Analysing and quantifying the effect of predictors of stroke direct costs in South Africa using quantile regression.

BACKGROUND

In South Africa (SA), stroke is the second highest cause of mortality and disability. Apart from being the main killer and cause of disability, stroke is an expensive disease to live with. Stroke costs include death and medical costs. Little is known about the stroke burden, particularly the stroke direct costs in SA. Identification of stroke costs predictors using appropriate statistical methods can help formulate appropriate health programs and policies aimed at reducing the stroke burden. Analysis of stroke costs have in the main, concentrated on mean regression, yet modelling with quantile regression (QR) is more appropriate than using mean regression. This is because the QR provides flexibility to analyse the stroke costs predictors corresponding to quantiles of interest. This study aims to estimate stroke direct costs, identify and quantify its predictors through QR analysis.

METHODS

Hospital-based data from 35,730 stroke cases were retrieved from selected private and public hospitals between January 2014 and December 2018. The model used, QR provides richer information about the predictors on costs. The prevalence-based approach was used to estimate the total stroke costs. Thus, stroke direct costs were estimated by taking into account the costs of all stroke patients admitted during the study period. QR analysis was used to assess the effect of each predictor on stroke costs distribution. Quantiles of stroke direct costs, with a focus on predictors, were modelled and the impact of predictors determined. QR plots of slopes were developed to visually examine the impact of the predictors across selected quantiles.

RESULTS

Of the 35,730 stroke cases, 22,183 were diabetic. The estimated total direct costs over five years were R7.3 trillion, with R2.6 billion from inpatient care. The economic stroke burden was found to increase in people with hypertension, heart problems, and diabetes. The age group 55-75 years had a bigger effect on costs distribution at the lower than upper quantiles.

CONCLUSIONS

The identified predictors can be used to raise awareness on modifiable predictors and promote campaigns for healthy dietary choices. Modelling costs predictors using multivariate QR models could be beneficial for addressing the stroke burden in SA.