refineR: A Novel Algorithm for Reference Interval Estimation from Real-World Data

Abstract

Reference intervals are essential for the interpretation of laboratory test results in medicine. We propose a novel indirect approach to estimate reference intervals from real-world data as an alternative to direct methods, which require samples from healthy individuals. The presented refineR algorithm separates the non-pathological distribution from the pathological distribution of observed test results using an inverse approach and identifies the model that best explains the non-pathological distribution. To evaluate its performance, we simulated test results from six common laboratory analytes with a varying location and fraction of pathological test results. Estimated reference intervals were compared to the ground truth, an alternative indirect method (kosmic), and the direct method (N = 120 and N = 400 samples). Overall, refineR achieved the lowest mean percentage error of all methods (2.77%). Analyzing the amount of reference intervals within ± 1 total error deviation from the ground truth, refineR (82.5%) was inferior to the direct method with N = 400 samples (90.1%), but outperformed kosmic (70.8%) and the direct method with N = 120 (67.4%). Additionally, reference intervals estimated from pediatric data were comparable to published direct method studies. In conclusion, the refineR algorithm enables precise estimation of reference intervals from real-world data and represents a viable complement to the direct method.

Establishment of Reference Interval and Aging Model of Homocysteine Using Real-World Data

“…All data were recorded in 2016 Microsoft Excel sheet (Microsoft, Redmond, WA, United States), then analyzed using packages ( 10 , 11 ) and codes ( 13 ) implemented in R language (version 4.0.5), as well as SPSS 25.0 Software (IBM Inc., Armonk, NY, United States). Tukey method was used to test outliers in subgroups considering differences in subgroups.…”

“…The process is simple, low cost and feasible, which is a great supplement to the direct sampling method. Currently, the main indirect sampling methods include Hoffmann method ( 7 ), Bhattacharya method ( 8 ), Expectation Maximization (EM) algorithm ( 9 ), kosmic algorithm ( 10 ), refineR algorithm ( 11 ) and so on. These algorithms are based on different principles to separate the mixed distribution and obtain the data distribution parameters of healthy individuals, each of which has advantages and disadvantages.…”

Age‐ and sex‐specific pediatric reference intervals for neutrophil‐to‐lymphocyte ratio, lymphocyte‐to‐monocyte ratio, and platelet‐to‐lymphocyte ratio

“…The refineR algorithm has been shown to outperform other indirect algorithms for datasets with a high fraction and overlap of pathological test results, for example, the kosmic algorithm. 12…”

Establishment of Reference Intervals for Thyroid-Associated Hormones Using refineR Algorithm in Chinese Population at High-Altitude Areas

“…Thus, it can be used for the establishment of RIs when healthy individuals cannot be defined and obtained using direct sampling. The refineR algorithm has been previously proposed for the problem of "data mining" RIs (21). An open-source R code for the establishment of RIs using refineR algorithm has been developed by Ammer (21) and its effectiveness has been also proved.…”

“…The refineR algorithm has been previously proposed for the problem of "data mining" RIs (21). An open-source R code for the establishment of RIs using refineR algorithm has been developed by Ammer (21) and its effectiveness has been also proved. The core of the algorithm is still parameter combination and optimal search.…”