This economic evaluation reports the prevalence, outcomes, and cost per case detected and successfully treated associated with simulated testing for Chlamydia trachomatis and Neisseria gonorrhoeae in young patients presenting at pediatric emergency departments.
Key Points
Question
What is the most cost-effective way to screen for Chlamydia trachomatis and Neisseria gonorrhoeae in adolescents and young adults seeking acute care at pediatric emergency departments (EDs)?
Findings
In this economic evaluation of a hypothetical population of 10 000 ED visits by individuals aged 15 to 21 years, targeted and universally offered screening strategies were more successful in detection and treatment of sexually transmitted infections than the no screening strategy, but they were more expensive. The incremental cost-effectiveness ratio for targeted screening vs no screening was $6444 and for universally offered screening vs targeted screening was $12 139.
Meaning
Findings from this study suggest that targeted screening and universally offered screening are both cost-effective strategies for identifying chlamydial and gonococcal infections in adolescents and young adults who access acute care at EDs.
Abstract
Importance
Adolescents and young adults compose almost 50% of all diagnosed sexually transmitted infection (STI) cases annually in the US. Given that these individuals frequently access health care through the emergency department (ED), the ED could be a strategic venue for examining the identification and treatment of STIs.
Objective
To examine the cost-effectiveness of screening strategies for Chlamydia trachomatis and Neisseria gonorrhoeae (chlamydia and gonorrhea) in adolescents and young adults who seek acute care at pediatric EDs.
Design, Setting, and Participants
This economic evaluation is a component of an ongoing, larger multicenter clinical trial at the Pediatric Emergency Care Applied Research Network. A decision analytic model, created using literature-based estimates for the key parameters, was developed to simulate the events and outcomes associated with 3 strategies for screening and testing chlamydial and gonococcal infections in individuals aged 15 to 21 years who sought acute care at pediatric EDs. Data sources included published (from January 1, 1997, to December 31, 2019) English-language articles indexed in MEDLINE, bibliographies in relevant articles, insurance claims data in the MarketScan database, and reimbursement payments from the Centers for Medicare and Medicaid Services. Because the events and outcomes were simulated, a hypothetical population of 10 000 ED visits by adolescents and young adults was used.
Interventions
The 3 screening strategies were (1) no screening, (2) targeted screening, and (3) universally offered screening. Targeted screening involved the completion of a sexual health survey, which yielded an estimated STI risk (at risk, high risk, or low risk).
Main Outcomes and Measures
Outcome metrics included cost (measured in 2019 US dollars) and the detection and successful treatment of STIs. The incremental cost-effectiveness ratio (ICER) of each strategy was calculated in a base case analysis. The ICER reflects the cost per case detected and successfully treated.
Results
A 3.6% prevalence of chlamydia and gonorrhea was applied to a hypothetical population of 10 000 ED visits by adolescents and young adults. Targeted screening resulted in the detection and successful treatment of 95 of 360 STI cases (26.4%) at a cost of $313 063, and universally offered screening identified and treated 112 of 360 STI cases (31.1%) at a cost of $515 503. The ICER for targeted screening vs no screening was $6444, and the ICER for universally offered screening vs targeted screening was $12 139.
Conclusions and Relevance
This economic evaluation found that targeted screening and universally offered screening compared with no screening appeared to be cost-effective strategies for identifying and treating chlamydial and gonococcal infections in adolescents and young adults who used the ED for acute care. Universally offered screening was associated with detecting and successfully treating a higher proportion of STIs in this population.
Introduction
In the United States, although adolescents and young adults represent only 25% of the sexually active population, they compose nearly 50% of all diagnosed sexually transmitted infection (STI) cases.1 Of the 20 million new cases of STIs each year, 10 million cases occur among adolescents and young adults.1 Failure to diagnose and treat STIs in a timely manner can result in reproductive morbidity, including pelvic inflammatory disease (PID), ectopic pregnancy, and infertility, and can facilitate HIV transmission. Adolescents and young adults frequently access health care services through the emergency department (ED). Thus, the ED could be a strategic venue for the diagnosis and treatment of STIs.
Despite the high rates of STIs among adolescents and young adults who seek health care in the ED,2 ED-based STI screening is not yet routine. Given the hectic and nonprivate nature of the ED, STI screening in this setting can be challenging. Unlike the opt-out HIV screening, which the Centers for Disease Control and Prevention (CDC) recommends for all patients aged 13 years or older in the ED setting,3 no CDC recommendations currently exist for ED-based Chlamydia trachomatis (chlamydia) and Neisseria gonorrhoeae (gonorrhea) screening. Such screening can be offered universally or to all adolescents and young adults using a targeted approach that is based on sexual experience and/or chief complaint.
In an ongoing multicenter clinical trial at 6 sites within the Pediatric Emergency Care Applied Research Network (PECARN), we are exploring a computerized sexual health survey that is integrated into the electronic health record to provide clinical decision support for chlamydial and gonococcal infection screening in pediatric EDs. One component of this PECARN project is the assessment of the cost-effectiveness of chlamydial and gonococcal screening strategies for adolescents and young adults visiting pediatric EDs for acute care. For this part of the trial, we constructed a decision analytic model using literature-based estimates for the key parameters. After the 4-year clinical trial is completed, we will replace the literature-based estimates with empirical data from the full 6-site study.
Methods
This economic evaluation was approved by the University of Utah Institutional Review Board. No informed consent was required because the analysis used literature-based data and did not involve human participation.
We developed a decision analytic model to simulate the events and outcomes associated with 3 chlamydial and gonococcal screening strategies for individuals aged 15 to 21 years seeking acute care at pediatric EDs. These strategies were (1) no screening (or usual care); (2) targeted screening, which used a sexual health survey to estimate STI risk (at risk, high risk, or low risk); and (3) universally offered screening. The outcome measures were cost (measured in 2019 US dollars) and the detection and successful treatment of STIs.
Review of Data from the Literature
We searched MEDLINE for English-language articles published from January 1, 1997, to December 31, 2019. The following keywords were used in the search: emergency service, hospital; screening; sexually transmitted diseases; prevalence; adolescent; young adult; chlamydia infections; Chlamydia trachomatis; gonorrhea; decision trees; decision support techniques; health surveys; decision making, cost-benefit analysis; computer assisted; sexual behavior; and health care costs. Three of us (M.H.E., J.L.R., M.K.G.) conducted the literature search and review. We also reviewed bibliographies in relevant studies.
Data from the CDC show marked variations in the incidence and prevalence of STIs by age category (Table 1).4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23 Case rates for both chlamydia and gonorrhea are highest among individuals aged 20 to 24 years, with roughly similar rates of gonorrhea in both female (702.6 per 100 000 population) and male (720.9 per 100 000 population) young adults.3,4,5 Reported cases of chlamydia are much higher in female (4064.6 per 100 000 population) than male (1784.5 per 100 000 population) young adults.4,5,9 If one assumes that the rate of adolescents and young adults who obtain acute care services from pediatric EDs reflects the rate in the broader population, then the estimated probability of STI among those who do not specifically seek care for STI symptoms is 3.6%. A literature review of screening in US EDs reported the prevalence of chlamydial and gonococcal infection as ranging between 2.0% and 15.0%, suggesting that those seeking acute care in EDs represent an enriched population with respect to STIs.15 However, many studies of the prevalence of chlamydia and gonorrhea in the ED population may be biased given that patients who agreed to be enrolled may be self-selected because of concerns about increased risk.10,11,15,24,25,26,27,28 In a study of universal screening among asymptomatic adolescents and young adults in an urban ED, 7.9% requested testing because of STI symptoms; the probability of STI among these patients was 25.0%.11 If the estimated overall probability of STI is 3.6% among young individuals seeking ED care for acute medical problems and not STI symptoms, and the probability of STI is 25.0% among those requesting testing because of STI symptoms or concerns, then the estimated probability of STI among the remaining asymptomatic ED patients would be 1.8%.
Table 1. Data Required in the Analysis.
| Variable | Base case value (clinically plausible range), % |
|---|---|
| Proportion of male ED patients | 0.50 |
| Prevalence of STI | 0.04a,b (0.03-0.14)c,d,e |
| Probability of chlamydial infection | |
| All patients | 0.03a |
| Female | 0.05a,f |
| Male | 0.02a,f |
| Relative risk of chlamydia in female vs male | 2.28 |
| Probability of gonococcal infection | |
| All patients | 0.003b,f |
| Female | 0.0003b,f |
| Male | 0.003b,f |
| Relative risk | |
| Gonorrhea in female vs male | 0.97 |
| STI among patients willing to be tested | 1.32g |
| STI among patients not willing to be tested | 0.53g |
| Proportion of ED patients requesting STI testing because of symptoms or worries | 0.08h (0.05-0.14) |
| Probability of STI among those | |
| Requesting STI testing because of symptoms or worries | 0.25h (0.21-0.32)i |
| Not requesting STI testing | 0.02 |
| Proportion agreeing to complete STI survey | 0.59g (0.59-0.80)d |
| Proportion agreeing to get STI testing under universally offered screening strategy | 0.59h (0.59-0.80) |
| Proportion receiving STI testing among those at risk or at high risk on STI survey | 0.52g (0.50-0.73) |
| Probability that clinician orders STI testing under universally offered screening strategy | 0.73e (0.70-0.99)j |
| Outcomes of STI testing | |
| Probability of epididymitis | |
| Male with untreated chlamydia | 0.02k,l |
| Male with untreated gonorrhea | 0.02k,l |
| Probability of PID | |
| Female with untreated chlamydia | 0.15k,l (0.10-0.20) |
| Female with untreated gonorrhea | 0.15k,l (0.10-0.20) |
| Probability that patient with diagnosed STI will receive treatment | 0.92m,n (0.80-1.00) |
| Probability of treatment success | |
| Chlamydia | 0.97a |
| Gonorrhea | 0.99b |
| Test characteristics: NAAT on urine specimen | |
| Sensitivity of gonococcal NAAT | 0.91 (0.78-1.00)o |
| Specificity of gonococcal NAAT | 0.97 (0.97-1.00)o |
| Sensitivity of chlamydial NAAT | 0.86 (0.72-1.00)o |
| Specificity of chlamydial NAAT | 0.97 (0.97-1.00)o |
| Test characteristics of STI screening tool | |
| High risk, using cutoff STI score ≥11 | |
| Sensitivity | 0.43c |
| Specificity | 0.76c |
| LR | 1.74 |
| At risk, using cutoff STI score 7-10 | |
| LR | 1.10 |
| Low risk, using cutoff STI score ≤6 | |
| LR | 0.56 |
| 2019 Cost, US $ | |
| Variable | |
| STI long-term complications | |
| Epididymitis | 499p,q |
| PID | 4261p,q |
| Screening test | |
| Gonococcal NAAT | 39k,l,r |
| Chlamydial NAAT | 39k,l,r |
| Antibiotic therapy | |
| Azithromycin, 1 g by mouth | 7s |
| Ceftriaxone sodium, 250 mg IM injection | 22s |
| ED visit for treatment of gonorrhea | 135t |
Abbreviations: APC, Ambulatory Payment Classification; CPT, Current Procedural Terminology; ED, emergency department; IM, intramuscular; LR, likelihood ratio; NAAT, nucleic acid amplification test; PID, pelvic inflammatory disease; STI, sexually transmitted infection.
Centers for Disease Control and Prevention,4 2015.
Centers for Disease Control and Prevention,5 2015.
Falasinnu et al,6 2014.
Reed et al,7 2017.
Mehta et al,8 2002.
Satterwhite et al,9 2013.
Ahmad et al,10 2014.
Goyal et al,11 2016.
Goyal et al,12 2012.
Data obtained February 20, 2020.
Owusu-Edusei et al,13 2013.
Owusu-Edusei et al,14 2013.
Jenkins et al,15 2013.
Burkins et al,16 2020.
Zakher et al,17 2014.
Gift and Owens,18 2006.
US Bureau of Labor Statistics,19 2019.
Centers for Medicare and Medicaid Services,20 2019.
US Department of Veterans Affairs,21 2019.
Although a number of studies have described multivariable STI risk prediction models, surveys, or questionnaires,10,11,24,25,27,28,29 to our knowledge, only 1 study has formally described test characteristics and area under the receiver operating characteristic curve (AUROC) data for derivation and validation cohorts.6 The AUROC was 0.74 in the derivation set of electronic medical records from 10 437 asymptomatic patients tested for chlamydia and/or gonorrhea between 2000 and 2012 at 2 STI clinics in Vancouver, British Columbia, Canada.6 In the validation set of 14 956 patients, the AUROC was 0.64.6 The STI risk prediction tool had a scoring range of up to 26, based on symptoms and behaviors. With a cutoff or positivity criterion of 11 or more, the tool’s sensitivity was 42.5% and specificity was 75.8%.6 We used the test characteristics of this STI risk prediction tool in our decision analysis, as we do not yet have complete data to calculate an ROC curve for the sexual health survey being used in the ongoing PECARN study.
Studies have suggested that adolescents and their guardians are supportive of ED-based screening for STIs.2,7,30,31 In a large study of computerized self-interviews for STI screening in ED settings, 800 of the 1337 eligible patients (59.8%) who were approached agreed to complete an STI survey, and the prevalence of STIs was 4.8% (38 of 800).10 This study did not obtain testing data on those who declined to complete the survey (n = 537). When the estimated overall STI prevalence of 3.6% (based on CDC data) was applied, 10 of the 1337 eligible patients were estimated to have an STI. Thus, the estimated prevalence of STIs among the 537 patients who declined to complete the survey was 0.02 (10 of 537). From these estimates, the relative risk of STI in those who completed the survey was 1.32 (0.05 of 0.04), and the relative risk of STI among those who declined to complete the survey was 0.53 (0.02 of 0.04) (Table 1).
In a study of universal screening for STIs among 553 enrolled patients, 326 (59.0%) agreed to have their urine tested for chlamydia and gonorrhea.11 A review of similar published studies of universal screening in pediatric EDs found that 59.0% to 80.0% of participants agreed to testing.27
The CDC-recommended method for testing for chlamydial and gonococcal infection is the nucleic acid amplification test (NAAT) of urine specimens.32 Diagnostic accuracy of NAAT is relatively good. Specificity for chlamydia and gonorrhea has been reported to be 97% or higher.17,33 Sensitivity for chlamydia ranges from 86% to 100%, and sensitivity for gonorrhea is between 91% and 100%.17,33
Precise long-term sequelae of both chlamydial and gonococcal infections are difficult to estimate. An analysis of direct medical costs in the US estimated that PID develops in 15.0% of untreated women with chlamydial or gonococcal infections, whereas epididymitis develops in 1.5% of untreated men with either STI.13 The recommended treatment for chlamydial infection is a single 1-g oral dose of azithromycin or 100 mg of doxycycline twice daily for 7 days.4 Treatment efficacy has been reported to be 97.0%.4 The recommended treatment for uncomplicated gonococcal infections is a single 250-mg dose of ceftriaxone sodium intramuscularly and a 1-g dose of oral azithromycin.5 The reported cure rate of this treatment for uncomplicated urogenital infections is 99.0%.5 However, STI management and ED care follow-up are challenging. Studies have reported a loss to follow-up rate of between 8.0% and 32.0%.15
Costs
Major costs for the 3 screening strategies that we considered included the direct costs of testing and treatment for those diagnosed with infection and the costs of sequelae associated with either an untreated infection or an unsuccessful treatment, including PID and epididymitis. Costs of long-term complications from PID and epididymitis, such as infertility, ectopic pregnancy, and chronic pelvic pain, were discounted at 3% per year. We performed the analyses from the health care system perspective and did not include indirect costs, such as those associated with patients’ lost time from work.
Details of the methods we used to obtain cost data are provided in the eMethods in the Supplement. Briefly, we reviewed insurance claims data in the MarketScan database (Truven Health Analytics, IBM) with Current Procedural Terminology codes for NAATs. Reimbursement payments from the Centers for Medicare and Medicaid Services were used as a proxy for costs. The 2019 cost for each NAAT performed was $39. We consulted the Federal Supply Schedules for the costs of medications, including antibiotic treatments for chlamydial and gonococcal infections.
Statistical Analysis
We used Decision Maker, version 3.0 (Lau et al34), to develop a decision analytic model, analyze decision trees, and perform sensitivity analyses.34 Figure 1 and the eMethods in the Supplement provide a detailed description of this decision model.
Figure 1. Decision Tree Model for Individuals Aged 15 to 21 Years Seeking Acute Care at the Pediatric Emergency Department (ED).
A, The 3 screening strategies shown at the square decision node are usual care (no screening), targeted screening, and universally offered screening. Round nodes represent chance events. Events following testing (represented by the testing diamond) for patients in each of the 3 strategies are shown in the testing subtree (B). Events following no testing (represented by the no testing diamond) are shown in the no testing subtree (C). The rectangular terminal nodes at the far right of the testing subtree represent the outcome metrics of cost and effectiveness for each terminus. NAAT indicates nucleic acid amplification test.
We performed a base case analysis to examine the incremental cost-effectiveness ratio (ICER) of strategies as they progressively increased in cost. The ICER reflects the cost per case detected and successfully treated. We conducted both deterministic and probabilistic sensitivity analyses. In the probabilistic sensitivity analyses, second-order Monte Carlo simulations were carried out using distributions for most parameter values in the decision model (eFigure 3 in the Supplement). In the deterministic sensitivity analyses, we examined parameter uncertainty or variation in setting, such as varying prevalence of STI in different geographical locations or settings (eg, urban vs suburban vs rural).
Results
Base Case Analysis
In the base case (Table 2), using an overall STI prevalence of 3.6% for a hypothetical population of 10 000 ED visits by adolescents and young adults, the no screening strategy was the least costly strategy ($190 409), targeted screening was the second most costly ($313 063), and universally offered screening was most costly ($515 503). As each strategy became more expensive, it also became more successful at detection and treatment. No screening resulted in the detection and treatment of 76 STI cases. Targeted screening detected and treated 95 STI cases, whereas universally offered screening detected and treated 112 STI cases. The ICER (cost per case detected and successfully treated) of targeted screening vs no screening was $6444, whereas the ICER of universally offered screening vs targeted screening was $12 139. As discussed, in a population of 10 000 ED visits with a mean overall prevalence of 3.6%, the total number of cases was 360.
Table 2. Results of Base Case Analysis.
| Strategy | Cost, $ | Successful detection and treatment, No. | Incremental cost, $ | Incremental success of detection and treatment, No. | ICER, $ |
|---|---|---|---|---|---|
| No screening | 19 | 0.01 | NA | NA | NA |
| Targeted screening | 31 | 0.01 | 12 | 0.002 | 6444 |
| Universally offered screening | 52 | 0.01 | 20 | 0.002 | 12 139 |
| Hypothetical population of 10 000 ED visits | |||||
| No screening | 190 409 | 76 | NA | NA | NA |
| Targeted screening | 313 063 | 95 | 122 655 | 19 | 6444 |
| Universally offered screening | 515 503 | 112 | 202 439 | 17 | 12 139 |
| Proportion of STIs detected and successfully treated, % | |||||
| No screening | 21.1 | NA | NA | NA | NA |
| Targeted screening | 26.4 | NA | NA | NA | NA |
| Universally offered screening | 31.1 | NA | NA | NA | NA |
Abbreviations: ICER, incremental cost-effectiveness ratio; NA, not applicable; STI, sexually transmitted infection.
Another way to consider these base case analysis results is by the proportion of patients with STIs who were identified and treated successfully. Thus, the no screening strategy resulted in the detection and successful treatment of 21.1% of STI cases (76 of 360), whereas targeted screening resulted in 26.4% (95 of 360) and universally offered screening resulted in 31.1% (112 of 360) of STI cases. The large proportion of STI cases detected and treated under the no screening strategy was associated with the substantial proportion of ED patients who presented with complaints or concerns about STIs and requested testing.
Deterministic Sensitivity Analyses
One of the most important model parameters that might differ from one ED to another was the prevalence of STI among young individuals visiting the ED for acute care. In the base case, we used an overall prevalence of 3.6%. However, this prevalence is known to vary widely according to both patient and geographical factors.15 As shown in Figure 2, the ICERs for both targeted screening vs no screening ($262 621 vs $5712) and for universally offered screening vs targeted screening (ICER: $491 099 vs $10 771; STI probability: 0.001% vs 0.04%) decreased quickly as the prevalence of STI in the population seeking ED care increased. Another critical parameter was the probability that patients diagnosed with an STI sought follow-up visits for appropriate treatment. As shown in eFigure 1 in the Supplement, the ICERs for both targeted screening vs no screening ($572 833 vs $6582) and for universally offered screening vs targeted screening (ICER: $1 051 899 vs $12 392; probability of getting treated: 0.01% vs 0.90%) remained low from the base case value of 0.92 until the probability of getting treated decreased below 0.2, after which it increased sharply.
Figure 2. One-Way Sensitivity Analysis of Probability of Sexually Transmitted Infection (STI).
The 2 screening curves are virtually superimposed above a probability of 5%. Above a probability of approximately 2%, the incremental cost-effectiveness ratio (ICER) is minimally affected by the prevalence of STI among patients presenting to pediatric emergency departments for acute care. Below a probability of 2%, the ICER increases steeply. The base case value is 3.6%.
Figure 3 features a tornado diagram, which summarizes the results of a series of 1-way sensitivity analyses that we performed. As shown, the parameters with the greatest association with results are at the top of the diagram, whereas those with the least association are at the bottom. Parameters with the greatest association with results within their clinical plausible ranges included the probability of getting tested for symptoms or worries about STI (0.08%; range, 0.05%-0.14%), the probability of STI (0.04%; range, 0.03%-0.14%), and the probability of STI among symptomatic or worried patients (0.25%; range, 0.21%-0.32%) (Table 1).
Figure 3. Tornado Diagram of the 1-Way Sensitivity Analysis for Universally Offered vs Targeted Screening.
The incremental cost-effectiveness ratio (ICER) in $ per sexually transmitted infection (STI) successfully identified and treated ranges from 0 to $50 000. For each parameter, the upper and lower limits of the sensitivity analysis are based on the clinically reasonable range. The base case result is marked by the dotted line at the center of the tornado plot. Parameters at the top of the figure (wide mouth of the tornado) have a larger impact on the ICER within their clinically reasonable range. NAAT indicates nucleic acid amplification test; and PID, pelvic inflammatory disease.
In addition, we explored moving the operating point of the STI survey along its ROC curve by performing a sensitivity analysis on the score cutoff for the 26-point scale. As shown in eFigure 2 in the Supplement, the ICER for targeted screening vs no screening decreased (ICER: $8941 vs $496) as the cutoff for high risk of STI increased (cutoff: 1 vs 25). This finding corresponds to a higher specificity of the tool at the cost of decreasing sensitivity. Thus, the ICER was lowest in the absence of false-positive results. However, at a specificity of 100%, the sensitivity was only 0.6%, resulting in the lowest rate of success in detection and treatment for the targeted screening strategy: 76 cases were detected and successfully treated (the same number as with the no screening strategy) in a hypothetical population of 10 000 ED visits by adolescents and young adults (eTable 1 in the Supplement). The cost and successful detection and treatment of the universally offered screening strategy were not associated with the test characteristics of the sexual health survey used in the targeted screening strategy. However, the ICER of universally offered screening compared with targeted screening varied as the cost and successful detection and treatment of targeted screening changed.
We also examined a scenario in which an STI survey with much better performance characteristics was used. If the sensitivity and specificity of the survey tool were 99%, targeted screening would be less costly but more successful in detection and treatment than no screening (eTable 2 in the Supplement). The ICER for universally offered screening vs targeted screening would now be considerably higher at $32 584 per case identified and successfully treated.
Discussion
We found that targeted screening and universally offered screening had a reasonable cost per case detected and successfully treated, with $6444 for target screening and $12 139 for universal screening. Both strategies were associated with the detection and treatment of STI cases in a larger proportion of patients visiting the ED than the no screening strategies, with 26.4% of cases detected and treated owing to target screening and 31.1% of cases with universal screening vs 21.1% with no screening. In a hypothetical cohort of 10 000 ED visits by young individuals, assuming a 3.6% prevalence of chlamydial or gonococcal infection, targeted screening resulted in the detection and treatment of STI in 95 of 360 cases, whereas universally offered screening resulted in the detection and treatment of STI in 112 of 360 cases.
Substantial factors that limited the success in STI detection and treatment under both targeted and universal strategies were the proportion of patients who agreed to complete the sexual history survey used to estimate STI risk in the targeted screening strategy and the proportion of patients who agreed to undergo testing either after being identified as high risk by the survey or under the universally offered screening strategy. In addition, clinicians needed to order testing of urine specimens after agreement from patients, and patients needed to either return to the ED for treatment of gonorrhea or obtain an oral prescription for the treatment of chlamydia.
Details of the screening strategies in the present decision model mirror the protocol of the ongoing PECARN multicenter clinical trial. Thus, testing for chlamydia and gonorrhea was performed with urine NAAT assays rather than with point-of-care tests. The test characteristics of NAATs and the ease of obtaining specimens and specimen transport are better than any other testing modality available for the diagnosis of chlamydia and gonorrhea and are recommended by the CDC.32 However, this modality obviates the ability to treat on site during the ED visit and introduces the added complication of follow-up and adherence with treatment. We assumed that 92% of patients who were diagnosed with an STI would get appropriate treatment according to a systematic review of EDs between 1995 and 2010.15 However, as the probability of receiving treatment decreased to less than 20%, the cost per case detected and successfully treated for either screening strategy increased steeply.
Prevalence of chlamydia and gonorrhea is a major factor in detection and successful treatment and in the cost-effectiveness associated with screening. We used a conservative estimate of 3.6% STI prevalence, based on surveillance data from the CDC for adolescents aged 15 to 21 years. We assumed that the prevalence of chlamydia and gonorrhea among young individuals who used the ED for a variety of acute care issues would reflect the overall population prevalence in the US for this age group. Sensitivity analyses showed that the cost per case detected and successfully treated decreased as the prevalence of STIs increased among those who sought ED care. Other cost-effectiveness analyses have assumed chlamydia and gonorrhea prevalence as high as 13.6% for an ED population.8 In the present analysis, the ICERs would decrease to $1061 for targeted screening and to $2078 for universally offered screening if a prevalence of 13.6% were applied. Although almost all of the pediatric EDs in the 6-site PECARN study treat patients up to age 21 years, the STI prevalence may be lower in EDs that treat patients up to age 18 years only, given that the highest prevalence subgroup is between the ages of 20 and 24 years. It is possible that EDs with lower age limits or those in other geographical regions may have a lower STI prevalence. Thus, the decisions about optimal screening strategies could be nuanced, to account for the prevalence at individual centers. If the prevalence at a particular center were known to be less than 1%, then targeted screening would become the more cost-effective option. If prevalence were less than 0.5%, then even targeted screening would no longer be highly cost-effective. A caveat about these specific thresholds is that they are defined as highly cost-effective on the basis of a societal willingness-to-pay threshold of $50 000 per quality-adjusted life-year. No willingness-to-pay thresholds were established for our utility metric of cost per case identified and successfully treated.
The proportion of patients who come to the ED for acute care who get tested for chlamydia and gonorrhea because of symptoms or worries has implications for the cost-effectiveness of screening strategies. We used a base case estimate of 7.9% from a previous study in an urban ED setting.11 The larger this number, the less successful were any screening strategies, because greater proportions of possible cases were detected outside of the screening programs.
The multicenter clinical trial on which we modeled the present analysis differs from other published cost-effectiveness analyses in several ways. This study examined screening of both female and male adolescents and young adults, whereas previous studies generally have focused on young women.35,36 Some analyses have focused on the diagnosis and treatment of gonorrhea alone,35 but the present study included both chlamydia and gonorrhea. The ED setting of this work is also different from the setting in other studies, such as jails and family planning clinics, which have reported higher STI prevalence.36,37
Limitations
This study has limitations. The major limitation is the use of literature-based estimates for base case parameter values and for the test characteristics of an STI risk prediction survey until the completion of the 6-site PECARN clinical trial. We tried to address the uncertainty in parameter estimates by conducting exhaustive deterministic and probabilistic sensitivity analyses. After the completion of the 4-year pragmatic trial, we will replace the literature-based parameter estimates with empirical data from the study and reanalyze the results.
Conclusions
Adolescents and young adults frequently access health care through ED visits. In this economic evaluation, the pediatric ED served as a strategic venue for examining the diagnosis and treatment of STI cases in this patient population. A decision analytic model was developed to simulate the events and outcomes associated with 3 screening strategies for chlamydial and gonococcal infection and to assess the cost-effectiveness of each strategy. Targeted screening and universally offered screening compared with no screening appeared to be cost-effective strategies for detecting and treating STIs in this population, and universally offered screening appeared to identify and successfully treat a higher proportion of STI cases.
eMethods. Detailed Description of Decision Model
eTable 1. Cost and Effectiveness of Targeted Screening Strategy as a Function of STI Risk Prediction Algorithm Cutoff for a Cohort of 10 000 Adolescents
eTable 2. Results of Alternate Scenario Analysis With High Performance STI Risk Prediction Algorithm
eFigure 1. One-Way Sensitivity Analysis of the Probability of Getting Treated
eFigure 2. Exploration of Optimal Operating Point on STI Survey Tool
eFigure 3. Cost-effectiveness Acceptability Curve
eReferences.
References
- 1.Centers for Disease Control and Prevention. Sexually transmitted diseases: adolescents and young adults. Last reviewed December 2017. Accessed November 20, 2019. https://www.cdc.gov/std/life-stages-populations/adolescents-youngadults.htm
- 2.Finelli L, Schillinger JA, Wasserheit JN. Are emergency departments the next frontier for sexually transmitted disease screening? Sex Transm Dis. 2001;28(1):40-42. doi: 10.1097/00007435-200101000-00009 [DOI] [PubMed] [Google Scholar]
- 3.Centers for Disease Control and Prevention. Incidence, prevalence, and cost of sexually transmitted infections in the United States. Published 2014. Accessed November 20, 2019. https://www.cdc.gov/nchhstp/newsroom/docs/factsheets/sti-incidence-prevalence-cost-fact-sheet-508.pdf
- 4.Centers for Disease Control and Prevention. Chlamydial infections. Last reviewed June 4, 2015. Accessed November 15, 2019. https://www.cdc.gov/std/tg2015/chlamydia.htm
- 5.Centers for Disease Control and Prevention. Gonococcal infections. Last reviewed June 4, 2015. Accessed November 15, 2019. https://www.cdc.gov/std/tg2015/gonorrhea.htm
- 6.Falasinnu T, Gilbert M, Gustafson P, Shoveller J. Deriving and validating a risk estimation tool for screening asymptomatic chlamydia and gonorrhea. Sex Transm Dis. 2014;41(12):706-712. doi: 10.1097/OLQ.0000000000000205 [DOI] [PubMed] [Google Scholar]
- 7.Reed JL, Punches BE, Taylor RG, Macaluso M, Alessandrini EA, Kahn JA. A qualitative analysis of adolescent and caregiver acceptability of universally offered gonorrhea and chlamydia screening in the pediatric emergency department. Ann Emerg Med. 2017;70(6):787-796.e2. doi: 10.1016/j.annemergmed.2017.04.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mehta SD, Bishai D, Howell MR, Rothman RE, Quinn TC, Zenilman JM. Cost-effectiveness of five strategies for gonorrhea and chlamydia control among female and male emergency department patients. Sex Transm Dis. 2002;29(2):83-91. doi: 10.1097/00007435-200202000-00004 [DOI] [PubMed] [Google Scholar]
- 9.Satterwhite CL, Torrone E, Meites E, et al. Sexually transmitted infections among US women and men: prevalence and incidence estimates, 2008. Sex Transm Dis. 2013;40(3):187-193. doi: 10.1097/OLQ.0b013e318286bb53 [DOI] [PubMed] [Google Scholar]
- 10.Ahmad FA, Jeffe DB, Plax K, et al. Computerized self-interviews improve chlamydia and gonorrhea testing among youth in the emergency department. Ann Emerg Med. 2014;64(4):376-384. doi: 10.1016/j.annemergmed.2014.01.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Goyal MK, Teach SJ, Badolato GM, Trent M, Chamberlain JM. Universal screening for sexually transmitted infections among asymptomatic adolescents in an urban emergency department: high acceptance but low prevalence of infection. J Pediatr. 2016;171:128-132. doi: 10.1016/j.jpeds.2016.01.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Goyal M, Hayes K, Mollen C. Sexually transmitted infection prevalence in symptomatic adolescent emergency department patients. Pediatr Emerg Care. 2012;28(12):1277-1280. doi: 10.1097/PEC.0b013e3182767d7c [DOI] [PubMed] [Google Scholar]
- 13.Owusu-Edusei K Jr, Chesson HW, Gift TL, et al. The estimated direct medical cost of selected sexually transmitted infections in the United States, 2008. Sex Transm Dis. 2013;40(3):197-201. doi: 10.1097/OLQ.0b013e318285c6d2 [DOI] [PubMed] [Google Scholar]
- 14.Owusu-Edusei K Jr, Nguyen HT, Gift TL. Utilization and cost of diagnostic methods for sexually transmitted infection screening among insured American youth, 2008. Sex Transm Dis. 2013;40(5):354-361. doi: 10.1097/OLQ.0b013e318285c58f [DOI] [PubMed] [Google Scholar]
- 15.Jenkins WD, Zahnd W, Kovach R, Kissinger P. Chlamydia and gonorrhea screening in United States emergency departments. J Emerg Med. 2013;44(2):558-567. doi: 10.1016/j.jemermed.2012.08.022 [DOI] [PubMed] [Google Scholar]
- 16.Burkins J, DeMott JM, Slocum GW, Gottlieb M, Peksa GD. Factors associated with unsuccessful follow-up in patients undertreated for gonorrhea and chlamydia infections. Am J Emerg Med. 2020;38(4):715-719. doi: 10.1016/j.ajem.2019.06.007 [DOI] [PubMed] [Google Scholar]
- 17.Zakher B, Cantor AG, Pappas M, Daeges M, Nelson HD. Screening for gonorrhea and chlamydia: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2014;161(12):884-893. doi: 10.7326/M14-1022 [DOI] [PubMed] [Google Scholar]
- 18.Gift TL, Owens CJ. The direct medical cost of epididymitis and orchitis: evidence from a study of insurance claims. Sex Transm Dis. 2006;33(10)(suppl):S84-S88. doi: 10.1097/01.olq.0000235149.41948.fa [DOI] [PubMed] [Google Scholar]
- 19.US Bureau of Labor Statistics. Consumer Price Index. Accessed December 1, 2019. https://www.bls.gov/cpi/data.htm
- 20.Centers for Medicare and Medicaid Services Clinical laboratory fee schedule files: 19CLABQ3. Accessed December 1, 2019. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ClinicalLabFeeSched/Clinical-Laboratory-Fee-Schedule-Files-Items/19CLABQ3
- 21.US Department of Veterans Affairs, Office of Procurement, Acquisition and Logistics (OPAL). Pharmaceutical prices. Published 2019. Accessed November 15, 2019. https://www.va.gov/opal/nac/fss/pharmPrices.asp
- 22.Centers for Medicare and Medicaid Services Physician Fee Schedule. 2019 National Physician Fee Schedule relative value file. Accessed January 16, 2019. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/PhysicianFeeSched/
- 23.Centers for Medicare and Medicaid Services Hospital Outpatient Annual Policy Files. 2019 Hospital outpatient APC payments. Accessed January 16, 2019. https://www.cms.gov/apps/ama/license.asp?file=/Medicare/Medicare-Fee-for-Service-Payment/HospitalOutpatientPPS/Downloads/2019-OPPS-APC-Offset-File.zip
- 24.Bonar EE, Walton MA, Caldwell MT, Whiteside LK, Barry KL, Cunningham RM. Sexually transmitted infection history among adolescents presenting to the emergency department. J Emerg Med. 2015;49(5):613-622. doi: 10.1016/j.jemermed.2015.02.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Goyal MK, Fein JA, Badolato GM, et al. A computerized sexual health survey improves testing for sexually transmitted infection in a pediatric emergency department. J Pediatr. 2017;183:147-152.e1. doi: 10.1016/j.jpeds.2016.12.045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mehta SD, Hall J, Lyss SB, Skolnik PR, Pealer LN, Kharasch S. Adult and pediatric emergency department sexually transmitted disease and HIV screening: programmatic overview and outcomes. Acad Emerg Med. 2007;14(3):250-258. doi: 10.1197/j.aem.2006.10.106 [DOI] [PubMed] [Google Scholar]
- 27.Schneider K, FitzGerald M, Byczkowski T, Reed J. Screening for asymptomatic gonorrhea and chlamydia in the pediatric emergency department. Sex Transm Dis. 2016;43(4):209-215. doi: 10.1097/OLQ.0000000000000424 [DOI] [PubMed] [Google Scholar]
- 28.Uppal A, Chou KJ. Screening adolescents for sexually transmitted infections in the pediatric emergency department. Pediatr Emerg Care. 2015;31(1):20-24. doi: 10.1097/PEC.0000000000000322 [DOI] [PubMed] [Google Scholar]
- 29.Marrazzo JM, Celum CL, Hillis SD, Fine D, DeLisle S, Handsfield HH. Performance and cost-effectiveness of selective screening criteria for Chlamydia trachomatis infection in women. Implications for a national chlamydia control strategy. Sex Transm Dis. 1997;24(3):131-141. doi: 10.1097/00007435-199703000-00003 [DOI] [PubMed] [Google Scholar]
- 30.Goyal MK, Shea JA, Hayes KL, et al. Development of a sexual health screening tool for adolescent emergency department patients. Acad Emerg Med. 2016;23(7):809-815. doi: 10.1111/acem.12994 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Monroe KW, Weiss HL, Jones M, Hook EW III. Acceptability of urine screening for Neisseria gonorrheae and Chlamydia trachomatis in adolescents at an urban emergency department. Sex Transm Dis. 2003;30(11):850-853. doi: 10.1097/01.OLQ.0000086600.71690.14 [DOI] [PubMed] [Google Scholar]
- 32.Papp JR, Schachter J, Gaydos CA, Van Der Pol B. Recommendations for the laboratory-based detection of Chlamydia trachomatis and Neisseria gonorrhoeae — 2014. MMWR Morb Mortal Wkly Rep. Published March 14, 2014. Accessed December 12, 2019. https://www.cdc.gov/mmwr/preview/mmwrhtml/rr6302a1.htm
- 33.US Preventive Services Task Force Final recommendation statement: chlamydia and gonorrhea. Published September 22, 2014. Accessed December 12, 2019. https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/chlamydia-and-gonorrhea-screening
- 34.Lau J, Kassirer JP, Pauker SG. Decision Maker 3.0. Improved decision analysis by personal computer. Med Decis Making. 1983;3(1):39-43. doi: 10.1177/0272989X8300300110 [DOI] [PubMed] [Google Scholar]
- 35.Aledort JE, Hook EW III, Weinstein MC, Goldie SJ. The cost effectiveness of gonorrhea screening in urban emergency departments. Sex Transm Dis. 2005;32(7):425-436. doi: 10.1097/01.olq.0000154501.22566.fa [DOI] [PubMed] [Google Scholar]
- 36.Howell MR, Quinn TC, Gaydos CA. Screening for Chlamydia trachomatis in asymptomatic women attending family planning clinics. A cost-effectiveness analysis of three strategies. Ann Intern Med. 1998;128(4):277-284. doi: 10.7326/0003-4819-128-4-199802150-00005 [DOI] [PubMed] [Google Scholar]
- 37.Kraut-Becher JR, Gift TL, Haddix AC, Irwin KL, Greifinger RB. Cost-effectiveness of universal screening for chlamydia and gonorrhea in US jails. J Urban Health. 2004;81(3):453-471. doi: 10.1093/jurban/jth130 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Detailed Description of Decision Model
eTable 1. Cost and Effectiveness of Targeted Screening Strategy as a Function of STI Risk Prediction Algorithm Cutoff for a Cohort of 10 000 Adolescents
eTable 2. Results of Alternate Scenario Analysis With High Performance STI Risk Prediction Algorithm
eFigure 1. One-Way Sensitivity Analysis of the Probability of Getting Treated
eFigure 2. Exploration of Optimal Operating Point on STI Survey Tool
eFigure 3. Cost-effectiveness Acceptability Curve
eReferences.