Francophone and Lusophone Africa: what changes when English is not the language
Francophone Africa medical travel, and the Portuguese speaking flow beside it, fails for Indian hospitals at the artefacts rather than at the marketing. The enquiry, the clinical record, the estimate, the consent form and the discharge summary all assume English. This piece works through each one, who should own it, why interpreting must be rostered rather than borrowed, and why capability has to come before content.
There is a moment, familiar to anyone who has run an international desk in India, when a message arrives in French and the counsellor pastes it into a translation tool, pastes the output back in English, and presses send. The family reads a reply that is grammatically fine and completely impersonal, and half of them never write again. Nobody logs that as a lost case. It is recorded as an enquiry that did not respond.
Francophone africa medical travel, and the smaller Portuguese speaking flow alongside it, is the part of the market where Indian hospitals are least present and least equipped. Not because the demand is hard to reach, but because every artefact in the journey assumes English: the enquiry form, the estimate, the consent document, the ward signage, the discharge summary, the follow up message. Change the language of one of them and nothing improves. Change all of them and a market opens.
So this piece is organised around the artefacts rather than around the countries. Each one has a different failure mode, a different cost to fix, and a different person inside your hospital who has to own it.
What francophone africa medical travel changes in the desk
The countries involved stretch from Senegal, Mali, Guinea and Côte d’Ivoire across to Cameroon, Gabon, Congo and the Democratic Republic of the Congo, with Portuguese used in Angola, Mozambique and elsewhere. They differ from each other in almost every way that matters commercially. What they share, from the point of view of an Indian hospital, is that the working language of medicine, administration and family conversation is not English.
That single fact changes four things. It changes who can staff the desk, because a counsellor who speaks the language is not a nice addition but the minimum requirement. It changes the speed of every exchange, because clarification loops are longer. It changes the risk profile, because consent and clinical instructions carry real consequences when understanding is partial. And it changes where demand forms, because families search, watch and ask in their own language, on platforms and in phrasing that an English language content plan never touches.
Most groups respond by hiring one French speaker and calling the market covered. That person becomes a single point of failure, works impossible hours, translates everything for everyone, and leaves within a year. The alternative is to treat language as a property of the process rather than of an individual, which is the argument made at length in the piece on building a desk that works in more than one language.
The enquiry arrives as a voice note
Start with how contact is actually made. In these markets a large share of enquiries arrive as recorded voice messages rather than typed text, for a practical reason: typing a medical description on a phone keyboard is slow and error prone, and speaking is not. Families also send photographs of documents rather than files.
Most Indian desks are built to handle typed enquiries. A voice note in French sits in a queue until someone who understands it is free, which may be hours. By then the family has sent the same message to two other hospitals. The fix is not clever software. It is a rule that voice messages are triaged within the same response window as text, with an acknowledgement in the family’s language sent immediately even if the substantive reply takes longer.
Write that acknowledgement once, properly, with a native speaker, and reuse it. The same applies to the ten or twelve messages your desk sends in every case: what we need from you, we have received your reports, the doctor has reviewed them, here is your estimate, here is how to pay, here is what to bring, here is your arrival plan. Prepared properly once, they cover most of the traffic and they free your scarce language capacity for the conversations that genuinely need a person.
Reports in French, and the translation you must not improvise
The clinical file is the second artefact, and it is the one with the highest stakes. Reports, discharge notes and prescriptions written in French or Portuguese use their own conventions, and medicines are frequently known by names that differ from those used here.
Two rules keep this safe. First, translation of clinical documents is a defined task with a named responsible person, not something a counsellor does between calls with a phone app. Second, the treating consultant should always be told which parts of the record were translated and by whom, so that they can ask for the original where something is unclear. That is a documentation discipline, not a clinical one, and it belongs to the international desk rather than to the doctor.
The same applies in reverse, and this is the part most hospitals ignore completely. Everything you send back with the patient will be read by a doctor at home who works in French or Portuguese. A discharge summary in dense English, full of abbreviations, is close to useless to them. Having the key elements of the summary available in the home language is one of the highest value things an international desk can build, and it is what turns a single case into a referring relationship.
The estimate and the consent form
These two documents share a problem: the patient signs them, and signing is not the same as understanding.
The estimate should exist in the family’s language, not as a courtesy but because the questions families ask about money are the questions that decide the case. What is included. What is not. What happens if the stay is longer. Who pays for the attendant. When each portion is due. A family reading that in a second language will understand most of it, agree, and then be surprised later, which is the most expensive outcome available to you.
Consent is different in kind. A consent conversation conducted through an unqualified interpreter, or a form signed in a language the patient reads poorly, is a governance problem before it is a service problem. It should be conducted with the arranged interpreter, documented as such in the record, and the record should name who interpreted. I would treat that as non negotiable in any hospital taking international work, and I would expect the medical director rather than the growth team to own the standard.
Interpreting is a capability, not a person
The common model, where a staff member who happens to speak French is called from another department whenever needed, fails in three predictable ways. It is unavailable at night and on leave days. It has no record, so nobody can later establish what was explained. And it quietly burns out the person doing it, who is also expected to do their actual job.
Treat interpreting as a rostered service with defined cover hours, a defined escalation path for clinical conversations, and a log. Whether that capacity sits in house, on call, or as scheduled availability arranged in advance depends on your volume, and the choice is worth making deliberately rather than by default. The full argument for this sits in treating interpreters as a service rather than a favour.
One more point that gets missed. Interpreting need is heaviest in the parts of the stay that nobody plans for: the ward at night, the physiotherapy session, the conversation about a delayed discharge, the moment a family is told a plan has changed. Rostering only for consultations covers the easiest hours and leaves the hardest ones uncovered.
Portuguese speaking markets are not a smaller version
It is tempting to fold Angola, Mozambique and the other Portuguese speaking countries into a single non English bucket. They behave differently. Routes differ, with connections often running through Europe or the Gulf rather than through the same hubs that serve West Africa. The referral patterns differ. Corporate and institutional funding appears in some of these markets in ways it does not in others.
Language itself is not interchangeable either. Portuguese material translated from French, or worse translated from English by someone who mainly speaks Spanish, reads badly to a native speaker and signals exactly the carelessness you are trying to avoid. If you are not going to do Portuguese properly, it is more honest to leave it alone and concentrate on French speaking markets until you have the capacity.
The decision of which markets to serve properly, rather than which to list on a website, is a genuine strategic choice and deserves the same discipline as any other product decision inside a hospital group. Medical value travel rewards depth in a few corridors far more than presence in many.
Routes, hubs and the working day
Travel from West and Central Africa to India routes through a hub, commonly in the Gulf, sometimes through North Africa, Europe or East Africa. Journeys are long, connections can involve an overnight wait, and arrivals into Indian cities frequently land at night. From Southern African Portuguese speaking countries, routing often goes through a different set of hubs again, with its own pattern of departure days.
The operational consequences are the ones you would expect and still probably have not designed for: airport reception running at the hours flights actually arrive, a phone number that works for someone with no local connectivity, and discharge dates checked against the days services on that route actually operate. The time difference from West Africa also means your desk’s morning is the family’s early morning, which is fine for scheduling calls if you plan it and irritating for everyone if you do not.
On entry requirements, the only defensible position is that rules change and must be confirmed for each case with the relevant authority and with the hospital’s own international team. Do not publish lists or timelines. They go out of date silently and the family who plans around them experiences a failure they will attribute to you.
Where the demand actually forms online
Families in these markets research in their own language, and the phrases they use are not translations of English phrases. They watch video, they ask in group chats, and increasingly they ask assistants that answer conversationally, which means your information has to be findable and consistent rather than merely well designed.
A country page in French that is a machine translation of your English page will rank poorly and convert worse. A page written by someone who speaks the language, answering the questions this specific market asks, with a named contact and a realistic description of the process, does both jobs. That is the case argued in country landing pages that rank and convert, and the language dimension makes it sharper, not softer. It is also worth auditing what assistants currently say about your hospital in these languages, which the AI search visibility audit gives you a structure for.
Underneath all of it sits the same principle that governs every part of this work: the family cannot verify you, so consistency is the product. Same facts, same format, same named owner, in their language, every time. That is why the international funnel is better managed as a digital product with an owner than as a marketing campaign with a budget.
Where to begin
Pick one language and do it completely before adding a second. If that is French, the sequence is straightforward. Write the standard message set properly with a native speaker: acknowledgement, document request, estimate covering note, payment instructions, pre arrival information, arrival plan. That is a week of work and it covers most of your message volume.
Then define who translates clinical documents, and write it into the case process so it is never improvised. Then put interpreting on a roster with named cover for nights and weekends, and make the consent standard explicit with the medical director’s signature on it.
Then, and only then, build the country pages. Doing the content first is the common mistake: you generate enquiries you cannot answer well, and a family that gets a poor reply in their own language trusts you less than one who never wrote at all. Capability first, demand second. That order is unpopular with everyone who has a campaign to launch, and it is the only sequence I have seen work.
Questions people ask
It is the flow of patients from French speaking African countries to Indian hospitals, with a smaller Portuguese speaking flow alongside it. Operationally it is defined less by geography than by language: every artefact in the journey, from the first reply to the discharge summary, has to exist in a language your desk probably does not work in, and partial coverage produces worse outcomes than none.
Because that person becomes a single point of failure. They cover only the hours they work, translate for everyone, hold all the market knowledge, and usually leave within a year. Language has to be a property of the process: prepared standard messages, a defined clinical translation route, rostered interpreting and written material in the language. Individuals then handle the conversations that genuinely need judgement.
Choose one language and cover it completely before adding another. Start with the standard message set written by a native speaker, because those messages carry most of the traffic. Then define who translates clinical documents and who interprets for consent. Building country pages before that capability exists generates enquiries you answer badly, which does more damage than not appearing in the market at all.
Consent. A form signed in a language the patient reads poorly, or a conversation conducted through an untrained interpreter, is a governance failure before it is a service failure. The standard should be owned by the medical director, require the arranged interpreter for consent discussions, and record in the case file who interpreted. That is a low cost control and the single most important one in this segment.
Often as recorded voice messages and photographs of documents rather than typed text and files, because speaking is faster than typing a medical description on a phone. Desks built for typed enquiries leave those messages in a queue until a language speaker is free, by which time the family has written elsewhere. Acknowledge immediately in the family’s language even when the substantive reply takes longer.
That the investment is capability rather than campaign spend: prepared material, translation responsibility, rostered interpreting cover. Those costs are predictable and largely fixed. The mistake is funding demand generation first, which produces enquiries the desk cannot serve and a cost per arrival that looks terrible for reasons that have nothing to do with the media buy.
By a named responsible person as a defined task, never by a counsellor using a phone app between calls. The treating consultant should always be told which parts of the record were translated and by whom, so they can ask for the original where something reads oddly. Conventions and medicine naming differ, and the point of the discipline is to make uncertainty visible rather than to hide it.
Because it is read by a doctor at home who works in French or Portuguese and who will manage the patient after they fly back. A dense English summary full of abbreviations is close to unusable for them. Providing the key elements in the home language is what turns one case into a referring relationship, and very few Indian hospitals currently do it.
No. Routes, referral patterns and funding structures differ, and the languages are not interchangeable. Portuguese material translated at second hand reads badly to a native speaker and signals carelessness. If you cannot do it properly, it is more honest to leave it and concentrate on French speaking markets until capacity exists. Depth in a few corridors beats presence in many.
A case record that stores voice messages and images against the case rather than in someone’s personal phone, a template library in each language, and the ability to record which language a patient uses and who interpreted. It should also record consent for who may receive clinical information, since these cases often involve relatives in other countries reading the file.
As a rostered service with defined hours, an escalation route for clinical conversations and a log of who interpreted what. Cover the hard hours, not just consultations: the ward at night, a delayed discharge, a change of plan. Whether capacity sits in house or is arranged on call depends on volume, but the choice should be deliberate rather than an accident of who happens to be available.
The standard message set is about a week of work with a native speaker. Translation responsibility and the consent standard can be agreed within a month. Rostered interpreting depends on volume and hiring. Demand response follows content, which follows capability, so expect two or three quarters before arrivals move. Doing it in that order is slower to start and far cheaper overall.
Which languages the hospital actually covers, what the cover hours are, how many cases involved an arranged interpreter for consent, and enquiries against arrivals for each language market separately. Presenting a single international line hides the fact that a market you have listed but cannot serve is dragging down the performance of one you serve well.
Only the statement that rules change and must be confirmed for each case with the relevant authority and with the hospital’s international desk. Do not publish categories, documents, timelines or fees in any language. Translated rules go stale exactly as fast as English ones, and a family planning around an outdated page will experience the failure as your failure.
