This study offered important insight into weight estimation for children with special needs aged 2–18 years whenever the standard weight measurement was impossible. Due to the physical constraints and cognitive disability among children with special needs, alternative weight measurements are required to predict their actual weight. In this study, the Cattermole and Mercy weight estimation methods were comparable, and surprisingly, both methods were predictors of the actual weight. The Mercy weight estimate method (partial weight of HL + partial weight of MUAC) was more accurate in predicting the real weight than the Cattermole weight estimation method. Findings from this study were in agreement with a previous study conducted by Talib et al. [15] in the United States of America, which demonstrated Mercy as the best option to estimate weight in Down syndrome children aged 18 years and below compared with other methods, including the Cattermole, Broselow tape and APLS methods.

In addition, the agreement within 20% of the actual weight between the current study and the previous study [15] was comparable. The Mercy method in this study demonstrated the closest proportion of children’s estimated weight within 20% of actual weight with the Mercy method in a previous study (85% vs. 88%). Similarly, the Cattermole method in this study proposed a near proportion of children’s estimated weight within 20% of actual weight with the Cattermole method in the previous study (48% vs. 40%). Nevertheless, the Mercy method in the current study was found to overestimate the actual weight, while the Mercy method in the previous study [15] was reported to underestimate the actual weight (RE 0.287 kg vs. − 1.4 kg).

The Mercy method performed better than the Cattermole method in this study because the two-dimensional systems were always far superior in accuracy to the one-dimensional system [19]. The Mercy method incorporates two-dimensional systems (HL and MUAC), which results in more accurate weight estimation than the Cattermole method, which only relies on MUAC [15, 11, 19]. To date, no study has reported that one-dimensional systems are more accurate than two-dimensional systems [19].

Furthermore, the finding of low accuracy of the Cattermole method in this study was in line with the studies conducted among Korean children by Choi et al. [21] and Suh et al. [20], which revealed that the Cattermole method was the worst accurate method and was only highly precise in children aged 6 to 14 years. Moreover, the accuracy of the Cattermole weight estimation method to predict weight in children with special needs is unclear. The existing studies related to Cattermole estimation weight were conducted among normal children [18, 21, 20] and were reliable in school-age children [18].

Instead of special needs children, the Mercy weight estimation method was also accurate in predicting weight in normal children [22–25]. Additionally, the Mercy approach proved more accurate than other methods in a wider range of ages [11]. Hence, it has been observed that the Mercy weight estimation method was more applicable in both normal and special needs children with a wide age range.

Several limitations to this study need to be acknowledged. This study did not engage with normal children or other special needs children diagnosed beyond Down syndrome, autism/ADHD and cerebral palsy. In addition, the current study merely focused on children with special needs aged 2–18 years, and findings from this study could not be extrapolated to other age populations, such as adults and elderly individuals. Moreover, the accuracy of weight estimation methods decreases with increasing age [23]. Furthermore, this study only involved two weight estimation methods, namely the Cattermole and Mercy methods, that could predict the weight of the children in this study. The accuracy of other weight estimation methods could not be determined in the current study. In addition, there is no reference standard or benchmark for assessing the accuracy of the weight estimation methods. Therefore, further local research is required to explore the accuracy of other weight estimation methods.

In conclusion, the humeral length and mid-upper arm circumference were the most robust factors for predicting actual weight in children with special needs. The most apparent finding that emerged from this study was that the Mercy weight estimation method performed well in Malaysian special needs children, similar to that shown in the Western population. Hence, the Mercy weight estimation method is recommended to predict the actual weight in Malaysian special needs children aged 2 to 18 years.