Artificial intelligence (AI) is being piloted at Lebanon’s Al Hamshari Hospital in a project that could help stretched medics provide better care for huge numbers of Palestinian refugees.
The hospital, near Lebanon’s largest Palestinian refugee camp, Ein el Hilweh, is using software from UK and Qatar-based healthtech startup Rhazes AI, which provides end-to-end agentic AI for clinicians. The natural language AI tool provides specialist-level help with clinical decisions, and helps them with paperwork and admin tasks.
Medics at the hospital, operated by the Palestine Red Crescent Society, are dealing with huge numbers of refugees, who do not have access to the Lebanon healthcare system.
Each month, about 4,000 refugees are being treated at the hospital, which has just 80 beds, 56 doctors and 31 nurses. Medics there often perform more than 400 surgeries a month during times of crisis, and it is the only medical facility with a working dialysis unit serving the southern Lebanon refugee camps.
About one million Palestinians have sought refuge in Lebanon, and because they are excluded from the national healthcare system, hospitals like Al Hamshari care for them.
“It is under a huge amount of pressure,” according to Zaid Al-Fagih, co-founder and CEO of Rhazes AI. “It caters for a population that has no other source of healthcare.”
The hospital is staffed primarily by Palestinian doctors who often see about 60 patients per day. Al-Fagih said for every hour a doctor spends with a patient, “they spend two hours doing paperwork”. “AI is a potential solution to this, and it is badly needed,” he told Computer Weekly. “But AI currently only exists in point solutions for specific modalities and so on.”
Testing opportunity
The Lebanon hospital is an opportunity to test the software in an environment outside of highly resourced western centres. “AI in healthcare is usually used in highly resourced environments such as the UK, US and UAE,” said Al-Fagih. “It is actually very rarely studied in low-resource environments, which is probably where the biggest marginal gain can be made.”
The agentic platform adapts to low-resource environments without requiring extensive hospital infrastructure.
The pilot at Al Hamshari Hospital, which is funded by Rhazes AI, is like a clinical trial, according to Al-Fagih. “Half the doctors are using our tool, and the other half won’t, and we will be comparing the outcomes,” he said.
Rhazes supports doctors end-to-end by transcribing consultations in real time, assisting with diagnostic reasoning and generating evidence-based management plans. The tool also automates documentation processes, helping doctors to create structured summaries, admission notes, billing codes and insights from the patient record. This reduces administrative burden and enables more efficient care.
“They are using it to write their notes, they are using it to get help on clinical decisions, and to help them access evidence-based guidelines,” he said. “This is something they never had access to because it is very low-budget and they would have to use open-source guidelines you can find on Google, where tailoring it to their specific environment is very difficult.”
Al-Fagih said resources are limited at the hospital, with one computer per floor and no electronic patient system, with note-taking and clinical decision-making support paper-based. A lot of tasks at the hospital are undertaken in a very “rudimentary way”, he said.
It is still too early to see any results of the pilot, but success will be measured by stats including those on the number of patients seen over a given period, patient outcomes such as death compared with admissions, better record keeping, and improved staff and patient satisfaction.
Rola Soboh, a Rhazes AI associate implementing the pilot, said: “I have supported multiple research and humanitarian projects focused on refugee health and well-being in Lebanon, and this project is deeply personal. This hospital isn’t just a building, it’s a lifeline.
“Doctors here don’t just treat patients, they carry entire communities,” she said. “So, when we talk about easing their load, it’s not just administrative, it’s emotional, physical, everything. To see cutting-edge technology actually serve people like this, in a place that’s so often forgotten, gives me a sense of real hope.”
Data has presented the biggest challenge, according to Al-Fagih. “It is actually very difficult to get proper data on how well we are doing in that environment where everything was done using paper previously,” he said. “Also, there are no computers, so we had to make it work using doctors’ own devices in a compliant way.”
This is the first trial in a low-resource environment. Trials have already been run in the UK and UAE, where it is now used.
