Towards a nuanced view of diagnostic test properties: an application to transfusion transmitted risk estimation

Bingham, Jeremy (2021-03)

Thesis (MSc)--Stellenbosch University, 2021.

Thesis

ENGLISH ABSTRACT: Laboratory screening (rather than pathogen inactivation) is likely to remain, for the foreseeable future, the primary means of ensuring the safety of blood products from transfusion transmissible viruses such as Hepatitis B, Hepatitis C, and Human Immunodeficiency Virus (HIV). Depending on the tests used, there is generically some ‘residual risk’ of transfusion transmitted infection, as no test can guarantee detection of all potentially infectious material. Previouslydescribed risk estimation approaches 1) mostly treat detectability and infectiousness as categories rather than continuously tunable; 2) disregard sources of variability and their correlation; and 3) are not generalizable to arbitrary detection biomarkers – making it difficult to generate estimates of residual risk without extensive programmatic monitoring. We describe a broad framework for modelling test performance which incorporates hitherto neglected sources of variability in parameters governing the infectiousness and detectability of transfusion-transmissible pathogens. We utilise models based on this framework to demonstrate the relationship between test performance and residual risk for various assumptions about the biomarker/infectiousness relationship, and illustrate how the same framework may be used to inform modelling efforts in related fields - such as infection dating and incidence estimation - which rely on realistic representations of test performance. The key findings from our scenario modelling demonstrate: 1) Diminishing returns on increased screening sensitivity not evident in less flexible models; 2) increasing inter-subject variability in detectability and infectiousness leads to increasing residual risk in our general model, but lower risk estimates than in a previously described and widely used semi-mechanistic model. These effects are stronger when the average delay between infectiousness and detectability is short. Planning blood product screening algorithms in light of simulations using our models can generate robust expectations of residual risk over a wide range of test performance and product risk levels. We outline when simpler models may be relied upon, and when additional nuance must be considered.

AFRIKAANSE OPSOMMING: Laboratoriumtoetsing (eerder as patogeeninaktivering) sal waarskynlik in die afsienbare toekoms die primêre manier bly om die veiligheid van bloedprodukte teen oordraagbare virusse soos hepatitis B en C en menslike immuniteitsgebreksvirus (MIV) te verseker. Afhangend van die toetse wat gebruikte word, is daar ’n ’oorblywende risiko’ van infeksie deur oortapping, aangesien geen toets die opsporing van alle moontlike aansteeklike materiaal kan waarborg nie. Voorheen beskryfde risiko-beramingsbenaderings 1) behandel opspoorbaarheid en aansteeklikheid meestal as kategorieë eerder as kontinue eienskappe; 2) bronne van variansie en die korrelasie daarvan word verontagsaam; en 3) kan nie veralgemeen word vir arbitrêre opsporing van biomerkers nie, wat dit moeilik maak om ramings van oorblywende risiko te genereer sonder uitgebreide programmatiese waarneming. Ons beskryf ’n breë raamwerk vir die modellering van die diagnostiese akkuraatheid van toetse. Dit sluit in bronne van variansie in die parameters wat die aansteeklikheid en opspoorbaarheid van oortapping-oordraagbare patogene beheer. Ons gebruik modelle wat gebaseer is op hierdie raamwerk om die verband tussen toetsprestasie en oorblywende risiko vir verskillende aannames oor die biomerker / aansteeklikheidsverhouding aan te toon, en illustreer hoe dieselfde raamwerk gebruik kan word om modelleringspogings in verwante velde in te lig. Voorbeelde sluit in die beraming van infeksie datums en die beraming van infeksie insidensie, wat albei staatmaak op realistiese beramings van die akkuraatheid van diagnostiese toetse. Die belangrikste bevindings uit ons scenario-modellering toon: 1) Afnemende opbrengste met verhoogde sifting-sensitiwiteit, wat nie duidelik is in minder soepel modelle nie; 2) groter variansie tussen individue in opspoorbaarheid en aansteeklikheid lei tot groter oorblywende risiko in ons basis model, maar laer risiko beramings as in ’n voorheen beskryfde en wyd gebruikte semi-meganistiese model. Hierdie effekte is sterker wanneer die gemiddelde periode tussen aansteeklikheid en opspoorbaarheid kort is. Bloedproduktesiftingsalgoritmes wat bepaal word deur simulasies met ons modelle, kan sterk verwagtinge van die residuele risiko oor ’n wye reeks toetsprestasies en produkrisikovlakke genereer. Ons gee ’n uiteensetting van wanneer eenvoudiger modelle gebruik kan word en wanneer addisionele nuanses in ag geneem moet word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/110212
This item appears in the following collections: