Study design and sampling
The study was conducted in two laboratories, one in Belgium and one in Italy.
In Belgium, GIN-positive and negative faecal samples were collected from Belgian Blue cattle stabled at the experimental farm of the Faculty of Veterinary Medicine (Ghent University). Positive samples were collected from calves (6 months-old) experimentally infected with 50,000 third-stage larvae (L3) of Ostertagia ostertagi (n = 2 calves) or Cooperia oncophora (n = 2 calves), whilst negative samples were collected from uninfected adult (> 24 months-old) housed cattle (n = 2 calves).
In Italy, GIN positive and negative faecal samples were collected from Podolian adult cattle (> 24 months-old) in a commercial farm located in the Salerno Province, Campania region. Positive samples were collected from cattle at pasture, naturally infected by different species of GINs, whilst negative samples were collected from stabled cattle. Each sample was analysed in five replicates by the FLOTAC basic technique  with an analytical sensitivity of 1 egg per gram (EPG) of faeces to determine the presence/absence of GIN eggs.
Both in Belgium and in Italy, the positive cattle were used as donors for the extraction of GIN eggs from faeces, using a mass recovery method, i.e. a method that employs 4 sieves of different mesh size (1 mm, 250 μm, 212 μm and 38 μm) in order to separate the eggs from the faeces, as described in Bosco et al. . Eggs were recovered by washing the 38 μm sieve with tap water, and centrifuging the eluate for 3 min at 4000× g. To concentrate the GIN eggs, the supernatant was removed by a water pump and the pellet was resuspended in 5 ml of a 40% sucrose solution.
After centrifugation for 3 min at 4000× g, the supernatant was transferred to a new tube, diluted with an equal volume of tap water and centrifuged again for 3 min at 4000× g. The supernatant was removed to reduce the final volume of the egg preparation to 5 ml. Then, 10 aliquots of 0.1 ml each were taken, after a thorough homogenization of egg preparation into two tubes for 10 times (avoiding foam formation) for each aliquot to provide precise counting of eggs . Finally, the number of eggs was counted at 100× magnification.
The egg suspensions were added to five confirmed negative faecal samples of 200 g each to obtain five samples with different EPG levels: 10, 50, 100, 200 and 500 EPG. Each sample was analysed, using saturated sodium chloride solution (specific gravity = 1.200), by two FEC techniques: Mini-FLOTAC  and a modified McMaster  technique at two reading levels, i.e. grids and chambers. In total, 12 replicates were used for each method and for each EPG level.
From each homogenised faecal sample, for each EPG level, 60 g were weighed for the Mini-FLOTAC technique (5 g for each replicate; dilution ratio = 1:10; reading volume = 2 ml; analytical sensitivity = 5 EPG) and 36 g for the McMaster technique (3 g for each replicate), reading the two grids (dilution ratio = 1:15; reading volume = 0.30 ml; analytical sensitivity = 50 EPG) and the two chambers (reading volume = 1 ml; analytical sensitivity = 15 EPG). All samples were prepared, analysed and read at 100× magnification by the same expert operator in Belgium and Italy.
In Italy, a pooled faecal culture was performed in order to identify the nematodes to the genus level, following the protocol described in MAFF . Developed third-stage larvae (L3) were identified using the morphological keys proposed by van Wyk & Mayhew . Identification and percentages of nematodes by genus were conducted on 100 L3; if a sample had 100 or less L3 present, all larvae were identified.
EPG values for each technique and for each GIN infection level were calculated by multiplying the raw counts by the appropriate multiplication factor (e.g. 5 for Mini-FLOTAC, 50 for McMaster grids and 15 for McMaster chambers) and then, the mean of the replicate counts for each sample was calculated. The sensitivity of each method was estimated using the following formula: [(total number of positive samples observed/12, i.e. total number of replicate spiked samples performed for each method and for each level of contamination) × 100].
To evaluate the precision of each method, a coefficient of variation (CV) [(standard deviation/mean egg count) × 100] was calculated for each set of replicate counts for each method and level of EPGs. Furthermore, the accuracy of each method was determined by the percentage (%) of egg recovery calculated for each level of contamination, using the following formula: % egg recovery = [(observed FEC/ true FEC) × 100]. Boxplots (indicating median, percentiles and outliers) were used to show the precision and accuracy of each technique for each of the five levels of egg contamination. The non-parametric Kruskal-Wallis test with Dunn’s post-hoc test were used to compare all the observed values to the true FEC value for each technique and for each level of contamination.
Finally, a logistic regression model was developed in order to evaluate the predicted accuracy of each technique. The Mann-Whitney comparison test was used to compare the GIN egg recovery rates by Mini-FLOTAC and McMaster (reproducibility) in the two different laboratories (Belgium and Italy) using different samples from cattle experimentally (Belgium) or naturally infected by different GIN species (Italy), of different ages (calves in Belgium vs adult cattle in Italy) and breed (Belgian Blue vs Podolian).
All statistical analyses were performed in GraphPad Prism v.8 (Graph Pad Software, San Diego, CA, USA). Significance testing was set at P < 0.05.