Diagnostic methods to determine microbiology of postpartum endometritis in South Asia: laboratory methods protocol used in the Postpartum Sepsis Study: a prospective cohort study

Blood was collected for complete blood count, culture, and to determine the presence or absence of malarial parasites. The purpose of each of these tests was unique and has been described in Table 1.

To evaluate for leukocytosis and leukopenia as possible markers of sepsis [7], blood counts were determined by flow cytometry at central laboratories affiliated with field sites. Blood samples of at least 1 ml were collected in ethylene diamine tetra-acetate (EDTA) tubes and transported to laboratories at 2–8 °C within 6 h of collection.

Thin and thick blood smears were made and examined at central laboratories to identify malarial parasites and to quantify and speciate Plasmodia, if present.

Aerobic and anaerobic blood cultures were performed in the BACTEC 9240® instrument with BACTEC Aerobic Plus® and Anaerobic F Lytic bottles. Five to ten ml of blood was collected in each bottle and transported to respective central laboratories for incubation and processing. Those flagged positive by the instrument were removed and gram stain and agar subculture was performed.

Endometrial tissue was collected using Tao sampling brushes by Cook Medical [8]. Sterile brushes (Fig. 1) were opened tail-end first immediately after insertion of the Cusco speculum and visualization of the cervix. Manufacturer’s instructions were followed to collect the endometrial tissue on the brush. Tao brushes are designed to collect endometrial tissue for cytology and histopathology. In using the same brush for culture, we modified the instructions to avoid vaginal contamination. The rounded tip at the sampling end was snapped off with sterile scissors. The brush was then cut into 2 halves and one half was used for endometrial culture and the other reserved for molecular detection of additional potential pathogens (bacteria not readily cultivable). A supplementary video of collection procedures as a mock demonstration is available as Supplementary File 1.

One half of the brush was aseptically cut and placed into a thioglycolate broth medium with cooked meat to facilitate recovery of strict anaerobes. This tube was transported to the central laboratory at 25 °C within 6 h of collection. Upon receipt in the laboratory, the broth was vortexed at high speed to dislodge tissue from the brush and used to inoculate chocolate, MacConkey, and 5 % sheep blood agar (SBA) media. To recover strict anaerobes, pre-reduced Brucella agar with kanamycin and vancomycin supplements was inoculated with the thioglycolate broth [9]. Agar incubation temperatures, sources, and culture methods and result parameters are described in detail in Table 2. Results were recorded semi-quantitatively as light (few colonies), medium, or heavy growth. Identification of gram-negative organisms was carried out with commercial API systems (BioMerieux) strips, (API 20E for enteric gram negative bacteria and API NE for non-fermenting gram negative bacteria). Identification of gram positive bacteria was based on the Clinical Laboratory Standards Institute (CLSI)-recommended tests [10], such as tube coagulase for S.aureus and bacitracin disk and Lancefield grouping for S.pyogenes and S.agalactiae. Commercial systems such as API Strep and API-NH (BioMerieux) were used for fastidious organisms, such as alpha haemolytic, nutritionally variant Streptococci, and Haemophilus spp.

The other half of the endometrial brush was introduced into a Universal Transport Medium (UTM) tube (Copan) and transported to the laboratory at 2–8 °C.

The presence of Chlamydia trachomatis, Mycoplasma spp, and Ureaplasma spp was detected by molecular methods. For specimens collected in Pakistan, the Sacace Biotechnologies (Como, Italy) kit [11] was used to identify the presence of C.trachomatis, M.genitalium, M.hominis, and U.urealyticum at Aga Khan University’s Infectious Disease Research Laboratory. Endometrial specimens obtained in Bangladesh were shipped to University of Alabama (UAB) for culture of mycoplasma and ureaplasma and PCR detection for Mycoplasma and Ureaplasma using previously published methods [12].

In Pakistan, high vaginal swabs were collected after endometrial brush specimens in an attempt to exclude vaginal contaminating flora and to evaluate for presence of vaginitis on the basis of presence or absence of clue cells. Samples were collected from the high vaginal wall in Ames transport medium (Medical Wire Transwab Ames, UK) and sent to the laboratory. Samples were plated on chocolate agar, Gardnerella agar (with 5 % human blood), Maconkey agar, and Sabouraud’s agar. Results were recorded semi-quantitatively as light, medium, or heavy growth.

Clean catch urine was collected to determine the etiology of urinary tract infection (UTI) in patients with likely UTI (significant growth of uropathogen with positive nitrite and leukocyte esterase on dipstick test) [13].

Dipstick tests, indicating the presence or absence of both leukocyte esterase and nitrite [12] were performed to identify supportive evidence of UTI. Quantitative cultures of urine were performed for growth of uropathogens on SBA and MaConkey’s agar media, and ≥ 10³ colony-forming unit (CFU) of bacteria per mL of urine were considered significant. Identification of uropathogens was carried out using either the minimal sulfide-indole-motility-citrate-urease-triple sugar iron (SIMCUT) method or commercial API (BioMerieux) strips.

Susceptibility testing by the disk diffusion method was carried out only for aerobic or facultatively anaerobic pathogens in accordance with CLSI guidelines. Where disk diffusion was not the preferred method, broth dilution minimum inhibitory concentrations (MICs) were carried out to confirm susceptibility (eg for methicillin resistant S.aureus, vancomycin MICs were performed by the broth dilution method) [14].

At the Karachi site, dengue IgM and Rapid ICT malaria (PanBio) tests were additionally performed to detect dengue fever, and to improve detection of submicroscopic malaria infections. The city suffered major dengue outbreaks since 1994 [15], and continues to experience annual outbreaks [16].