Thursday December 15 2016
MRI scans could cut down on false positive results
"Detailed MRI scans should be offered to some women in pregnancy to help spot brain defects in the developing baby, say researchers," BBC News reports. A UK study suggests combining an MRI scan with ultrasounds could prevent misdiagnosis.
Current guidelines recommend that pregnant women are given at least two ultrasound scans, at 8 to 14 weeks and then between 18 and 21 weeks to check for abnormalities, including brain defects.
But the researchers were concerned that the diagnostic accuracy of ultrasound scans is limited; estimated at around 70%. This could lead to some women terminating their pregnancy due to thinking they will have a miscarriage or stillbirth, when in fact the pregnancy was healthy.
So, researchers wanted to see if using an additional MRI scan when a brain defect was suspected would lead to an improvement in diagnosis rates.
They considered 570 cases and found a 25% improvement in diagnosis rates when an MRI scan was used in addition to the ultrasound scan. They also found that almost all women included in the study thought it was a good idea.
The researchers propose that any pregnant woman whose ultrasound scan suggests her baby may have a brain defect should have an MRI scan as well for a more accurate diagnosis.
Where did the story come from?
This UK based study was carried out by researchers from The University of Sheffield, the University of Newcastle, the University of Birmingham, Birmingham Women's Foundation Trust, and Leeds Teaching Hospitals NHS Trust and was funded by the National Institute for Health Research Health Technology Assessment Programme.
The study was published in the peer-reviewed medical journal The Lancet and is open-access, meaning it is free to read online.
BBC News' reporting of the story was accurate, making the point that the extra test is most useful in borderline cases when doctors are uncertain of the outcome.
What kind of research was this?
This was a large multi-centre prospective cohort study in the UK, including 16 foetal medicine centres.
The study included 570 cases of potential foetal brain abnormality that were detected by ultrasound.
The cases were then assessed to see if in-utero MRI (iuMRI) scans (scans used during pregnancy) when used in addition to ultrasound improved diagnostic accuracy.
Prospective cohort studies are useful for assessing diagnostic accuracy as the results of any testing can then be compared against the actual outcome.
What did the research involve?
Researchers recruited women aged 16 or over between 2011 and 2014. In total, 570 cases were included whose unborn baby had a foetal brain abnormality detected by ultrasound.
Cases were divided into two groups depending when in pregnancy the initial ultrasound scan took place:
- between 18 and 24 weeks (n=369)
- at or after 24 weeks (n=201)
Women were recruited by being offered an iuMRI scan after the ultrasound suggested a brain abnormality.
After the ultrasound scan, doctors were asked to record their certainty of diagnosis for each brain abnormality from very unsure (10% certain) to highly confident (90% certain).
Less than 14 days afterwards, participants had an iuMRI scan at one of six sites across the UK.
The radiologist was aware of the level of certainty recorded by the ultrasound expert before the iuMRI was done. The radiologist was then required to comment on the diagnosis made with the ultrasound and add extra diagnoses where appropriate.
Diagnostic accuracy was assessed separately for the group who had the initial ultrasound scan between 18 and 24 weeks and the group who had it at 24 weeks or later.
This was done either by using brain scans for babies who were delivered during a viable pregnancy or by autopsy or post-mortem MRI in cases of termination of pregnancy, stillbirth or neonatal death.
What were the basic results?
When using the iuMRI scan in addition to the standard ultrasound, diagnostic accuracy was improved by:
- 23% (95% confidence interval (CI) = 18 to 27) in the 18 weeks to less than 24 weeks group
- 29% (95% CI 23 to 36) in the 24 weeks and older group
The overall diagnostic accuracy was 68% for ultrasound and 93% when combined with iuMRI – a difference of 25% (CI= 21 to 29).
Diagnoses were reported with high confidence on ultrasound in 465 of 570 cases compared with 544 of 570 cases when combined with iuMRI.
The MRI scans provided additional diagnostic information in 49% of cases, changed prognostic information (information regarding outcomes) in at least 20% of cases and led to changes in clinical management in more than one in three cases.
At least 95% of women who took part in the study said they would have an iuMRI scan if a future pregnancy was also complicated by a foetal brain abnormality.
How did the researchers interpret the results?
The researchers conclude that "our results indicate a 23% absolute increase in diagnostic accuracy when iuMRI is used to supplement ultrasound imaging in the 18 weeks to younger than 24 weeks gestational age group of fetuses and a 29% increase in the 24 weeks or older fetuses.
"Diagnostic confidence is also improved when iuMRI is used to assess prenatal fetal neuropathology as an adjunct to ultrasound."
They add that "the increased diagnostic accuracy and confidence results in changes in counselling and clinical management in a high proportion of cases. These factors, in conjunction with high patient acceptability, lead us to propose that any fetus with a suspected brain abnormality on ultrasound should have iuMRI before definitive counselling."
Overall, the study provides evidence for a link between having an iuMRI when a brain abnormality has been detected in the foetus and improved diagnostic accuracy. This improvement led to a change in management in a significant minority of cases.
This study provides strong evidence for adding this extra scan in cases when brain abnormalities are detected. Also, most women involved in the study thought this extra scan was a good idea if brain abnormalities were detected in the ultrasound scan.
However, there are some minor limitations to this pragmatic study which are acknowledged by the researchers:
- The radiologists assessing the iuMRI scans had already seen the notes from the ultrasound, which may have led to confirmation bias, where they are more likely to agree with the opinion already put forward.
- The iuMRI scan was carried out up to 14 days after the ultrasound. The time difference between examinations is an important factor for diagnostic accuracy as the foetal brain is growing and maturing rapidly. Most brain abnormalities will be easier to detect in more mature foetal brains. This makes it difficult to directly compare the accuracy of diagnosis from the ultrasound with that of the iuMRI scan. At this later stage, ultrasound scans may have also been more accurate.
- The majority of cases were taken from one site and the demographic of women in this area may differ compared to other areas and therefore results may not apply to other areas in the UK.