I interviewed Allison Ryan, a geneticist who specifically works on human reproduction, namely preventing and minimizing the risk of genetic diseases in newborns. Her job is divided into two main categories: embryos implanted in mothers, so that the mother can select the best embryos to minimize genetic diseases, and fetus who were conceived naturally. Allison Ryan designed the tests that are used in her company, she also reviews tests results and helps mothers consult on specific issues.
Because Ms. Ryan worked on both the design and the review/consultation side of genetic testing, the ethical dilemmas she faced were two fold. I thus asked her to reflect on ethical challenges during the design stage, and in the review/consultation stage.
The design stage was very enlightening, and surprisingly very general and applies to many other fields of engineering as well. Ms. Ryan describes the difficulty of weighing the reliability of giving an answer versus the reliability of the answer being correct. Putting too much emphasis on one side over the other would cause problems: put too much emphasis on being correct 100% of the time, and one would never reach a decision; however, put too much emphasis on giving an answer, and then one runs the risk on giving more incorrect diagnosis that is acceptable. Designing the tests then requires one to set up a specification between the probability of getting an answer versus the reliability of the answer being correct, an example from Ms. Ryan herself being a tests that produces results 90% of the time, with the answer having an error 1 in 1000 times.
For consulting, one ethical issue is privacy. Sometimes, famous people appear in genetic clinics, and if certain celebrities are pregnant, one would assume that newspaper tabloids would be very interested in being the first to know this information. However, under the Health Insurance Portability and Accountability Act, which geneticists must comply to, states that patients must be treated with discretion.
I pressed for more specific examples, and to my joy, Ms. Ryan was able to present me an example that allowed me to see the ethical dilemas of both the design and the consulting stages of her work. She gave me the example of a hormonal test for Down Syndrome, the test she designed, poorly assessed whether a child had down syndrome; in fact, when the test came positive for Down Syndrome, the chance of the child actually having Down Syndrome was sub-fifty percent. However, when the test came negative for Down Syndrome, it was all but certain that the child would have Down Syndrome. Although this test would appear sub-par, it is an acceptable test because it is a low cost preliminary tests. If the test came negative, then the mother would be spared from more expansive tests, and if the test comes in positive, then the mother would take more accurate, albeit costly tests.
This is interesting because it proves that accuracy and best-results are not always the most ethical choice: a paradox at first glance no doubt. As engineers we feel obligated to give the best results and the best care for the people who use our products, but best isn’t always right. I have taken many computer science classes, and one of the things I learned from these classes is that there is no such thing as a free lunch. If one method is unequivocally better than another, there is usually some trade off. This trade-off could be cost, time, adverse side effects, etc. Say if Ms. Ryan did not have this test, and only worked an a test that was as close as 100% accurate as technologically possible. The costs of the tests would rise, as Ms. Ryan herself stated. Because of this many mothers would pay much more to have these tests for genetic defects. This becomes very apparent when we take the fact that most people are not afflicted with Down Syndrome. Thus many mothers would only need the cheap test that was very accurate at negative diagnosis, but if Ms. Ryan only cared about 100% accuracy tests, then all mothers would have to pay for the expensive tests. Then there are the two intentionally unethical paths. Doctors who do not wish to continue testing, and just give a negative diagnosis without a high degree of accuracy–after all, if a genetic disease has a .001% chance of occuring, then the doctor can just say “negative” all the time and be wrong .001% of the time. Then there is the doctor who always suggests the most expensive and accurate tests first, even when cheaper first steps are available. This could be to either collect more money from expensive tests, or to save time so the doctor only needs to test a patient once instead of potentially many times if they come up positive.
Ms. Ryan states that the biggest tool to help them make ethical decisions, is a specific field called Genetic Counseling. These Counsels recommend actions to mothers, and even help Ms. Ryan help reach decisions as well. I was satisfied with this answer, and was honestly surprised that an entire field was devoted to maintaining the ethical relationship between doctor and patient.
I learned many things from Ms. Ryan, the balance of reliability versus accuracy, and that sometimes the balance between reliability and accuracy means many different tests of varying reliability and accuracy, especially since cost and time are variables in making good ethical decisions. I will take Ms. Ryan’s experiences in mind as I continue my studies in engineering.