https://orcid.org/0000-0003-2097-8949
Figures
Abstract
Objective
To provide a summary of the economic and methodological evidence on capturing antimicrobial resistance (AMR) associated costs for curable sexually transmitted infections (STIs). To explore approaches for incorporating the cost of AMR within an economic model evaluating different treatment strategies for gonorrhoea, as a case study.
Methods
A systematic review protocol was registered on PROSPERO (CRD42022298232). MEDLINE, EMBASE, CINAHL, Cochrane Library, International Health Technology Assessment Database, National Health Service Economic Evaluation Database, and EconLit databases were searched up to August 2022. Included studies were analysed, quality assessed and findings synthesised narratively. Additionally, an economic evaluation which incorporated AMR was undertaken using a decision tree model and primary data from a randomised clinical trial comparing gentamicin therapy with standard treatment (ceftriaxone). AMR was incorporated into the evaluation using three approaches—integrating the additional costs of treating resistant infections, conducting a threshold analysis, and accounting for the societal cost of resistance for the antibiotic consumed.
Results
Twelve studies were included in the systematic review with the majority focussed on AMR in gonorrhoea. The cost of ceftriaxone resistant gonorrhoea and the cost of ceftriaxone sparing strategies were significant and related to the direct medical costs from persistent gonorrhoea infections, sequelae of untreated infections, gonorrhoea attributable-HIV transmission and AMR testing. However, AMR definition, the collection and incorporation of AMR associated costs, and the perspectives adopted were inconsistent or limited. Using the review findings, different approaches were explored for incorporating AMR into an economic evaluation comparing gentamicin to ceftriaxone for gonorrhoea treatment. Although the initial analysis showed that ceftriaxone was the cheaper treatment, gentamicin became cost-neutral if the clinical efficacy of ceftriaxone reduced from 98% to 92%. By incorporating societal costs of antibiotic use, gentamicin became cost-neutral if the cost of ceftriaxone treatment increased from £4.60 to £8.44 per patient.
Conclusions
Inclusion of AMR into economic evaluations may substantially influence estimates of cost-effectiveness and affect subsequent treatment recommendations for gonorrhoea and other STIs. However, robust data on the cost of AMR and a standardised approach for conducting economic evaluations for STI treatment which incorporate AMR are lacking, and requires further developmental research.
Citation: Ayinde O, Ross JDC, Jackson L (2023) Economic evaluation of antimicrobial resistance in curable sexually transmitted infections; a systematic review and a case study. PLoS ONE 18(10): e0292273. https://doi.org/10.1371/journal.pone.0292273
Editor: Kwame Kumi Asare, University of Cape Coast, GHANA
Received: March 7, 2023; Accepted: September 15, 2023; Published: October 19, 2023
Copyright: © 2023 Ayinde et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: JDCR reports personal fees from GSK Pharma and Hologic Diagnostics, as well as ownership of shares in GSK Pharma and Astrazeneca Pharma; and is author of the UK and European Guidelines on Pelvic Inflammatory Disease; is a Member of the European Sexually Transmitted Infections Guidelines Editorial Board. He is an NIHR Journals Editor and associate editor of Sexually Transmitted Infections journal. He is an officer of the International Union against Sexually Transmitted Infections (treasurer). LJ and OA report no conflicts of interest. This does not alter our adherence to PLOS ONE policies on sharing data and materials.
Introduction
Chlamydia, gonorrhoea, syphilis, and trichomoniasis are the most prevalent sexually transmitted infections (STIs) globally, with a reported incidence of 374 million infections in 2020 [1]. Though mostly curable, many of these infections are becoming increasingly resistant to first-line treatments [2, 3]. For instance, gonorrhoea, the world’s second most prevalent bacterial STI has progressively developed resistance over the last 80 years to a wide variety of antibiotic regimens including sulfonamides, penicillins, tetracyclines, macrolides, and fluoroquinolones [4]. In many countries, first line therapy for gonorrhoea is limited to extended-spectrum cephalosporins, such as cefixime and ceftriaxone [5]. Consequently, the World Health Organization (WHO) has identified gonorrhoea as one of the top 12 priority pathogens for new antibiotic research and development [6].
The pipeline for new drugs to treat emerging resistance is limited due to high development costs and challenges around formulation, regulation, and profitability [7]. Therefore, there is an urgent need for strategies to limit the development of antimicrobial resistance (AMR). These include the use of existing ‘older’ antibiotic treatments (where possible) or changes in care pathways to slow the spread of AMR. A number of studies have assessed the efficacy of alternative antibiotics [8, 9], and explored the use of AMR guided therapy and the role of antibiotic stewardship [10–12]. However, there is little evidence regarding the costs associated with AMR for curable STIs and how this affects the cost-effectiveness of interventions designed to reduce AMR [13–16]. This limits comprehensive evaluation of AMR control strategies and the cost implications for patient management, which are important to clinicians, commissioners and policy makers when developing new management guidance [17, 18]. In addition, data on the economic impact of AMR is needed to direct targeted investment into future drug development.
Therefore, we (i) performed a systematic review to appraise the economic evidence relating to AMR for curable STIs, and provide a comprehensive overview of the methods currently used to incorporate AMR into economic evaluations of treatments in patients with curable sexually transmitted infections, and (ii) used a case study to explore how AMR could be incorporated into an economic evaluation using a decision-analytic model incorporating data from a large pragmatic multicentre randomised clinical trial (RCT) which recruited patients with gonorrhoea.
Methods
Systematic review
A systematic review was conducted following the guidelines of the Centre for Review and Dissemination [19] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [20] to answer the research question—What evidence is available on the costs associated with AMR in curable STIs and what methods have been adopted to include such costs in economic evaluations? A systematic review protocol was developed and registered with PROSPERO at the CRD, University of York (Registration No CRD42022298232)— https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022298232.
Eligibility criteria.
Studies were considered eligible for review if they met the following criteria:
- P–the population consisted of men or women of any age with a curable STI (gonorrhoea, chlamydia, syphilis, trichomoniasis, Mycoplasma genitalium)
- I–the intervention was treatment of the STI
- C–the comparator was licensed or unlicensed pharmacological or non-pharmacological treatments of the STI
- O–outcomes were reported in terms of costs, economic evaluations and health outcomes. We also evaluated approaches to modelling, modelling assumptions and proxy outcomes associated with AMR.
There was no restriction to the study setting or the publication date. However, publications were restricted to those in the English language.
Search strategy.
Scoping searches were initially carried out to refine the search strategy. Thereafter, the following databases were searched from inception to August 2022. MEDLINE, EMBASE, British Nursing Index and Cumulative Index Nursing and Allied Health Literature, Cochrane Library, International Health Technology Assessment Database, National Health Service Economic Evaluation Database, and EconLit. In addition, relevant websites related to STIs and economic evaluations were searched; the National Institute for Health and Care Excellence, Cost-Effectiveness Analysis (CEA) Registry, Research Papers in Economics, WHO, Public Health England, Centers for Disease Control and Prevention (CDC), the European Centre for Disease Prevention and Control, and the British Association for Sexual Health and HIV. Reference lists of included selected studies were hand searched. The database search strategy is reported in S1 File.
Study selection.
All identified records were transferred to EndNote referencing manager (V.X9) for management and categorisation. A two-stage process as outlined by Roberts [21] was used to select studies. In stage I, titles and abstract were screened and assigned into categories A to G. In stage II, full text articles of studies categorised A to C, were further categorised into 1–8 (S2 File). The identification and initial categorisation were performed by two reviewers. A third reviewer checked subsets of the selection process (screening, eligibility and inclusion) to confirm the categorisation of studies. Studies categorised A to C and grouped into 1 to 3 were included in the review (Fig 1).
Data extraction and synthesis.
Data extraction was performed independently by two reviewers on all included studies. Disagreements were resolved by discussion and by a third independent reviewer. The extracted data were tabulated (data extraction form, S1 Table) and synthesised narratively. A narrative synthesis was adopted this being the most appropriate approach for bringing together studies with heterogeneous methodologies [19].
Quality assessment.
The quality of included studies was assessed using an adapted version of the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist [22] and the checklist of cost of illness [23] for economic evaluations and cost studies respectively (S3 File). Quality assessment was used to inform the synthesis but no studies were excluded on the basis of quality.
Economic evaluation
For the second element of this study, we used the findings of the systematic review to explore different methods to incorporate the costs associated with AMR within an economic evaluation. For this component we adapted an existing model developed for a RCT concerned with different treatment strategies for gonorrhoea. The methods and results of the ‘Gentamicin compared with ceftriaxone for the treatment of gonorrhoea RCT (GToG)’ RCT are reported elsewhere [24]. In brief, a blinded, non-inferiority RCT was conducted in 14 sexual health clinics in England. 720 adult sexual health clinic attendees with uncomplicated gonorrhoea were randomised 1:1 to receive either gentamicin 240 mg or ceftriaxone 500 mg, both administered as a single intramuscular injection. All participants also received 1 g oral azithromycin. The primary outcome was clearance of Neisseria gonorrhoeae at all initially test positive sites two weeks after treatment. The GToG RCT was approved by Health Research Authority South Central–Oxford C Research Ethics Committee (14/SC/1030). The trial was registered prior to start of recruitment (ISRCTN51783227).
Model structure.
A simple decision tree model was developed using TreeAge Pro 2016 (TreeAge Software Inc., Williamstown, MA, USA). The structure was informed by the trial objectives and patient pathways indicated by the clinical data. Patients entered the model at the point of randomisation when they were assigned to receive the alternative treatment (gentamicin) or the standard treatment (ceftriaxone). Following the initial course of antibiotic treatment, patients either received additional NHS care (e.g. General Practitioner—GP visit) or did not access care. At two weeks post-treatment, patients were either cleared of Neisseria gonorrhoeae (confirmed by a negative nucleic acid amplification test [NAAT]) or they were not cleared and required further treatment (Fig 2). The economic analysis focused on establishing if gentamicin compared to ceftriaxone was cost neutral in the treatment of gonorrhoea, which was deemed to be most relevant for a non-inferiority trial. The analysis was undertaken from the perspective of the health service (NHS). For the initial economic evaluation, the time horizon was two weeks and restricted to the follow-up time within the RCT.
Data sources.
Clinical data on the primary outcome (microbial resolution), resource use and costs were collected prospectively via trial reporting mechanisms (Tables 1 and 2). Data on additional resource use after initial treatment and prior to the 2-week test of cure was also collected prospectively (Table 3). Additional data relating to unit costs and scenarios on further treatment due to non-clearance of infection at the 2-week test of cure were sourced from the literature (Table 4).
Initial analysis.
The cost per patient successfully treated (measured in terms of microbial clearance of Neisseria gonorrhoeae at all infected sites) was estimated. As the trial was concerned with the immediate post-treatment period (two weeks), discounting was not undertaken. All costs are given in £UK for 2020/2021. A probabilistic sensitivity analysis was undertaken to explore uncertainties. All parameters were varied simultaneously sampling multiple sets of parameter values from defined probability distributions. A Monte Carlo simulation was used to sample from the distributions; this involved 1000 repeated random draws to analyse how variation in the parameters used in the model would affect the results. For binomial data, beta distributions were used, and gamma distributions were used for costs, in line with recommendations for specifying distributions for parameters [27].
Secondary analysis—accounting for antimicrobial resistance.
Building on the results of the systematic review of the literature, we employed three approaches to account for AMR—(i) including the additional costs associated with treating resistant infections [13], (ii) using a threshold analysis to assess the level of resistance required to impact on cost to potentially change the decision [15], and (iii) estimating and including the societal cost of resistance for each antibiotic consumed [15, 18]. These approaches were analysed in the context of alternative antibiotic treatments for gonorrhoea.
- Additional costs associated with treating resistance to ceftriaxone: We identified an appropriate treatment pathway which would be followed if an infection was resistant to ceftriaxone (Fig 3). This was informed by current and proposed guidelines for the treatment of gonorrhea at the time [28, 29]. The additional cost included cost of additional clinic visits, and additional courses of antibiotics (following guidance, we assumed that gentamicin would be used as treatment if the infection was resistant to ceftriaxone) (Table 5).
- Threshold analysis on the level of ceftriaxone resistance necessary for gentamicin to be cost-neutral: this involved varying the value of a parameter that was important to the analysis and assessing what value estimate would cause a programme or intervention to be cost-effective or not cost-effective (or cost-neutral in this case) [30]. The level of gonorrhoea ceftriaxone resistance was varied as this was judged to be critical and would affect the context in which decision-making around treatment takes place.
- Including the societal cost of resistance: We applied the societal costs associated with AMR to the ceftriaxone treatment arm only. This is because the extended-spectrum cephalosporins, cefixime and ceftriaxone are the only remaining monotherapy that can be used effectively to treat gonorrhoea. Therefore, in reality the costs associated with resistance in ceftriaxone would be higher for society. The costs associated with ceftriaxone resistance for each course of antibiotic consumed as reported by Shrestha et al. [18] ranged between $1 and $25.6. These estimates were converted to UK pounds and inflated.
For this secondary analysis, gonorrhea AMR was defined as persistent gonorrhea infection determined by a positive gonorrhea NAAT test at the 2-week test of cure at anatomical sites that were previously tested. Further assumptions made for the initial and secondary analyses are reported in S4 File.
Results
Systematic review findings
Study selection.
3,402 studies were identified from the database search and 232 duplicate records were removed. Using the 2-stage process to screen and identify eligible studies, in Stage I titles and abstracts of 3170 records were screened for relevance, of these, 3001 articles were excluded and 169 relevant full-text articles were further assessed for eligibility in Stage II. Following full text analysis, 159 were excluded and 12 full text articles—10 identified from the database search and an additional 2 full text articles identified by hand searching were included in the review (Fig 1).
Study characteristics.
The characteristics of the included studies are summarised in Table 6. Overall, 7/12 studies were economic evaluations of which most employed a cost-effectiveness design (6/7). 5/12 studies were cost studies including a national action plan and cost-analysis, cost minimisation and cost modelling studies. The study population was heterogeneous, including adolescents and adults, men and women, national populations and specific groups such as sex workers, pregnant women or men who have sex with men (MSM). 7/12 included studies which originated from the United States and 11/12 were specific to gonorrhoea infection.
The aims of the included studies varied, including estimating the costs or cost-effectiveness of AMR testing strategies (6/12), cost projections for emerging AMR (2/12), and evaluating interventions for STI management with AMR as an adjunct to the analysis (4/12). Almost all of the included studies (11/12) reported only direct medical costs (Table 6).
Review of the cost of AMR.
The economic consequences of AMR identified from the review can be presented in four main themes—(i) the projected cost of emerging gonorrhoea cephalosporin resistance, (ii) cost of dual therapy in managing gonorrhoea AMR, (iii) cost associated with AMR testing and targeted antibiotic treatment and (iv) threshold parameters associated with AMR testing.
- Costs of emerging gonorrhoea AMR to cephalosporin: Two cost studies from the United States found the cost of emerging gonorrhoea AMR to first-line cephalosporin therapy over a 10-year period could be substantial (Table 7). Cost projections from the CDC suggests that widespread gonorrhoea cephalosporin resistance resulted in additional cases of pelvic inflammatory disease, epididymitis and HIV infections with a cumulative direct medical cost of at least $235 million [36]. Similarly, Chesson et al. [38] estimated that emerging gonorrhoea ceftriaxone resistance (rising from 2% to 15%) could cost $378.2 million, resulting from 1,157,100 additional gonorrhoea infections and 579 gonorrhoea-attributable HIV infections (Table 7).
- Costs of dual therapy in managing gonorrhoea AMR: In response to the evolving gonorrhoea ceftriaxone resistance, some guidelines recommend a dual antibiotic treatment regimen [41]. Xiridou et al. [37] investigated the cost-effectiveness of dual therapy (ceftriaxone plus azithromycin) compared to ceftriaxone monotherapy using a transmission model, and found that dual therapy slowed down the spread of resistance (5% resistance threshold) by at least 15 years in MSM, but resulted in additional treatment cost over 60 years (cumulative ICER- cost/QALY of €9.74 x 108 and €14866 for 10 and 60 years respectively) (Table 8). The analysis was based on the cost of medical consultation, testing, treatment, quality of life loss for symptomatic gonorrhoea, duration of infection and 10% prevalence of gonorrhoea complications in symptomatic patients. The cost-effectiveness estimate was sensitive to the costs of consultations, tests and the weighting for quality of life loss. When there was initial azithromycin resistance (5%), dual therapy was not cost-effective and unlikely to preserve ceftriaxone use.
- Cost of gonorrhoea AMR testing and targeted antibiotic treatment: The costs of AMR testing and antibiotic stewardship was investigated in five studies. Three studies [11, 12, 39] compared AMR testing (point-of-care-test POCT and AMR screening) to improve antibiotic stewardship with dual therapy (ceftriaxone plus azithromycin) for gonorrhoea. These studies consistently found that AMR testing and treatment cost more compared to dual therapy (Table 8). In one study, the additional cost for optimal treatment (using a ceftriaxone sparing regimen) was at least £414.7, and all strategies which involved dual resistance testing (ciprofloxacin and azithromycin) were dominated by standard of care. The results were associated with a base case cost of £29.00 and £31.90 for single and dual resistance testing respectively [39]. Similarly, Turner et al. [11] found gonorrhoea AMR POCT (gonorrhoea ciprofloxacin and/or penicillin resistance) led to an additional cost of £34 million per annum compared to the standard of care, assuming AMR POCT added £25 to the first-line testing cost and that 66% of ceftriaxone treatments could be replaced by ciprofloxacin annually. A screening test for gonorrhoea ciprofloxacin resistance (assessing DNA gyrase—gyrA) followed by targeted therapy with ciprofloxacin or ceftriaxone plus azithromycin was estimated to cost an additional $54.40 (minimum $12.40) per patient compared to the standard of care (2-drug ceftriaxone and azithromycin therapy) in a retrospective study [12]. These costs were based on a test cost of $100.50 (test and labour), 35.0% prevalence of ciprofloxacin resistance, and 30.3% incidence of indeterminate genotype test results. In contrast, prior to the introduction of ceftriaxone as first line treatment for gonorrhoea treatment, gonorrhoea AMR testing and targeted therapy was found to be cost-effective [32, 34]. Beta-lactamase (an enzyme associated with resistance to ß-lactam antibiotics such as penicillin) screening and antibiotic stewardship was cost-effective compared to empirical treatment with ceftriaxone, assuming penicillin resistance at 5% and a $0.50 per ß-lactamase screening test [32]. Roy et al. [34] also found that culture-based testing strategies (strategies 1 and 3) were optimal (lowest cost per patient successfully treated) at lower levels of gonorrhoea prevalence (≤ 5%)—Table 8.
- Threshold parameters associated with gonorrhoea AMR testing: Two studies reported on threshold analyses in relation to AMR. Wynn and Klausner [40] found the breakeven price for using a gyrA assay to detect ciprofloxacin resistance followed by targeted antibiotic treatment compared to standard dual therapy was $50 per gyrA test for the treatment of asymptomatic patients being screened for STIs, assuming a gonorrhoea prevalence of 2%, ciprofloxacin susceptibility of 70%, and gyrA assay sensitivity and specificity of 98 and 99%. The breakeven cost was sensitive to the prevalence of ciprofloxacin susceptibility, the cost of standard of care, and the frequency of indeterminant gryA test findings, but insensitive to gonorrhoea prevalence (2% or 8%). A threshold gonorrhoea prevalence of 2.5% was reported by Phillips et al. [31] for reducing the direct medical costs of gonorrhoea management in women when comparing routine gonorrhoea culture and susceptibility testing ($9 per culture test) with ‘no’ testing. However, the threshold prevalence was sensitive to the cost of culture tests and risk of adverse sequelae after treatment.
Review of the methodological approaches for inclusion of AMR in economic evaluations.
The following key methodological elements and strategies were associated with incorporating AMR into cost/economic evaluations.
- Definition of gonorrhoea AMR: AMR was defined in broad terms in some studies and more specifically in others (Tables 7 & 8). Broad definitions included gonorrhoea cephalosporin resistance [36], ceftriaxone resistance [38], ciprofloxacin resistance or penicillin resistance [11], and gonorrhoea treatment failure [35, 37]. When AMR was defined in this manner it was often unclear what specific method (treatment failure or antibiotic susceptibility testing with minimum inhibitory concentration breakpoints) was used to establish resistance because AMR was frequently linked to published epidemiological estimates from sentinel surveillance programmes without providing additional detail. Other studies used gonorrhoea resistance specific genetic/molecular phenotypes such as mutant gyrA for ciprofloxacin resistance [12, 39, 40], 23SrRNA and mtrCDE for azithromycin resistance [39] or culture for penicillinase producing Neisseria gonorrhoeae [31] to define resistance.
- Method of gonorrhoea AMR incorporation into the economic evaluation: The most common approach for incorporating AMR was to account for the additional cost of AMR testing and/or treatment [11, 12, 37, 39]. Threshold analyses were also reported to identify the breakeven cost of using AMR POCT [40] and the threshold prevalence of gonorrhoea for using culture based testing [31]. Mostly, a baseline prevalence of resistance to certain drugs in a base case scenario was incorporated in the main model, and then altered in a sensitivity analysis [32–34, 38, 40]. Others made assumptions around the baseline AMR test accuracy [12, 39] and cost per test [12, 32], and varied these in the sensitivity analysis [32, 39]. One study utilised a more complex methodology by creating two scenarios—one with and the other without baseline azithromycin resistance, and developed a gonorrhoea transmission model which was then used to inform the economic model [37].
- Gonorrhoea AMR assumptions: Variable but inconsistent assumptions about resistance were made, including the prevalence of gonorrhoea AMR (in 9/13 studies [11, 32–34, 36–38, 40, 42]), what therapy would be used when AMR was present (in 4/12 studies [12, 32, 34, 37]), and the degree of clinical effectiveness for modelled antibiotics (in 2/12 studies [35, 39]).
- Study reported limitations: Of the reported study limitations, the transferability of epidemiological data on AMR prevalence in one area to other geographical locations, and uncertainty in predicting future AMR prevalence were acknowledged [38, 39]. The use of static economic models which do not take into account the complexities of STI transmission dynamics were common [11, 33, 38]. The limited settings in which AMR POCT can be utilised due to cost and required technical expertise were acknowledged [39] and the lack of data on how introducing new treatment strategies might affect the evolution of gonorrhoea AMR was also identified [11].
- Study perspectives and horizons: Of the eight studies which reported a perspective, all took the perspective of the provider (healthcare system, sector or government). The study horizons ranged from the time of treatment (most commonly) to a 60 year time period (Tables 7 & 8).
Quality assessment.
The reporting quality of both the cost and economic evaluation studies varied. For cost studies, a sensitivity analysis was included in 2/5 studies, and the descriptions of the data sources, analytical methods, valuation techniques utilised and the method used to estimate the costs associated with AMR was limited in one of these [36] (S3 File). For the economic evaluations, the study population, choice of model and assumptions were appropriately reported, however information on study perspective, horizon, discount rate and study limitations was often limited (S3 File).
Case study accounting for AMR cost in the treatment of gonorrhoea
Initial analysis without accounting for AMR.
In the GToG trial a higher proportion of patients treated with ceftriaxone compared with gentamicin had microbiological cure at the 2-week follow up (98% vs 91%). The average cost per patient treated with gentamicin was £13.25, compared with £9.41 for those treated with ceftriaxone. The higher cost of gentamicin treatment was due to the cost of additional consultations and treatment of patients with persistent infection at the 2-week follow up. Treatment with gentamicin was therefore not non-inferior to ceftriaxone and it was not cost-neutral (Table 9). Fig 4 shows the results of the probabilistic sensitivity analysis involving 1000 simulations. Most of the points were in the top left-hand quadrant, indicating that treatment with ceftriaxone dominated treatment with gentamicin confirming that gentamicin is not shown to be non-inferior and is unlikely to be cost-neutral.
Analysis accounting for AMR.
On Table 10, we show the results when accounting for AMR in the economic evaluation of an alternative treatment such as gentamicin compared to ceftriaxone, as the current standard of care, for the treatment of gonorrhoea.
- The potential additional cost of treating a patient with ceftriaxone resistant gonorrhoea was estimated to be £190.54 (Table 5). This cost was applied to the trial data, assuming that all those who experienced a treatment failure in the ceftriaxone arm had a resistant strain of gonorrhoea. As expected, including the potential costs of resistance, for those who were not successfully treated during their initial treatment, increased the overall costs per patient treated in this arm (Table 10). However, as we assumed that only those who had an initial treatment failure would experience the additional costs associated with resistance, ceftriaxone remained the cheaper treatment.
- In a threshold analysis we varied the rates of clearance of infection for ceftriaxone. This demonstrated that clearance rates for gentamicin would need to be higher than those for ceftriaxone for the treatment to be cost-neutral, due to the higher initial treatment costs associated with gentamicin treatment. If the clearance rate for ceftriaxone dropped to 92% then gentamicin became a cost-neutral treatment.
- To assess the societal cost of resistance we applied the range of costs estimated for ceftriaxone resistance from a previous study [18] to those in the ceftriaxone arm of the trial which resulted in the average cost per patient ranging from £10.21 to £29.94. Our analysis determined that the cost of ceftriaxone would need to increase to £8.44 in order for treatment with gentamicin to be cost neutral.
Discussion
We conducted a systematic review of the economic evidence for measuring AMR in curable STIs and developed a preliminary economic evaluation model which incorporates gonorrhoea AMR to assess an alternative antimicrobial regimen for gonorrhoea treatment. We found a small number of studies were eligible for inclusion, suggesting that the existing evidence base is limited. The majority of the included studies (11/12) related to gonorrhoea, possibly reflecting the high prevalence and clinical importance of gonorrhoea resistance [4].
Cost estimates from the United States indicate that the cost of gonorrhoea resistance to current first-line treatment (ceftriaxone) is substantial, equating to $235 to $378.2 million in direct medical costs over 10 years [36, 38]. These estimates are likely to be significantly higher in low and middle income settings [18], and if indirect costs and the societal implications of AMR were included. Gonorrhoea resistance testing and improved antibiotic stewardship were frequently reported as strategies to extend the time period over which ceftriaxone remains effective by restricting its widespread use. A 70% reduction in ceftriaxone use could potentially be achieved using molecular based AMR point-of-care testing to guide the use of quinolone therapy but this approach was not found to be cost-effective when compared to ceftriaxone/azithromycin dual therapy [11, 12, 39, 40]. Cost-effectiveness was reduced further if there was an increase in the prevalence of quinolone resistance or cost of point-of-care tests, or if AMR test performance was lower than predicted. However, these analyses were restricted to evaluating direct medical costs, had a short time horizon, used static models, and did not include the broader societal impact of AMR which limits their interpretation.
We found the methodology used in existing cost/economic evaluations varied considerably, including in terms of how AMR was defined, assumed projections of AMR prevalence, costs included, and the perspective adopted. How AMR is defined presents a specific challenge since molecular, microbiological and clinical definitions of resistance can vary significantly and with no single ‘gold standard’ measure [43]. The modelling assumptions around AMR in gonorrhoea were mainly related to the prevalence of resistance, cost of treatment and proposed treatment pathway when resistance occurred. AMR prevalence estimates were obtained from a variety of sources which could have been influenced by laboratory methodology, frequency of testing and/or the underlying healthcare delivery system [44]. Cost estimates were sensitive to changes in AMR prevalence in most studies highlighting the importance of surveillance systems which are representative of the general population and utilise a robust methodology. The cost of treatment was sourced from either primary and/or secondary sources, but the assumptions around patient management pathways when AMR was suspected varied significantly with no commonly accepted ‘best’ approach for managing patients who had failed first line therapy. The most frequent perspective adopted was that of the healthcare provider, but this may underestimate the total cost of resistance [16] by not including the full cost of morbidity, loss of income, reduced productivity and use of antibiotic prophylaxis. A more consistent and comprehensive approach to incorporate patient and societal costs of AMR is therefore desirable [14, 16, 42].
Our case study explored different approaches to incorporate AMR using data from a recent large multicentre RCT in which patients were treated with ceftriaxone or gentamicin [24]. We conducted a threshold analysis and found gentamicin treatment to be cost neutral if the failure rate for ceftriaxone increased to 10% (from 2%) which is consistent with reports by Wynn and Klausner [40] who found the cost neutrality for ciprofloxacin resistance genetic testing was dependent on the prevalence of ciprofloxacin resistance but not gonorrhoea prevalence. A limitation of most previous studies was the lack of inclusion of societal costs of AMR. Given that ceftriaxone is the only remaining reliably effective therapy for gonorrhoea, and that continued use increases the risk of subsequent ceftriaxone AMR development, we assessed the societal cost—i.e. direct and indirect costs, resistance modulating factors and rate of consumption of antibiotics that drive resistance [18]. We identified that if the cost of ceftriaxone was increased to £8.44 from £4.60, treatment with gentamicin would become cost-neutral. By accounting for gonorrhoea AMR via multiple approaches and using prospective data, it is therefore possible to provide additional useful information for decision-makers about when alternative antibiotics might be considered as a replacement for standard treatment.
We recognise a number of potential limitations. The systematic review was restricted to studies published in English, and the search terms chosen were based on scoping searches and prior knowledge of the literature. The use of additional search terms may have increased the number of records returned but would have made the number of records requiring review unfeasible and was considered unlikely to identify other highly relevant studies. For the economic evaluation of gentamicin as an alternative to ceftriaxone, a static rather than dynamic model structure was adopted. Although a dynamic model would have allowed the wider impacts of resistance to be considered, it was beyond the scope of this exploratory study which aimed to assess different approaches for incorporating the effects of AMR. A further limitation was the short follow up period associated with the clinical trial, which did not allow data to be gathered on the longer-term management of patients who had failed treatment. We addressed this by assuming that management would follow current clinical management pathways. Also, as our case study was conducted in a high income setting, the findings may not be generalizable to low income settings but contributes to the debate on how future economic evaluations could more fully incorporate the economic impact of AMR for curable STIs.
There is a dearth of data in relation to the economic analysis of AMR for curable STIs. However, by incorporating AMR costs (including in our exemplar study on alternative treatment for gonorrhoea) more robust interpretations in relation to the costs/cost effectiveness of new or alternative treatment strategies for curable STIs can be made. The current evidence relating to the economic impact of AMR for curable STIs is generally limited to direct, aggregated costs over a short period, obtained from high income countries, is largely specific to gonorrhoea AMR, and often lacks a societal perspective. There is no standardised approach for the measurement of AMR in patients with curable STIs which results in uncertainty when reporting on the economic impact of AMR and makes comparisons between studies difficult. Further research is required to inform guidance on optimal approaches to capture AMR costs for curable STIs (e.g. how AMR is defined and how AMR associated costs are adequately captured) and methodically incorporate such costs into economic evaluations.
Supporting information
S3 File. Quality assessment of included studies.
https://doi.org/10.1371/journal.pone.0292273.s003
(DOCX)
S4 File. Assumptions for the economic evaluation comparing gentamicin to ceftriaxone for the treatment of gonorrhoea.
https://doi.org/10.1371/journal.pone.0292273.s004
(DOCX)
Acknowledgments
The authors thank the whole GToG team, Pollyanna Ford, as her MSc dissertation helped lay the foundations for aspects of this paper, and Lena Schnitzler for providing the English translation for the Müller et al. [23] checklist for cost-of-illness.
References
- 1.
WHO. Global progress report on HIV, viral hepatitis and sexually transmitted infections, 2021. WHO, 2021 15 July 2021.
- 2. Tien V, Punjabi C, Holubar MK. Antimicrobial resistance in sexually transmitted infections. J Travel Med. 2020;27(1). Epub 2019/12/17. pmid:31840758.
- 3. Sena AC, Lensing S, Rompalo A, Taylor SN, Martin DH, Lopez LM, et al. Chlamydia trachomatis, Mycoplasma genitalium, and Trichomonas vaginalis infections in men with nongonococcal urethritis: predictors and persistence after therapy. J Infect Dis. 2012;206(3):357–65. Epub 2012/05/23. pmid:22615318; PubMed Central PMCID: PMC3490700.
- 4. Unemo M, Shafer WM. Antimicrobial resistance in Neisseria gonorrhoeae in the 21st century: past, evolution, and future. Clin Microbiol Rev. 2014;27(3):587–613. Epub 2014/07/02. pmid:24982323; PubMed Central PMCID: PMC4135894.
- 5. Unemo M, Lahra MM, Escher M, Eremin S, Cole MJ, Galarza P, et al. WHO global antimicrobial resistance surveillance for Neisseria gonorrhoeae 2017–18: a retrospective observational study. Lancet Microbe. 2021;2(11):e627–e36. Epub 2022/05/12. pmid:35544082.
- 6. Tacconelli E, Carrara E, Savoldi A, Harbarth S, Mendelson M, Monnet DL, et al. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis. 2018;18(3):318–27. Epub 2017/12/26. pmid:29276051.
- 7. Piddock LJ. The crisis of no new antibiotics—what is the way forward? Lancet Infect Dis. 2012;12(3):249–53. Epub 2011/11/22. pmid:22101066.
- 8. de Vries HJC, de Laat M, Jongen VW, Heijman T, Wind CM, Boyd A, et al. Efficacy of ertapenem, gentamicin, fosfomycin, and ceftriaxone for the treatment of anogenital gonorrhoea (NABOGO): a randomised, non-inferiority trial. Lancet Infect Dis. 2022;22(5):706–17. Epub 2022/01/23. pmid:35065063.
- 9. Rob F, Klubalova B, Nycova E, Hercogova J, Unemo M. Gentamicin 240 mg plus azithromycin 2 g vs. ceftriaxone 500 mg plus azithromycin 2 g for treatment of rectal and pharyngeal gonorrhoea: a randomized controlled trial. Clin Microbiol Infect. 2020;26(2):207–12. Epub 2019/08/17. pmid:31419483.
- 10. Lee CR, Cho IH, Jeong BC, Lee SH. Strategies to minimize antibiotic resistance. Int J Environ Res Public Health. 2013;10(9):4274–305. Epub 2013/09/17. pmid:24036486; PubMed Central PMCID: PMC3799537.
- 11. Turner KM, Christensen H, Adams EJ, McAdams D, Fifer H, McDonnell A, et al. Analysis of the potential for point-of-care test to enable individualised treatment of infections caused by antimicrobial-resistant and susceptible strains of Neisseria gonorrhoeae: a modelling study. BMJ Open. 2017;7(6):e015447. Epub 2017/06/16. pmid:28615273; PubMed Central PMCID: PMC5734280.
- 12. Allan-Blitz LT, Hemarajata P, Humphries RM, Wynn A, Segura ER, Klausner JD. A Cost Analysis of Gyrase A Testing and Targeted Ciprofloxacin Therapy Versus Recommended 2-Drug Therapy for Neisseria gonorrhoeae Infection. Sex Transm Dis. 2018;45(2):87–91. Epub 2018/01/13. pmid:29329176; PubMed Central PMCID: PMC6879096.
- 13. Smith R, Coast J. The true cost of antimicrobial resistance. BMJ. 2013;346:f1493. Epub 2013/03/13. pmid:23479660.
- 14. Holmes EAF, Hughes DA. Challenges for Economic Evaluation of Health Care Strategies to Contain Antimicrobial Resistance. Antibiotics (Basel). 2019;8(4):166. Epub 2019/10/02. pmid:31569624; PubMed Central PMCID: PMC6963561.
- 15. Oppong R, Smith RD, Little P, Verheij T, Butler CC, Goossens H, et al. Cost effectiveness of amoxicillin for lower respiratory tract infections in primary care: an economic evaluation accounting for the cost of antimicrobial resistance. Br J Gen Pract. 2016;66(650):e633–9. Epub 2016/07/13. pmid:27402969; PubMed Central PMCID: PMC5198702.
- 16. Jit M, Ng DHL, Luangasanatip N, Sandmann F, Atkins KE, Robotham JV, et al. Quantifying the economic cost of antibiotic resistance and the impact of related interventions: rapid methodological review, conceptual framework and recommendations for future studies. BMC Med. 2020;18(1):38. Epub 2020/03/07. pmid:32138748; PubMed Central PMCID: PMC7059710.
- 17. Coast J, Smith RD, Millar MR. Superbugs: should antimicrobial resistance be included as a cost in economic evaluation? Health Econ. 1996;5(3):217–26. Epub 1996/05/01. pmid:8817296.
- 18. Shrestha P, Cooper BS, Coast J, Oppong R, Do Thi Thuy N, Phodha T, et al. Enumerating the economic cost of antimicrobial resistance per antibiotic consumed to inform the evaluation of interventions affecting their use. Antimicrob Resist Infect Control. 2018;7(1):98. Epub 2018/08/18. pmid:30116525; PubMed Central PMCID: PMC6085682.
- 19.
Akers J, Aguiar-Ibáñez R, Baba-Akbari A. Systematic reviews: CRD’s guidance for undertaking reviews in health care: York, UK: Centre for Reviews and Dissemination, University of York; 2009.
- 20. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. PLoS Med. 2021;18(3):e1003583. Epub 2021/03/30. pmid:33780438.
- 21. Roberts T, Henderson J, Mugford M, Bricker L, Neilson J, Garcia J. Antenatal ultrasound screening for fetal abnormalities: a systematic review of studies of cost and cost effectiveness. BJOG. 2002;109(1):44–56. Epub 2002/02/15. pmid:11843373.
- 22. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. BMC Med. 2013;11(1):80. Epub 2013/03/28. pmid:23531108; PubMed Central PMCID: PMC3607979.
- 23. Müller D, Stock S, Dintsios C-M, Chernyak N, Gerber-Grote A, Gloede TD, et al. Checkliste zur Erstellung und Bewertung von Krankheitskostenstudien. Gesundheitswesen. 2018;80(08/09):744–53. Epub 18.05.2017.
- 24. Ross JDC, Brittain C, Cole M, Dewsnap C, Harding J, Hepburn T, et al. Gentamicin compared with ceftriaxone for the treatment of gonorrhoea (G-ToG): a randomised non-inferiority trial. Lancet. 2019;393(10190):2511–20. Epub 2019/05/06. pmid:31056291; PubMed Central PMCID: PMC6620599.
- 25. BNF. Drugs A to Z 2022 [30/11/2022]. Available from: https://bnf.nice.org.uk/drugs/.
- 26. Jones K, Burns A. Unit costs of health and social care 2021. Kent: Personal Social Services Research Unit, 2021.
- 27. Briggs AH, Weinstein MC, Fenwick EA, Karnon J, Sculpher MJ, Paltiel AD, et al. Model parameter estimation and uncertainty analysis: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force Working Group-6. Med Decis Making. 2012;32(5):722–32. Epub 2012/09/20. pmid:22990087.
- 28. Bignell C, Fitzgerald M, Guideline Development Group British Association for Sexual Health and HIV. UK national guideline for the management of gonorrhoea in adults, 2011. Int J STD AIDS. 2011;22(10):541–7. Epub 2011/10/15.
- 29. Fifer H, Saunders J, Soni S, Sadiq ST, FitzGerald M. 2018 UK national guideline for the management of infection with Neisseria gonorrhoeae. Int J STD AIDS. 2020;31(1):4–15. Epub 2019/12/25. pmid:31870237.
- 30.
Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. Oxford: Oxford: Oxford University Press; 2015.
- 31. Phillips RS, Safran C, Aronson MD, Taylor WC. Should women be tested for gonococcal infection of the cervix during routine gynecologic visits? An economic appraisal. Am J Med. 1989;86(3):297–302. Epub 1989/03/01. pmid:2493192.
- 32. Nettleman MD, Smith V, Moyer NP. Penicillin resistant Neisseria gonorrhoeae in low prevalence areas: implications for cost-effective management. Sex Transm Dis. 1990;17(4):175–80. Epub 1990/10/01. pmid:2124730.
- 33. Crabbe F, Vuylsteke B, de Clerck M, Laga M. Cost-effectiveness of management strategies for acute urethritis in the developing world. Trop Med Int Health. 2000;5(9):640–7. Epub 2000/10/24. pmid:11044279.
- 34. Roy K, Wang SA, Meltzer MI. Optimizing treatment of antimicrobial-resistant Neisseria gonorrhoeae. Emerg Infect Dis. 2005;11(8):1265–73. Epub 2005/08/17. pmid:16102317; PubMed Central PMCID: PMC3320483.
- 35. Price MA, Stewart SR, Miller WC, Behets F, Dow WH, Martinson FE, et al. The cost-effectiveness of treating male trichomoniasis to avert HIV transmission in men seeking sexually transmitted disease care in Malawi. J Acquir Immune Defic Syndr. 2006;43(2):202–9. Epub 2006/09/05. pmid:16951650.
- 36. CDC. National action plan for combatting antibiotic resistance, March 2015. 2015.
- 37. Xiridou M, Lugner A, de Vries HJ, van Bergen JE, Gotz HM, van Benthem BH, et al. Cost-Effectiveness of Dual Antimicrobial Therapy for Gonococcal Infections Among Men Who Have Sex With Men in the Netherlands. Sex Transm Dis. 2016;43(9):542–8. Epub 2016/08/12. pmid:27513379.
- 38. Chesson HW, Kirkcaldy RD, Gift TL, Owusu-Edusei K Jr, Weinstock HS. An Illustration of the Potential Health and Economic Benefits of Combating Antibiotic-Resistant Gonorrhea. Sex Transm Dis. 2018;45(4):250–3. Epub 2018/02/22. pmid:29465709; PubMed Central PMCID: PMC6724164.
- 39. Harding-Esch EM, Huntington SE, Harvey MJ, Weston G, Broad CE, Adams EJ, et al. Antimicrobial resistance point-of-care testing for gonorrhoea treatment regimens: cost-effectiveness and impact on ceftriaxone use of five hypothetical strategies compared with standard care in England sexual health clinics. Euro Surveill. 2020;25(43). Epub 2020/10/31. pmid:33124553; PubMed Central PMCID: PMC7596918.
- 40. Wynn A, Klausner JD. Addressing Neisseria gonorrhoeae Treatment Resistance With the DNA Gyrase A Assay: An Economic Study, United States. Sex Transm Dis. 2020;47(2):111–3. Epub 2019/11/07. pmid:31688726.
- 41. Unemo M, Ross J, Serwin AB, Gomberg M, Cusini M, Jensen JS. 2020 European guideline for the diagnosis and treatment of gonorrhoea in adults. Int J STD AIDS. 2020:956462420949126. Epub 2020/10/31. pmid:33121366.
- 42. Barlow E, Morton A, Megiddo I, Colson A. Optimal subscription models to pay for antibiotics. Social science & medicine (1982). 2022;298:114818. pmid:35247782
- 43. Chandler CIR. Current accounts of antimicrobial resistance: stabilisation, individualisation and antibiotics as infrastructure. Palgrave Commun. 2019;5(1):53. Epub 2019/06/04. pmid:31157116; PubMed Central PMCID: PMC6542671.
- 44.
Ait Ouakrim D, Cassini A, Cecchini M, Plachouras D. The health and economic burden of antimicrobial resistance. In: Mossialos E, Anderson M, Cecchini M, editors. Challenges to Tackling Antimicrobial Resistance: Economic and Policy Responses. European Observatory on Health Systems and Policies. Cambridge: Cambridge University Press; 2020. p. 23–44.