A retrospective review of computed tomography pulmonary angiography image quality and the impact on diagnostic outcome at a tertiary South African hospital

Holtzhausen, Jeanette

A retrospective review of computed tomography pulmonary angiography image quality and the impact on diagnostic outcome at a tertiary South African hospital

Master Thesis

2022

Abstract

Background: Computed Tomography Pulmonary Angiography (CTPA) is a key diagnostic imaging modality for pulmonary embolism. These studies are technically challenging to perform. Degraded image quality may result from inadequate pulmonary artery contrast opacification, motion- or streak artefact as well as patient factors. Literature suggests that poor quality scans could lead to indeterminate outcomes and suboptimal clinical decisions with risk of increased mortality. Objective: The study aimed to benchmark the image quality and diagnostic outcomes of CTPA studies in the setting of a tertiary Southern African hospital. The relationships between CTPA image quality and diagnostic and clinical outcomes, as well as related variables such as health risk factors and effective dose, were also explored. Methods: A retrospective cross-sectional study evaluated consecutive CTPA studies performed at Groote Schuur hospital, Cape Town, South Africa, over a six-month period from 1 July 2018 to 31 December 2018. All studies performed for suspected acute or chronic pulmonary embolism (PE) in patients 18 years and older were included. Records were reviewed regarding image quality and diagnostic and clinical outcomes. Correlation tests were performed between continuous variables and chisquare tests among categorical variables. Results: During the study period, 231 CTPA studies were performed, of which 226 were included. The sample comprised 69 % females and 31 % males, with median age of 45 years (range 19-84 years). In 204 (90.3 %) of studies, adequate contrast opacification ≥ 211 HU was obtained. Inadequate contrast opacification was present in 9.7% of cases, in line with previous research. Motion and/ or streak artefacts were present in 45.6%. PE was confirmed in 22% and excluded in 65 % of cases. The number of scans with indeterminate diagnostic results only comprised 30 out of the 226 scans reviewed, however, the percentage was higher than previously reported (13.3% vs mean of 6.4 % in published literature). Amongst these, inadequate contrast opacification occurred in 15 (50 %) of studies and artefacts degraded image quality in 24 (80 %). Patients with a diagnosis of PE had higher mortality, compared to patients with negative and indeterminate scans. Clinicians interpreted indeterminate scans as negative, however, this did not impact adversely on mortality. Conclusions It was encouraging that the percentage of studies with adequate contrast opacification met published bench-marks. Although the higher-than-expected percentage of indeterminate studies may partially be explained by the prevalence of artefacts, it requires further investigation. This did not, however, translate into adverse mortality outcomes.

Keywords

Diagnostic Radiology

Reference:

Collections

Masters

Full item page