An average score of 9.06 is no breach of the rules: why Filter Service did not obtain the suspension of the hand-care contract award
Filter Service sought the extreme-urgency suspension of the award by AZ Sint-Jan Brugge-Oostende of a framework agreement for hand-care products to Dialex Biomedica, its core complaint being that the quality assessment did not follow the announced scoring method — but the Council of State held that an average of 57 test scores logically need not result in one of the three fixed scores, and that the plea was serious in none of its branches.
What happened?
AZ Sint-Jan Brugge-Oostende launched a public supply contract for a framework agreement for hand-care products (specifications APO-2022/01), placed through a negotiated procedure with prior call for competition. The award criteria and their weight were: price (45), product quality (35, assessed by means of a test form), plan of approach and timing (10), and after-sales service (10). After assessment, Dialex Biomedica scored 84.07 % and Filter Service 82.40 %. On 21 December 2022 the hospital awarded the contract to Dialex Biomedica. On 5 January 2023 Filter Service sought suspension under extreme urgency, raising a single plea with four branches. In a first branch it invoked the principle patere legem quam ipse fecisti: the sub-criterion 'drying time at a volume of 3 ML' had allegedly not been assessed according to the announced mathematical scoring method, since instead of one of the three fixed scores (0, 5 or 10 points) it received 9.06/10 and the chosen bidder 8.08/10; moreover the sub-criterion 'condition of the skin' had allegedly been wrongly assessed partly on the product's smell, with contradictory reasons. In a second branch it complained of a breach of the transparency principle: it received only 1 out of 5 for 'possible changeover of dispensers' because its bottles did not fit the dispensers in the operating theatre of AZ Sint-Lucas, whereas those specific features had not been defined in the contract documents. In a third branch it accused the hospital of carelessness and a manifest error of assessment in the third and fourth award criteria, and in a fourth branch of a breach of the formal duty to state reasons. The Council of State — which in interim relief only examines whether a serious plea or a manifest illegality is present — rejected every branch. As to quality, the specifications (point I.16 'Provision of samples') showed that large quantities of samples were tested by a broad group of staff and doctors in both hospitals; 57 people filled in an evaluation form, with fully blinded samples labelled 'firm A' and 'firm B' so as not to bias the assessment. For the 'drying time' sub-criterion each evaluator did award one of the three announced scores (0, 5 or 10); the fact that the final score, as an average of all 57 scores given, logically does not result in one of those three values did not, according to the Council, mean that the assessment method had been departed from. That testing was done by several people was, moreover, expressly stated in the specifications. For the 'condition of the skin' sub-criterion, the hospital made it plausible prima facie that 'smell' could be relevant — not the smell of the product itself, but the smell the skin gives off after use — and the alleged contradictions were the logical consequence of various evaluators each formulating their experiences in their own words. The remaining branches shared the same fate: Filter Service did not make it plausible that its tender had been carelessly examined or that a manifest error of assessment had been made, and the mere fact that it could formulate substantive criticism of the reasons already showed that it could meaningfully defend itself, so that the formal duty to state reasons was satisfied. The Council concluded that the single plea was serious in none of its branches and rejected the application, even though the auditor had given a contrary opinion. Dialex Biomedica's intervention was admitted. Filter Service was ordered to pay the costs of the application — a roll fee of 200 euros, a contribution of 24 euros and a procedural indemnity of 770 euros to the hospital — and the intervening party the costs of its intervention, set at a roll fee of 150 euros.
Why does this matter?
Qualitative award criteria scored by a test panel often provoke precisely the complaint Filter Service made: the final score is not a 'clean' value from the announced grid, so the authority supposedly departed from its own method. This judgment clearly refutes that reasoning. Where the specifications provide that a large number of testers each award one of the fixed scores, the final score is necessarily an average — and an average rarely coincides with one of the discrete values. That is no breach of patere legem, but the arithmetical result of a correctly applied method. The judgment also clarifies two other recurring points of dispute. First, that a seemingly 'odd' characteristic such as smell can indeed fall under a quality criterion if it is reasonably connected to it — here the smell the skin gives off after use, as an aspect of the condition of the skin. Second, that apparently contradictory remarks in an award report are inherent to an assessment by many different people each choosing their own words, and do not in themselves constitute a defect of reasoning. Finally, the judgment recalls that whoever levels extensive substantive criticism of the reasons in its application thereby demonstrates that the reasoning allowed it to defend itself — an argument that usually neutralises a plea about the formal duty to state reasons. That the auditor saw it differently underlines that these are real and contestable assessments, not foregone conclusions.
The lesson
If you are an unsuccessful bidder considering a plea against the scoring of a quality criterion, first examine how the specifications organise the assessment. If they provide for a test panel awarding a fixed score per product, then a final average score that does not coincide with one of those fixed values is no breach of patere legem but a normal arithmetical result — rarely a basis for a serious plea. Better to target a demonstrable irregularity: a criterion filled in only afterwards, an assessment element not announced in advance, or a real, identifiable error of assessment. Realise too that substantive criticism of the reasons undermines your own plea about the formal duty to state reasons, because it proves you could defend yourself. If you are the authority, this judgment confirms the value of a well-designed quality assessment: state expressly in the specifications that testing is done by a panel, blind the samples to avoid bias, and document each evaluator's scores and their reasoning. That way a final average score stands firm, even if it does not coincide with an announced value.
Ask yourself
Do you know whether the specifications have quality scored by a panel awarding a fixed score per product — and do you realise that a final average score then logically need not coincide with one of those fixed values, without this breaching patere legem? Can a seemingly unusual characteristic (such as smell) reasonably fall under the quality criterion concerned? Are the apparent contradictions in the report not simply the result of several testers each choosing their own words? And do you realise that extensive substantive criticism of the reasons usually neutralises your own plea about the formal duty to state reasons, because it shows you could meaningfully defend yourself?
About this database
The Council of State (Raad van State / Conseil d'État) is Belgium's supreme administrative court. In disputes over public procurement — from contract awards to tenderer exclusions — the Council of State is the final arbiter. The rulings in this database are summarised by TenderWolf in plain language, with practical lessons for tenderers and contracting authorities. View all rulings →