c) Decisions criticising the three-point test 

In T 910/03 the board criticised the second condition of the essentiality test (omitted feature not, as such, indispensable for the function of the invention in the light of the technical problem it serves to solve). The board criticised this approach in the light of G 2/98 (OJ 2001, 413) for being tantamount to making a distinction between technical features which were related to the function and the effect of the invention and technical features which were not. In the case at issue, the board came to the conclusion that nowhere in the application documents as filed as a whole was it stated, suggested, or hinted at, that the presence of the deleted element was optional, or that the element, or elements, could be omitted.

In case T 2311/10 which concerned an intermediate generalisation, the board expressed the view that the three-point or essentiality test was unhelpful or even misleading (see also T 1840/11, T 2095/11, T 2095/12). In particular, the board highlighted, with regard to the second criterion, that the test of T 331/87 necessarily had to relate to the problem derivable from the application as it concerned the disclosure of the application. This was however not always observed.

In T 1852/13 the board concluded that the three-point essentiality test developed in T 331/87 should no longer be used (similarly T 830/16): in view of that earlier decision's wording ("may not"), logic alone dictated that it could not be congruent with the "gold standard" (G 2/10). Indeed, even the board that had developed the test had acknowledged that it might be met, yet Art. 123(2) EPC still infringed. Although the essentiality test could be a useful indicator in certain cases, the "gold standard" was the only test that counted (see T 755/12). Moreover the Enlarged Board, in G 2/98, ruling on when a priority claim was valid, had rejected the essentiality-based approach taken in T 73/88 (Snackfood), laying down instead a condition analogous to the "gold standard", and its concern that evaluations of essentiality might be arbitrary applied to amendments too. The board therefore agreed with T 910/03 that the conclusion to be drawn from G 2/98 was that the essentiality test should no longer be applied. Another advantage of the "gold standard" over the essentiality test was that it was a single test for all types of amendment.