> In a 1993 study from France, [...] Of the 478 positive (pregnant) urine samples distributed, 230 were falsely interpreted as negative (sensitivity, 48%)
> Valanis and Perlman [...] The incidence of false-negative results in this study was 24.3%.
While the study does show that pregnancy tests in 1993 were very inaccurate, there is nothing in the study that suggests the inaccuracy was due to people having problems interpreting the result of the test. They are only hypothesizing about reasons for the inaccuracy, and they think that the biggest factor is that women did not wait long enough, which is pretty much what I said.
I strongly doubt that these computer based tests are in any way better than analog tests, especially since the electronics just add another potential point of failure, and they do nothing to address all the other issues with pregnancy tests.
[citation needed]
https://archive.is/20081206110632/http://archfami.ama-assn.o...
> In a 1993 study from France, [...] Of the 478 positive (pregnant) urine samples distributed, 230 were falsely interpreted as negative (sensitivity, 48%)
> Valanis and Perlman [...] The incidence of false-negative results in this study was 24.3%.