Picture a woman, ten weeks into her pregnancy, when the clinic rings. The blood test has come back positive for Down syndrome. She spends the next eleven days in something close to grief. Then the diagnostic test arrives. Her baby is fine. Nothing was wrong at any point, except her understanding of what that first phone call meant.
That scenario is not rare, and it is not anyone’s fault. Most parents weighing up an NIPT test in Singapore have been told the test is around 99 per cent accurate, and that figure is broadly right. The problem is that 99 per cent accurate does not mean what almost everyone assumes it means. It is one of those numbers that is completely true and still manages to mislead.
The Number Nobody Quotes
When a laboratory says a screening test is 99 per cent accurate, it is usually describing how good the test is at spotting the condition when the condition is there. That is a useful number before you take the test. It is the wrong number afterwards.
The number you actually want, once a result comes back high risk, is a different one. It answers a narrower question: of all the women who got this result, how many were carrying an affected baby?
The American College of Obstetricians and Gynecologists puts it in plain arithmetic. Among women in a low-risk group who receive a positive cell-free DNA result, “out of 100 women with a positive result, 33 will have babies with Down syndrome, and 67 will not”. In a higher-risk group, the same positive result means 83 out of 100. Same test. Same laboratory. Two very different phone calls.
Why Rarity Changes Everything
The reason is not a flaw in the technology. It is arithmetic, and it catches out clinicians as often as patients.
Down syndrome is uncommon. In any large group of pregnancies, the number of unaffected pregnancies is enormous next to the number of affected ones. So even a test that is wrong very rarely will produce a fair number of false alarms, simply because it has so many more unaffected pregnancies to be wrong about.
Think of a stadium with one person in it wearing a red hat. A search method that is 99 per cent reliable will find that person. It will also wave at a handful of people in orange. Down on the pitch, that handful looks like a crowd.
What A Blood Test Cannot See
There is a second thing worth knowing, and it has nothing to do with statistics.
NIPT reads fragments of placental DNA floating in the mother’s blood. It is looking at chromosomes. It is not looking at the baby. A first trimester ultrasound assessment, usually carried out between 11 weeks and 13 weeks plus six days, does close to the opposite. It measures the nuchal translucency behind the baby’s neck, checks the nasal bone, and can pick up some major structural findings that no blood sample would ever reveal.
Parents often ask which test is better. That question has no clean answer, because the two are not looking at the same thing. One reads a code. The other looks at a body.
A Rule Worth Borrowing
If you take one thing from this, make it a habit rather than a fact. Treat every screening result as a question, never as an answer.
In practice, that means:
- A low-risk result lowers the odds. It does not rule anything out.
- A high-risk result raises the odds. It does not confirm anything.
- Only a diagnostic test, such as amniocentesis or chorionic villus sampling, can give a definite answer.
- Ask for the predictive value for your age and risk group, not just the headline accuracy.
- Ask which conditions the panel covers, because panels are not all the same.
The Mistake That Costs The Most
The expensive mistake is acting on the screening call instead of waiting for the confirmation. Sometimes the cost is eleven days of grief. Occasionally it is a decision that cannot be taken back.
People make this mistake for an understandable reason. In every other corner of medicine, a positive result is close to a verdict. A positive swab means you have the infection. We carry that instinct into a room where it does not belong, and the word “positive” lands with a weight the test never intended it to carry.
So ask the second question before you need it. Before the blood is drawn, ask what a high-risk result would actually mean for someone in your position, and what the next step would be. Ask it while you are calm, because the day of the phone call is the worst possible day to be learning how a screening test works. The number 99 will still be true. You will simply know what it is describing, and, more usefully, what it is not.
