Performance of Patient-collected Specimens for Neisseria gonorrhoeae Culture Free

Abstract

Background

Neisseria gonorrhoeae culture is necessary to determine antimicrobial resistance, but typically requires specimen collection by clinicians. We sought to determine the sensitivity of patient-collected specimens for N. gonorrhoeae culture.

Methods

We performed N. gonorrhoeae cultures on paired clinician- and patient-collected specimens from the pharynx (n = 93), rectum (n = 88), endocervix/vagina (n = 89), and urethra/urine (n = 46). We calculated the percent concordance and the kappa statistics for paired-specimen results, and determined the test sensitivity for each specimen type using positivity of either specimen in a pair as a gold standard defining the presence of true infection.

Results

At least 1 specimen was positive in 26%, 31%, 61%, and 3% of paired samples in the pharynx, rectum, urethra/urine, and endocervix/vagina, respectively. Patient- and clinician-collected results were highly concordant at the pharynx (95%; kappa = 0.85), rectum (99%; kappa = 0.97), urethra/urine (83%; kappa = 0.87), and endocervix/vagina (100%; kappa = 1.0; P ≤ .005 for all comparisons). Patient-collected pharyngeal and rectal swabs and urine were 92%, 96%, and 96% sensitive, while clinician-collected specimens at these anatomic sites were 87.5%, 100%, and 94% sensitive (P > .05 for all comparisons). Among 24 urine specimens held for 4–22 hours after collection, 100% yielded concordant N. gonorrhoeae culture results, compared to immediate processing.

Conclusions

Patient- and clinician-collected specimens are comparably sensitive for N. gonorrhoeae culture. These findings suggest that patient-collected specimens could be used to expand the availability of gonococcal antimicrobial resistance testing for both clinical and surveillance purposes.

The use of Neisseria gonorrhoeae culture has declined dramatically over the last 2 decades, since the introduction of nucleic acid amplification tests (NAAT) for the diagnosis of N. gonorrhoeae. NAAT offers many advantages over culture for the detection of gonorrhea, including improved sensitivity, specimen shelf stability at room temperature for upwards of 30 days, and relative ease of specimen collection, which allows patients to self-collect specimens [1, 2]. However, at present, defining antimicrobial susceptibility and resistance patterns of N. gonorrhoeae typically still requires culture. This ideally involves having clinicians collect specimens from the urethra, endocervix, pharynx, or rectum, with direct plating of specimens onto selective agar at the bedside and immediate placement of the agar plates into a CO2-rich environment at 37°C. These collection requirements create significant barriers to specimen collection, since they require supplies and infrastructure that are not available in most clinical settings. Moreover, the process can be both time consuming and invasive for the patient.

At the same time, rates of antimicrobial-resistant gonorrhea are rising. In 2018, nearly 5% of isolates collected as part of the Centers for Disease Control and Prevention’s (CDC) long-standing Gonococcal Isolate Surveillance Project (GISP) were resistant to azithromycin [3], 1 of the 2 main drugs used to treat gonorrhea in the United States [4]. However, GISP surveillance reports are based on <1% of all gonorrhea cases in the United States annually [3], and GISP does not include specimens from women or from extra-genital sites of infection. “Strengthening US Response to Resistant Gonorrhea” (SURGG) is a CDC project that aims to increase N. gonorrhoeae culture capacity, diversify the populations from which surveillance cultures are collected, and develop and evaluate health departments’ capacity to rapidly respond to antimicrobial-resistant cases. As part of this initiative, we sought to decrease barriers to specimen collection for gonococcal culture. In particular, we wanted to determine whether clinician-collected specimens were really necessary for N. gonorrhoeae culture. Here, we report results of an evaluation of the sensitivity of patient-collected specimens for gonococcal culture from diverse anatomic sites.

METHODS

Population and Study Procedures

The project recruited participants from patients seeking care at the Public Health–Seattle and King County Sexually Transmitted Disease (STD) Clinic between 29 May 2018 and 9 April 2019. Clinic staff asked patients to participate if they met clinical criteria for N. gonorrhoeae culture for the CDC surveillance program, SURRG. These criteria included: (1) a NAAT-positive screening test for gonorrhea in a person who had not yet been treated for their positive test; (2) seeking care because of a sexual exposure to a person diagnosed with gonorrhea; (3) a diagnosis of mucopurulent cervicitis or symptoms of pelvic inflammatory disease; and (4) symptomatic urethritis with intracellular Gram-negative diplococci seen on microscopy of a Gram-stained specimen. After hearing about the project, patients were asked to participate and were free to refuse participation with no impact on their clinical care. As a public health evaluation project designed to advance a surveillance objective, the project was exempt from Institutional Review Board approval.

Specimen collection for participants reflected standards of care in the STD clinic. For men who have sex with men who were evaluated for symptoms of STD or as contacts to gonorrhea, testing included NAAT and the culture of anatomic sites exposed in the prior 2 months (pharynx and urethra) and 12 months (rectum). Women with symptoms consistent with cervicitis or pelvic inflammatory disease, men who have sex with women only (MSW) who were evaluated for urethritis, and both women and MSW seeking care as contacts to gonorrhea were tested for genital tract infections with a vaginal NAAT and endocervical culture (women) or a urine NAAT and culture at the urethra (men). For women and MSW, pharyngeal testing was performed based on clinician discretion. We did not routinely screen asymptomatic men for urethral infection. Persons returning to clinic for treatment of an infection identified through screening underwent culture testing only at the site of their identified infection.

For patients who met the inclusion criteria and agreed to participate, clinicians collected clinical specimens as outlined above before collecting specimens for the evaluation project. This included a swab for N. gonorrhoeae culture plated directly onto modified Thayer Martin (MTM) agar and a NAAT (Aptima Combo 2). Both clinician-collected and patient-collected pharyngeal and rectal validation specimens were collected in the BBL CultureSwab plus Amies gel with charcoal [5]. The order of swab collection, either clinician-collected first followed by patient-collected or vice versa, was randomly assigned. We used randomly generated, preprinted, deidentified labels to indicate to clinicians the order in which to collect project specimens. Male contacts and those with symptomatic urethritis provided urine as the patient-collected specimen after undergoing a clinician-collected urethral swab plated on MTM agar. Because women required a pelvic exam for a clinician-collected endocervical culture, the order of specimen collection was determined by the clinician to optimize the clinic flow. Specimens collected from women included a clinical endocervical swab on MTM agar, a clinician-collected endocervical swab in the BBL collection kit, and a patient-collected vaginal swab in the BBL collection kit. We aimed to collect 100 paired specimens from each anatomic site, hypothesizing a 90% concordance, which would provide a 95% confidence interval (CI) of 82–95%. Urine and BBL CultureSwab specimens were transported to the University of Washington Neisseria Reference Laboratory for processing.

In April 2019, due to low numbers of female patients at the Public Health–Seattle and King County STD Clinic, an additional SURRG grantee site, Guilford County Public Health in Greensboro, NC, began collecting paired specimens from women. Women attending the STD clinic and undergoing a routine gonococcal culture collected for clinical purposes were invited to participate, similar to the procedures outlined above. The order of patient and provider specimen collection was randomized by visit time. That is, each day clinicians designated either the morning or afternoon clinic session to collect endocervical specimens first. Women self-collected a vaginal culture using the BBL collection kit, and the clinician-collected endocervical specimen was plated on MTM.

Laboratory Procedures

The University of Washington Neisseria Reference Laboratory processed BBL CultureSwab and urine specimens for N. gonorrhoeae culture within 1 hour of specimen collection, and recultured urine specimens between 4 and 22 hours later, after storage at room temperature, to determine the impact of delayed urine processing on N. gonorrhoeae viability. BBL CultureSwab specimens were inoculated on MTM agar plates (Thermo Fisher Scientific Remel, Lenexa, KS) and streaked for isolation. We cultured urine sediment for N. gonorrhoeae, as described previously [6, 7]. Briefly, we centrifuged 40 ml aliquot of urine at 4000 × G for 15 min. We carefully decanted the supernatant fluid, used Dacron swabs to inoculate urine sediment on MTM agar plates (Thermo Fisher Scientific Remel, Lenexa, KS), and streaked the plates to isolate individual bacterial colonies. The plates were incubated at 36°C in 5% CO2. All Thayer-Martin culture plates were examined for colonies typical of N. gonorrhoeae daily for up to 72 hrs. Cultures were identified and confirmed as N. gonorrhoeae by characteristic gonococcal colonial morphology, standard biochemical tests, and 3 species-confirmatory tests: an acid detection test, the PhadeBact GC monoclonal test (MKL Diagnostics AB, Sollentuna, Sweden), and API NH (bioMérieux, Marcy l’Etoile, France), as described previously [8]. Clinical culture specimens—that is, those plated directly on MTM at the bedside—were processed at the Public Health–Seattle and King Count Laboratory using standard methods.

The Guilford County Public Health SURRG Laboratory processed BBL vaginal CultureSwab and MTM-plated specimens for N. gonorrhoeae culture within 1 hour at the onsite public health laboratory. The plates were incubated at 36°C in 5% CO2. All culture plates were examined for colonies typical of N. gonorrhoeae daily for up to 48 hrs. Cultures were identified and confirmed as N. gonorrhoeae by characteristic gonococcal colonial morphology, Gram stain, oxidase reaction, BactiCard Neisseria (Thermo Fisher Scientific Remel, Lenexa, KS), and API NH identification system.

Statistical Analyses

We compared paired clinician- and patient-collected culture results using 2-by-2 tables and calculated the percent concordance, associated Kappa statistic, and 95% CI for each anatomic site. At the pharynx and rectum, we compared cultures collected by both the patient and clinician using the BBL swab kit. At the female urogenital tract, we compared Seattle women who had a clinician-collected endocervical swab and patient-collected vaginal swab, both using BBL collection kits, and North Carolina women who had a patient-collected vaginal BBL swab and a clinician-collected swab plated on MTM. The recovery of N. gonorrhoeae by urine culture was compared to a clinician-collected urethral swab plated directly onto MTM plates. We interpreted the Kappa Statistic of ≤0.59 as unreliable or weak,  >0.59–0.79 as moderate, and >0.79 as strong and very reliable [9]. Additionally, we calculated the sensitivity and 95% CIs of both patient-collected and clinician-collected specimens using a gold standard of either specimen testing positive for N. gonorrhoeae. For urine specimens, we also evaluated the recovery of N. gonorrhoeae after delayed processing 4–18 hours after specimen collection, compared to specimens that underwent immediate processing. All analyses were conducted in Stata version 15.0 (StataCorps, College Station, TX).

RESULTS

Study Participants and Specimens Provided

A total of 258 individuals provided 93 paired pharyngeal, 88 paired rectal, 46 paired urethral/urine, and 89 paired vaginal specimens. Most (n = 209; 81%) individuals provided specimens from only 1 anatomic site. There were 40 (16%) and 9 (3.5%) participants who provided specimens from 2 and 3 anatomic sites, respectively.

Performance of Self-collected Pharyngeal Specimens

Of the 93 paired pharyngeal specimens included in the analysis, 24 (26%) tested positive by either the patient- or clinician-collected swab (Table 1). There were 19 infections (79%) that were positive by both collection methods. There were 5 pairs with discordant results: 2 clinician-collected specimens were positive when the patient-collected specimen was negative, and 3 patient-collected specimens were positive when the clinician-collected specimen was negative. In these 5 cases, the accompanying NAAT, either from the same day or <10 days prior, were positive. Overall, the percent concordance between patient- and clinician-collected culture specimens was 95% (95% CI, 88–98%) and the Kappa statistic was strong at 0.85 (95% CI, .72–.98; Table 1). The sensitivity of patient-collected pharyngeal specimens (Table 2) for gonococcal culture was 92% (95% CI, 73–99%)

Performance of Self-collected Rectal Specimens

A total of 88 paired rectal specimens were collected. There were 26 pairs (30%) that were positive by both patient- and clinician-collected methods. Only 1 pair had a discordant result, and was positive by clinician collection yet negative by patient collection. This corresponds to 99% concordance (95% CI, 94–100%), and a Kappa statistic of 0.97 (95% CI, .92–1.00). Patient-collected rectal specimens for the culture of N. gonorrhoeae were 96% sensitive (95% CI, 81–100%).

Performance of Male Urine

Of the 46 paired urine and urethral cultures, 28 (61%) were positive by at least 1 method, 25 (54%) were positive by both methods, 1 (2%) was positive only by urethral swab culture, and 2 (4%) were positive by urine but not by urethral culture. All of the discordant results (n = 3) were Gram-stain positive for Gram-negative diplococci and NAAT positive for N. gonorrhoeae on the same day. Overall, concordance was 93% (95% CI, 82–99%), with very strong agreement (Kappa, 0.87; 95% CI, .72–1.0). Urine for N. gonorrhoeae culture was 97% sensitive (95% CI, 83–100%). There were 24 urine samples that were saved and recultured between 4 and 8 hours (n = 17; 71%) and between 16 and 22 hours (n = 7; 29%) after collection. Comparing culture results from the immediate urine processing to delayed processing resulted in 100% concordance (95% CI, 86–100%).

Performance of Self-collected Vaginal Specimens

Out of a total of 89 pairs of patient-collected vaginal specimens and clinician-collected endocervical specimens, 10 were collected in Seattle and 79 were collected in North Carolina. All 3 cases of gonorrhea in the female genital tract were positive by both collection methods (100% concordance; 95% CI, 96–100%; Kappa statistic, 1.0): 1 of these cases occurred in North Carolina and 2 were in the Seattle cohort.

DISCUSSION

Removing barriers to N. gonorrhoeae culture collection is an important step to increase the use of culture in an era of growing antimicrobial-resistant gonorrhea. We found that clinician-collected and patient-collected rectal and pharyngeal specimens were comparably sensitive, and that urine and urethral swab specimens were likewise equivalent in the detection of N. gonorrhoeae by culture. Moreover, delayed specimen processing did not compromise the sensitivity of urine culture, suggesting that these specimens were relatively robust, and that test performance is not dependent on the availability of very rapid processing. These findings demonstrate that obtaining specimens for culture and antimicrobial-resistance testing for N. gonorrhoeae can be substantially simplified, allowing for an expansion of the number of persons tested for both for surveillance and clinical purposes.

At present, GISP analyzes the antimicrobial susceptibility profiles of <1% of all diagnosed cases of N. gonorrhoeae in the United States, and this data forms the basis of the national treatment guidelines for gonorrhea [3]. At the same time, GISP excludes isolates from women and extragenital sites, which may have different patterns of susceptibility. Given the current, and appropriate, reliance on NAAT as the primary diagnostic method for gonorrhea, the most obvious way to rapidly increase the proportion of cases that undergo antimicrobial surveillance would be through molecular methods. However, that technology is not yet ready to replace culture-based surveillance. Moreover, even when molecular surveillance can be used as the primary surveillance of antimicrobial-resistant N. gonorrhoeae, it must be paired with phenotypic results in order to monitor the emergence of novel genetic mutations. Some experts have estimated that in order to identify novel mutants, antimicrobial-resistance surveillance systems would need to include the phenotypic characterization of at least 3% of incident cases [10]. To achieve this level would require more than tripling the number of cultures currently performed as part of GISP. Patient-collected specimens for culture are just 1 step in the process of scaling up and diversifying N. gonorrhoeae antimicrobial surveillance.

A growing body of literature supports the use of self-collected specimens for gonorrhea and chlamydia testing using NAAT. In addition to saving busy clinics and patients time, patients prefer self-collection to clinician collection [11]. Given that there are virtually no differences in these methodologies from the perspective of the patient, we can assume that patients would prefer to self-collect cultures, though we did not capture this data. We can, however, report that nursing staff find self-collected cultures acceptable. Following this validation, we encouraged 1 of our SURRG community partner clinics to employ patient-collected specimens. In this clinic, the nursing team is responsible for collecting N. gonorrhoeae cultures at the time of gonorrhea treatment. Since the adoption of patient-collected cultures, the clinic has more than tripled the average number of cultures collected per month. With a self-collected culture program, providers can remove the need for a clinician visit for a speculum exam or urethral swab at the time of treatment, which is particularly appealing for those individuals who screened NAAT positive. Moreover, a self-collected culture program can be routinized, linked to ceftriaxone administration, and fully implemented by the nursing staff, which in turn increases the likelihood of culture collection.

Beyond culture’s role in antimicrobial-resistance surveillance, increasing access to N. gonorrhoeae cultures in primary care clinics and emergency departments would be valuable for both the clinical management of suspected treatment failure and the rapid detection of clinically significant resistance. The CDC currently recommends that those patients with gonorrhea whose symptoms do not abate at 3–5 days following recommended treatment with a third-generation cephalosporin and who deny reexposure to gonorrhea should undergo repeat testing with culture [12]. However, access to N. gonorrhoeae cultures is often limited to specialty STD clinics. Our hope is that by reducing barriers to obtaining cultures, more clinical settings will employ them.

The primary limitation of our findings relates to the small number of women with gonococcal infection included in the project. Only 3 women we tested were infected with N. gonorrhoeae. Given this small number, we are not able to draw conclusions about the adequacy of self-collected vaginal swabs compared to endocervical cultures. Additionally, the robust performance of patient-collected cultures reported in this paper may be due in part to our laboratory’s extensive experience working with N. gonorrhoeae, which is a notoriously fastidious organism. Reproduction of these results is warranted prior to more widespread use of this specimen collection method.

In summary, patient-collected specimens for the culture of N. gonorrhoeae are feasible and are comparably sensitive to more traditional, clinician-collected specimens. This new method of obtaining specimens for N. gonorrhoeae culture provides opportunities for enhanced antimicrobial-resistant gonorrhea surveillance, and the simplification of specimen collection for clinical purposes.

Notes

Acknowledgments. The authors thank the Public Health–Seattle and King County Sexually Transmitted Disease Clinic clinicians; Allison Rollins and Rushlenne Pascual from the University of Washington Neisseria Reference Laboratory; and Cindy Toler, Oana Dobre-Buonya, and Vonda Pabon from Guilford County Health Department (GCHD) laboratory and clinical services.

Financial support. This project was supported by funding from the U.S. Centers for Disease Control and Prevention’s (CDC) Epidemiology and Laboratory Capacity for the Prevention and Control of Infectious Diseases (ELC) Cooperative Agreement [CK19-1904] under the Strengthening the U.S. Response to Resistant Gonorrhea (SURRG) project.

Potential conflicts of interest. M. R. G. reports grants from Hologic, outside the submitted work. L. A. B. reports the in-kind donation of test kits for studies from Hologic, not related to this work; grants from SpeeDX, outside the submitted work; and consultancy work from Nabriva, unrelated to this work. C. J. M. reports grant BTZ116577 from GSK and grants from Atlas Genetics/Binx, outside the submitted work. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References