Setting research priorities for global pandemic preparedness: An international consensus and comparison with ChatGPT's output.
Song P., Adeloye D., Acharya Y., Bojude DA., Ali S., Alibudbud R., Bastien S., Becerra-Posada F., Berecki M., Bodomo A., Borrescio-Higa F., Buchtova M., Campbell H., Chan KY., Cheema S., Chopra M., Cipta DA., Castro LD., Ganasegeran K., Gebre T., Glasnović A., Graham CJ., Igwesi-Chidobe C., Iversen PO., Jadoon B., Lanza G., Macdonald C., Park C., Islam MM., Mshelia S., Nair H., Ng ZX., Htay MNN., Akinyemi KO., Parisi M., Patel S., Peprah P., Polasek O., Riha R., Rotarou ES., Sacks E., Sharov K., Stankov S., Supriyatiningsih W., Sutan R., Tomlinson M., Tsai AC., Tsimpida D., Vento S., Glasnović JV., Vokey LB., Wang L., Wazny K., Xu J., Yoshida S., Zhang Y., Cao J., Zhu Y., Sheikh A., Rudan I., International Society of Global Health (ISoGH) None.
BackgroundIn this priority-setting exercise, we sought to identify leading research priorities needed for strengthening future pandemic preparedness and response across countries.MethodsThe International Society of Global Health (ISoGH) used the Child Health and Nutrition Research Initiative (CHNRI) method to identify research priorities for future pandemic preparedness. Eighty experts in global health, translational and clinical research identified 163 research ideas, of which 42 experts then scored based on five pre-defined criteria. We calculated intermediate criterion-specific scores and overall research priority scores from the mean of individual scores for each research idea. We used a bootstrap (n = 1000) to compute the 95% confidence intervals.ResultsKey priorities included strengthening health systems, rapid vaccine and treatment production, improving international cooperation, and enhancing surveillance efficiency. Other priorities included learning from the coronavirus disease 2019 (COVID-19) pandemic, managing supply chains, identifying planning gaps, and promoting equitable interventions. We compared this CHNRI-based outcome with the 14 research priorities generated and ranked by ChatGPT, encountering both striking similarities and clear differences.ConclusionsPriority setting processes based on human crowdsourcing - such as the CHNRI method - and the output provided by ChatGPT are both valuable, as they complement and strengthen each other. The priorities identified by ChatGPT were more grounded in theory, while those identified by CHNRI were guided by recent practical experiences. Addressing these priorities, along with improvements in health planning, equitable community-based interventions, and the capacity of primary health care, is vital for better pandemic preparedness and response in many settings.