Accuracy and precision of intravitreally injected ranibizumab doses: an experimental study

The aim of this study was to analyse the accuracy, inaccuracy and precision of the doses of ranibizumab injections and to compare preadjusted dose measurements in syringes with differential dose adjustment during injection.

Ranibizumab was injected into artificial eyes with a reservoir by the Lucentis set under laboratory conditions. The applied dose was evaluated by weighing syringes and reservoirs before and after injection with an analytical balance (resolution: 0.1 mg). Measurements were corrected by the density factor of Lucentis. 120 measurements were performed in 6 series. Dose application was performed in 3 different modes: 1) after preadjustment of the desired dose in the syringes with normal (PA) and 2) with forced injection (PAF), and 3) by differential volume reduction from an overdose filled syringe. Quality parameters were defined as: accuracy = mean % of differences between measured and intended dose, inaccuracy = mean of absolute accuracy data, and precision = percentage of standard deviation in relation to the mean of the series.

The mean injection of PA mode (intended 0.5 mg) was 0.5059 (95 % confidence interval [CI]: 0.4903 - 0.5214). PA resulted in a significantly (p = 0.002) lower dose (mean: 0.4542, CI: 0.4339 - 0.4744) than DI. The mean dose in PAF mode was significantly (p < 0.001) higher (0.5335, CI: 0.5130 - 0.5540) compared to PA. Results in the 0.3 mg series in DI and PA injection were relatively close to the target (DI mean: 0.2900, CI: 0.2792 - 0.3008, PA mean: 0.2912, CI: 0.2773 - 0.3051, p > 0.05). PAF yielded a significantly higher dose (each p < 0.001) than PA and DI (mean: 0.3878, CI: 0.3748 - 0.4008). Accuracy exceeded the 5 % level in the 0.5 PA (- 9.1 %), 0.5 PAF (+ 6.7 %) and by + 29.3 % in the 0.3 PAF series. Inaccuracy extended the 5 % range in all series and was highest in the 0.5 PA (10.1 %), 0.5 PAF (9.7 %) and the 0.3 PAF series (29.3 %). Precision data presented as relatively homogeneous and ranged between 7.0 (0.5 DI) and 10.9 % (0.3 PA).

Under optimized laboratory conditions the use of Lucentis injection sets results in a relatively bad controlled application of ranibizumab doses. Preloading of syringes to the level of the intended injection should be abandoned and substituted by a differential injection mode. It seems recommendable to use smaller syringes than those provided.