An unusual findingin an experiment by two Boston-area doctors is raising questions about a new generation of prenatal screening tests used by an increasing number of pregnant women to predict the risk a fetus has Down syndrome or other genetic conditions.

A pair of obstetricians at Boston Maternal Fetal Medicine who specialize in high-risk pregnancies submitted blood samples from two women who were not pregnant to five commercial labs that perform these new screens, claiming the samples were from women who were 12 weeks pregnant. The study was funded by Ariosa Diagnostics Inc., one of the testing companies.

Three of the labs returned results that indicated the presence of a normal female fetus, even though the samples they received were from non-pregnant women.

The results, according to the doctors who conducted the experiment, underscore the need for better oversight of the screens, which are catching on rapidly in the U.S. and globally.

“Quality controls are zero,’’ said Dr. Tamara Takoudes, co-author of a letter to the editor describing the experiment, which has been published online in Ultrasound in Obstetrics & Gynecology. “As physicians we have to be more leery about the products we are using.”

Concerned that patients and providers have been treating these popular new screens as “bullet-proof clearance” for pregnancies, co-author Dr. Benjamin Hamar said he and Takoudes decided to design a project to highlight issues with the way several of these companies conduct their testing. A difference between testing companies is how much fetal fraction – the amount of placental DNA in the mother’s blood, an indication of the presence of a fetus – is needed to provide a robust screen result. Ariosa, Natera, and Sequenom measure fetal fraction as part of their screens, but Illumina does not, according to companies and published articles.

The three companies sending back screen results indicating a female fetus were: Sequenom, Inc., which offers the MaterniT21 PLUS screen; Illumina Inc., which offers verifi; and LabCorp, which uses the same sequencing technology as Illumina under the name informaSeq. Illumina and Sequenom both said the study was flawed because the screens are not designed to test for pregnancy; rather, they look for specific chromosomal abnormalities. LabCorp did not return a request for comment.

The screens, known as non-invasive prenatal tests (NIPT), analyze placental DNA circulating in a mother’s bloodstream to pinpoint potential chromosomal conditions, such as Down syndrome and other more severe conditions such as Patau or Edwards syndromes, which are associated with severe birth defects and early death.

Studies have shown NIPT to perform better than traditional screens doctors have historically offered. But the screens are not diagnostic in the way an amniocentesis, which directly samples fetal tissue, is. The U.S. Food and Drug Administration does not regulate the labs offering NIPT, although the agency recently announced its intention to begin regulating a broad array of diagnostic tests that would include NIPT. The new rules, if approved, are expected to take several years to implement.

Sequenom returned the oddest result because it reported a fetal fraction, which indicates the presence of a fetus, for both women, along with a gender.

Blood from a non-pregnant female, the company said in a statement to NECIR, is “unsuitable for testing” because it doesn’t allow for a true test of the screen’s performance. They said their method for assessing fetal fraction has been robustly tested though did not address why the non-pregnant women had a reportable fetal fraction.

Illumina called the study “simply not valid” and said it was akin to submitting a female sample for a prostate cancer screen and receiving a normal result.

Tristan Orpin, senior vice president for Illumina’s reproductive and genetic health business, said in an email that Illumina’s results reported no abnormal chromosome count and the presence of two X chromosomes. The authors, he said, “drew their own conclusion that this meant ‘female fetus.’ ”

Ariosa correctly reported no results, citing insufficient fetal DNA. The company declined to comment, as did Roche, the pharmaceutical company acquiring Ariosa. Arisoa was blind to which samples were the non-pregnant women, Takoudes and Hamar said, because their practice sends out hundreds of samples each month for screening to the company.

Natera, Inc., which offers the Panorama screen, also correctly reported no results.

Melissa Stosic, Natera’s medical education and clinical affairs director, was quick to differentiate her company from competitors.

“From now on, doctors and actual pregnant patients must think twice when receiving a negative result with female sex from one of those three labs,” she said, referring to Sequenom, Illumina, and LabCorp.

Both Takoudes and Hamar said their findings should prompt health care providers to be more vigilant, because when fetal fraction is not tested for or reported, pregnant women with very low fetal fraction might get a “normal” result when their fetus is, in fact, affected.

Dr. Anthony Gregg, vice president of clinical genetics at the American College of Medical Genetics and Genomics and a professor at the University of Florida, took issue with the way the experiment was set up.

“This is not a tool that has ever been intended for non-pregnant patients,” he said. “[Takoudes and Hamar] gave two samples from inappropriate patients.”

Takoudes, who said she would continue to use these screens but advocate better regulation, disagreed with the criticism, saying, “it’s an inappropriate result, not an inappropriate patient.”