This is why we need RCTs, total-body CT edition

The Lancet:

Background: Published work suggests a survival benefit for patients with trauma who undergo total-body CT scanning during the initial trauma assessment; however, level 1 evidence is absent. We aimed to assess the effect of total-body CT scanning compared with the standard work-up on in-hospital mortality in patients with trauma.

Methods: We undertook an international, multicentre, randomised controlled trial at four hospitals in the Netherlands and one in Switzerland. Patients aged 18 years or older with trauma with compromised vital parameters, clinical suspicion of life-threatening injuries, or severe injury were randomly assigned (1:1) by ALEA randomisation to immediate total-body CT scanning or to a standard work-up with conventional imaging supplemented with selective CT scanning. Neither doctors nor patients were masked to treatment allocation. The primary endpoint was in-hospital mortality, analysed in the intention-to-treat population and in subgroups of patients with polytrauma and those with traumatic brain injury. The χ2 test was used to assess differences in mortality.

A number of observational studies have shown that total-body CT scans are beneficial when working up a trauma victim. So we do them. We know observational studies aren’t what we need, but we do them.

Enter this RCT from four hospitals in the Netherlands and Switzerland. Adults who had undergone trauma were randomized on a 1:1 basis to get either immediate total-body CT scans or a standard workup with conventional imagine and CT scanning when individually needed. Clearly doctors weren’t blinded, and neither were patients. The outcome of interest were in-hospital mortality.

Over nearly 3 years almost 5500 patients were assessed, and just over 1400 were randomized. Almost 1100 of them were included in the primary analysis. In-hospital mortality was 16% in the total-body CT group and 16% in the control group. So… no difference. There were also no significant differences in mortality in a subgroup of patients with polytrauma and a subgroup with traumatic brain injury.

Three serious adverse events occurred in the total-body CT group and one in the standard group. Those patients all died.

This was an intention-to-treat analysis from an RCT. It had a power calculation:

539 patients per group were needed for detection of a difference in mortality of 5% with a power of 80% and a two-sided alpha of 5%

So it was powered to see a difference of 5%. They found a no difference at all, let alone a statistically significant one.

I’m fine hearing a debate on these findings. But they contradict prior, weaker studies. Total-body CT involves a lot of radiation, and it’s not without harm. It also costs money. Debate is fine… in fact it’s necessary. Don’t sweep this under the rug. Don’t ignore it. Don’t pretend it’s “just one study” equal to the others. We do that too often. We need to stop.


P.S. This doesn’t mean that some patients don’t warrant total-body CT. This is just about doing one as a knee-jerk reaction for all trauma patients.

Hidden information below


Email Address*