ANTI RETROVIRAL LIFE-SAVING THERAPY

A CURVE FITTING APPROACH IN PAKISTAN

Abstract

Objectives: Effective implementation of mathematical and statistical models
maximizes the likelihood of understanding the trajectory, level and pattern of incurable diseases
and their therapies, so that precise demand strategies for costly lifesaving therapies for future,
resource allocation and timely intervention could be ensured, a significant concern for less
developed countries like Pakistan. Methods: The record of monthly number of patients on
lifesaving anti-Retroviral Therapy for more than 18 HIV treatment centers in Pakistan was
acquired for the period January 2011 to July 2013.A set of eleven curve fitting models namely
linear, quadratic, cubic, logarithmic, inverse ,exponential growth model, logistics-curve and
compound models was carried out for prediction. Results: After the execution of various curve
fitting models by taking the ANOVA approach along with coefficient of determination and the
forecast accuracy measures namely, mean percentage error (MAPE), mean absolute error
(MAE) and mean square error (MSE) in the selection of final efficient model. Cubic model was
selected for forecasting the monthly anti-retroviral therapy cases for all categories i.e. overall,
male female and children. On the basis of final selected model 3% increment is expected
number of paticuts on anti-retroviral therapy annually in HIV treatment center in all Pakistan.
Male patients are expected to increase 14% who will get lifesaving anti-retroviral therapies.
While the annually expected reduction of 17% and 42% in female and children cases taking
anti-retroviral therapy is expcted in future respectively. Conclusions: The overall number of
people on anti-retroviral therapy are expected to increase in Pakistan. Males are expected to
have greater risk as compared to female and children in Pakistan.