A semi-quantitative multiplex PCR assay for the diagnosis of bacterial vaginosis (BV) was evaluated in a prospective study in a population of Dutch women with complaints of abnormal vaginal discharge. The PCR targets Gardnerella vaginalis, Atopobium vaginae, Megasphaera phylotype 1, Lactobacillus crispatus and Lactobacillus iners. Together with a short questionnaire, a vaginal swab for PCR and vaginal smear for microscopy were taken by their general practitioner or gynaecologist. Data from 151 women (median age 32) were available. Nugent Score (NS) was used to classify the samples and 83 samples were classified as normal (NS 0–3), 13 as intermediate (NS 4–6), and 55 as bacterial vaginosis (NS 7–10). In women with a NS of 7–10, PCR detected Gardnerella vaginalis, Atopobium vaginae and Megasphaera phylotype 1 in respectively, 96 %, 87 % and 60 %, whereas in women with a NS of 1–3 these species were detected in 27 %, 6 % and 2 % (P <0.001). A ratio of Lactobacillus crispatus over Lactobacillus iners of <1 (as calculated from the quantification cycle value (Cq)) was present in women with a NS of 7–10 in 66 % versus 33 % in women with a NS of 1–3 (P <0.001). The BV-PCR displayed a sensitivity of 92 % and specificity of 96 % with a positive predictive value of 94 % and a negative predictive value of 95 %. The Lactobacillus-index improved the correct classification of samples where only one of the other bacterial species was detected. Compared to the Nugent Score this multiplex qPCR offers a convenient tool for performing observer independent diagnosis of BV.

The performance of a multiplex real-time PCR for the detection of Blastocystis, Dientamoeba fragilis, Giardia lamblia, Cryptosporidium species and Entamoeba species in faecal samples was evaluated in an observational prospective study. Paediatric patients (0-18 years) presenting with gastrointestinal symptoms and suspected of having enteroparasitic disease were included. A questionnaire on gastrointestinal symptoms and the chosen treatment was completed at the start of the study and after 6 weeks. Of 163 paediatric patients (mean age, 7.8 years), 114 (70%) had a PCR-positive faecal sample. D. fragilis was detected most frequently, in 101 patients, followed by Blastocystis in 49. In faecal samples of 47 patients, more than one protozoan was detected, mainly the combination of D. fragilis and Blastocystis. Reported gastrointestinal symptoms were abdominal pain (78%), nausea (30%), and altered bowel habits (28%). Eighty-nine of the PCR-positive patients were treated with antibiotics. A significant reduction in abdominal pain was observed both in treated and in untreated patients. This study demonstrated that multiplex real-time PCR detects a high percentage of intestinal protozoa in paediatric patients with gastrointestinal symptoms. However, interpretation and determination of the clinical relevance of a positive PCR result in this population are still difficult.

Detection of intestinal protozoa by PCR methods has been described as being sensitive and specific, and as improving the diagnostic yield. Here we present the outcome of the transition from microscopy to molecular screening for detection of a select group of intestinal protozoa in faeces in our laboratory. Introduction of molecular screening for intestinal protozoa resulted in higher sensitivity, reduced hands-on-time, reduced time-to-results, leading to improved diagnostic efficiency.