If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The objective of this article is to assess the availability and validity of economic evaluations of carotid artery stenosis (CS) diagnosis and treatment.
Design
Systematic review of economic evaluations of the diagnosis and treatment of CS.
Methods
Systematic review of full economic evaluations published in Medline and Google Scholar up until 28 February 2012. Based on economic checklists (Evers and Philips), the identified studies were classified as high, medium, or low quality.
Results
Twenty-three evaluations were identified. The study quality ranged from 26% to 84% of all achievable points (Evers). Seven studies were of high, eight of medium and eight of low quality. No comparison was made between carotid angioplasty and stenting (CAS) and best medical treatment (BMT). For subjects with severe stenosis, comparisons of carotid endarterectomy (CEA) and BMT were also missing. Three of five studies dealing with pre-operative imaging found that duplex Doppler ultrasound (US) was cost-effective compared with carotid angiogram (AG).
Conclusions
There is a huge lack of high-quality studies and of studies that confirm published results. Also, for a given study quality, the most cost-effective treatment strategy is still unknown in some cases (‘CAS’ vs. ‘BMT’, ‘US combined with magnetic resonance angiography supplemented with AG’ vs. ‘US combined with computer tomography angiography’).
This study identifies research gaps in economic evidence in the context of carotid artery stenosis diagnosis and treatment. Settings in which the most cost-effective treatment strategy is still unknown were identified. We recommend filling these gaps in economic evidence. In the long run, this may lead to the more efficient use of available resources.
Carotid artery stenosis (CS) is an important cause of stroke; 20% of all ischaemic strokes are caused by severe carotid stenosis.
There are several methods for the detection and treatment of CS. These include carotid endarterectomy (CEA), carotid angioplasty and stenting (CAS) and multiple imaging technologies.
Although new and costly methods for the detection and treatment of CS are being developed, it is unclear whether the additional costs are justified in terms of both effectiveness and cost-effectiveness.
Nowadays, economic evaluations are frequently conducted to assess the economic impact of health interventions. A full economic evaluation is defined as an analysis that, first, compares at least two alternative strategies and, second, considers both costs and consequences.
The main types of full economic evaluations are cost-effectiveness, cost–utility and cost–benefit analysis. Furthermore, there is cost-minimisation analysis, based on the precondition that the considered treatment strategies do not differ with respect to the health outcome.
Furthermore, economic evaluations are performed from a particular perspective, most commonly the societal perspective. Economic evaluations are often divided into model-based evaluations and original data analyses (ODAs). ODAs require fewer assumptions but are restricted to the follow-up period. However, models usually combine information from multiple sources.
Several checklists exist to assess the quality of economic analyses.
These assess, for example, whether the model structure is suitable, whether data sources are appropriate and whether the uncertainty of the results has been considered appropriately.
A simple way of assessing parameter uncertainty is deterministic sensitivity analysis (DSA). In DSA, results are re-calculated based on explicitly specified parameters. To assess the overall effect of parameter uncertainty, probabilistic sensitivity analysis (PSA) has been developed. In PSA, all uncertain parameters are sampled simultaneously from distributions that are supposed to represent the true parameter uncertainty. This results in a probability distribution of the model outcomes.
Recent developments in health economic evaluation include displaying results as cost-effectiveness acceptability curves, because confidence intervals for incremental cost-effectiveness ratios are often not suitable.
Common instruments to measure quality of life are the ‘EuroQol – 5 Dimensions’ (EQ–5D), the Health Utilities Index (HUI) and the 36-item short-form health survey (SF-36).
A Multi-perspective cost-effectiveness analysis comparing rivaroxaban with enoxaparin sodium for thromboprophylaxis after total hip and knee replacement in the German healthcare setting.
The objectives of this study are, first, to identify the economic evidence (i.e., the availability and validity of economic evaluations) for interventions in the prevention, diagnosis and treatment of CS. Second, we aim to identify settings where economic evidence is lacking in making a decision about which strategy should be performed when considering both cost and consequences.
Materials and Methods
We performed a systematic review on the health economic evidence of CS prevention, diagnosis and treatment. Only full economic evaluations were considered. We restricted our search to the search engines PubMed and Google Scholar, as these search engines cover a wide range of literature relevant to the scientific audience. Based on the search terms ‘carotid artery stenosis’, ‘carotid angioplasty’, ‘duplex ultrasound’, etc., a search algorithm was built (for the full research strategy, see Appendix). In addition, we screened the references of the identified relevant articles. Two investigators (AS and MJ) examined the titles and abstracts of the potentially eligible articles. Once the articles were chosen, the inclusion of the articles was discussed in cases of differences of opinion (MJ, AS and BS). In cases of queries, discussions were carried out among the reviewers (AS, MJ and BS). The systematic review was performed on February 28, 2012 (last update). A detailed report of the methodological assessment is available from the authors on request.
Only original papers were included; comments on papers, systematic reviews, meta-analysis and protocols were excluded. Furthermore, the articles had to be abstracted in English.
We classified the studies into cost-minimisation analyses, cost-effectiveness analyses, cost–utility analyses and cost–benefit analyses.
Articles that did not explicitly state the perspective were classified according to the reported costs. All included studies were assessed according to Evers' checklist;
All papers were furthermore classified as high, medium or low quality, depending on their quality rating and on the clinical evidence. To be rated of high quality, the clinical evidence of the main health effect needed to be based on at least one randomised controlled trial (RCT) or on a meta-analysis of RCTs of sufficient quality. Furthermore, at least 50% of all Evers' and 50% of all Philips' criteria needed to be fulfilled, and finally, the health outcome needed to be rated as appropriate. If the health outcome used or the clinical evidence was rated as inappropriate, or if less than 40% of all Evers' or all Philips' criteria were fulfilled, studies were rated as low quality. All studies that met the minimum requirements, but did not meet all the criteria to be rated as high, were classified as medium quality.
To interpret the overall strength of the health economic evidence, we used the following scheme. Strong evidence required two or more studies of high quality, moderately strong evidence required two or more studies of medium or high quality and limited evidence required at least one study of medium or high quality. Finally, insufficient evidence represents the situation in which there are no studies available, when all available studies have low quality or when the available studies of the highest quality provide contradictory conclusions. To judge whether studies provide contradictory conclusions, we considered the conclusions drawn within the study.
Results
Initially, 570 studies from PubMed and 6020 studies from Google Scholar were identified by the search algorithm (Fig. 1). Only 49 studies were assessed as a full text after removing the duplicates and screening the records. From these 49 studies, 26 studies were excluded (see Fig. 1), mainly because they were not a full economic evaluation (19 studies) or did not focus on CS (4 studies). Altogether, 23 studies were included. More than two-thirds of the studies (17 studies)
Outcome, observer reliability and patient preferences if CTA, MRA or Doppler ultrasound were used, individually or together, instead of digital subtraction angiography before carotid endarterectomy.
The overall quality assessment yielded seven studies of high, eight of medium and eight of low quality (Table 1, Table 2, Table 3). Methodological quality according to Evers' list (all economic evaluations) ranged from 26% to 84% with an average of 61% (Table 1, Table 2, Table 3). The model quality according to Philips' list ranged from 31% to 74% of all achievable points (average 52%) (Table 4). Although structural aspects have been considered best (on average 68%), data identification and synthesis ranked second (49%) and uncertainty assessment and consistency ranked lowest (36%) (Table 4).
Table 1Economic evaluation of carotid stenosis treatment.
Authors
Year
Country
Evaluation type
Model type
Perspectives
Time horizon
Treatment
Health outcome
Cost/cost-effectiveness (main result)
Major data source
Assessment of health utilities
Males (%)
Age mean range (years)
Evers' list score (%)
Evidence level
Mahoney et al.
2011
USA
CUA
ODA, RCT
Third party payer
1 year
CAS vs. CEA
QoL, life expectancy, QALY
ICER: $6555/QALY gained for stenting
SAPPHIRE trial
EQ–5D (Dolan's value set), time trade-off
68
72
79
High
Young et al.
2010
USA
CUA
MM
Third party payer
Lifetime
CAS vs. CEA
POR, QALY, stroke, death, MI
CEA dominant
Gurm et al.: meta-analysis, Luebeke et al.: meta-analysis, SPACE trial, EVA-3S trial, SAPPHIRE trial, NASCET, ECST, ACST
Systematic review of utility values (based on EuroQoL)
NS
70
63
High
Janssen et al.
2008
NL
CUA
MM
Societal
10 years
CAS vs. CEA
Major stroke rates, long-term survival, QALY
CEA dominant
ECST trial, Ederle et al.: Cochrane Review, Wohley et al.: Review
Systematic review of utility values (no unique method)
NS
NS
53
Medium
Park et al.
2006
USA
CM
ODA, CT
Third party payer
30 days
CAS vs. CEA
Perioperative mortality, MI stroke, and death
CEA dominant
ODA
Not applicable
53
71
63
Low
Kilaru et al.
2003
USA
CUA
MM
Third party payer
Lifetime
CAS vs. CEA
QALY, major and minor stroke
CEA dominant
NASCET
Rating scale
NS
70 (50–90)
58
High
Henriksson et al.
2008
SE
CUA
MM
Societal
Lifetime
CEA vs. BMT
QALY
ICER: €34,557/QALY gained for CEA
ACST
EQ–5D, HUI 2&3, time trade-off
NS
70
68
High
Patel et al.
1999
USA
CUA
MM
Third party payer
Lifetime
CEA vs. BMT
QALY, POR of stroke or death, medical and surgical stroke risk
ICER: $4462/QALY gained for CEA
NASCET
Rating scale
NS
66 (60–90)
84
High
Cronenwett et al.
1997
USA
CUA
MM
Societal
Lifetime
CEA vs. BMT
Major stroke, minor stroke, death, QALY
ICER: $8000/QALY gained for CEA
ACAS, NASCET, ECST, and Veterans Administrative Cooperative Study
Assumptions based on previous models
66
67
84
High
Kuntz and Kent
1996
USA
CUA
MM
Societal
Lifetime
CEA vs. BMT
QALY, morbidity, mortality
ICER: $4100/QALY gained for CEA
NASCET, ACAS
Assumptions
100
65
58
Medium
ACAS: Asymptomatic Carotid Atherosclerosis Study; ACST: Asymptomatic Carotid Surgery Trial; BMT: best medical therapy; CAS: carotid angioplasty and stents; CEA: carotid endarterectomy; CM: cost minimization; CT: controlled trial; CUA: cost–utility analysis; ECST: European Carotid Surgery Trial; EQ–5D: EuroQol – 5 Dimensions; EuroQoL: European Quality of Life Scale; EVA-3S trial: The Endarterectomy vs. Angioplasty in Patients with Symptomatic Severe Carotid Stenosis Trial; HUI: Health Utilities Index; ICER: incremental cost-effectiveness ratio; MI: myocardial infarction; MM: Markov model; NASCET: The North American Symptomatic Carotid Endarterectomy Trial; NL: Netherlands; NS: not stated; ODA: original data analysis; POR: perioperative rate; QALE: quality-adjusted life expectancy; QALY: quality-adjusted life–year; QoL: quality of life; RCT: randomized controlled trial; SAPPHIRE: Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy Trial; SE: Sweden; SPACE: Stent-Protected Angioplasty vs. Carotid Endarterectomy Trial; USA: United States of America.
(Table 1, Table 5). Besides the available treatment options (CEA, CAS and best medical treatment (BMT)), the studies also differed with respect to the target population (symptomatic vs. asymptomatic patients, moderate vs. severe stenosis, high risk groups). Strong evidence was found only for two settings: first, symptomatic patients with severe stenosis treated with CEA vs. CAS and, second, asymptomatic patients with moderate stenosis treated with CEA vs. BMT. For the remaining comparisons, either only one high-quality study was available or economic evaluations were not available at all (Table 5). When comparing CEA vs. CAS in ‘asymptomatic patients with severe stenosis’ and in ‘symptomatic patients with moderate stenosis,’ we considered differences in the target population (high risk vs. average risk) when classifying the results, as they yielded different conclusions.
Table 5Health economic evaluation of the treatment of carotid artery stenosis based on subgroup analysis.
Carotid endarterectomy vs. carotid angioplasty with stenting
Carotid endarterectomy vs. best medical treatment
Carotid angioplasty with stenting vs. best medical treatment
In brackets: study quality; BMT: best medical treatment; CE: cost-effective; CAS: carotid angioplasty with stenting; CEA: carotid endarterectomy; DOM: dominant.
basically agreed that CEA is cost-effective for patients with moderate stenosis. However, three reported a variation in the result with respect to age.
No study was found comparing CAS with BMT (insufficient evidence). Comparisons of CEA and BMT are lacking for patients with severe stenosis (insufficient evidence). However, there is moderate evidence comparing CEA with BMT in symptomatic patients with moderate stenosis and strong evidence comparing CEA with BMT in asymptomatic patients with moderate stenosis. In these cases, CEA is regarded as cost-effective compared with BMT.
Although CEA was compared with both CAS and BMT, no comparisons were made between CAS and BMT (insufficient evidence). Furthermore, no comparisons were made between CEA and CAS for asymptomatic moderate stenosis patients.
Economic evaluation of CEA supporting procedures
CEA is the standard treatment for carotid stenosis and, like other surgeries, it is accompanied by different procedures and investigations. Ten studies are related to such procedures and investigations (Table 2). They were divided into three groups: studies comparing pre-operative investigation; studies dealing with intra-operative procedures; and studies on post-operative ultrasound surveillance.
The standard pre-operative imaging is the carotid angiogram (AG), and there were five studies that compared potential alternatives with AG.
Potential comparators were ‘US’,‘AG’, ‘MRA’, ‘computed tomography angiography’ (CTA), and the combined strategies ‘US + CTA’,‘US + MRA’ and ‘US + MRA supplemented with AG’. There was no strong evidence for any of the pairwise comparisons, as the evidence in none of the studies was based on RCTs. Moderate evidence was available for the comparison of ‘MRA’ vs. ‘US’ (two studies of medium quality concluded that MRA was dominant). For the remaining comparisons, there was limited evidence, as there was one study of medium quality available for each comparison. Furthermore, several comparisons were missing (insufficient evidence).
conducted a comparison of four imaging strategies: ‘US’, ‘magnetic resonance angiography’ (MRA), ‘AG’, and a ‘combination of US and MRA supplemented with AG’. Their conclusion was that the combination of ‘US and MRA supplemented with AG’ is cost-effective compared with the other strategies. The results of comparisons with other strategies could not be extracted because they were not reported.
compared ‘intra-operative US’, ‘completion AG’ and ‘no intra-operative imaging’. However, as this study was of low quality, these findings were rated as providing insufficient evidence.
Finally, with respect to post-operative US surveillance, our review identified two studies dealing with post-operative surveillance.
Patel et al. set up a model from the third-party payer perspective to check the timing of surveillance US or doing no surveillance of the patient after CEA, and they concluded that no surveillance is cost-effective.
Post et al., on the other hand, checked different surveillance strategies, and they concluded that a symptom-guided follow-up strategy is the most cost-effective.
Both these studies were of medium quality, resulting in moderate evidence.
Economic evaluation of screening asymptomatic subjects for carotid stenosis
A further group of studies evaluates screening programs for carotid stenosis (Table 3). The main health effect of the screening models was not based on RCTs, but on the sensitivity and specificity of the test. In consequence, all screening studies were classified as of medium quality. The studies by Lee et al. and Yin and Carpenter compared the following strategies: ‘screen with US, conform by AG and do CEA’, ‘screen with US and do CEA’ and ‘no screening’.
Lee et al. concluded that no screening is cost-effective (societal perspective) compared with the alternatives, whereas Yin and Carpenter concluded that ‘screening with US, conforming by AG and doing CEA’ is the most cost-effective strategy (third-party payer perspective).
Bluth et al. performed a multiple comparison of screening strategies, including ‘power Doppler imaging’, ‘standard duplex Doppler’, ‘MRA’ and ‘AG’. They concluded that power Doppler imaging is cost-effective.
Overall, the evaluations of screening yielded contradictory conclusions. It is unclear whether and how often screening should be carried out and which screening procedure should be used. Thus, the evidence with respect to screening for asymptomatic carotid stenosis patients was judged to be insufficient.
Discussion
We systematically reviewed the literature for economic evaluations of CS treatment, screening and prevention. Although there are many previous systematic reviews in the context of CS,
Medical (nonsurgical) intervention alone is now best for prevention of stroke associated with asymptomatic severe carotid stenosis: results of a systematic review and analysis.
Medical (nonsurgical) intervention alone is now best for prevention of stroke associated with asymptomatic severe carotid stenosis: results of a systematic review and analysis.
Within our systematic review, we identified a substantial lack of economic evidence. Strong evidence was found only for two settings, and several comparisons were not available at all. In the context of carotid stenosis treatment, the study quality and the insufficient number of studies were not the only problems; to make a treatment decision based on the published evidence, the following comparisons were missing: CEA vs. CAS in asymptomatic patients with moderate stenosis; CEA vs. BMT in symptomatic and asymptomatic patients with severe stenosis; and CEA vs. BMT in symptomatic patients with moderate stenosis.
In the context of pre-endarterectomy investigations, none of the observed comparisons was based on RCTs. Thus, as economic analyses cannot provide a higher evidence level than the underlying clinical studies, there was no study of high quality. However, for a given evidence level, most comparisons have been studied only once. Moderate evidence was only provided when comparing MRA vs. US; for the remaining comparisons, there was only limited evidence (one study of medium quality) or insufficient evidence (no study). Overall, from the economic point of view, it is impossible to judge sufficiently which pre-endarterectomy investigations should be performed. When ignoring the evidence level of the single comparisons, one could conclude that either ‘combining US and MRA supplemented with AG’ or ‘combining US with CTA’ is preferable; however, a comparison of these two alternatives is also missing.
In the economic evaluation of carotid intra-endarterectomy procedures, there was a high proportion of low-quality studies. Although the existing studies concluded that local anaesthesia was superior to general anaesthesia, a further high-quality study would be needed to validate the previous findings. The study comparing ‘US’ vs. ‘US plus completion AG’ should also be re-evaluated, following the requirements of Philips' and Evers' checklists more precisely.
According to post-endarterectomy US surveillance, further studies would be desirable to validate the view that symptom-guided US surveillance or no US surveillance is the most cost-effective.
With respect to screening, no RCTs exist that were able to demonstrate a potential health gain. However, based on prevalence data, accuracy data from diagnostic tests and the benefit observed in asymptomatic patients, models found that once-in-a-lifetime screening in a high-prevalence population is cost-effective. Given that once-in-a-lifetime screening is cost-effective, it is still unclear at what age this screening should be performed. To achieve strong evidence, it would be wise, first, to design an ‘optimal’ screening based on the given data, and then to conduct an RCT to validate the modelling results.
One limitation of this review is that we may not have identified all the relevant studies via our search algorithm. Furthermore, many different outcome measures, such as QALYs, strokes avoided, event-free survival, etc., have been used within the economic evaluations, which complicates comparison. However, even costs per QALY are limited when comparing results, because the methodology used to derive QALYs may vary substantially. Even if there were common standards on how to calculate QALYs, further factors affect the costs per QALY ratio, such as whether unrelated medical costs are included, for which setting (country, perspective, etc.) the analysis was performed and other methodological choices.
One problem observed within several economic evaluations was a limited methodology of estimating quality of life. As economic evaluations are often based on secondary data, authors were not entirely free in choosing the utility measurement methodology. Sometimes it could not be specified which instrument (e.g., EQ–5D, HUI or SF-36) was used. Neither could it be specified whether utilities corresponded to the time-trade-off approach, the standard gamble or the visual analogue scale. Instead, utility estimates were blended or even corresponded to an educated guess.
A further limitation is the criteria that we set up to decide whether there is strong, moderate, limited or insufficient evidence. The requirement of at least two economic evaluations to conclude strong or moderate evidence was to guarantee reproducibility. However, others might be satisfied with only one high-quality study. Furthermore, in some cases, there might be ethical reasons for not conducting RCTs and, in consequence, strong evidence, as defined in our study, would never be achieved. The absence of strong evidence in such cases cannot be judged to be a gap in health economic evidence. One has to bear in mind that the goal of decision analysis is to make the best decision on the evidence available.
In conclusion, in this review, we identified limited evidence on economic evaluations for CS. There is a huge lack of high-quality studies, caused by either economic evaluation methodology or low clinical evidence. Furthermore, there is a lack of studies validating the published results, as higher evidence levels require reproducibility. But also, if the evidence level of the analyses is disregarded, comparisons between alternative strategies are missing. These are comparisons between ‘CAS’ and ‘BMT’ for the treatment of CS and between ‘US combined with MRA supplemented with AG’ and ‘US combined with CTA’ for the pre-endarterectomy imaging technique. Furthermore, there is a gap in subgroup analyses, especially according to factors such as gender, age or chronic diseases such as hypertension, diabetes, ischaemic heart disease or cerebrovascular diseases. There is also missing evidence concerning countries outside the United States of America, as it is well known that the results of economic evaluations depend on the setting, including the health system and the country.
We recommend filling these gaps in economic evidence. In the long run, this may lead to the more efficient use of available resources.
Appendix.
The detailed search algorithm, as applied for both PubMed and Google Scholar, is as follows:
‘carotid artery stenosis’ OR ‘carotid angioplasty’ OR ‘carotid ultrasound’ OR ‘carotid endarterectomy’ OR ‘carotid artery stenting’ OR ‘magnetic resonance angiography’ OR ‘carotid stenosis’) AND (‘cost-effectiveness’ OR ‘Markov model’ OR ‘cost–utility’ OR ‘decision-analytic’ OR ‘cost–benefit’ OR ‘cost-effective’ OR ‘costs’ OR ‘cost comparison’ OR ‘economic impact’.
Funding
None.
Conflict of Interest
None declared.
References
Safian R.D.
Treatment strategies for carotid stenosis in patients at increased risk for surgery.
A Multi-perspective cost-effectiveness analysis comparing rivaroxaban with enoxaparin sodium for thromboprophylaxis after total hip and knee replacement in the German healthcare setting.
Outcome, observer reliability and patient preferences if CTA, MRA or Doppler ultrasound were used, individually or together, instead of digital subtraction angiography before carotid endarterectomy.
Medical (nonsurgical) intervention alone is now best for prevention of stroke associated with asymptomatic severe carotid stenosis: results of a systematic review and analysis.
To submit a comment for a journal article, please use the space above and note the following:
We will review submitted comments as soon as possible, striving for within two business days.
This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
We require that commenters identify themselves with names and affiliations.
00597-7&id=gr1.jpg){kind=link}