ClinicaLingo

Blog — Clinical Spanish

Stroke assessment in Spanish: why “time is brain” has a translation problem

The most consequential question in stroke care — last-known-well — fails in Spanish in a specific and predictable way. The phrases for FAST, NIHSS, and tPA consent are learnable in an afternoon. The failure modes are not obvious until you’ve watched one happen. Here are the three ways stroke assessment breaks down across the language barrier, and the exact conversation mechanics that close each gap.

The short version: “¿Cuándo empezó esto?” is not the same as “¿Cuándo fue la última vez que lo vio completamente bien?” These two questions get different answers and the difference can close the tPA window. Lead tPA consent with what happens without the medication before presenting the bleeding risk — a family member who hears “sangrado en el cerebro” first will refuse on that risk alone. Phrase reference: Stroke assessment in Spanish — FAST, NIHSS, tPA consent, thrombectomy.

The woman who woke up wrong

Elena Gutiérrez is 67. Her husband Rodrigo calls 911 at 8:14 AM. He tells the dispatcher: “se despertó, estaba confundida, no le entiendo lo que dice.” She woke up confused. She’s not making sense. The EMT arrives, logs the call at 8:22 AM, and documents last-known-well as 6:00 AM — the time Elena got up, per Rodrigo.

By the time the ED nurse gets the radio report, the math is favorable: 2 hours and 14 minutes since last-known-well. tPA window is 4.5 hours. There is time. The team activates a stroke alert.

What happens next is not unusual. The ED physician, with a phone interpreter, asks Rodrigo directly: “¿Cuándo notó que algo estaba mal?” Rodrigo answers: “Cuando se levantó a las seis y empezó a hablar raro.” When she got up at six and started talking strangely.

Except the physician also asks: “¿Y usted la vio antes de que se levantara?” Had he seen her before she got up? Rodrigo pauses. Yes. He was already awake at 5:30 AM. He saw Elena sitting on the edge of the bed, reaching toward the nightstand, touching things without picking anything up. He thought she was just tired. He did not say anything to her. He did not know she could not find her words. He went to make coffee.

The real last-known-well is 11:00 PM the night before — when they went to sleep and Elena was fine. Nine hours and fourteen minutes ago. The tPA window closed at 3:30 AM.

Rodrigo was not withholding information. He answered the question he was asked. The question was: when did you notice something was wrong? The answer was 6:00 AM, when the wrongness was obvious. The question that needed to be asked was different.

Three ways stroke assessment fails in Spanish

1. The last-known-well question is not the same as “when did symptoms start”

This is the most clinically costly language gap in stroke care. It happens not because anyone uses the wrong Spanish words, but because the two questions sound similar and get answered from different reference points.

“¿Cuándo empezó esto?” (When did this start?) and “¿A qué hora notó que algo estaba mal?” (What time did you notice something was wrong?) both get an answer about when things were bad enough to notice. For Rodrigo, that was 6:00 AM. For a family whose mother had subtle word-finding difficulty for twenty minutes before the arm dropped, it will be when the arm dropped. For a family who called 911 four hours into a wake-up stroke, it will be when they called.

The question that actually certifies the tPA window is different:

“¿Cuándo fue la última vez que lo/la vio completamente bien — hablando normal, moviéndose normal, como de costumbre?”

(When was the last time you saw him/her completely normal — speaking normally, moving normally, the way they usually are?)

This is not the same question as “when did it start.” It asks the witness to certify a specific moment of normality, not to describe when they recognized abnormality. Those two moments are often separated by minutes or by hours, and the difference determines whether a 67-year-old woman gets tPA or palliative care.

The wake-up stroke pattern requires an additional specific question:

“¿A qué hora se durmieron anoche? ¿Y en ese momento ella estaba bien — hablando normal, moviéndose normal?”

(What time did you go to sleep last night? And at that point was she okay — speaking normally, moving normally?)

In a wake-up stroke, last-known-well is the last moment of witnessed normalcy — which is bedtime if the patient went to bed neurologically intact. EMT teams, triage staff, and ED nurses who ask “when did symptoms start?” instead of “when was the last time she was normal?” will consistently document a last-known-well that is hours later than the true last-known-well.

The family shame pattern and how to break it

There is a second layer to the last-known-well problem that no Spanish phrase class covers: the family member who knows the real answer and gives a later one.

Rodrigo saw Elena at 5:30 AM on the edge of the bed, reaching and not finding. He did not intervene. He is standing in the ED now being asked when she was last normal. The accurate answer is 11:00 PM. The answer that doesn’t require him to explain why he made coffee instead of calling 911 is 6:00 AM.

This is not unique to Spanish-speaking patients or Latin American families. It is a universal pattern of protective self-presentation that clinicians encounter every day. What differs in a Spanish-language encounter is that the family member has fewer cues to tell them that the clinical team is not assigning blame — and more barriers to volunteering context that wasn’t explicitly asked for.

Two adjustments close this gap. First: ask every family member present separately before accepting an answer. The daughter may know something the husband does not, or may frame it differently. Second: add an explicit no-blame framing before the question, not after:

“Nadie hizo nada malo. Lo que necesitamos es la hora más precisa posible para saber qué tratamiento le podemos dar. ¿Cuándo fue la última vez — y me refiero a la última, no al primer momento en que algo le pareció raro — que usted la vio completamente normal?”

(Nobody did anything wrong. What we need is the most accurate time possible to know what treatment we can give her. When was the last time — and I mean the last time, not the first moment something seemed off — that you saw her completely normal?)

“Nadie hizo nada malo” does specific clinical work. It removes the guilt frame from the question before the witness has a chance to construct a protective answer around it.

2. tPA consent collapses when “bleeding in the brain” lands first

The standard tPA consent sequence presents the drug, then what it can do, then the risk, then the consent ask. This sequence works well when the person giving consent has a baseline model of benefit-risk tradeoff in medicine — the implicit understanding that all treatments have risks and the question is whether the benefit justifies them.

For a family member standing in an ED at 8:30 in the morning watching their 67-year-old mother not be able to speak, that baseline model is often not operative. What is operative is fear. A frightened person hearing “sangrado dentro del cerebro — en seis de cada cien personas” does not perform a benefit-risk calculation. They hear: this drug might kill her. The answer is no.

The adjustment is structural, not linguistic. Lead with what happens without the medication before presenting the risk of the medication:

  1. What is happening to the brain right now, without treatment. “En este momento hay un coágulo bloqueando el flujo de sangre a una parte del cerebro de su mamá. Cada minuto que ese coágulo está ahí, esa parte del cerebro está muriendo. Sin tratamiento, la probabilidad de que quede con la debilidad o sin poder hablar de forma permanente es muy alta.”
    (Right now there is a clot blocking blood flow to a part of your mother’s brain. Every minute that clot is there, that part of the brain is dying. Without treatment, the probability that she will be left with weakness or unable to speak permanently is very high.)
  2. What the medication can do. “Hay un medicamento que puede disolver ese coágulo. Si funciona — y funciona en la mayoría de los pacientes — su mamá puede recuperar el habla y el movimiento que perdió esta mañana.”
    (There is a medication that can dissolve that clot. If it works — and it works for most patients — your mother can recover the speech and movement she lost this morning.)
  3. The risk of the medication, compared explicitly to the alternative. “El riesgo del medicamento es sangrado dentro del cerebro — eso ocurre en aproximadamente seis de cada cien personas que lo reciben. Sin el medicamento, el riesgo de daño permanente es mucho más alto que ese seis por ciento. Le estamos pidiendo que compare: daño permanente casi seguro sin el medicamento, versus un riesgo de seis en cien con él.”
    (The risk of the medication is bleeding inside the brain — that happens in approximately six out of every one hundred people who receive it. Without the medication, the risk of permanent damage is much higher than that six percent. We are asking you to compare: almost certain permanent damage without the medication, versus a six-in-one-hundred risk with it.)
  4. The consent ask, with the time constraint named. “Necesitamos su respuesta ahora mismo — el medicamento solo funciona dentro de cierta ventana de tiempo, y esa ventana se está cerrando. ¿Nos da su permiso para dárselo?”
    (We need your answer right now — the medication only works within a certain time window, and that window is closing. Do you give us permission to give it to her?)

The reason this sequence works where the standard sequence sometimes fails: the family member hears the outcome of not acting before they hear the risk of acting. “Daño permanente casi seguro sin el medicamento” establishes the reference point. “Seis en cien” is evaluated against that reference point, not against a baseline of “she was fine at dinner last night.”

3. The family making the decision for an incapacitated patient

When Elena cannot consent and Rodrigo is the surrogate, a dynamic enters the room that does not appear in any Spanish-for-nurses curriculum: Rodrigo is deciding whether his wife lives or has permanent disability, in a language that is not his primary language, forty-five minutes after calling 911, standing next to a security officer, while someone holds a clipboard.

Before asking for consent, three things need to happen:

Confirm who is deciding

“¿Usted es el familiar más cercano? ¿Hay alguien más — un hijo adulto, otro familiar — que debería estar aquí para esta decisión? Pregunto porque lo que le estoy pidiendo es muy importante y quiero que usted tenga el apoyo que necesita.”

(Are you the closest family member? Is there anyone else — an adult child, another family member — who should be here for this decision? I ask because what I’m asking of you is very important and I want you to have the support you need.)

The question about whether others should be present is not a delay tactic. It is a recognition that the surrogate’s ability to decide well is affected by whether they feel alone. A family where three people say yes is a more reliable consent than one where a frightened husband says it alone because no one thought to ask if there was a daughter in the waiting room.

Name the role explicitly before asking for a decision

“Como su esposo, usted es la persona que va a tomar esta decisión por ella ahora mismo. Eso significa que voy a darle toda la información que tengo, y entonces voy a pedirle que me diga qué cree que ella hubiera querido. No qué quiere usted — qué hubiera querido ella. ¿La conoce bien? ¿Ha hablado alguna vez de estas cosas?”

(As her husband, you are the person who is going to make this decision for her right now. That means I am going to give you all the information I have, and then I am going to ask you to tell me what you think she would have wanted. Not what you want — what she would have wanted. Do you know her well? Have you ever talked about these things?)

The reframe from “what do you want” to “what would she have wanted” is standard substituted-judgment language and it is clinically significant. A husband deciding for his wife on the basis of his own fear is making a different decision than one deciding on the basis of what he knows she would choose. Many families, when asked this way, will say: she would have wanted every chance. That answer is not available if the question is never asked.

When the family says no

A refusal is almost always a fear response to incomplete information, not a considered rejection of the medication. The clinical response is to find out what the refusal is actually about.

“Escucho que le preocupa el riesgo de sangrado. Quiero asegurarme de que tenga toda la información antes de decidir. ¿Puedo preguntarle qué parte le preocupa más?”

(I hear that the bleeding risk concerns you. I want to make sure you have all the information before deciding. Can I ask what specifically worries you most?)

The answer to that question tells you what the refusal is actually about. Common patterns:

None of these responses override a considered refusal. They address specific failure modes in the consent conversation that produce uninformed refusals. The difference between “I refuse because I understand the tradeoff and reject it” and “I refuse because I heard ‘bleeding in the brain’ and panicked” is the entire job of informed consent.

What the SEO page covers and what this post is about

The companion page at Stroke assessment in Spanish is a phrase reference: FAST commands, Cincinnati Stroke Scale items, all eleven NIHSS items in Spanish, time-last-known-well questions, CT consent, tPA informed consent, thrombectomy explanation, and ICU admission. It is structured for bedside use — the order you will use the phrases, with the clinical notes you need to use them correctly.

This post is about the failure modes that occur before and after those phrases — the conversation-level mechanics that determine whether the phrases land or not. The distinction matters because a clinician who memorizes the tPA consent phrase but delivers it in the standard sequence (what the drug is → what it does → risk → consent ask) will still lose a frightened family that hears “sangrado en el cerebro” as a sentence, not as the smaller side of a comparison.

Both are necessary. The phrases are the vocabulary. The sequence and framing are the grammar.

One more thing: the interpreter arrives after the tPA decision

The stroke protocol is built around time windows that rarely align with interpreter availability. In most US EDs, a qualified phone interpreter takes three to eight minutes to connect. In a stroke alert, three minutes is the difference between acting and explaining to the family why you waited.

The post at The interpreter is on hold for eleven minutes covers this in full. The short version for stroke: there is a set of bridging actions — FAST, Cincinnati Stroke Scale, CT consent, time-last-known-well — that you can conduct in Spanish before the interpreter arrives without compromising informed consent for tPA. tPA consent itself requires a qualified interpreter whenever possible. The judgment call you will face is whether “whenever possible” means “not yet” or “this is the window.”

That judgment is clinical, not linguistic. What the phrases give you is the ability to be doing something productive during the wait instead of nothing.

FAQs — stroke assessment in Spanish

What is the correct way to ask last-known-well in Spanish?

“¿Cuándo fue la última vez que lo/la vio completamente bien — hablando normal, moviéndose normal, como de costumbre?” This is different from “¿Cuándo empezó esto?” — the first certifies normalcy; the second gets an answer about when things were bad enough to notice. For wake-up strokes: “¿A qué hora se durmieron anoche? ¿Y en ese momento estaba bien?” Last-known-well for a wake-up stroke is bedtime. Ask every family member separately before accepting an answer, and lead with “Nadie hizo nada malo” to pre-empt the shame response that produces a later time than the true last-known-well.

How do I explain tPA to a Spanish-speaking family when the patient cannot consent?

Lead with what happens without the medication before presenting the risk: “Sin este medicamento, la probabilidad de que quede con la debilidad permanente es muy alta.” Then the medication: “El medicamento puede disolver el coágulo y recuperar el habla y el movimiento.” Then the risk in comparison: “El riesgo del medicamento es sangrado dentro del cerebro — seis en cien personas. Sin el medicamento, el daño permanente es mucho más probable que ese seis por ciento.” A family that hears the bleeding risk before the alternative will often refuse on the bleeding risk alone.

How do I do FAST in Spanish at the bedside?

Face: “Sonríame. Muéstreme los dientes.” Watch for asymmetry. Arms: “Levante los dos brazos. Cierre los ojos y manténgalos arriba por diez segundos.” Speech: “Repita esta frase: el cielo está despejado hoy.” Time: “¿Cuándo fue la última vez que lo/la vio completamente normal?” The T is time-last-known-well, not time-to-call. Never leave the bedside without a clock time certified by the most reliable witness present. Full NIHSS questions and thrombectomy framing: Stroke assessment in Spanish.

What do I do when the family refuses tPA in Spanish?

Start by finding out what the refusal is actually about: “Escucho que le preocupa el riesgo. ¿Puedo preguntarle qué parte le preocupa más?” Common patterns: they heard “catheter to the brain” and applied it to the IV medication (clarify the route); they have community-circulated fear from someone else’s bad outcome (name the comparison explicitly); they want to wait for another family member (name the time window); they are paralyzed and can’t decide (reframe to substituted judgment: “¿Qué cree que ella hubiera querido?”). A refusal is almost always incomplete information, not a considered rejection of the treatment.

How do I explain thrombectomy in Spanish when the family has never heard of it?

Lead with the physical metaphor: “Los médicos meten un tubo muy fino por una arteria en la ingle y lo suben hasta el cerebro, donde sacan el coágulo con un dispositivo especial — como una aspiradora muy pequeña.” Distinguish from tPA: “El medicamento disuelve el coágulo. Este procedimiento lo saca físicamente.” Then stakes: “Cada minuto que el cerebro está sin circulación se pierden millones de células nerviosas. ¿Me da su permiso para proceder?” Full phrase reference: Stroke assessment in Spanish.

Related resources