Time-Series Forecasting of Hemodialysis Population in the State of Qatar by 2030.

BACKGROUND

There are few statistics on dialysis-dependent individuals with end-stage kidney disease (ESKD) in Qatar. Having access to this information can aid in better understanding the dialysis development model, aiding higher-level services in future planning. In order to give data for creating preventive efforts, we thus propose a time-series with a definitive endogenous model to predict ESKD patients requiring dialysis.

METHODS

In this study, we used four mathematical equations linear, exponential, logarithmic decimal, and polynomial regression, to make predictions using historical data from 2012 to 2021. These equations were evaluated based on time-series analysis, and their prediction performance was assessed using the mean absolute percentage error (MAPE), coefficient of determination (R), and mean absolute deviation (MAD). Because it remained largely steady for the population at risk of ESKD in this investigation, we did not consider the population growth factor to be changeable. (FIFA World Cup 2022 preparation workforce associated growth was in healthy and young workers that did not influence ESKD prevalence).

RESULT

The polynomial has a high R of 0.99 and is consequently the best match for the prevalence dialysis data, according to numerical findings. Thus, the MAPE is 2.28, and the MAD is 9.87%, revealing a small prediction error with good accuracy and variability. The polynomial algorithm is the simplest and best-calculated projection model, according to these results. The number of dialysis patients in Qatar is anticipated to increase to 1037 (95% CI, 974-1126) in 2022, 1245 (95% CI, 911-1518) in 2025, and 1611 (95% CI, 1378-1954) in 2030, with a 5.67% average yearly percentage change between 2022 and 2030.

CONCLUSION

Our research offers straightforward and precise mathematical models for predicting the number of patients in Qatar who will require dialysis in the future. We discovered that the polynomial technique outperformed other methods. Future planning for the need for dialysis services can benefit from this forecasting.