BacteriologyCost-effectiveness of nucleic acid amplification testing to guide treatment for vaginitis: a decision-modeling analysis
Introduction
Vaginitis is one of the most common gynecologic conditions, accounting for over 10 million physician office visits each year in the United States (Kent, 1991; Paladine and Desai, 2018; Sobel et al., n.d.). The 3 most common infectious causes of vaginitis in the primary care setting are bacterial vaginosis (BV; 22% to 50% of cases), vulvovaginal candidiasis (VVC; 17% to 39% of cases), and Trichomonas vaginalis (TV; 4% to 35% of cases), with about 30% of the cases having an indeterminate cause (Anderson et al., 2004; Hainer and Gibson, 2011). If left untreated, vaginitis can increase a woman's risk of other health problems. For example, BV has been associated with pelvic inflammatory disease and increased susceptibility to HIV infection (Onderdonk et al., 2016).
Drugs that target the causative microorganism are highly effective, resolving symptoms in more than 80% of the patients treated (Carr et al., 2005). Therefore, accurate identification of the causative microorganisms is important to guide the prescription of targeted antimicrobial therapy. Accurate diagnosis and treatment with the appropriately targeted therapy not only relieve symptoms but also provide other potential health benefits such as reducing the risk of acquiring sexually transmitted diseases (Centers for Disease Control and Prevention, n.d.). Accurate diagnosis and treatment could also reduce the costs associated with unresolved symptoms and complications. In North America, the annual economic burden of treating symptomatic BV is estimated at US$1.3 billion, a cost that would be even higher if the cost of BV-associated complications was included (Peebles et al., 2019).
Options for assessing the cause of vaginitis include clinical and microscopic examination (CME), nucleic acid amplification testing (NAAT), and nonamplified nucleic acid probe (probe) testing (Andrea and Chapin, 2011; Cartwright et al., 2013; Gaydos et al., 2017; Richter et al., 2019; Schwebke et al., 2018; Thompson et al., 2020). CME relies on the evaluation of clinical symptoms, measurement of vaginal pH, amine (whiff) test, and KOH or wet mount microscopy. NAAT and probe testing can simultaneously identify the 3 types of microbial pathogens (bacteria, Candida spp., and Trichomonas vaginalis)—NAAT uses real-time polymerase chain reaction (PCR) for amplification of pathogen DNA targets, while the DNA probe–based test hybridizes nucleic acid probes to unamplified pathogen DNA in vaginal fluid specimens. CME has been traditionally used in the clinic, but the sensitivity for microbial pathogens is low, ranging from 58% for VVC to 77% for BV (Schwebke et al., 2018). CME also requires expertise in microscopy and an on-site microscope, making CME impractical in many primary care settings. Probe testing has higher sensitivity for BV (~90%), but the sensitivity is 58% for VVC and 46% for TV (Cartwright et al., 2013). In contrast, NAAT has 97% sensitivity for all 3 types of microbial pathogens. Overall, these 3 diagnostic methods have relatively high specificity, except that probe testing has only modest specificity for BV (Cartwright et al., 2013).
Given that CME, NAAT, and probe testing differ in both diagnostic accuracy and cost, a better understanding of the cost-effectiveness of test-and-treat scenarios based on these 3 diagnostic methods could help inform healthcare decisions. Ackerman et al. recently analyzed commercial healthcare insurance claims during the 6-month period after patients sought treatment for vaginitis (Ackerman et al., 2019). They found that all-cause healthcare resource utilization and all-cause direct costs were lower for the group that used NAAT than for the group that used probe testing. However, since this was a retrospective analysis and patients were not randomized between NAAT and probe testing, differences in patient or physician characteristics may have influenced the results (Sherman et al., 2016). Furthermore, the analysis of health insurance data may not have always captured CME that was done in addition to a NAAT or probe test because healthcare providers may not take the time to bill for the small fee allowed for the performance of a CME. Therefore, we have used a decision-modeling analysis to evaluate the cost-effectiveness of a test-and-treat scenario based on NAAT compared with a scenario based on probe testing. We also evaluated the cost-effectiveness of using either NAAT or probe testing as a follow-up test after a negative or indeterminate CME diagnosis compared with using CME alone. This study was conducted from a payer perspective in the United States.
Section snippets
Study population and model
The hypothetical population modeled in this study consisted of 1000 women aged 18 years and older who were seeking treatment for vaginitis: the size of the cohort was chosen so that the number of patients with symptom resolution could be reported in whole numbers—the cohort size has no effect on the incremental cost-effectiveness ratio (ICER). The age range of the patients was chosen based on the clinical studies of the vaginitis tests cited in our analysis (Cartwright et al., 2013; Schwebke et
NAAT versus probe testing: base case
In the base-case model of a hypothetical cohort of 1000 vaginitis patients, the test-and-treat scenario based on NAAT resulted in 140 more symptom resolutions than did the test-and-treat scenario based on probe testing (Table 2). These 140 incremental symptom resolutions were achieved at a cost of $210 each, that is, the ICER was $210 per incremental symptom resolution. The average cost of symptom resolutions achieved was $871 in the test-and-treat scenario based on probe testing and $720 in
Discussion
These models indicate that NAAT-based test-and-treat scenarios increase the number of patients who achieve resolution of vaginitis symptoms compared with test-and-treat scenarios based only on CME and probe testing. The cost of incremental symptom resolutions due to the use of a NAAT-based test-and-treat scenario is lower than the average cost of symptom resolutions achieved by test-and-treat scenarios based on either probe testing or both CME and probe testing. And since both probe testing and
Conclusion
This analysis indicates that NAAT-based test-and-treat scenarios for patients with vaginitis could cost-effectively improve patient health outcomes.
Acknowledgments
The authors thank Kathleen Troeger, Adrian Vilalta, and Jeff Radcliff for helpful comments on the manuscript.
Funding
No specific funding directly applied to this study. However, Quest Diagnostics provided employment to the authors.
Conflict of interest
The authors are employees of Quest Diagnostics.
References (25)
Epidemiology of vaginitis
Am J Obstet Gynecol
(1991)- et al.
Value for money? Array genomic hybridization for diagnostic testing for genetic causes of intellectual disability
Am J Hum Genet
(2010) - et al.
Cost-effectiveness of rapid diagnostic assays that perform directly on blood samples for the diagnosis of septic shock
Diagn Microbiol Infect Dis
(2019) - et al.
Health care utilization and costs following amplified versus non-amplified molecular probe testing for symptomatic patients with suspected vulvovaginitis: a US commercial payer population
Clinicoecon Outcomes Res
(2019) - Aetna. Diagnosis of vaginitis. Medical clinical policy bulletins number: 0643...
- et al.
Evaluation of vaginal complaints
JAMA
(2004) - et al.
Comparison of Aptima Trichomonas vaginalis transcription-mediated amplification assay and BD affirm VPIII for detection of T. vaginalis in symptomatic women: performance parameters and epidemiological implications
J Clin Microbiol
(2011) - et al.
"Shotgun" versus sequential testing. Cost-effectiveness of diagnostic strategies for vaginitis
J Gen Intern Med
(2005) - et al.
Comparison of nucleic acid amplification assays with BD affirm VPIII for diagnosis of vaginitis in symptomatic women
J Clin Microbiol
(2013) - Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines....