EUVIMEDEUVIMED entdecken
ValidierungsstudieOpen Access · CC BY 4.0

Forecasting the Dialysis Burden in Japan: Validation‐Based Projections of Prevalence and Incidence Through 2050

Hatice Şahin iD, Berrak Itır Aylı iD, Nüket Paksoy iD

Therapeutic Apheresis and Dialysis · 2026-12 · 30 (6) · 925-934

DOI: 10.1002/1744-9987.70183 ↗

Kurz gefasst

Validierungsstudie in der Zeitschrift Therapeutic Apheresis and Dialysis, online erschienen am 06.07.2026. Verfasst von Hatice Şahin und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Nephrology, Etlik City Hospital, Ankara, Türkiye. Hauptthemen laut MEDLINE: Renal Dialysis, COVID-19, Kidney Failure, Chronic. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Japan has one of the highest dialysis prevalence rates worldwide and a shrinking, aging population. Whether dialysis burden has entered a sustained post‐peak phase or whether recent declines partly reflect pandemic‐related disruptions remains uncertain.

Methods

We analyzed nationwide data from the Japanese Society for Dialysis Therapy Renal Data Registry (1983–2022) to project dialysis prevalence and incidence through 2050. Seven forecasting approaches were implemented, including classical time‐series models, demographic regression models, machine‐learning algorithms, and a performance‐weighted ensemble. Models were trained on 1983–2017 data and validated against a holdout period (2018–2022) encompassing the historical peak in dialysis prevalence in order to forecast accuracy using RMSE, MAE, and MAPE.

Results

Dialysis prevalence increased from 52 505 patients in 1983 to a peak of 349 700 in 2021, followed by a decline in 2022, though the contribution of COVID‐19‐related excess mortality to this inflection cannot be excluded. During validation, the ensemble model demonstrated the highest predictive accuracy for prevalence and incidence (MAPE 0.60% and 0.48% respectively). Individual models produced divergent long‐term trajectories, ranging from substantial decline to continued growth. The ensemble model projected a gradual decrease in dialysis prevalence to approximately 322 000 patients by 2050 (−7.2% vs. 2022) and a 24.4% reduction in annual dialysis incidence.

Conclusions

Dialysis prevalence in Japan shows signs of a potential shift toward a downward trend with future burden increasingly driven by demographic change rather than rising incidence. Ensemble forecasting grounded in empirical validation provides balanced projections and highlights substantial uncertainty across modeling paradigms. Japan's experience offers important insights for other aging societies planning future kidney care services.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Renal DialysisCOVID-19Kidney Failure, ChronicJapanHumansIncidenceForecastingPrevalenceRegistriesFemaleMaleAgedMiddle AgedPrediction Algorithms

Stichwörter der Autor:innen

Japandialysisdisease projectionend stage renal diseaseensemble modelingforecastingtime‐series analysis

Autor:innen und Institutionen

  • Department of Nephrology, Etlik City Hospital, Ankara, Türkiye.
  • University of Westminster, London, UK.
  • Department of Public Health, Hacettepe University Faculty of Medicine, Ankara, Türkiye.

ORCID-Profile: Hatice Şahin, Berrak Itır Aylı, Nüket Paksoy

Zitieren

APA

Şahin, H., Aylı, B. I., & Paksoy, N. (2026). Forecasting the Dialysis Burden in Japan: Validation‐Based Projections of Prevalence and Incidence Through 2050. Therapeutic Apheresis and Dialysis, 30(6), 925-934. https://doi.org/10.1002/1744-9987.70183

Vancouver

Şahin H, Aylı BI, Paksoy N. Forecasting the Dialysis Burden in Japan: Validation‐Based Projections of Prevalence and Incidence Through 2050. Therapeutic Apheresis and Dialysis. 2026;30(6):925-934. doi:10.1002/1744-9987.70183

BibTeX
@article{ahin2026,
  author = {Şahin, Hatice and Aylı, Berrak Itır and Paksoy, Nüket},
  title = {Forecasting the Dialysis Burden in Japan: Validation‐Based Projections of Prevalence and Incidence Through 2050},
  journal = {Therapeutic Apheresis and Dialysis},
  year = {2026},
  volume = {30},
  number = {6},
  pages = {925--934},
  doi = {10.1002/1744-9987.70183}
}
Kostenlos starten →