Carrier Screening With an Imperfect Test
A couple seeks counseling. The woman's brother has cystic fibrosis (CF); her partner has no family history and a negative CF carrier screen. The lab panel detects 90% of pathogenic CFTR variants in people of European ancestry. Population carrier frequency is about 1 in 25 in that group. The woman is 28 and healthy; her partner is 30 and healthy. They ask, 'Since his test is negative, is our chance of having an affected child basically zero?' The counselor knows the woman's risk is not the population risk because of her affected sibling. The couple is planning a pregnancy within the year and wants a number, not a vague reassurance. A follow-up expanded panel is available but costs $300 and has a 2–3 week turnaround. The counselor must give a quantitative estimate and explain its limits.
1. Calculate the couple's approximate risk of an affected child before and after the partner's negative screen. Show each step and state your assumptions.
Read the worked answer
Assumptions: CF is autosomal recessive; the woman is unaffected, so with an affected sibling her carrier probability is 2/3; the partner's prior carrier probability is the population value, 1/25; the panel detects 90% of carriers and is assumed specific; no consanguinity.
Before screening:
- Woman carrier: 2/3
- Partner carrier: 1/25
- Affected child if both carriers: 1/4
- Risk = (2/3)(1/25)(1/4) = 2/300 = 1/150 ≈ 0.67%.
After a negative screen: the partner's posterior carrier probability is (0.10 × 1/25) / [(0.10 × 1/25) + (1 × 24/25)] = 0.004 / 0.964 ≈ 0.00415 ≈ 1/241. Then child risk = (2/3)(1/241)(1/4) ≈ 1/1,446 ≈ 0.07%. The negative screen lowers risk about tenfold, but it does not make it zero.
2. The couple asks whether the negative result means zero risk. Explain the limits of the screen in terms they can understand.
Read the worked answer
A negative screen lowers probability; it cannot prove absence. The panel misses about 10% of pathogenic CFTR variants, and rare variants outside the panel can still be carried. The partner could therefore be a carrier even with a negative result. The residual chance of an affected child is about 1 in 1,450 under the stated assumptions - low, but not zero. It is also higher than the general population risk because the woman's sibling with CF makes her carrier risk 2/3, not 1/25. The counselor should avoid saying 'basically zero' and instead say: 'Your chance is much lower than before the test, but a small residual risk remains because no carrier screen is perfect.' This framing supports informed choice without overstating certainty.
3. What additional information would most change your estimate, and how would you frame the decision about the $300 expanded panel?
Read the worked answer
Most influential information: the partner's ancestry (panel sensitivity differs by population), consanguinity, and whether the affected brother's two CFTR variants were identified. If the brother's variants are known, targeted testing of the woman could confirm whether she carries one; if she is not a carrier, the child risk is near zero regardless of the partner. An expanded panel that includes sequencing and deletion/duplication analysis would reduce the partner's residual carrier probability, but variants of uncertain significance could complicate interpretation. Frame the $300 panel as buying uncertainty reduction, not certainty. Ask what result the couple would act on and whether the 2–3 week turnaround fits their timeline. Present the numerical range and the limits nondirectively.