For human IL-6 ELISA kits used to measure serum samples, are there large differences in recovery rates among different brands?

I have used both imported and domestic products. In my personal experience, for serum IL-6 at low levels, the differences in readings between brands are more obvious than at high levels. Spike recovery commonly fluctuates between 80% and 120%, but it is more scattered across different individual serum samples. Before switching brands, I suggest you take 5-8 pooled serum samples and test two or three small-package products simultaneously, and check the coverage of the highest and lowest points of the standard curve and the actual measured values of the samples. I recommend focusing on the one whose linear range can cover the actual IL-6 levels in your samples, especially since IL-6 in inflammation models may be very high or very low; one with a wide linear range and good separation at the low end is more suitable.

In our lab, when measuring human serum IL-6, we mainly look at two things: whether the standard curve stably covers the low values down to a few pg/mL, and whether serum samples require dilution. Some brands list a very wide linear range, but the CV at the low end is large; some require pre-dilution. If your 80 samples are not expected to have high IL-6, it is advisable to choose one with a low lower limit of quantification, for example, a batch that can reach 2-5 pg/mL at the low end. When performing spike recovery, use real pooled serum at three levels, and do not make the spike amount too high, otherwise it will mask matrix effects. Before switching brands, first ask technical support for the protocol and sample dilution recommendations; if that does not work, measure in parallel at the same time.

IL-6 spike recovery may vary between different brands, especially in serum matrix. A spike recovery of around 80% does not necessarily mean the kit is inadequate; endogenous interference in serum and matrix effects in low-value samples can both lower recovery. It is advisable not to focus only on single-point recovery. Instead, use pooled serum for spiking at 3 concentration levels, then check whether the serial dilution linearity is parallel; if the CV at the low-value point is high, first check whether the low-end points of the kit standard curve are stable. When selecting a kit, look for those whose instructions or COA clearly provide validation data in human serum matrix and intra-assay/inter-assay CV, which is more reliable than simply asking "what is the recovery rate."

Your sample size of 80 is not small, so I suggest not committing to a single brand all at once. You can first ask the distributor for small sizes or trial samples, and use the same batch of mixed serum to run three experiments: spike recovery, dilution linearity, and intra-assay CV. For inter-assay variation, check the QC in the instructions. Generally, an intra-assay CV below 10% and an inter-assay CV below 15% are common specifications, but don't rely only on the manufacturer's data—running one batch yourself will show you. If you later need to publish papers or perform pharmacodynamic statistics, a large inter-assay variation is more troublesome than a low recovery rate in a single batch. When purchasing, confirm whether they can guarantee that all 80 samples are covered by the same lot number, or at least avoid mixing lots midway.

Domestic batch-to-batch differences sometimes lie not in recovery but in standard curve drift. You mentioned elevated CV in replicate wells; first rule out operational and plate-washing issues, such as whether serum samples are hemolyzed or have high lipid levels. Samples need to be centrifuged at 10,000 g for 10 minutes before testing. During pipetting, avoid exceeding 20 minutes per plate, and for serum samples, it is best to use low-protein-binding tips. For selection, I recommend asking the manufacturer for spike recovery validation reports using human serum, rather than only looking at standard recovery; some brands show good standard recovery, but actual serum samples may read low. Both imported and leading domestic brands have channels supporting small-sample trials, so you can compare them simultaneously.