Abstract
We bring together insights from the sociology of diagnosis and the sociology of ignorance to examine the early diagnostic unfolding of 'Long COVID' (LC). Originally described by patient activists, researchers set out to ponder its unwieldy clinical boundaries. Using a scoping review method in tandem with qualitative content analytic techniques, we analyse medicine's initial struggles to construct a LC diagnosis. Paying attention to the dynamics of ignorance, we highlight three consequential conceptual manoeuvres in the early classifications of LC: causal agnosticism concerning the relationship between COVID-19 and LC, evasion of lumping LC with similar conditions; and the predictable splitting off of medically explainable cases from the LC designation. These manoeuvres are not maleficent, inept or unreasonable. They are practical but impactful responses to the classificatory dilemmas present in the construction of diagnoses amidst ignorance. Although there are unique aspects to LC, we suggest that its early fate is nevertheless emblematic of medicine's diagnostic standardisation processes more generally. To varying degrees, diagnoses are ignorance management strategies; they create a pathway through the uncertainty at the core of disease realities. However, while diagnoses circumscribe some types of ignorance, they produce others through the creation of blind spots and paths not taken.