The efficacy of PCR was gauged at times against culture-positive cases, and I don't think it did all that well. I think its range was 10 to 60%?
Well not as found in the review I referenced above. Here are a couple of quotes.
When skin biopsy specimens are first cultured and the culture
material is analyzed by PCR, the test sensitivity should be
higher because of the multiplication of organisms in culture. In
one such study in which a nested PCR targeting the
ospA gene
was used, 8 of 10 patients with EM were positive; the interval
between the tick bite and the presentation of the patients
varied between 1 week and 3 months. The sensitivity of PCR
(80%) was considerably higher than that of culture (14%) or
serologic testing (77%) (116).
Recently, plasma from 76 patients with EM was analyzed in
a prospective, blinded study of PCR (51). Spirochetemia could
be documented in 14 patients (18.4%) by PCR (Table 4); only
4 (5.3%) were culture positive. Positive PCR results correlated
with clinical evidence of disseminated disease: 10 of 33 patients
with systemic symptoms (30.3%) were positive compared with
4 of 43 without such evidence (9.3%).
The sensitivity of culture is lower than that of PCR in all
studies. Even when PCR was performed on cultures of skin
biopsy specimens from patients with EM, culture was less sensitive
than PCR (161). After 2 weeks of culture, PCR was
reactive in 19 of 20 patients compared with 14 of 20 patients by
microscopic inspection.
No positive cultures have ever been found in CSF samples
from patients with late disease (93), and cultures from SF
succeeded in only a few patients with LA (123). Therefore, in
a patient with a negative culture and a positive PCR result, the
likelihood of a false-negative culture is much higher than that
of a false-positive PCR result.