Premium
Indeterminate CT pulmonary angiogram: Why and does it matter?
Author(s) -
Yeo Ju Hee,
Zhou Lifeng,
Lim Remy
Publication year - 2017
Publication title -
journal of medical imaging and radiation oncology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.31
H-Index - 43
eISSN - 1754-9485
pISSN - 1754-9477
DOI - 10.1111/1754-9485.12500
Subject(s) - medicine , indeterminate , hounsfield scale , logistic regression , univariate analysis , radiology , univariate , nuclear medicine , multivariate analysis , computed tomography , multivariate statistics , statistics , mathematics , pure mathematics
This study assessed the rate of indeterminate CT pulmonary angiogram ( CTPA ) in a general hospital and evaluated potential contributing factors to an indeterminate CTPA and patients’ eventual clinical outcome. Methods Four hundred and three consecutive CTPA and their finalised report from 01/08/2012 to 05/12/2012 at Waitemata DHB were reviewed retrospectively. Patient demographics, scan parameters and average Hounsfield unit ( HU ) of the main pulmonary trunk ( PT ) and likely cause of indeterminate study were documented. Studies were categorised into diagnostic, suboptimal or non‐diagnostic studies. Univariate and multiple regression analyses were employed. Results Six per cent ( n = 24) of the studies were deemed indeterminate which included suboptimal and non‐diagnostic studies. Seven pregnant patients were scanned during the study period and four of the seven studies were deemed indeterminate. In univariate analysis, predictors of indeterminate studies included ‘weight’ ( P = 0.022), ‘average HU of PT ’ ( P < 0.0001) and ‘effective dose’ ( P = 0.0003). In the multivariate logistic regression model, only ‘average HU of PT ’ was associated with indeterminate studies ( OR = 0.99, 95% CI : 0.99, 1.00). Twelve of 24 indeterminate studies had suboptimal pulmonary enhancement with causes including suboptimal timing of contrast bolus, hyperdynamic state due to pregnancy and transient interruption of contrast. Three patients had excessive noise due to body habitus. Eight studies had motion artefacts and one patient had left lower lobe pneumonia which decreased the accuracy. A third of the indeterminate studies were clinically considered as negative by referring clinicians. Conclusions Pulmonary trunk average Hounsfield unit is a predictor of indeterminate CT pulmonary angiogram.