Participants, specimen and data collection, and quality control
We performed a cross-sectional study of 500 adult patients presenting for care to Mbarara Regional Referral Hospital (MRRH), a 608-bed hospital located in Mbarara, approximately 265 km southwest of Uganda’s capital Kampala. The patient population resides in a predominantly rural, agricultural area. Between January to October 2015, one hundred patients were enrolled consecutively from each of five hospital departments: inpatient surgery ward, in-patient medicine ward, in-patient maternity ward, general out-patient department, and the out-patient Immune Suppression Syndrome (ISS) Clinic, which exclusively serves patients living with HIV until the enrollment target was reached and then enrollment was begun in the next department. Adult participants ≥18 years of age cared for at any of the five hospital departments were eligible for enrollment. Exclusion criteria were age < 18 years, inability to speak English or the local language Runyankole, and inability for the participant or their next-of-kin to provide informed consent. Participants included were part of a clinical study assessing prevalence and correlates of MRSA and MSSA nasal carriage at a Ugandan regional referral hospital [10]. Study participants had their bilateral anterior nares swabbed with two sterile Dacron-tipped dual-swabs representing a combined sample of both nares. Each Dacron-tipped collection device has two swabs giving a total of four swabs. Three were used for the MRSA detection methods, and the remaining swab was kept refrigerated for retesting. Swabs were transported same-day to the Epicentre laboratory on-site for refrigeration and processing within 24 h. Quality control procedures were performed using commercially available bacteria strains ATCC 33591 MRSA and ATCC 25923 MSSA on each new lot of testing kits. All samples producing invalid and erroneous results were repeated once. Demographic characteristics were captured using investigator-designed questionnaires.
BD CHROMagar MRSA II
One swab (Copan Tran-system Liquid Amies double swab) was inoculated directly onto nutrient-enriched selective agar media (CHROMagar MRSA II, BD Diagnostic Systems, Sparks, USA) and incubated under aerobic conditions at 37 °C for 24 h according to the manufacturer’s instructions. After 24 h, each plate was examined for colony growth and color. All mauve colonies underwent further identification including Gram stain and coagulase testing to be confirmed as MRSA. Culture-negative plates were further incubated and examined at 48 h and discarded if negative.
Xpert SA nasal complete
The second swab was processed for the Cepheid Xpert SA Nasal Complete assay (Cepheid, Sunnyvale, USA) according to the manufacturer’s instructions and run on the Cepheid Xpert SA Complete platform. Fluorescent signals of target DNA, SA-spa/MRSA-mecA and mecA and SCCmec) were measured and results were provided automatically by the GeneXpert machine. Results were reported positive for S. aureus if the spa gene was detected above threshold limits, and samples were reported MRSA-positive if spa, mecA and SCCmec genes were all detected above threshold limits. The minimum cycle threshold (Ct) detection limit for all genes was a Ct of 10, and the maximum Ct detection limit for spa, mecA and SCCmec was 35, 36, and 38, respectively.
Hain GenoQuick MRSA
The third swab was run on the Hain GenoQuick MRSA assay (Hain Life Science, Nehren, Germany) according to the manufacturer’s instructions. A Lack of MRSA-positive samples in the first 100 swabs and documented 30-fold differences in level of detection led study investigators to optimize the Hain GenoQuick MRSA amplification procedure [11]. The remaining 400 samples were tested under a revised protocol where double the amount of DNA lysate was used, improving the test’s level of detection. As a result of logistical shortage of Hain GenoQuick kits, only 70 of the initial 100 samples were retested using the revised protocol, and 470 total sample results are reported. Results were then interpreted as positive or negative for MRSA for the assay according to the manufacturer’s instructions.
Data entry and analysis
Demographic data and microbiology results were manually entered into a REDCap database hosted at Partners Healthcare in Boston, USA [12]. Analyses were performed in Stata version 12 (StataCorp, College Station, USA). As our primary interest was in assessing test sensitivity as the most relevant measure for infection control purposes, we created a composite reference standard (CRS), defining a sample as positive by the CRS if it tested positive by one or more individual tests: 1) BD CHROMagar, 2) Cepheid Xpert SA Nasal Complete, 3) Hain GenoQuick. We then calculated sensitivity of each test compared to the CRS. We carried out activity/task analysis by observation of these three test methods to define outcomes, including cost, time, and logistical needs.
