We performed a field evaluation of polymerase chain reaction (PCR)-based en
zyme-linked immunosorbent assays (ELISA) for the diagnosis of malaria. A co
mmercially available PCR-ELISA microplate hybridization (MPH) assay was use
d. Blood specimens were collected from 300 volunteers seeking care at malar
ia clinics in Thailand. Examination of 200 high power fields by Giemsa-stai
ned thick and thin smear (GTTS) revealed 51 P. falciparum (Pf), 45 P. vivax
(Pv), seven mixed Pf-Pv infections. These plus a random sample of 48 GTTS-
negative specimens were selected for this study. All 151 specimens were pro
cessed for parasite DNA extraction and assayed by PCR-MPH. The target DNA s
equence of the 18S small subunit ribosomal RNA (SSUrRNA) gene was amplified
by PCR and hybridized with species-specific probes for Pf, Pv, P. malariae
(Pm) and P. ovale (Po) immobilized in the wells of the microtiter plate an
d detected by colorimetric assay. Colour development was assessed at an opt
ical density (OD) of 405 nm. An absorbance reading of greater than or equal
to 0.1 was used as a positive cut-off. In comparison with GTTS results, PC
R-MPH sensitivity was 91.4% (53/58, 95% CI 84.2-98.6) for Pf, 94.2% (49/52,
87.9-100) for Pv and specificity was 95.8% (46/48, 95% CI 90.2-100). There
was statistically significant positive correlation between parasite densit
ies less than or equal to 7000/mul blood and absorbance reading, suggestive
of PCR-MPH being semiquantitative. PCR-MPH also detected additional Pf and
Pv cases as well as Pm and Po.