Back to results

Department of Statistical Sciences

Longitudinal analysis of platelet count data

Abstract

dc:description.abstract

Platelet transfusions are critical in managing bleeding risks in patients with low platelet counts or dysfunctional platelets. This research explores the dynamics of platelet count levels. The primary aim is to understand when and why platelet products are failing, by investigating differences in platelet count trajectories among donor groups, exploring seasonality, identifying donor clusters with similar behaviours, and establishing connections between platelet count dynamics and product failures. Using longitudinal data from the South African National Blood Service (SANBS), I employed linear mixed-effect models to analyse platelet count trajectories and latent class mixed models to uncover donor clusters with distinct patterns. The findings reveal evidence of seasonal fluctuations in platelet counts, with highs in winter months, though deviations were observed in specific branch zones. Functional principal component analysis (FPCA) further confirmed these seasonal patterns and revealed inter-year variability. Critical to this study is the identification of two primary donor clusters, one with stable or elevated platelet counts and another showing a declining trend post-2018. Notably, these clusters did not significantly correlate with demographic factors like gender or location, suggesting other factors influencing platelet dynamics. The research also uncovered parallels between donor clusters and branch zones, highlighting variability in platelet profiles and product pass rates, particularly during periods of observed declines. This research provides insights into the temporal dynamics of platelet counts and their role in the quality and reliability of platelet products. By understanding these dynamics, we can better identify the factors contributing to product failures, ultimately improving the safety and efficacy of platelet transfusion practices.

Degree

thesis:*
Grantor dc:publisher.institution
Department of Statistical Sciences
Year dc:date.issued
2025

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Marcus, Mahdi
Advisor dc:contributor.advisor
  • Gumedze, Freedom

Subjects

dc:subject × 1

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/11427/42403
OAI identifier oai:identifier
oai:open.uct.ac.za:11427/42403

Chain of custody

source
Harvested from
University of Cape Town
Base URL
open.uct.ac.za/oai/request
Last updated
2026-07-22
Source record
OAI-PMH GetRecord
citation

Marcus, Mahdi. Longitudinal analysis of platelet count data. Department of Statistical Sciences, 2025. http://hdl.handle.net/11427/42403