Sexually Transmitted Diseases  >  Screening Tests  >  Page 3

Screening Tests To Detect Chlamydia trachomatis and Neisseria gonorrhoeae Infections - 2002

Previous Page

Table of Contents

Next Page

Section on this page:

• Selecting Screening Tests

Selecting Screening Tests

Multiple considerations affect the selection of a screening test. Test sensitivity is emphasized to minimize occurrence of false-negative tests, which can result in complications of untreated infection and ongoing transmission. However, additional considerations might lead to the selection of a different test (see Additional Considerations in Selecting a Screening Test ). For example, a goal of maximizing test sensitivity to avoid missing the opportunity to treat infected persons might warrant tolerating a limited number of false-positive diagnoses. However, consideration must also be given to reducing the rate and consequences of false-positive tests and to cost (see Methods To Enhance Performance or Reduce Costs) . This report focuses on screening applications of tests for C. trachomatis and N. gonorrhoeae infections. This report also contains a listing of additional indications for C. trachomatis testing and recommendations for choice of test and type of specimen (Appendix A); similar information for N. gonorrhoeae is also included (Appendix B).

Performance Perspective for Selecting Screening Tests

On the basis of sensitivity, ease of specimen collection, and ability to assess antimicrobial susceptibility (N. gonorrhoeae), recommendations for screening women and men for C. trachomatis and N. gonorrhoeae genitourinary tract infections are outlined in this section. Recommended screening tests will, compared with less sensitive tests, minimize the risk for disease sequelae and continued transmission of infections as a result of false-negative screening tests. Clinicians should be aware of the potential for adverse consequences caused by a false-positive test result (e.g., substantial psychosocial or legal consequences); patients with positive results should be counseled regarding the potential for false-positive results, and additional testing should be considered. Because such a result might itself be falsely negative and therapies for C. trachomatis and N. gonorrhoeae are safe and effective, treatment might be offered while awaiting results from additional testing or even if an additional test is negative. At lower prevalences, consideration should be given to routine additional testing after a positive screening test. Testing strategies have been proposed that can increase specificity and constrain costs by combining use of NAATs and non-NAATs. A particularly promising strategy called gray-zone testing involves screening with a non-NAAT and using a lower cutoff value than that established by the manufacturer as a criterion for a positive result. A NAAT is then performed as an additional test when the non-NAAT screening test results are in a zone above the new cutoff value. This strategy could achieve greater sensitivity than would be provided by using a non-NAAT by itself and greater specificity than would be provided by the separate use of either a non-NAAT or a NAAT, and cost less than using a NAAT by itself. This strategy might be useful for screening among lower prevalence populations for which both PPV and cost of detecting an infection are of increased concern. This strategy warrants further evaluation (see Additional Considerations in Selecting a Screening Test and also, Methods To Enhance Performance or Reduce Costs) .

Screening Women for C. trachomatis and N. gonorrhoeae Genitourinary Tract Infections

The following recommendations are based on sensitivity, ease of specimen collection, and ability to assess antimicrobial susceptibility (N. gonorrhoeae) (Box 1). Additional concerns, including cost and PPV, are addressed in subsequent sections (see (see Additional Considerations in Selecting a Screening Test and also, Methods To Enhance Performance or Reduce Costs) .

BOX 1. Tests used for screening women for genitourinary tract infection

C. trachomatis.

1. A C. trachomatis NAAT performed on an endocervical swab specimen provides the highest sensitivity and might be preferred if a pelvic examination is acceptable; otherwise, a NAAT can be performed on urine. The sensitivity of NAATs when using urine to detect C. trachomatis in women is similar, or only slightly inferior, to their sensitivity when using endocervical swabs (78 ,86). By using urine as a specimen, screening persons in such venues as family planning clinics can now be extended beyond women undergoing pelvic examinations. Screening can be performed also in nontraditional venues where pelvic examinations are not performed (e.g., schools, jails and detention centers, and street HIV or STD outreach programs).
2. Although less sensitive than NAATs, unamplified nucleic acid hybridization tests, EIAs, and DFA performed on an endocervical swab specimen are acceptable for screening (see Methods To Enhance Performance or Reduce Costs) .
3. C. trachomatis culture performed on an endocervical swab specimen is a suitable test for screening. Specificity of culture tests is high; however, sensitivity is less than for NAATs and variable because of technical complexity, lack of standardization, and the challenge of maintaining viable organisms.

No tests have been cleared by FDA for use with vaginal specimens. However, studies have assessed using vaginal specimens for screening by NAATs (37--49), including vaginal specimens collected by the patient (37 ,38 ,40 ,42 ,44 ,47--49). The results of these studies are promising. Unless manufacturers obtain FDA clearance for this intended use, individual laboratories testing vaginal specimens will need to conduct a study to establish, rather than just to verify, test performance characteristics to satisfy CLIA requirements (Appendix C). The sensitivity of non-NAATs with urine or vaginal swab specimens is suboptimal. In the case of EIAs, specificity with vaginal swab and urine specimens is also lower than with endocervical swab specimens (65). Such uses of non-NAATs are not recommended.

N. gonorrhoeae.

1. An N. gonorrhoeae culture test performed on an endocervical swab specimen might be preferred on the basis of ease of additional testing and on the continuing need for antimicrobial-resistance monitoring. When culture sensitivity might be compromised because of problems in maintaining appropriate transport and storage conditions, a NAAT or nucleic acid hybridization test can be performed on an endocervical swab specimen.
2. As with C. trachomatis, screening for N. gonorrhoeae can be extended to nontraditional clinical settings by using a NAAT performed on urine. The sensitivity of NAATs to detect N. gonorrhoeae might be less when using urine than when using an endocervical swab specimen (54--56). This decrease might be greater for certain NAATs than others (54--56).

As with C. trachomatis, no tests are FDA-cleared for use with vaginal specimens, although a study that assessed using patient-obtained vaginal specimens for screening by NAATs yielded promising results (50). Unless manufacturers obtain FDA clearance for this intended use, laboratories testing vaginal specimens will need to conduct a study to establish, rather than just to verify, test performance characteristics to satisfy CLIA requirements (Appendix C). The sensitivity of non-NAAT tests to detect N. gonorrhoeae with urine or vaginal swab specimens is suboptimal. Such uses of non-NAATs are not recommended.

Screening Men for C. trachomatis and N. gonorrhoeae Urethral Infections

This section includes recommendations for tests for screening men for C. trachomatis and N. gonorrhoeae urethral infections. These recommendations are based on sensitivity, ease of specimen collection, and ability to assess antimicrobial susceptibility (N. gonorrhoeae) (Box 2). Additional concerns, including cost and PPV, are addressed in subsequent sections (see Additional Considerations in Selecting a Screening Test and Methods to Enhance Performance or Reduce Costs).

BOX 2. Tests used for screening men for urethral infection

C. trachomatis.

1. A NAAT for C. trachomatis performed on an intraurethral swab or urine specimen is the preferred test. According to STD specialists, NAATs might be more sensitive for detecting C. trachomatis infection of asymptomatic men when performed on an intraurethral swab specimen than on urine, although limited published evaluations demonstrate no superiority for swab specimens (77 ,87 ,88). However, intraurethral swab specimens are often unacceptable to males who do not have symptoms of urethritis (89--91). NAATs performed on urine to detect C. trachomatis have adequate sensitivity to be used for screening. The sensitivity of non-NAATs to detect C. trachomatis in urine from men is too low for them to be recommended for screening.
2. A non-NAAT or culture for C. trachomatis performed on an intraurethral swab specimen is acceptable.

N. gonorrhoeae.

1. Culture performed on an intraurethral swab specimen is preferred in settings where collecting such specimens is acceptable and transport and storage conditions are suitable for culture. Collection of exudate for N. gonorrhoeae culture at the urethral meatus without insertion of the swab into the urethra is sufficient when exudate is present. Culture performed on an intraurethral swab specimen is preferred because it facilitates additional testing, which is critical for monitoring for antimicrobial resistance. When culture sensitivity might be compromised because of problems in maintaining appropriate transport and storage conditions, a NAAT or nucleic acid hybridization test can be performed on an intraurethral swab specimen.
2. A NAAT or nucleic acid hybridization test performed on an intraurethral swab specimen or a NAAT performed on urine are acceptable tests for screening. Intraurethral swab specimens are often unacceptable to males who do not have symptoms of urethritis (89--91). In such cases, NAATs performed on urine to detect N. gonorrhoeae have adequate sensitivity for use in screening. Only a limited number of published studies have addressed whether NAATs perform better with urethral swab specimens than with urine in males with asymptomatic infection. One study of asymptomatic males reported that a NAAT
3. was substantially less sensitive with urine than with urethral swabs; furthermore, NAATs were less sensitive with either specimen among males with asymptomatic infection than with symptomatic infection (54, 56).

Screening Women or Men with Possible Rectal or Pharyngeal Exposure to C. trachomatis or N. gonorrhoeae Infection

Selecting tests for screening for C. trachomatis and N. gonorrhoeae rectal or pharyngeal infections is difficult because of limited experience with nonculture tests and increased potential for cross-reactivity with other organisms (Box 3).

BOX 3. Tests for screening women or men for rectal or pharyngeal infection

C. trachomatis.

1. Culture isolation is acceptable for detecting C. trachomatis in rectal or pharyngeal swab specimens. A C. trachomatis-MOMP--specific stain should be used with culture of rectal or pharyngeal specimens to detect C. trachomatis. Less-specific inclusion-detection methods by using EIA, iodine, Giemsa, or LPS-specific monoclonal antibodies, are not recommended (29,30).
2. DFA can be performed on rectal or pharyngeal swab specimens. A C. trachomatis-MOMP-specific stain should be used.

N. gonorrhoeae.

1. Culture isolation is well established for detecting N. gonorrhoeae in rectal and pharyngeal swab specimens. After isolation from rectal or pharyngeal specimens of typical oxidase-positive, Gram-negative diplococci on selective media, definitive identification of N. gonorrhoeae requires additional testing because of the common occurrence of other pathogenic and nonpathogenic Neisseria species in the pharynx and their occasional occurrence in the rectum (12 ,92 ,93) (see Methods To Enhance Performance or Reduce Costs) .

If the foregoing methods are not available, screening for C. trachomatis and N. gonorrhoeae in rectal or pharyngeal specimens is not recommended. Only limited evaluations have been published of nonculture tests to detect C. trachomatis or N. gonorrhoeae oropharyngeal or rectal infections (32 ,94 ,95).

Additional Considerations in Selecting a Screening Test

In addition to test sensitivity, ease of specimen collection, and assessment for antimicrobial susceptibility (N. gonorrhoeae), other considerations in choosing a screening test include 1) the relatively high cost of NAATs (i.e., economic considerations); 2) laboratory environmental changes necessary to implement NAATs; 3) the need for additional testing to support C. trachomatis or N. gonorrhoeae diagnoses; and 4) the likelihood of screening-test--positive persons returning for treatment. The ability of some tests to detect C. trachomatis and N. gonorrhoeae in the same specimen might also affect the choice of test (see Methods To Enhance Performance or Reduce Costs). Recommendations for transporting and storing specimens must also be considered (Appendix D).

Economic Analysis

Economic analyses can assist in selecting the optimal tests and strategy for C. trachomatis and N. gonorrhoeae screening. Moreover, economic analyses can be relevant for STD screening because, in the absence of symptoms or signs, the cost of screening tests, including the costs of labor and other nonkit costs, is often not reimbursed by insurance plans or other payers. Thus, knowing the costs and benefits of a screening program that might not be fully reimbursed can be critical to decision makers (Box 4).

Box 4. Performing an economic analysis

* Sources: Haddix AC, Teutsch SM, Shaffer PA, Duñet DO. Prevention effectiveness: a guide to decision analysis and economic evaluation. New York, NY: Oxford University Press, 1996. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New York, NY: Oxford University Press, 1996.

† Other common economic analyses include cost-utility and cost-benefit. Cost-utility analysis converts cost-effectiveness outcomes to quality-adjusted life years (QALYs) or disability-adjusted life years (DALYs). Cost-utility analyses are rare in sexually transmitted diseases (STD) because the QALY or DALY impacts of STD and their sequelae are unresearched. Cost-benefit analysis converts the outcomes achieved by a program to dollar values so that the effects of different programs (e.g., STD screening and environmental health) can be directly compared (Source: Haddix AC, Teutsch SM, Shaffer PA, Duñet DO. Prevention effectiveness: a guide to decision analysis and economic evaluation. New York, NY: Oxford University Press, 1996).

§ For example, if a program incurs $1,000 for screening and treating women for C. trachomatis and, therefore, avoids $2,000 in treatment costs for PID, the program is cost-saving. However, the program might still be cost-effective even if the program does not avert enough sequelae costs to be cost-saving. A program that has a net cost (program costs minus averted sequelae costs) of $500/case of PID averted might be considered cost-effective, compared with the first program, although it is more expensive, if it prevented more cases. The determination of whether such a program is cost-effective requires that a judgement be made regarding the value of the health outcomes in question. No universally applicable rules exist.

¶ The choice of perspective can influence the conclusions of a cost-effectiveness analysis. In the previous example, done from a health-care–system perspective, program costs were $1,000/woman, and averted PID costs were $2,000, making the program cost-saving. Certain PID costs (e.g., those for infertility treatments) routinely occur after the initial infection, and the program conducting the testing might never incur them. The program doing the testing might only reduce its costs for sequelae by $800 (for treatment of acute PID) for every $1,000 spent on screening. From the perspective of the individual program, screening might not be costsaving. A societal perspective provides the truest indication of the cost-effectiveness of a program, but a more limited perspective can point to the impact on a given provider or program, including the barriers to implementation of new tests or screening programs.

Laboratory Environment

The laboratory must consider how candidate tests match the laboratory environment. For example, laboratories must satisfy the requirements for physical space, separation of work areas (e.g., sample processing and amplification), and use of any special equipment (e.g., biosafety cabinets) as recommended by the test kit manufacturer. Differences in throughput among different NAATs must be considered. Additional considerations include technician time, turnaround time, technical difficulty, equipment costs, and time required to maintain equipment. Laboratories are encouraged to investigate these concerns and their impact on cost-effectiveness when choosing among the available testing methods.

Need for Additional Testing To Support C. trachomatis or N. gonorrhoeae Diagnoses

Efforts to maximize test sensitivity to avoid missing the opportunity to identify and treat infected persons might warrant tolerating a certain number of false-positive diagnoses. However, consideration must also be given to reducing the rate and consequences of false-positive tests. All tests, including culture for C. trachomatis, occasionally generate false-positive results. For these reasons, all positive tests are considered to be presumptive evidence of infection (Box 5). Regardless of health-care provider setting, patient population, and patient characteristics, a false-positive test result for C. trachomatis or N. gonorrhoeae can have adverse medical, social, and psychological impacts for a patient. In such cases, consideration should be given to performing an additional test to verify a positive screening test.

BOX 5. Additional testing and patient management after a positive screening test

The potential for false-positive test results complicates the interpretation of positive tests among patients in populations with a low prevalence of infection. This occurs because the proportion of total positive tests that are truly positive (i.e., PPV) is lower among such a population. Consideration should be given to routine additional testing for persons with positive C. trachomatis or N. gonorrhoeae screening tests when risk-factor information or actual surveys indicate that the prevalence is low, resulting in a lower PPV (e.g., <90%) (see C. trachomatis and N. gonorrhoeae Test Performance When Used for Screening).

Economic analysis of additional testing to increase the specificity of a screening test is problematic because the primary benefit of such additional testing is in averting costs that are intangible. These include psychosocial costs that can result from a false-positive test result. These costs vary according to setting and are not well-researched. The benefits of reducing these intangible costs must be weighed against the tangible direct costs of performing the confirmatory tests. Fortunately, these costs apply only to patients who initially test positive and frequently constitute only a limited proportion of total costs per positive test, particularly in low prevalence settings where the additional testing is typically needed (see Methods To Enhance Performance or Reduce Costs) .

Consideration of Point-of-Care Testing

Point-of-care tests for C. trachomatis screening are less sensitive than laboratory-based tests but should be considered in situations where screening-test--positive persons might fail to return for treatment or return after substantial delays. Point-of-care tests are not a cost-effective option if they are processed after the patient visit because they are relatively insensitive and require labor-intensive processing. Each health-care provider needs to compare the sensitivities, costs, and treatment rates for point-of-care and laboratory-based tests. Providers need to determine whether the opportunity to provide treatment to certain patients who would otherwise go untreated warrants the additional cost and less favorable sensitivity of point-of-care testing. FDA-cleared C. trachomatis and N. gonorrhoeae tests that can be performed rapidly enough to qualify as point-of-care tests must be performed in a CLIA-certified laboratory because they are classified under CLIA as moderate complexity tests (24).

Methods To Enhance Performance or Reduce Costs

Different approaches have been used to increase the efficiency of standard screening methods. Although selective screening is not a laboratory method, using selective screening criteria is included in the following discussion because the predictive values and cost to detect an infection are strongly influenced by infection prevalence. Another approach is to use a NAAT to test specimens that yield results from an EIA or unamplified nucleic acid probe test that fall in a zone around the cutoff (i.e., gray zone). This technique warrants further evaluation as a method to decrease the gap in sensitivity between NAATs and other tests without incurring the full additional cost of testing all specimens with a NAAT. Interest in pooling specimens for testing by a NAAT is similarly motivated. Augmenting screening tests with additional testing to improve test specificity is of increasing importance because C. trachomatis and N. gonorrhoeae prevalences have declined after the introduction of screening programs and because C. trachomatis screening has expanded into lower prevalence populations. Using test formats (e.g., nucleic acid probe tests or NAATs) that permit testing for both C. trachomatis and N. gonorrhoeae might reduce costs. Finally, the urine leukocyte esterase test, which has a low sensitivity but is inexpensive, has been used to select specimens for testing with a specific C. trachomatis or N. gonorrhoeae test when use of a more sensitive initial test was not feasible.

Selective Screening To Increase the Percentage of Positive Tests

Selecting persons for testing who are at high risk can increase the prevalence of infection among tested persons, thereby reducing screening costs to detect persons with C. trachomatis infection. Symptoms or signs are not included as screening criteria because they warrant immediate diagnostic testing. Evaluations of screening algorithms for C. trachomatis, including an algorithm recommended by CDC in 1993 (31), have been published (96--104). The third U.S. Preventive Services Task Force (USPSTF) reviewed this information and published C. trachomatis screening recommendations (105) (see http://www.ahrq.gov/clinic/uspstf/uspschlm.htm) (Appendix E). USPSTF also concluded that satisfactory urine screening tests for men had been introduced too recently for sufficient evidence to have been developed to make screening recommendations. Using criteria to select women or men to screen for N. gonorrhoeae infection has received only limited evaluation, although the prevalence of N. gonorrhoeae is usually lower than the prevalence of C. trachomatis. Thus, concerns regarding the cost to detect an infection and PPV might be greater for N. gonorrhoeae than for C. trachomatis (106).

Gray-Zone Testing To Improve Test Performance

To establish the positive-negative cutoff value for a diagnostic test, a manufacturer attempts to achieve the highest sensitivity while keeping the specificity at or near 100%. This is done by setting the cutoff at the lowest level possible while minimizing false-positive results. Certain test readings from truly infected persons will fall below this cutoff value (false-negatives). Often, a substantial proportion of the readings for persons with false-negative results cluster just below the cutoff, whereas the readings for persons with truly negative results are clustered farther below the cutoff. Consequently, a relatively narrow zone below the cutoff might exist within which the proportion of readings that are from truly infected persons is substantially higher than the proportion of readings farther below the cutoff. The ratio of true positives to true negatives in this zone is affected by the prevalence of infection as well as by test performance among the tested population. Similarly, false-positive results are clustered just above the cutoff, whereas the readings for persons with truly positive results are clustered farther above the cutoff. A relatively narrow zone above the cutoff might exist within which the proportion of readings that are from truly uninfected persons is substantially higher than the proportion of readings farther above the cutoff. When screening test results are in these zones, commonly called the negative and positive gray zones, respectively, a laboratory has the option of retesting specimens with another test.

By using a NAAT to retest specimens with gray-zone results from less sensitive and less expensive screening tests, the sensitivity and specificity of the screening test might be improved at less cost than switching entirely to the NAAT. However, such a restricted use of a NAAT might be less sensitive than testing the entire population with a NAAT. A NAAT is also theoretically preferable to a competition or blocking antibody assay for additional testing of specimens from persons with positive non-NAAT screening tests (see Additional Testing To Improve Test Specificity). Testing strategies that combine use of a NAAT by itself in higher prevalence populations and as an additional test after gray-zone non-NAAT results among lower prevalence populations warrant further evaluation. Laboratories should establish feasibility, cost-effectiveness, and gray-zone limits before implementation of the strategy.

Although gray-zone testing with a NAAT has been evaluated in multiple studies (107--113), these studies have not employed independent reference standards. In certain cases, the second test was performed on the specimen that was tested by the screening test. If the second test has not been cleared by FDA for use with that type of specimen, CLIA requires that the laboratory conduct a study to establish the performance of the second test when conducted on the screening test specimen (Appendix C). This regulatory requirement is critical because NAATs might generate more false-positive and false-negative results when performed on specimens collected for tests designed with less stringent protection against contamination and amplification enzyme inhibitors. Gray-zone testing must be evaluated in appropriately designed research studies by using an independent reference standard before a recommendation is made regarding its utility in routine practice.

Pooling Specimens To Reduce Costs

Because of the high sensitivity of NAATs, pooling specimens before testing for C. trachomatis and N. gonorrhoeae with LCR and PCR has been proposed as a method of reducing costs (114--120). Samples of individual specimens are first combined into a pool, which is then tested by a NAAT. If the pool is negative, all specimens forming the pool are reported as negative. If the pool is positive, a second aliquot of each specimen that contributed to the pool is tested individually. The potential cost-savings with pooling increases with decreasing prevalence of infection, because more specimens can be included in a pool at lower prevalences without increasing the probability of a pool testing positive. Available evidence indicates that pooling might be a cost-effective alternative to testing individual specimens with minimal if any loss of sensitivity or specificity (114--120). However, insufficient data exist from published, peer-reviewed studies that address the implementation or performance concerns raised by pooling to make a recommendation concerning testing of pooled specimens.

Theoretically, pooling could either increase or decrease the sensitivity and specificity of a NAAT compared with processing individual specimens. To be reported as positive, pooled specimens must be positive according to the pooled result and according to the individual test. This requirement should contribute an increase in specificity and a decrease in sensitivity compared with individual testing. In addition, pooling could reduce NAAT sensitivity because the majority of pooling protocols result in a reduced amount of an individual specimen being tested. Moreover, pooling can permit inhibitors present in a single specimen to cause specimens that would have been truly positive when tested individually to be falsely negative when tested in a pool. Such inhibition will go undetected if an internal control for inhibition is not included when testing pooled specimens. Conversely, dilution of specimens by pooling can decrease the effect of inhibitors. Pooling adds testing and data-recording steps that can result in errors and reporting of additional false-positive or false-negative results. In addition, pooling results in a higher number of specimens in a run being positive. This increases the chances for sub-optimal technique (e.g., by pipetting) to result in cross-contamination and false-positive results.

Published studies of pooling using PCR and LCR tests for C. trachomatis and N. gonorrhoeae are still limited, and the majority of such studies have been conducted without an independent reference standard. Nevertheless, published studies indicate that pooling 4--10 urine or endocervical specimens before testing can substantially reduce costs and improve throughput while maintaining the high performance characteristics associated with NAATs (114--120). Savings from reduced reagent costs have ranged from 40% to 60%. Because of the increased complexity of the pooling protocol, savings in personnel time are proportionately less than the savings in reagent costs.

In addition to these performance concerns, the following concerns should be addressed when considering use of pooling:

• what procedures to follow when all the individual specimens from a positive pool provide negative results (e.g., test for inhibition);
• use of internal amplification controls to detect inhibition;
• the need to inform the provider or patient that specimens are being tested by using a pooling protocol;
• how to report test results; and
• reimbursement by third-party payers.

Because pooling is a modification of FDA-cleared procedures, laboratories that implement pooling of patient specimens must satisfy CLIA requirements by establishing performance specifications (Appendix C).

Additional Testing To Improve Test Specificity

An additional test might be indicated for a person with a positive screening test result, if a false-positive result would have a serious adverse consequence (Box 5). Because treatments for C. trachomatis and N. gonorrhoeae are safe and relatively inexpensive, the person might wish to receive and complete treatment while additional testing is being done, or even if the additional test is negative. Routine additional testing to improve the predictive value of a positive screening test should be considered when the prevalence of either C. trachomatis or N. gonorrhoeae infection is low, resulting in a low PPV (e.g., <90%) (see C. trachomatis and N. gonorrhoeae Test Performance When Used for Screening and also, Additional Considerations in Selecting a Screening Test).

False-positive results might occur for multiple reasons, including the following:

• the nucleic acid target, antigen, or phenotype (e.g., colistin resistance, oxidase production) detected by the test is possessed by organisms in the specimen other than C. trachomatis or N. gonorrhoeae;
• the detection system generates a signal in the absence of target;
• the specimen becomes contaminated; and
• staff make data entry errors.

Approaches to detect false-positive results by applying an additional test can be ordered by preference on the basis of theoretical considerations (Box 6).

BOX 6. Approaches to additional testing, in order by theoretical consideration, after a positive screening test

Test a second specimen with a different test that uses a different target, antigen, or phenotype and a different format. Test the original specimen with a different test that uses a different target, antigen, or phenotype and a different format. Repeat the original test on the original specimen with a blocking antibody or competitive probe. Repeat the original test on the original specimen.

Theoretically, testing a second specimen with a different type of test is least likely to confirm a false-positive result, whereas repeating the original test on the original specimen is most likely to do so. Using a NAAT as an additional test after a positive non-NAAT test might be an effective and highly economical approach that deserves additional evaluation. However, except when culture is used to obtain an isolate, a non-NAAT should not be used as an additional test after a NAAT because of the lower sensitivity of non-NAATs. Such additional testing could be extended to persons with non-NAAT screening test results in the gray zone (see Gray-Zone Testing To Improve Test Performance). An advantage of isolating the organism by culture is flexibility in the choice of multiple additional testing procedures, and compared with NAATs, a reduced risk of error caused by contamination during initial or subsequent testing. Using an FDA-cleared test as an additional test to verify the positive results of a screening test does not by itself introduce additional requirements under CLIA (24). However, if the additional test is performed on the original screening test specimen and the additional test has not been cleared by FDA for use with that type of specimen, CLIA requirements for establishing performance characteristics (24) need to be satisfied (Appendix C).

Additional Testing After a Positive C. trachomatis Screening Test. Selection of an additional test to verify a positive C. trachomatis screening test is straightforward when a second specimen can be collected, as when a patient returns for screening test results or when the need for additional testing can be anticipated when the screening specimen is collected (Box 7). Using culture with a C. trachomatis-specific anti-MOMP stain as the additional test has the advantage of high specificity and flexibility in choice of additional testing, if warranted. However, culture lacks sensitivity and necessitates obtaining and maintaining a suitable sample either at the initial or a subsequent visit. This fact has made culture less desirable. In addition, finding a laboratory to perform culture can be problematic. If obtaining a separate specimen for additional testing is difficult, the additional test must perform adequately with the transport medium used for the screening test.

 

BOX 7. Additional testing after a positive Chlamydia trachomatis screening test

Using a blocking antibody format to verify a positive EIA screening test and using a competitive probe format to verify a positive nucleic acid probe screening test have been the additional tests most widely used. The blocking antibody or competitive probe additional test is usually performed on the original screening specimen. These formats have been widely used for additional testing, despite not being the theoretically preferred approaches. This is because approaches preferred on theoretical grounds introduce the logistical and economic challenges of establishing a format for the additional test that differs from that of the screening test. Collecting a separate specimen for the additional test also presents logistical difficulties and increases cost. Studies have not been conducted to compare application of such different approaches.

Because of the greater sensitivity of NAATs, a NAAT is the only recommended additional test to verify a result from another NAAT and is, potentially, a superior additional test to verify a non-NAAT positive C. trachomatis screening test. Except for using culture to obtain an isolate, a non-NAAT should not be used as an additional test after a NAAT because of the lower sensitivity of the non-NAAT. Using a NAAT as an additional test has received limited evaluation. In particular, determining that NAAT specificity and sensitivity are not reduced is critical if a NAAT is performed on specimens collected, transported, and possibly processed by using non-NAAT procedures that might be less stringent with respect to preventing contamination and inhibition of amplification enzymes.

Manufacturers of NAATs for C. trachomatis infection have developed alternate target (e.g., MOMP targets) versions of their commercial tests that the manufacturers have employed for resolving discrepant results in evaluation studies. Using the same format for the additional test as for the screening test offers logistical and economic advantages analogous to those offered by the blocking antibody and competitive probe approaches for non-NAAT tests. Such assays are not offered commercially by NAAT manufacturers but might be in the future.

Additional Testing After a Positive N. gonorrhoeae Screening Test. Methods to confirm that Gram-negative, oxidase-positive, bacteria isolated on a selective culture medium are N. gonorrhoeae have been reviewed (Box 8) (12 ,32). Acid production from carbohydrates and the nucleic acid probe culture confirmation test (AccuProbe,^® manufactured by Gen-Probe, San Diego, California) are the most sensitive and specific methods. PACE 2 is also FDA-cleared for identifying N. gonorrhoeae.

BOX 8. Additional testing after a positive Neisseria gonorrhoeae screening test

* Sources: Martin DH, Cammarata C, Van der Pol B, et al. Multicenter evaluation of AMPLICOR and automated COBAS AMPLICOR CT/NG tests for Neisseria gonorrhoeae. J Clin Microbiol 2000;38:3544–9. Van der Pol B, Martin DH, Schachter J, et al. Enhancing the specificity of the COBAS AMPLICOR CT/NG test for Neisseria gonorrhoeae by retesting specimens with equivocal results. J Clin Microbiol 2001;39:3092–8. Becton Dickinson. BDProbeTec™ ET Chlamydia trachomatis and Neisseria gonorrhoeae amplified DNA assays [Package insert]. Sparks, MD: Becton Dickinson, 1999.

Pure growth on subculture is required for acid production tests, but not for the AccuProbe test. AccuProbe is not known to cross-react with other organisms found in humans; however, variant strains of certain pathogenic and commensal Neisseria species can provide false-positive acid production results, and certain strains of N. gonorrhoeae might provide false-negative acid production results. Additional tests will be required to differentiate between species that might produce acid from glucose but not from maltose, sucrose, or lactose (Appendix F) or to identify variant gonococcal isolates that fail to produce acid from glucose.

Although nongonococcal pathogenic and commensal Neisseria species are commonly found in the oropharynx, they are unusual in genitourinary specimens. Either the acid production or the nucleic acid probe methods could be used for routine additional testing after presumptively positive screening cultures. Requiring that both types of tests be positive should guarantee an increased level of specificity. A detailed discussion of problems in differentiating N. gonorrhoeae from other Neisseria and related species is located at http://www.cdc.gov/std/gonorrhea/lab/ident.htm. If an isolate cannot be conclusively identified as N. gonorrhoeae at a local laboratory, the isolate should be sent to a reference laboratory (e.g., a city, county, or state public health laboratory) for confirmation, chiefly in cases of alleged sexual assault or rape.

Antibody tests (i.e., tests that detect gonococcal antigens) are not recommended for detection of N. gonorrhoeae. Coagglutination tests, which might cross-react with nongonococcal strains, are less sensitive or specific than acid production tests or the AccuProbe or PACE 2 tests. The monoclonal fluorescent antibody test for N. gonorrhoeae does not detect all gonococcal isolates and is not recommended as the primary test for detecting N. gonorrhoeae.

Additional testing after positive nonculture screening tests for N. gonorrhoeae involves the same approaches that are described in the previous section for C. trachomatis. Because verification of positive nonculture tests for N. gonorrhoeae has received only limited evaluation and certain NAATs might cross-react with nongonococcal Neisseria species (54 ,56 ,57), substantial effort is warranted to arrange for culture and the assistance of a reference laboratory if a false-positive diagnosis of N. gonorrhoeae would have serious adverse consequences, as in testing of children.

Reporting Test Results. When additional testing has been performed, the laboratory should report the results of both the screening test and the additional tests, as well as the overall interpretation. The laboratory has the responsibility to educate clinicians regarding the importance of all laboratory results, including both screening and additional test results. In particular, clinicians need to be aware of the limitations of the additional tests, including the possibility that they yield false-negative results when the screening test is positive. Because serious side effects from therapies for C. trachomatis and N. gonorrhoeae are uncommon, clinicians might recommend treatment after a positive screening test for a person at risk for infection, pending additional testing or even when a positive screening test is not verified by additional testing.

Cotesting for C. trachomatis and N. gonorrhoeae To Reduce Costs

Multiple tests permit testing for both organisms by using the same specimen. The prevalence of N. gonorrhoeae is less than C. trachomatis in the majority of areas of the United States; however, the prevalence of each varies widely, even within such limited areas as cities or counties. Usually, screening for N. gonorrhoeae will not be justified unless screening for C. trachomatis is also warranted. Decisions regarding screening for either or both organisms should not be made without a careful evaluation of the local epidemiology of N. gonorrhoeae and C. trachomatis.

Cotesting for C. trachomatis and N. gonorrhoeae by using tests specially designed for such cotesting should be considered, if transport conditions would reduce the sensitivity of N. gonorrhoeae culture or if using such tests reduces the cost. However, provision should be made to perform an additional test to improve test specificity whenever indicated (see Additional Testing To Improve Test Specificity) and to obtain isolates for antimicrobial susceptibility testing in the case of a repeated treatment failure (see Test of Cure, Treatment Failure, and Antimicrobial Resistance).

Using LET to Select for C. trachomatis or N. gonorrhoeae Tests

LET, which has a low sensitivity but is inexpensive, has been used to select specimens for testing with a specific C. trachomatis or N. gonorrhoeae test when universal testing with a more sensitive initial test is not feasible. Studies have demonstrated that using a positive LET to select patients for screening with a more sensitive and specific test can be cost-effective. Published studies have focused primarily on asymptomatic men and have reported that this approach is more cost-effective when the prevalence is low (89 ,91 ,121--124). One study (121) determined that the negative predictive value of LET among asymptomatic men aged ≥30 exceeded 98%, thereby providing support for no additional C. trachomatis testing among that population. Among men aged <30 years, whether a negative result is a valid basis on which to exclude that population from testing with more sensitive tests is uncertain. Because the test does not detect either C. trachomatis or N. gonorrhoeae, but rather nonspecific inflammatory enzymes, a positive LET should be followed by specific tests for C. trachomatis and N. gonorrhoeae. The importance of a positive LET that is followed by negative tests for C. trachomatis and N. gonorrhoeae is unknown, but could indicate infection with other organisms (e.g., Trichomonas vaginalis, Ureaplasma urealyticum, or Mycoplasma genitalium).

Previous Page

Table of Contents

Next Page