2.15, 2.16 - Genetic screening and ethical choices

Carriers and family probabilities

A person can pass on an allele associated with a recessive condition without having that condition themselves.

For cystic fibrosis, a carrier usually produces enough functioning CFTR protein to avoid having CF. Appearance alone therefore cannot distinguish a carrier from someone with two functioning alleles. Carrier testing examines DNA, for example DNA extracted from white blood cells in a blood sample, for relevant variants. If a disease-causing variant has already been identified in a family, testing can specifically look for it. Testing a partner then helps establish the possible allele combinations in a child. A negative result from a panel that checks only some variants does not exclude every possible disease-causing variant; the remaining uncertainty depends on what the test covered. This limitation is explicit in NHS England's CF testing guidance.

From parental genotypes to a probability

Use a simplified single-gene model: $C$ is a functioning allele and $c$ is a particular recessive disease-causing allele. Assume $cc$ always causes the condition, $CC$ and $Cc$ are unaffected, each parent transmits either allele with equal probability, fertilisation is random, and no new mutation occurs. These assumptions define the model; they are not a complete prediction of every real genetic condition.

For two confirmed carriers, $Cc\times Cc$, the four equally likely combinations are:

The affected probability is $\frac12\times\frac12=\frac14$, or 25%. The carrier probability is $\frac24=\frac12$, or 50%. The probability of being unaffected is $\frac34$, including both carriers and non-carriers. Each pregnancy is a new independent event under this model. An affected first child does not use up the chance of another affected child.

Family information can also change the denominator. In this exact family, both of Maya's biological parents are confirmed $Cc$, and Maya is known to be unaffected but has not been tested. Among the three equally likely unaffected combinations above, two are $Cc$. Maya's carrier probability is therefore $\frac23$, rather than $\frac12$. If her partner is confirmed $Cc$, their child's affected probability before Maya is tested is $\frac23\times\frac14=\frac16$. If testing confirms Maya is $Cc$, it becomes $\frac14$. The test changes the information available about an existing genotype, not the genotype itself.

Testing embryos before transfer

Identifying parental variants can make it possible to test embryos for a specified inherited condition before pregnancy is established.

The specification calls this pre-implantation genetic diagnosis (PGD). For a condition caused by a particular gene, the current term is pre-implantation genetic testing for monogenic disorders (PGT-M). This is distinct from tests for chromosome structural rearrangements or chromosome-number changes.

Eggs are collected and fertilised in a laboratory during in vitro fertilisation (IVF). After embryos develop, a small sample of cells is removed from an embryo and its DNA is tested for the relevant familial variant or variants. An embryo whose result does not indicate the target condition may be selected for transfer into the uterus, where implantation may occur. For a recessive condition, an unaffected carrier embryo is different from an affected embryo. Selection changes which embryo is transferred; it does not repair an allele in an embryo.

Testing may reduce the chance of establishing a pregnancy affected by the target condition and may avoid a later decision about termination for that condition. However, IVF is physically and emotionally demanding. There may be no suitable embryos, sampling can damage an embryo, testing can fail or give an incorrect result, and transfer does not guarantee implantation or a live birth. These limitations and the current terminology are described by the HFEA, reviewed 30 July 2026.

Screening and diagnosis

During pregnancy, it matters whether a result estimates a chance or directly investigates the suspected condition.

Prenatal screening may use blood measurements or ultrasound findings. For example, non-invasive prenatal testing (NIPT) is a blood-based screening test for specified chromosome conditions. It is not a test for every genetic condition. A higher-chance result can lead to an offer of further investigation; it is not itself a diagnosis. A lower-chance result reduces concern about the conditions screened for but cannot guarantee their absence. The NHS explains this distinction.

A diagnostic test investigates whether a particular condition or genetic change is present, using an appropriate laboratory test on a sample. Amniocentesis and CVS obtain samples for prenatal diagnosis. A known familial variant may justify an offer of targeted prenatal testing even without a preceding higher-chance population screen. In the specification, “genetic screening” is a broad heading covering carrier and embryo testing as well as prenatal investigations; keep the screening-versus-diagnosis distinction when describing an individual test.

Consider this constructed example, which illustrates test interpretation rather than a real screening programme. All 100 pregnancies have a screening result and a separate reliable diagnostic outcome for the same condition.

The 12 higher-chance results with the condition absent are false positives. The one lower-chance result with the condition present is a false negative. Screening can identify pregnancies for further investigation while leaving uncertainty in either result group. The table does not justify assuming the same performance for other conditions or populations.

CVS and amniocentesis

The two invasive sampling methods collect different material from the pregnancy so that laboratories can investigate DNA or chromosomes.

In chorionic villus sampling (CVS), a specialist removes a small sample of chorionic villi from the placenta, using ultrasound guidance. Access is through the abdomen with a needle or through the cervix with a tube or small forceps. The placenta develops from the early embryo and usually shares its genetic constitution, allowing the sample to provide information about the fetus. However, a chromosome change can sometimes be present in placental cells without being present in the fetus. This can make interpretation uncertain and lead to further testing.

In amniocentesis, an ultrasound-guided needle passes through the abdomen into the amniotic sac to remove a small volume of amniotic fluid, while avoiding the fetus. The fluid contains cells from the fetus. DNA from these cells can be tested for a known familial variant; appropriate chromosome tests can investigate specified chromosome conditions. Sampling and laboratory analysis are separate steps: the sample does not automatically reveal every genetic condition. NHS England describes both methods, and the NHS explains the placental interpretation limit.