Obstetric providers' experience with, interpretation of, and communication of NIPS test results

NIPS (Noninvasive prenatal screening) is rapidly rising in popularity in prenatal genetics clinics to screen for chromosome abnormalities, and has generated much discussion in the obstetrics community regarding concerns of its positive predictive value, patients' perception of the test, and providers' interpretation of test results. This study reports on obstetricians, genetic counselors, and midwives experience with the test, as well as their interpretation and communication of test results. This study has a specific focus on positive predictive value and how accurately providers are able to calculate it for NIPS test results. A survey was administered in person to obstetricians at a continuing education conference hosted by UCSF, via email to National Society of Genetic Counselors (NSGC) members, and via email to American College of Nurse-Midwives (ACNM) members. The survey consisted of 27 questions that covered three main areas: provider demographics, provider experience with NIPS, and statistics knowledge. Descriptive analysis was conducted for questions answered by all three types of providers, and ANOVA/t-test analysis was performed to determine if certain demographic and NIPS experience variables correlated with how well providers did on the statistics portion of the survey. Results revealed that genetic counselors were more comfortable with test related statistical concepts and performed significantly better on the statistics portion of the survey than ob-gyns (p0.001) and midwives (p0.001). Additionally, presenting a hypothetical scenario in natural frequency format rather than percentages seemed to make it easier for all three groups of providers to calculate positive predictive value. Interestingly, 16% of midwives viewed NIPS as diagnostic and committed the error of confusing sensitivity with positive predictive value, while all three groups of providers committed the error of subtracting false positive rate from sensitivity to calculate positive predictive value. These findings highlight important targets of educational intervention for all three groups of providers.