Reference intervals are vital for clinical laboratory test interpretation determining patient care (Katayev et al, 2010). These reference intervals are defined as the set of values in which 95% of the normal healthy population falls and therefore, the remaining 5% is considered as abnormal. These intervals should be intended to match the population with results that are going to be compared to this range. To develop a normal reference interval, a minimum of 120 samples are collected for analysis. The initial step in determining a reference range is to classify the population to which the reference is going to apply to. A large group of individuals within this population are tested and data is collected to be averaged for a range to be established (Huma et Waheed, 2013). There are many limitations, however, as methods used prove to be expensive, difficult to perform, are often significantly inaccurate, and non-reproducible (Katayev et al, 2010). The methods used are also time consuming due to having to obtain informed consent from the reference subjects.
This is further enhanced in trying to establish reference ranges for different ages groups in that population. Furthermore, inaccuracy and non-reproducibility play a role because very few laboratories perform their own reference range studies; instead, previous studies are used, where both the method and population are shown to be different. Moreover, a level of uncertainty arises in defining a healthy reference subject as some of the selected subjects could have subclinical diseases, leading to further inaccuracies (Katayev et al, 2010).