The study investigated the relationship between statistical outliers and the inconsistent values of the CD4 count recommended for starting the Antiretroviral Therapy (ART) by HIV-positive patients. Low CD4 counts imply a low immune system. It could be due to AIDS existence or closeness to death. An effective treatment to curb HIV impact is ART, which is recommended for low CD4 counts. However, countries differ in the values used. Developed nations recommend start of ART when the CD4 is still high in order to curb its development. Poor countries use very low CD4 counts. Some countries keep changing the CD4 value for this purpose. The problem is that when CD4 counts are too low, HIV may already be too advanced, making it difficult to save the patient from progression to full AIDS. There should be CD4 values derived using scientific methods to assist in standardizing the CD4 count value for ART commencement. Using a retrospective single-site cohort study design, the study analyses the CD4 counts using robust statistical methods together with conventional statistical methods to study outliers and then derive a compromise CD4 count value that could serve as the standard cut-off starting point for using ART. The CD4 count confidence limits showed outliers to be above 300 cells/mm3. The lower bound of 0 cannot happen to any living person. The 300 upper bound is a value within the manageable outliers that has not reached critical HIV state. However, this value is near risky CD4 count values. The CD4 count of 300cells/mm3 indicates deteriorating HIV. If it is set as the starting point of taking ARVs, the patients involved can be saved from reaching painful states of lower CD4 counts.HIV patients’ immune systems at this level can still be boosted without them showing physical weakness from eye inspection.