
STUDY GROUP
TSUI Edmund
JACKSON Nicholas
CHEN Judy
HOLLAND Gary
Purpose
The current approach for determining laser flare photometry (LFP) values is to obtain 7 LFP measurements and use the mean value after removal of the highest and lowest values. Our aim was to identify a simpler method to achieve the same standard using alternative approaches to establishing a final LFP value.
Methods
LFP was prospectively performed on 200 eyes (100 patients with uveitis) on the Kowa FM-700 laser flare photometer. Seven measurements were obtained on each eye. Eight statistical methods were used to calculate final LFP values, including comparisons of raw data, mean, removal of high/low values, median, and multiple outlier detection methods. To simulate fewer LFP measurements, analyses were repeated on the first 5 of 7 values. LFP values were separated into subgroups of <20, 20-50, and >50 photon units/msec for additional comparison. Intraclass correlation coefficient (ICC) was used to compare results.
Results
Mean age of patients was 49.5±19.1 years (72% female). Using all 7 LFP values, across all 8 methods and subgroups, the ICC for the final values did not vary meaningfully with a range of ICC from 0.74 to 1.00. When using the first 5 values to calculate final LFP values, the ICC across the statistical methods and subgroups did not vary meaningfully with a range of ICC from 0.81 to 1.00.
Conclusions
All statistical methods demonstrated final LFP values similar to those obtained from the current standard approach. This was also seen across all subgroups of LFP values. Our study suggests that obtaining fewer measurements and using simpler approaches for calculating final LFP values may be suitable for use in patient care and clinical research.