Are We Overdosing Women with CT Contrast? What Our Data Showed

Updated August 2nd, 2026

We found it by accident

The finding did not come from a study designed to look for it. We were working on something else entirely: whether haemoglobin can be quantified from a routine contrast-enhanced photon-counting CT, which we published in Radiology in 2022. Somewhere in those datasets the pattern showed up. Women were coming out brighter than men. Same protocol, same contrast volume, more iodine in the liver and in the blood pool.

That is the kind of observation you either write down or forget. We wrote it down, and then we went looking for it properly. The result is in European Radiology, and this is the plain-language version of it.

What we did

Two hundred and seventy-four patients with a clinically indicated thoracoabdominal CT in portal-venous phase, scanned on a dual-source photon-counting system between August 2021 and January 2022. One hundred and sixty-eight men, 106 women, mean age 68 years, mean BMI 26.

Everyone received a fixed volume of iopromide at 300 mgI/mL, either 100 or 120 mL, followed by a 30 mL saline chaser at 4 mL/s. Bolus-triggered, 45 seconds after the ascending aorta reached 120 HU. This is not an experimental protocol. It is what most departments do, which is exactly the point.

The reason the study was possible at all is spectral. On a photon-counting dataset you get a virtual non-contrast image and an iodine map from the same acquisition, so you can measure how much of the attenuation is iodine and how much was always there. Without that you would need a true non-contrast scan on every patient, which nobody is going to do for a research question. Regions of interest went into both liver lobes, the portal vein, the spleen, the left atrium, the left ventricle, the pulmonary trunk and the ascending and descending aorta, on 70 keV virtual monoenergetic images.

What came out

Eleven HU in one group, twelve in the other. In the univariable model, being female was worth 25 HU more blood attenuation, with a confidence interval from 20.2 to 30.8. For comparison, each additional year of age bought 0.34 HU.

The obvious objection is body size, and it does not survive the data. Match on BMI and the difference stays. Match on body weight and it stays. In the multivariable model, sex and weight both remained independent predictors while age and height dropped out.

Why

Fat has a poor blood supply and little interstitial space, so it does very little to dilute an iodine bolus. The contrast stays where it can be measured: in the vessels and the parenchyma. Women carry proportionally more fat and less muscle than men at the same weight. Dose by total body weight and you are effectively dosing a compartment that does not participate.

That is the mechanism, and it suggests its own fix. If the problem is that we are dosing by a compartment the contrast never reaches, dose by the compartment it actually occupies.

The fix is a formula from 1962

We recalculated everything against blood volume using the Nadler formula, which takes sex, height and weight and has been in use in intensive care for sixty years.

Correct for blood volume and sex stops being significant for blood attenuation, hepatic parenchyma and portal vein. The difference does not shrink. It disappears.

The spleen was the exception and stayed significant, which we think is a measurement problem rather than a physiological one. Splenic enhancement in the portal-venous phase is famously patchy, and a region of interest placed in it measures luck as much as iodine.

Why this is not a rounding error

Women have a higher rate of acute adverse reactions to iodinated contrast, and anaphylactoid reactions in particular skew heavily female. The reported risk of contrast-induced nephropathy is higher too. And the probability of a reaction rises with the volume administered.

So the population that is more likely to react is systematically receiving more iodine than it needs to reach diagnostic enhancement. That is not a subtle inequity. It is a dosing error with a known direction.

There are two smaller arguments on top. Iodinated contrast has been in short supply within recent memory, and it is expensive at the scale radiology uses it. And excreted contrast turns up in drinking water, because wastewater treatment does not remove it. Neither is the reason to change practice, but neither argues the other way.

What I would and would not claim

This is a retrospective single-centre analysis with an unbalanced sex ratio, 168 men to 106 women. We did not assess renal function, and there was no control group receiving individualised dosing. Cardiac output and blood pressure influence enhancement and were not measured, and sex differences in blood pressure are themselves well documented.

Earlier work disagrees. Svensson and colleagues also saw a sex difference in hepatic enhancement, but it lost significance after adjustment for weight, height and age, in a cohort of 100. Kidoh favoured body surface area over blood volume; Kondo favoured lean body weight; Davenport argued for plain weight adjustment. There is no consensus and I am not pretending our paper creates one.

What I will claim is narrower and I think secure. A fixed or weight-adapted contrast protocol produces higher enhancement in women than in men, the effect survives adjustment for body size, and a blood-volume-based dose removes it in liver, portal vein and blood pool. A prospective study is running to test whether that translates into practice.

What to do with this on Monday

If your department uses a fixed volume, that is the situation this paper describes. Calculating Nadler blood volume needs sex, height and weight, all of which are already on the request. The step nobody should skip is auditing your own enhancement values by sex before changing anything, because if your protocol already runs lean you may not have room to reduce.

And if you are about to inject, the thyroid question comes first. There is a decision tool for the thyroid workup before iodinated contrast on this site. The physics of why lower tube voltage raises iodine contrast is covered in the article on kVp.

Source

Becker J, Huber A, Bette S, Rubeck A, Arndt TT, Müller G, Risch F, Canalini L, Wollny C, Schwarz F, Scheurig-Muenkler C, Kroencke T, Decker JA. Are we systematically overdosing women? Revisiting standardized contrast protocols for thoracoabdominal CT scans. European Radiology 2025;35:3729–3738. doi:10.1007/s00330-024-11329-8