Acute retinal necrosis-related publication activity and global epidemiological indicators: an exploratory time-series analysis, 1990–2023

Background Acute retinal necrosis is a rare but rapidly progressive necrotizing retinitis primarily caused by herpes viruses and is frequently associated with poor visual outcomes. Because of its rarity, evolving diagnostic criteria, and regional variability, its global epidemiology remains poorly characterized. This study evaluated long-term global trends in acute retinal necrosis using publication-based proxy indicators and examined temporal associations with herpes virus infection and diabetes mellitus. Methods Annual PubMed records from 1990 to 2024 were searched to identify acute retinal necrosis–related publications and case reports. Publication counts were normalized to all indexed publications to generate proxy indicators. Global data on herpes virus infection and diabetes mellitus were obtained from the Global Burden of Disease database for 1990–2023. Temporal associations between the publication-based ARN indicators and the epidemiological indicators were assessed over the common observation period of 1990–2023 using correlation, lagged correlation, Granger-based temporal predictive analysis, and vector autoregression–based impulse response analyses. A structural vector autoregression model incorporating a dummy variable for the widespread adoption of polymerase chain reaction testing after 2000 was constructed, and Bayesian vector autoregression was used to evaluate shock responses and generate forecasts. Results A total of 1,209 acute retinal necrosis–related publications and 690 case reports were identified, increasing by 48.8% and 53.8%, respectively, between 1990 and 2024. Publication-based proxy indicators were negatively correlated with herpes virus and diabetes mellitus prevalence ( r = −0.59 to −0.73, p < 0.001) and with diabetes mellitus incidence ( r = −0.60 to −0.74, p ≤ 0.002) but positively correlated with herpes virus incidence ( r = 0.52–0.75, p < 0.001). Herpes virus incidence significantly predicted changes in the publication-based proxy indicator ( p < 0.001). The temporal analyses identified exploratory associations between herpes virus incidence and publication-based proxy indicators, whereas the vector autoregression model did not satisfy the stability condition and its cumulative impulse response estimates were therefore not interpreted. Bayesian analysis projected a modest upward trend in proxy indicators with increasing predictive uncertainty. Conclusion Publication-based proxy indicators provide a practical approach for investigating global trends in acute retinal necrosis when direct epidemiological data are limited. The findings suggest that herpes virus epidemiology and metabolic conditions such as diabetes mellitus may influence future publication activity. Clinical trial registration Not applicable.