Use AI for clinical decision support, patient education, telemedicine, and public health campaigns across Caribbean and Latin American health systems.
The Caribbean and Latin America face a distinctive set of healthcare challenges: a high burden of non-communicable diseases (NCDs) including diabetes, hypertension, and obesity; ongoing tropical disease pressures from dengue, Zika, and chikungunya; limited specialist availability in small island states; and persistent underfunding of public health systems. AI tools can help you address these challenges — whether you are a clinician, administrator, public health officer, or patient. This guide provides practical prompts for the most important use cases, alongside critical warnings about safe use.
AI is a clinical support tool — it does not replace physician judgment. Always have a qualified healthcare professional review AI-generated clinical recommendations before acting on them. Never use AI output as a substitute for examination, diagnosis, or treatment by a licensed clinician.
Use the following prompt to get structured differential diagnoses, recommended investigations, and referral criteria — framed around the conditions most prevalent in your country.
I am a [doctor/nurse practitioner] in [country].Patient: [age, sex, weight, height].Chief complaint: [describe].Vital signs: BP [X/X], HR [X], Temp [X], SpO2 [X]%.Relevant history: [chronic conditions, medications].Current medications: [list].Lab results (if available): [results].In the context of common conditions in [country/Caribbean region],provide: differential diagnoses (most to least likely),recommended investigations, initial management approach,red flag signs to watch for, and referral criteria.Note: this supports clinical decision-making, not replaces physician judgment.
Generate discharge summaries, referral letters, case note templates, and triage protocols aligned to your Ministry of Health standard.
Create [document type: discharge summary/referral letter/case notes template/triage protocol] for a [type of facility] in [country].Patient/clinical context: [describe].Required format: [Ministry of Health standard/regional standard].Language: [language].Ensure compliance with [country's health records regulations].Keep clinical language at [appropriate level for recipient].
Create patient-facing content in local languages using locally available foods and resources — covering NCDs, HIV, dengue, maternal health, and more.
Create patient education materials about [condition: Type 2 Diabetes/Hypertension/HIV/Dengue/Maternal health] for patients in [country].Patient literacy level: [basic/standard].Language: [English/Spanish/French/Haitian Creole/Papiamento].Format: [single page fact sheet/WhatsApp message series/poster].Include: what the condition is (simple terms),local risk factors and statistics for [country],symptoms to watch for, when to seek care,diet advice using LOCALLY AVAILABLE foods in [country],lifestyle changes practical for [country] context,available medications and where to get them in [country],and local support resources.
Always have patient education materials reviewed by a qualified clinician before distributing them. Ensure materials are appropriate for the literacy level and language of your target population — this includes Haitian Creole, Papiamento, and indigenous languages where relevant.
Design a telemedicine program that connects patients in your territory with off-island specialists — including technology requirements for low-bandwidth environments and a data privacy compliance framework.
Design a telemedicine program for [country/territory]to connect patients with specialists in [specialty].Current situation: [patients travel to Trinidad/Barbados/US for care].Technology available: [smartphone penetration X%, internet quality].Partner institutions: [list potential specialist partners].Target conditions: [most common referral diagnoses].Create: patient eligibility criteria, consultation workflow,technology requirements and alternatives for low-bandwidth,data privacy compliance framework, billing model,and expected impact on medical travel costs.
Generate full campaign concepts — including radio scripts, social media posts, and community health worker talking points — adapted for specific languages and cultural contexts.
Create a public health campaign for [country/region]targeting [disease/issue: NCDs/HIV/maternal mortality/vaccine hesitancy].Target population: [demographics].Key message: [message].Channels available: [TV/radio/social media/community health workers].Language: [language, include Creole/Patois variants if needed].Cultural considerations: [specific to country].Create: campaign concept, key messages for each channel,radio script (60 seconds), social media content (5 posts),community health worker talking points,and success metrics.
Always present AI as a tool that supports clinical decision-making. In patient-facing materials, be transparent that content was AI-assisted and has been reviewed by a healthcare professional.
Test for population diversity
Ensure any AI diagnostic or screening tools have been validated on diverse populations, including Afro-Caribbean and indigenous communities. Many AI models are trained predominantly on data from North American or European populations and may underperform for regional demographics.
Comply with national data protection laws
Patient data is subject to national laws including the Jamaica Data Protection Act, Brazil’s LGPD, and GDPR for French territories. Do not input identifiable patient data into general-purpose AI tools without confirming your compliance obligations.
Address algorithmic bias
Diagnostic AI trained on non-representative data may produce biased results for Caribbean and Latin American patients. Scrutinize outputs and flag unexpected patterns for clinical review.
Design for low-literacy populations
A significant portion of the regional population has limited health literacy. AI-generated patient materials should be reviewed for reading level and adapted for the local context — including the use of Creole, Patois, Papiamento, or indigenous language variants.
