorous scrutiny at the same time. Indeed, a process that identifies errors early can serve both aims. Endless rechecking, without a recorded reason or a conclusion, is not automatically a sign of vigilance. Sometimes it is simply a failure to make a decision.

An appeal mechanism is essential to that decision-making. The officer who first rejects a claim should not be the only person able to reconsider it. A provider needs a defined route, a reasonable response period and an explanation of the evidence considered. The authority, in turn, should be able to close repetitive challenges after a proper determination. Fairness is not an invitation to keep every dispute open indefinitely; it is a method for reaching a defensible conclusion.

Independent oversight should examine the whole process rather than celebrate only the amount recovered or rejected. Useful questions include whether similar claims received similar treatment, whether requests for documents were proportionate and whether payments followed the published priorities. A high rejection figure does not, on its own, establish good administration. Nor does a high payment figure prove that controls were adequate. Both numbers need the evidence beneath them.

There is a broader lesson for Kenya’s debate about public obligations. Language developed for sovereign borrowing should not be carelessly applied to every disputed invoice. A hospital claim, a loan agreement and an alleged loss through corruption raise different factual and legal questions. Calling an amount unjust does not establish the remedy. Before demanding repudiation, recovery or settlement, we should identify the transaction, the responsible institution and the basis on which the money is said to be owed.

That precision does not weaken political scrutiny. It makes scrutiny harder to evade. A minister can dismiss a sweeping slogan as exaggerated. It is harder to dismiss a dated record showing that a complete claim has received no decision, or that an accepted payment has been repeatedly promised without a funded plan. The strongest public challenge is often the most specific one: here is the file, here is the unresolved step, and here is the explanation still required.

The announced visits should now produce that kind of traceable record. Providers should prepare the requested evidence, and officials should document what they find. The next public update should distinguish visits planned from visits completed, claims checked from claims accepted, and accepted claims from payments made. It should also identify the unresolved work. Those distinctions are not bureaucratic decoration. They are how a promise becomes something that can be tested.

The process should also identify who takes responsibility when a response deadline passes. An escalation route with a named office would turn delay into an issue that can be managed. Without that route, even a well-designed claims portal can leave the provider repeatedly checking a status that nobody is required to explain.

A functioning state must be able to say both no and yes with reasons. It should refuse an unsupported demand without using that refusal to cast suspicion on every provider. It should honour a valid obligation without presenting payment as political generosity. The inspection schedule creates an opportunity to make that standard visible. The government should use it to establish a payment process that people can follow, rather than another queue whose destination only insiders understand.