A cost-effectiveness analysis of screening for intracranial aneurysms in persons with one first-degree relative with subarachnoid haemorrhage

Esther M Hopmans; Ynte M Ruigrok; Anne SE Bor; Gabriel JE Rinkel; Hendrik Koffijberg

doi:10.1177/2396987316674862

. 2016 Oct 19;1(4):320–329. doi: 10.1177/2396987316674862

A cost-effectiveness analysis of screening for intracranial aneurysms in persons with one first-degree relative with subarachnoid haemorrhage

Esther M Hopmans ¹, Ynte M Ruigrok ^1,^✉, Anne SE Bor ¹, Gabriel JE Rinkel ¹, Hendrik Koffijberg ^2,³

PMCID: PMC6301253 PMID: 31008294

Abstract

Introduction

Although persons with one first-degree relative with aneurysmal subarachnoid haemorrhage have an increased risk of aneurysm formation and aneurysmal subarachnoid haemorrhage, screening them for unruptured intracranial aneurysms was not beneficial in a modelling study from the 1990s. New data on the risk of aneurysmal subarachnoid haemorrhage in these persons and improved treatment techniques call for reassessment of the cost-effectiveness of screening.

Patients and methods

We used a cost-effectiveness analysis using a Markov model and Monte Carlo simulation comparing screening and preventive aneurysm treatment with no screening in persons with one first-degree relative with aneurysmal subarachnoid haemorrhage. We analyzed the impact on quality-adjusted life years, costs and net health benefit of single screening (at varying screening age) and serial screening (with varying screening age and intervals) using a cost-effectiveness threshold of €20,000/quality-adjusted life year.

Results

In 17 of the 24 strategies assessed, additional costs for screening for unruptured intracranial aneurysm were <€20,000 per quality-adjusted life year gained. The strategy with highest net health benefit was screening at age 40 and 55. Screening every five years from age 20 to 70 yielded the highest health benefits at the highest additional costs.

Discussion

Based on current risks of aneurysmal subarachnoid haemorrhage and complications of preventive treatment, several strategies to screen for unruptured intracranial aneurysm in persons with one first-degree relative with aneurysmal subarachnoid haemorrhage are cost effective compared with no screening, when applying a cost-effectiveness threshold of €20,000/quality-adjusted life year.

Conclusion

We recommend discussing with persons at risk the option of screening twice, at age 40 and 55, which will result overall in substantial health benefits at acceptable additional costs.

Keywords: Subarachnoid haemorrhage, cost-effectiveness, intracranial aneurysms, screening, first-degree relatives

Introduction

Despite improvement in treatment over the last decades, subarachnoid haemorrhage from a ruptured aneurysm (aneurysmal subarachnoid haemorrhage (aSAH)) still has a high case fatality and morbidity.¹ Because of the relatively young age of onset and poor outcome, aSAH causes a significant loss of productive life years in the general population.²

Non-invasive screening for unruptured intracranial aneurysms (UIA) with magnetic resonance angiography (MRA) enables early detection and preventive treatment of UIA, and thereby preventing future aSAH. However, screening is expensive and has negative effects on quality of life,³ and preventive aneurysm occlusion carries a risk of complications.^4,5 To overcome the downsides of screening, it should be targeted to groups of people at increased risk.

Around 3% of the population has a first-degree relative (FDR) with an aSAH, and these persons are at increased risk of having an UIA and an increased lifetime risk of aSAH compared to the general population.^6–8 Screening is cost effective in persons with two or more affected FDR.^9,10 In contrast, for persons with only one affected FDR, a modelling study from the 1990s showed that although a single MRA screening for UIA increased life expectancy slightly, this increase was at the cost of reduced life years in good health and thus screening in this population was not expected to be effective.⁷ Since that study, new data have shown that persons with one affected FDR have a higher risk of aSAH than assumed in the modelling study⁶ and remain at risk in the first 15 years after a negative screening.⁸ Also, endovascular techniques for UIA occlusion have become available, and complication rates have decreased over time, both for surgical and endovascular treatment.^4,5 These new data call for reassessment of screening for and preventive treatment of UIA in persons with one FDR with aSAH.

Methods

Markov model

We used a Markov decision-analytic model with Monte Carlo simulation to assess the overall benefits of screening by analysing the health outcomes and costs of screening for UIA compared to no screening in persons with one affected FDR with an aSAH.¹¹ The model was built using TreeAge Pro software (version 2015, TreeAge software, Williamstown, MA) and was adapted from a previously published model on the cost-effectiveness of screening persons with two affected FDR with aSAH.⁹

All hypothetical persons enter the model at age 15 in the state ‘healthy without UIA’ and each time cycle reflects one year of life. Figure 1 and supplementary Figure I (all supplementary material is available online with this article, eso.sagepub.com) show a simplified version of the model. Patients can become disabled or die because of aSAH or of complications of preventive UIA treatment. Patients may also die of other causes, based on age-specific mortality rates in the general Dutch population (http://statline.cbs.nl). Disabled persons are assumed to become nursing home dependent or to be discharged to a rehabilitation centre. Recovery from being disabled to healthy is possible only in the first year after becoming disabled. Health outcomes and costs were recorded in the simulation until all persons had died.

Screening was assessed based on MRA to detect UIA. In case of UIA, preventive treatment with clipping or coiling was simulated for UIA >5 mm. Patients with smaller UIA entered the state ‘healthy with small known UIA’. We considered 58% of all de novo UIA too small (i.e. < 5 mm) for treatment.^7,12 These small UIA have an annual risk of rupture of 0.5–1% (considering that UIA in patients with a positive family history have a higher rupture risk than UIA in patients without such a history)^13,14 and a risk of growth of 3.3% per year.^14,15 These small UIA were followed up with MRA every year for the first two years and every two years thereafter. UIA that then enlarged were always treated. Large UIA have a risk of rupture of 1–2% per year, again taking into account the higher rupture risk in familial UIA.^14,16

Screening strategies

Costs and health outcomes were assessed for 24 different screening strategies varying in age of starting and stopping the screening and in screening interval. We analysed strategies with a single screening (at 30, 35, 40, 45, 50 or 55 years old) and with serial screening with varying screening age (starting screening at 20, 30 or 40 years old, ending screening at 60 or 70 years old) and varying screening intervals (5 -, 10 - and 15-year interval). The minimum screening interval was set to five years.

Transition probabilities

We assigned transition probabilities using data available from literature. Distributions and ranges were defined for each model input parameter to reflect parameter uncertainty (Table 1; for table including references see supplementary Table I).

Table 1.

Probabilities and distributions used in our Markov model.

	Probability	Type of distribution	95% CI /range
Aneurysm development (de novo) (per year)^a	0.00165	Normal	0.001649–0.001651
Aneurysm growth (from < 5 mm to > 5 mm, per year)^a	0.033	Triangular	0.03–0.081
Risk of aneurysm being small ( < 5 mm) (per event)^a	0.583	Beta	0.424–0.738
Rupture risk for large aneurysms (>5 mm) (per year)^a	0.014	Uniform	0.011–0.016
Rupture risk for small aneurysms (per year)^a	0.0075	Uniform	0.005–0.010
Risk of mortality in nursing home during first year^a	0.228	Beta	0.149–0.321
Risk of mortality in nursing home after first year (per year)^a	0.152	Beta	0.033–0.271
Probability leave nursing home/rehabilitation centre in good condition (per year)^a	0.261	Beta	0.177–0.357
Risk death other causes (per year)^a	Age dependent		–
Risk death aSAH (per event)	0.261		0.242–0.274
– 20–50 y^a
– 50–65 y^a	0.295		0.279–0.310
– >65 y^a	0.471		0.453–0.489
Risk disability aSAH (per event)^a	0.090	Beta	0.065–0.097
Risk death preventive coiling (per event)^b	0.008	Beta	0.0057–0.0105
Risk disability preventive coiling (per event)^b	0.014	Beta	0.0063–0.0254
Risk death preventive clipping (per event)^b	0.012	Beta	0.0093–0.0153
Risk disability preventive clipping (per event)^b	0.025	Beta	0.0029–0.0585
Risk MRA false positive (only <5 mm), first scan (per event)^b	0.056	Beta	0.0231–0.1014
Risk MRA false positive next scans (per event)^b	0.015	Range	0.01–0.02
Risk MRA false negative (per event)^b	0.073	Beta	0.0356–0.122
SMR after aSAH (per year)^a	2.200	Normal	2.100–2.300
Probability coiling as treatment for unruptured aneurysm (per event)^b	0.384	Beta	0.332–0·438
Utilities	Utility	Distribution	95% CI
Utility unscreened population (healthy)^a	0.839	Normal	0.832–0.847
Utility screen negatives (reassured persons)^b	0.846	Triangular	0.839–0.854
Utility screen positive (anxious persons)^b	0.805	Triangular	0.798–0.813
Utility recovered after SAH/rehabilitation^a	0.819	Triangular	0.812–0.827
Utility disabled (living in nursing home/rehabilitation centre)^a	0.4	Triangular	0.393–0.408
Death^a	0	–	–
Costs	Costs
Cost screening (per event)^b	€492
Cost disabled (in nursing home) (per year)^a	€ 92,082
Cost death (per event)^a	€ –
Cost SAH treatment (per event)^a	€ 30,555
Cost preventive clipping (per event)^b	€ 9744
Cost preventive coiling (per event)^b	€11,368

Open in a new tab

SAH: aneurysmal subarachnoid haemorrhage; MRA: magnetic resonance angiography; SMR: standardised mortality ratio.

Parameter used in both arms.

Parameter used in screening arm.

Development, growth and rupture of UIA

Evidence on the rate of UIA development and rupture for the population with one affected FDR is limited. Therefore, we calibrated the probability of UIA formation and rupture risk in our model to ensure that UIA prevalence and aSAH incidence as estimated from the model were similar to those reported in literature.

The calibration resulted in a probability of 0.165% per year for UIA development and 1–2% per year for UIA rupture. In the natural history scenario of our Markov model, this resulted in a lifetime risk of approximately 10% of developing an UIA, and a 3% lifetime risk of aSAH, which is similar to the risk found in Dutch and Swedish studies.^6,17 Details on the calibration procedure and the evidence used for development, growth and rupture of UIA are provided in the supplementary material.

Costs

Costs were derived from the Dutch manual for costing,¹⁸ local data of Utrecht University Medical Centre and literature¹⁹ (Table 1). Costs of screening included MRA and outpatient consultation with a neurologist and specialised nurse. Costs for treatment of UIA included hospitalisation, follow-up in the outpatient clinic and imaging. All costs were updated using Dutch consumer price indices (http://statline.cbs.nl).

Utilities

Health-related quality of life values (utilities) were assigned to all health states and quality-adjusted life years (QALYs) were used to estimate health outcomes. Death was given a utility value of 0. We used data on quality of life in aSAH survivors and persons screened for UIA from the literature.^3,20,21 Utilities are temporarily higher after a negative screening for UIA and lower after a positive screening,³ which includes screening with false-positive results (see also supplementary data).

Persons who returned to a healthy state after aSAH or a complication from preventive treatment and persons who returned home after recovery in a nursing home or rehabilitation centre all had a reduced utility for the rest of their lives.²¹

Discount rates of 4% for costs and 1.5% for effects were applied according to Dutch guidelines.¹⁸ An overview of the assumptions underlying our analysis is provided in the supplementary material.

Cost-effectiveness analysis

To estimate the cost-effectiveness of screening we calculated the difference in QALYs and costs, and the incremental cost-effectiveness ratio for all screening strategies compared to no screening. A cost-effectiveness threshold of €20,000 per QALY gained was applied.²² In the Netherlands health policy makers consider a range of cost-effectiveness thresholds from €20,000 to €80,000 per QALY depending on context (i.e. type of technology, severity of disease, budget impact, etc.). Therefore, €20,000 per QALY can be used as a save, or conservative, estimate of this threshold: interventions with cost-effectiveness below this lower threshold will certainly be viewed as having favourable cost-effectiveness. As several strategies resulted in very small QALY differences, we also calculated the net health benefit (NHB) (difference in QALYs minus the difference in costs divided by the cost-effectiveness threshold). A positive NHB indicates that screening is cost effective, and the strategy with the highest NHB would be optimal.

To visualise the potential effect of screening on the number of UIA and episodes of aSAH during life we also simulated a single cohort of 10,000 patients and analysed the individual outcomes in this cohort, for three different screening strategies.

Sensitivity analysis

To assess the combined uncertainty embedded in all probabilities and utility values, we performed probabilistic sensitivity analyses with Monte Carlo simulation, using 1000 hypothetical cohorts of 5000 persons each. We determined the probability of each screening strategy being cost effective compared to no screening, and determined the optimal screening strategy when all strategies are compared for a range of cost-effectiveness threshold values. Results were visualised in a bar plot instead of cost-effectiveness acceptability curve to accommodate all 24 strategies and in a cost-effectiveness acceptability frontier. Finally, we performed a value of information analysis, investigating the value of collecting additional data, using the online Sheffield Accelerated Value of Information (SAVI) tool.²³ We first estimated the expected value of perfect information (EVPI), that is the value from improved decision making when all uncertainty would be resolved. The total population EVPI was calculated based on a conservative estimate of number of two new relatives eligible for screening following each new aSAH (∼1500 episodes annually in the Netherlands²⁴), leading to a target population of 3000 persons per year. We also estimated the expected value of partial perfect information (EVPPI), that is the value of resolving all uncertainty in a single, or group of, model parameter(s). Results were visualised in a bar plot.

Results

Screening for UIA in persons with one affected FDR with aSAH was expected to be cost effective (<€20.000/QALY compared to no screening) for 17 of the 24 strategies analysed (Table 2 and supplementary Figure II).

Table 2.

Outcomes of the different screening strategies compared to outcomes of no screening.

Starting and	Screening	Change in	Incremental	Incremental
stopping age	frequency	aSAH risk	health benefits	costs	ICER	NHB
(years)	(years)	(%)	(QALYs)	(€)	€/QALY	(QALYs)	DESCRIPTION
20–60	5	−1.92%	0.076	1941	25,472	−0.021	Repeated screening (9x in total)
20–60	10	−1.74%	0.056	1146	20,527	−0.001	Repeated screening (5x in total)
20–60	15	−1.44%	0.043	826	19,363	0.001	Repeated screening (3x in total)
20–70	5	−2.14%	0.084	2063	24,703	−0.020	Repeated screening (11x in total)
20–70	10	−1.93%	0.060	1209	20,078	0.000	Repeated screening (6x in total)
20–70	15	−1.69%	0.048	905	18,752	0.003	Repeated screening (4x in total)
30–60	5	−1.78%	0.060	1210	20,070	0.000	Repeated screening (7x in total)
30–60	10	−1.63%	0.044	746	17,017	0.007	Repeated screening (4x in total)
30–60	15	−1.48%	0.035	585	16,664	0.006	Repeated screening (3x in total)
30–70	5	−2.00%	0.066	1332	20,113	0.000	Repeated screening (9x in total)
30–70	10	−1.82%	0.044	807	18,488	0.003	Repeated screening (5x in total)
30–70	15	−1.60%	0.036	599	16,842	0.006	Repeated screening (3x in total)
40–60	5	−1.47%	0.038	697	18,212	0.003	Repeated screening (5x in total)
40–60	10	−1.35%	0.030	453	14,924	0.008	Repeated screening (3x in total)
40–60	15	−1.21%	0.027	348	12,763	0.010	Repeated screening (2x in total)
40–70	5	−1.69%	0.045	819	18,263	0.004	Repeated screening (7x in total)
40–70	10	−1.55%	0.035	516	14,880	0.009	Repeated screening (4x in total)
40–70	15	−1.40%	0.028	410	14,501	0.008	Repeated screening (3 x in total)
30	Once	−0.47%	0.014	298	21,146	−0.001	Single screening strategy
35	Once	−0.54%	0.016	238	14,964	0.004	Single screening strategy
40	Once	−0.60%	0.013	205	15,905	0.003	Single screening strategy
45	Once	−0.62%	0.011	176	15,685	0.002	Single screening strategy
50	Once	−0.61%	0.011	146	12,831	0.004	Single screening strategy
55	Once	−0.56%	0.010	124	12,328	0.004	Single screening strategy
55	Once	−0.56%	0.024	459	18,877	0.001	S1: no discounting
55	Once	−0.56%	0.009	124	13,392	0.003	S2: psychological effect lasts 1/4 year
55	Once	−0.56%	0.010	126	13,007	0.003	S3: likelihood coiling equal 0.20

Open in a new tab

aSAH: aneurysmal subarachnoid haemorrhage; ICER: incremental cost-effectiveness ratio; NHB: Net Health benefit; QALY: quality-adjusted life year. Change in lifetime aSAH risk, incremental health benefits (in QALYs) and costs (in €) concern the difference for screening strategies compared to no screening. Costs and effects are based on a discount rate of 4% for costs and 1.5% for effects. The NHB is based on a cost-effectiveness threshold of €20,000 per QALY.

All screening strategies improved health outcomes but were also more expensive than no screening. More frequent screening and screening over longer age ranges increased health benefits and costs.

Our simulation cohort with 10,000 persons on the effectiveness of screening in terms of untreated UIA and aSAH incidence showed that the impact of screening on lowering the number of UIA becomes apparent around the age of 30 (Figure 2).

Figure 2. — The figure shows the impact of screening on untreated UIA and aSAH incidence compared to natural history simulated in our Markov model with a single cohort of 10,000 patients. Three examples of different screening strategies are shown. Panel (a) reflects screening from age 20 to 70 every five years, panel (b) reflects screening from age 30 to 60 every 15 years and panel (c) a single screening at age 45. After every screening moment there is a visible decrease in number of untreated UIA. Frequent screening produces repeated small reductions in untreated UIAs (a) and a constant lower aSAH number. Less intensive screening programmes have much larger impact per screening moment on untreated UIA prevalence and can be highly effective in aSAH prevention when screening is performed shortly before the peak incidence of aSAH at age 50 (c). As panels (a) to (c) reflect analyses of three separate cohorts, the number of untreated UIAs and aSAHs in natural history as a function of age is similar but not identical, due to random variation. aSAH: aneurysmal subarachnoid haemorrhage; UIA: unruptured intracranial aneurysms.

Table 2 indicates that screening every five years from age 20 to 70 yielded the highest health benefits at the highest additional cost (0.084 QALY for +€ 2063, per screened person). Single screening strategies all yielded very small health benefits at relatively low costs. For a cost-effectiveness threshold of €20,000 per QALY the strategy with the highest NHB was screening every 15 years starting at age 40 and stopping at age 60, which translates to screening at age 40 and 55.

The probability that any of the 24 screening strategies is cost effective compared to no screening is shown in supplementary Figure III. Here it is apparent that although 17 out of 24 strategies are expected to be cost effective, this cannot be established with certainty due to uncertainty in model outcomes, and since most strategies have relatively low probability of being cost effective for a cost-effectiveness threshold of €20,000 per QALY. Conversely, for a cost-effectiveness threshold of €100,000 per QALY multiple screening strategies with five-year frequency strategies have around 70% chance of being cost effective.

When all screening strategies are compared, with no screening and with each other, a cost-effectiveness acceptability frontier also shows that screening once (at age 55) or twice (at age 40 and 55) is expected to be best, in terms of NHB, for low cost-effectiveness threshold values (< €20,000 per QALY), whereas the most extensive screening strategy (from age 20 to 70 every five years) is expected to be best for high cost-effectiveness threshold values (> €45,000 per QALY) (supplementary Figure IV). For threshold values in between, screening from age 40 to 70 every 10 years, or from age 30 to 70 every five years is expected to be best. However, the probabilities that these strategies are cost effective, compared to each other, is quite low (<15% for threshold values >€7500). Conversely, the probabilities that they are cost effective compared to no screening is much higher higher.

The value of information analysis indicated that the per person EVPI was equal to €1892 for a 10-year decision horizon. Annual inclusion of 3000 individuals in the screening programme amounts to a population EVPI value of 56.8 million Euros (or, equivalently, 2,838 QALYs, for a cost-effectiveness threshold of €20,000 per QALY) in the Netherlands. Further analysis of uncertainty in individual model parameters and groups of parameters revealed that, of all parameters in the model, collecting additional information on the parameter “risk of aneurysm rupture” would be most valuable (EVPPI €211 per person) (supplementary Figure V). When assessing groups of parameters, collecting additional data on all parameters related to treatment complications appears most valuable (EVPPI €331 per person). However, collecting additional data on aneurysm development and rupture, consequences of aSAH, MRA accuracy and treatment options would all be valuable according to this value of information analysis.

Discussion

Our cost-effectiveness analysis shows that several screening strategies for UIA are expected to be cost effective in persons with one affected FDR with aSAH. Depending on the cost-effectiveness threshold applied, screening once, twice or even up to 11 times between age 20 and 70 years may be optimal. Although current evidence shows that several screening strategies are likely to be cost effective compared to no screening, it does not yet allow us to conclude with certainty which screening strategy is the best. This is exemplified by our value of information analysis which suggests that collecting additional data on aneurysm development and rupture, consequences of aSAH, MRA accuracy and treatment options is valuable to support the decision on screening for UIA in this population.

The conclusion from our current cost-effectiveness analysis contrasts with the recommendation based on our previous modelling study. Our previous modelling study concluded that implementation of screening for persons with one affected FDR was not warranted.⁷ There are several reasons why the conclusions have changed. Catheter angiography is no longer needed for accurate diagnosis and thus complications of diagnostic catheter angiography were no longer considered. Also, the risk of complications from preventive treatment has decreased over time.^4,5,25 Moreover, previously it was assumed that the risk of aSAH after screening and successful preventive treatment was negligible. Studies performed after our modelling studies in the 1990s showed that, first, patients who have survived an episode of aSAH are at increased risk of new episodes of aSAH,²⁶ second, patients with a positive family history for SAH have a substantial risk of developing aneurysm after one or more negative screens,²⁷ third, persons with one FDR with aSAH with a single negative screening for UIA still are at risk of an aSAH in the initial 15 years after the negative screen.⁸ All these data explain why the results of the current study differ from that of 25 years ago.

Our study has several limitations. First, limited data are available on UIA formation, growth and rupture for the population of persons with one affected FDR with aSAH and we had to make assumptions for these probabilities. However, the chances of finding an aneurysm in this population,⁷ the risk of aSAH^6,28 and to a lesser extent the risk of aSAH after a negative screen in this population⁸ are well known, and outcomes of our model match these observational data. Second, we were not able to incorporate important predictors for rupture risk including aneurysm characteristics as size and location and patient characteristics as smoking and hypertension.¹³ Thus, we cannot conclude whether screening is more or less effective according to presence or absence of these risk factors. Third, although we used all available data on quality of life of persons with UIA,^3,20,21,29 there is lack of data on improvement of quality of life by reassurance from a negative screen for persons with one affected FDR with aSAH. More data on these factors will allow more accurate and more individually tailored advice on screening but will not alter the main conclusions of our study. A fourth limitation is that we did not take into account the costs and complication risk of retreatment after preventive coiling needed by reopening of the aneurysm. However, because this retreatment concerns only 10% of patients,³⁰ and complication risk of retreatment is very small,³¹ this will not have influenced our results to a large extent. In addition, we did not incorporate new treatment modalities such as balloon-assisted coiling or the use of other new devices.³² Fifth, as healthcare facilities differ over countries and over different hospitals, a specialised nurse for aneurysm care may not be present in every clinical setting. Costs of screening might therefore be slightly inflated in clinical settings not supported by such a specialised nurse. Last, as we could not account for uncertainty in cost estimates, due to lack of evidence, the uncertainty in cost differences may be underestimated in our results.

We used a cost-effectiveness threshold of €20,000 but other countries use different thresholds or refuse to determine a threshold because of ethical considerations.³³ The probabilities and especially costs are based on Dutch numbers and it is known that costs for treatment may vary substantially between countries.^2,19,34 Therefore, interpretation of the presented NHB results is complex and has to be adapted to the distinct perspectives of different countries. Furthermore, in a cost-effectiveness analysis the health effects and costs of considered strategies are not directly related to the feasibility of their implementation. Although high frequency screening strategies are expected to be optimal for cost-effectiveness threshold values exceeding €20,000 per QALY, these may not be easy to implement, given the number of required MRAs. Clinical trials would be the best method to assess effectiveness of screening but the low aSAH incidence in this population and thus large number of FDRs and the extensive follow-up time needed make a randomised trial unfeasible. Moreover, many of the FDRs that have seen the impact of aSAH within their family wish to have preventive screening and treatment, making randomisation difficult if not impossible. In such circumstances, simulation models are an appropriate alternative.³⁵ Thus, the current data are the best available data to base recommendation upon regarding screening and preventive treatment of aneurysms to prevent aSAH with its inherent high case fatality and morbidity.¹

Given that several screening strategies are likely, but not certain, to be cost effective compared to no screening further studies may focus collecting additional data for more precise estimates of probabilities and risks, and on identifying the most optimal screening strategy. Our results indicate that additional data can be valuable, in particular on the risk of treatment complications, aneurysm development, growth and rupture.

For clinical practice, our results suggest that for persons with one affected FDR with SAH screening twice, at age 40 and 55, should be considered. It should be discussed with such persons that this is the most optimal screening strategy resulting in substantial health benefits at acceptable additional costs based on the current data. The uncertainties revealed by our studies can be included in this patient encounter. For policy makers our study provides data on which screening strategies should be accepted. Serial screening yielded the highest health benefits; however, frequent serial screening also had the highest additional costs and will only be cost effective for high cost-effectiveness threshold value.

Supplementary Material

Supplementary material

online_supplement_1.pdf^{(410.3KB, pdf)}

Acknowledgements

The authors would like to thank Dr MJH Wermer, MD, for her comments on a previous version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: YMR was supported by a clinical fellowship grant of the Netherlands Organization for Scientific Research (NWO) (# 40-00703-98-13533). HK has received a personal grant from NWO, The Netherlands Organization for Scientific Research (# 916.11.126), for studying methods enhancing the economic evaluation of diagnostic tests and screening strategies.

Ethical approval

N/A.

Informed consent

N/A.

Guarantor

YMR.

Contributorship

EMH drafted the article. EMH, YMR, ASEB, GJER and HK participated in the data acquisition, analysis or interpretation of data. Statistical analyses were performed by EMH, ASEB and HK. GJER obtained funding. YMR, GJER and HK supervised the study and approved the final version of the manuscript before submission. All authors acknowledge and agree on the content of the manuscript.

References

1.Nieuwkamp DJ, Setz LE, Algra A, et al. Changes in case fatality of aneurysmal subarachnoid haemorrhage over time, according to age, sex, and region: a meta-analysis. Lancet Neurol 2009; 8: 635–642. [DOI] [PubMed] [Google Scholar]
2.Rivero-Arias O, Gray A, Wolstenholme J. Burden of disease and costs of aneurysmal subarachnoid haemorrhage (aSAH) in the United Kingdom. Cost Eff Resour Alloc 2010; 8: 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Wermer MJ, van der Schaaf IC, Van Nunen P, et al. Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke 2005; 36: 836–840. [DOI] [PubMed] [Google Scholar]
4.Alshekhlee A, Mehta S, Edgell RC, et al. Hospital mortality and complications of electively clipped or coiled unruptured intracranial aneurysm. Stroke 2010; 41: 1471–1476. [DOI] [PubMed] [Google Scholar]
5.Jalbert JJ, Isaacs AJ, Kamel H, et al. Clipping and coiling of unruptured intracranial aneurysms among Medicare beneficiaries, 2000 to 2010. Stroke 2015; 46: 2452–2457. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Bor AS, Rinkel GJ, Adami J, et al. Risk of subarachnoid haemorrhage according to number of affected relatives: a population based case-control study. Brain 2008; 131: 2662–2665. [DOI] [PubMed] [Google Scholar]
7.Magnetic Resonance Angiography in Relatives of Patients with Subarachnoid Hemorrhage Study Group. Risks and benefits of screening for intracranial aneurysms in first-degree relatives of patients with sporadic subarachnoid hemorrhage. N Engl J Med 1999; 341: 1344–1350. [DOI] [PubMed] [Google Scholar]
8.Rasing I, Ruigrok YM, Greebe P, et al. Long-term risk of aneurysmal subarachnoid hemorrhage after a negative aneurysm screen. Neurology 2015; 84: 912–917. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Bor AS, Koffijberg H, Wermer MJ, et al. Optimal screening strategy for familial intracranial aneurysms: a cost-effectiveness analysis. Neurology 2010; 74: 1671–1679. [DOI] [PubMed] [Google Scholar]
10.Brown RD, Jr, Huston J, Hornung R, et al. Screening for brain aneurysm in the Familial Intracranial Aneurysm study: frequency and predictors of lesion detection. J Neurosurg 2008; 108: 1132–1138. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993; 13: 322–338. [DOI] [PubMed] [Google Scholar]
12.Vlak MH, Algra A, Brandenburg R, et al. Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: a systematic review and meta-analysis. Lancet Neurol 2011; 10: 626–636. [DOI] [PubMed] [Google Scholar]
13.Greving JP, Wermer MJ, Brown RD, Jr, et al. Development of the PHASES score for prediction of risk of rupture of intracranial aneurysms: a pooled analysis of six prospective cohort studies. Lancet Neurol 2014; 13: 59–66. [DOI] [PubMed] [Google Scholar]
14.Broderick JP, Brown RD, Jr, Sauerbeck L, et al. Greater rupture risk for familial as compared to sporadic unruptured intracranial aneurysms. Stroke 2009; 40: 1952–1957. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Bor AS, Tiel Groenestege AT, terBrugge KG, et al. Clinical, radiological, and flow-related risk factors for growth of untreated, unruptured intracranial aneurysms. Stroke 2015; 46: 42–48. [DOI] [PubMed] [Google Scholar]
16.Wermer MJ, van der Schaaf IC, Algra A, et al. Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics: an updated meta analysis. Stroke 2007; 38: 1404–1410. [DOI] [PubMed] [Google Scholar]
17.Bromberg JE, Rinkel GJ, Algra A, et al. Subarachnoid haemorrhage in first and second degree relatives of patients with subarachnoid haemorrhage. BMJ 1995; 311: 288–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Swan Tan S, Bouwmans-Frijters C and Hakkaart-van Roijen L. Handleiding voor kostenonderzoek: methoden en referentieprijzen voor economische evaluaties in de gezondheidszorg (In Dutch) Tijds. gezondheids.wetenschappen 2012; 90: 367.
19.Halkes PH, Wermer MJ, Rinkel GJ, et al. Direct costs of surgical clipping and endovascular coiling of unruptured intracranial aneurysms. Cerebrovasc Dis 2006; 22: 40–45. [DOI] [PubMed] [Google Scholar]
20.Buijs JE, Greebe P, Rinkel GJ. Quality of life, anxiety, and depression in patients with an unruptured intracranial aneurysm with or without aneurysm occlusion. Neurosurgery 2012; 70: 868–872. [DOI] [PubMed] [Google Scholar]
21.von Vogelsang AC, Burstrom K, Wengstrom Y, et al. Health-related quality of life 10 years after intracranial aneurysm rupture: a retrospective cohort study using EQ-5D. Neurosurgery 2013; 72: 397–406. [DOI] [PubMed] [Google Scholar]
22.Zwaap J, Knies S, van der Meijden C, et al. Kosteneffectiviteit in de praktijk. Report in Dutch: advice to the Ministry of Health (#2015076142). National Health Care Institute, https://www.zorginstituutnederland.nl/binaries/content/documents/zinl-www/actueel/nieuws/2015/zorginstituut-stelt-rapport-%E2%80%98kosteneffectiviteit-in-de-praktijk%E2%80%99-vast/zorginstituut-stelt-rapport-%E2%80%98kosteneffectiviteit-in-de-praktijk%E2%80%99-vast/zinl%3ADocument/1506-kosteneffectiviteit-in-de-praktijk/Kosteneffectiviteit+in+de+praktijk.pdf (accessed 26 June 2015).
23.Strong M, Oakley JE, Brennan A. Estimating multi-parameter partial expected value of perfect information from a probabilistic sensitivity analysis sample: a non-parametric regression approach. Med Decis Making 2014; 34: 311–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.de Rooij NK, Linn FH, van der Plas JA, et al. Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends. J Neurol Neurosurg Psychiatry 2007; 78: 1365–1372. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Qureshi AI, Chaudhry SA, Tekle WG, et al. Comparison of long-term outcomes associated with endovascular treatment vs surgical treatment among Medicare beneficiaries with unruptured intracranial aneurysms. Neurosurgery 2014; 75: 380–387. [DOI] [PubMed] [Google Scholar]
26.Wermer MJ, Greebe P, Algra A, et al. Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005; 36: 2394–2399. [DOI] [PubMed] [Google Scholar]
27.Bor AS, Rinkel GJ, van Norden J, et al. Long-term, serial screening for intracranial aneurysms in individuals with a family history of aneurysmal subarachnoid haemorrhage: a cohort study. Lancet Neurol 2014; 13: 385–392. [DOI] [PubMed] [Google Scholar]
28.Teasdale GM, Wardlaw JM, White PM, et al. The familial risk of subarachnoid haemorrhage. Brain 2005; 128: 1677–1685. [DOI] [PubMed] [Google Scholar]
29.Aaronson NK, Muller M, Cohen PD, et al. Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations. J Clin Epidemiol 1998; 51: 1055–1068. [DOI] [PubMed] [Google Scholar]
30.Ferns SP, Sprengers ME, van Rooij WJ, et al. Coiling of intracranial aneurysms: a systematic review on initial occlusion and reopening and retreatment rates. Stroke 2009; 40: e523–e529. [DOI] [PubMed] [Google Scholar]
31.Slob MJ, Sluzewski M, van Rooij WJ, et al. Additional coiling of previously coiled cerebral aneurysms: clinical and angiographic results. Am J Neuroradiol 2004; 25: 1373–1376. [PMC free article] [PubMed] [Google Scholar]
32.Colby GP, Lin LM, Paul AR, et al. Cost comparison of endovascular treatment of anterior circulation aneurysms with the pipeline embolization device and stent-assisted coiling. Neurosurgery 2012; 71: 944–950. [DOI] [PubMed] [Google Scholar]
33.Ligtenberg G. Kosteneffectiviteit in de zorg, www.rijksoverheid.nl (2013, accessed 26 June 2015).
34.Roos YB, Dijkgraaf MG, Albrecht KW, et al. Direct costs of modern treatment of aneurysmal subarachnoid hemorrhage in the first year after diagnosis. Stroke 2002; 33: 1595–1599. [DOI] [PubMed] [Google Scholar]
35.Schaafsma JD, van der Graaf Y, Rinkel GJ, et al. Decision analysis to complete diagnostic research by closing the gap between test characteristics and cost-effectiveness. J Clin Epidemiol 2009; 62: 1248–1252. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material

online_supplement_1.pdf^{(410.3KB, pdf)}

[bibr1-2396987316674862] 1.Nieuwkamp DJ, Setz LE, Algra A, et al. Changes in case fatality of aneurysmal subarachnoid haemorrhage over time, according to age, sex, and region: a meta-analysis. Lancet Neurol 2009; 8: 635–642. [DOI] [PubMed] [Google Scholar]

[bibr2-2396987316674862] 2.Rivero-Arias O, Gray A, Wolstenholme J. Burden of disease and costs of aneurysmal subarachnoid haemorrhage (aSAH) in the United Kingdom. Cost Eff Resour Alloc 2010; 8: 6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr3-2396987316674862] 3.Wermer MJ, van der Schaaf IC, Van Nunen P, et al. Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke 2005; 36: 836–840. [DOI] [PubMed] [Google Scholar]

[bibr4-2396987316674862] 4.Alshekhlee A, Mehta S, Edgell RC, et al. Hospital mortality and complications of electively clipped or coiled unruptured intracranial aneurysm. Stroke 2010; 41: 1471–1476. [DOI] [PubMed] [Google Scholar]

[bibr5-2396987316674862] 5.Jalbert JJ, Isaacs AJ, Kamel H, et al. Clipping and coiling of unruptured intracranial aneurysms among Medicare beneficiaries, 2000 to 2010. Stroke 2015; 46: 2452–2457. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr6-2396987316674862] 6.Bor AS, Rinkel GJ, Adami J, et al. Risk of subarachnoid haemorrhage according to number of affected relatives: a population based case-control study. Brain 2008; 131: 2662–2665. [DOI] [PubMed] [Google Scholar]

[bibr7-2396987316674862] 7.Magnetic Resonance Angiography in Relatives of Patients with Subarachnoid Hemorrhage Study Group. Risks and benefits of screening for intracranial aneurysms in first-degree relatives of patients with sporadic subarachnoid hemorrhage. N Engl J Med 1999; 341: 1344–1350. [DOI] [PubMed] [Google Scholar]

[bibr8-2396987316674862] 8.Rasing I, Ruigrok YM, Greebe P, et al. Long-term risk of aneurysmal subarachnoid hemorrhage after a negative aneurysm screen. Neurology 2015; 84: 912–917. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr9-2396987316674862] 9.Bor AS, Koffijberg H, Wermer MJ, et al. Optimal screening strategy for familial intracranial aneurysms: a cost-effectiveness analysis. Neurology 2010; 74: 1671–1679. [DOI] [PubMed] [Google Scholar]

[bibr10-2396987316674862] 10.Brown RD, Jr, Huston J, Hornung R, et al. Screening for brain aneurysm in the Familial Intracranial Aneurysm study: frequency and predictors of lesion detection. J Neurosurg 2008; 108: 1132–1138. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr11-2396987316674862] 11.Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993; 13: 322–338. [DOI] [PubMed] [Google Scholar]

[bibr12-2396987316674862] 12.Vlak MH, Algra A, Brandenburg R, et al. Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: a systematic review and meta-analysis. Lancet Neurol 2011; 10: 626–636. [DOI] [PubMed] [Google Scholar]

[bibr13-2396987316674862] 13.Greving JP, Wermer MJ, Brown RD, Jr, et al. Development of the PHASES score for prediction of risk of rupture of intracranial aneurysms: a pooled analysis of six prospective cohort studies. Lancet Neurol 2014; 13: 59–66. [DOI] [PubMed] [Google Scholar]

[bibr14-2396987316674862] 14.Broderick JP, Brown RD, Jr, Sauerbeck L, et al. Greater rupture risk for familial as compared to sporadic unruptured intracranial aneurysms. Stroke 2009; 40: 1952–1957. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr15-2396987316674862] 15.Bor AS, Tiel Groenestege AT, terBrugge KG, et al. Clinical, radiological, and flow-related risk factors for growth of untreated, unruptured intracranial aneurysms. Stroke 2015; 46: 42–48. [DOI] [PubMed] [Google Scholar]

[bibr16-2396987316674862] 16.Wermer MJ, van der Schaaf IC, Algra A, et al. Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics: an updated meta analysis. Stroke 2007; 38: 1404–1410. [DOI] [PubMed] [Google Scholar]

[bibr17-2396987316674862] 17.Bromberg JE, Rinkel GJ, Algra A, et al. Subarachnoid haemorrhage in first and second degree relatives of patients with subarachnoid haemorrhage. BMJ 1995; 311: 288–289. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr18-2396987316674862] 18.Swan Tan S, Bouwmans-Frijters C and Hakkaart-van Roijen L. Handleiding voor kostenonderzoek: methoden en referentieprijzen voor economische evaluaties in de gezondheidszorg (In Dutch) Tijds. gezondheids.wetenschappen 2012; 90: 367.

[bibr19-2396987316674862] 19.Halkes PH, Wermer MJ, Rinkel GJ, et al. Direct costs of surgical clipping and endovascular coiling of unruptured intracranial aneurysms. Cerebrovasc Dis 2006; 22: 40–45. [DOI] [PubMed] [Google Scholar]

[bibr20-2396987316674862] 20.Buijs JE, Greebe P, Rinkel GJ. Quality of life, anxiety, and depression in patients with an unruptured intracranial aneurysm with or without aneurysm occlusion. Neurosurgery 2012; 70: 868–872. [DOI] [PubMed] [Google Scholar]

[bibr21-2396987316674862] 21.von Vogelsang AC, Burstrom K, Wengstrom Y, et al. Health-related quality of life 10 years after intracranial aneurysm rupture: a retrospective cohort study using EQ-5D. Neurosurgery 2013; 72: 397–406. [DOI] [PubMed] [Google Scholar]

[bibr22-2396987316674862] 22.Zwaap J, Knies S, van der Meijden C, et al. Kosteneffectiviteit in de praktijk. Report in Dutch: advice to the Ministry of Health (#2015076142). National Health Care Institute, https://www.zorginstituutnederland.nl/binaries/content/documents/zinl-www/actueel/nieuws/2015/zorginstituut-stelt-rapport-%E2%80%98kosteneffectiviteit-in-de-praktijk%E2%80%99-vast/zorginstituut-stelt-rapport-%E2%80%98kosteneffectiviteit-in-de-praktijk%E2%80%99-vast/zinl%3ADocument/1506-kosteneffectiviteit-in-de-praktijk/Kosteneffectiviteit+in+de+praktijk.pdf (accessed 26 June 2015).

[bibr23-2396987316674862] 23.Strong M, Oakley JE, Brennan A. Estimating multi-parameter partial expected value of perfect information from a probabilistic sensitivity analysis sample: a non-parametric regression approach. Med Decis Making 2014; 34: 311–326. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr24-2396987316674862] 24.de Rooij NK, Linn FH, van der Plas JA, et al. Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends. J Neurol Neurosurg Psychiatry 2007; 78: 1365–1372. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bibr25-2396987316674862] 25.Qureshi AI, Chaudhry SA, Tekle WG, et al. Comparison of long-term outcomes associated with endovascular treatment vs surgical treatment among Medicare beneficiaries with unruptured intracranial aneurysms. Neurosurgery 2014; 75: 380–387. [DOI] [PubMed] [Google Scholar]

[bibr26-2396987316674862] 26.Wermer MJ, Greebe P, Algra A, et al. Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005; 36: 2394–2399. [DOI] [PubMed] [Google Scholar]

[bibr27-2396987316674862] 27.Bor AS, Rinkel GJ, van Norden J, et al. Long-term, serial screening for intracranial aneurysms in individuals with a family history of aneurysmal subarachnoid haemorrhage: a cohort study. Lancet Neurol 2014; 13: 385–392. [DOI] [PubMed] [Google Scholar]

[bibr28-2396987316674862] 28.Teasdale GM, Wardlaw JM, White PM, et al. The familial risk of subarachnoid haemorrhage. Brain 2005; 128: 1677–1685. [DOI] [PubMed] [Google Scholar]

[bibr29-2396987316674862] 29.Aaronson NK, Muller M, Cohen PD, et al. Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations. J Clin Epidemiol 1998; 51: 1055–1068. [DOI] [PubMed] [Google Scholar]

[bibr30-2396987316674862] 30.Ferns SP, Sprengers ME, van Rooij WJ, et al. Coiling of intracranial aneurysms: a systematic review on initial occlusion and reopening and retreatment rates. Stroke 2009; 40: e523–e529. [DOI] [PubMed] [Google Scholar]

[bibr31-2396987316674862] 31.Slob MJ, Sluzewski M, van Rooij WJ, et al. Additional coiling of previously coiled cerebral aneurysms: clinical and angiographic results. Am J Neuroradiol 2004; 25: 1373–1376. [PMC free article] [PubMed] [Google Scholar]

[bibr32-2396987316674862] 32.Colby GP, Lin LM, Paul AR, et al. Cost comparison of endovascular treatment of anterior circulation aneurysms with the pipeline embolization device and stent-assisted coiling. Neurosurgery 2012; 71: 944–950. [DOI] [PubMed] [Google Scholar]

[bibr33-2396987316674862] 33.Ligtenberg G. Kosteneffectiviteit in de zorg, www.rijksoverheid.nl (2013, accessed 26 June 2015).

[bibr34-2396987316674862] 34.Roos YB, Dijkgraaf MG, Albrecht KW, et al. Direct costs of modern treatment of aneurysmal subarachnoid hemorrhage in the first year after diagnosis. Stroke 2002; 33: 1595–1599. [DOI] [PubMed] [Google Scholar]

[bibr35-2396987316674862] 35.Schaafsma JD, van der Graaf Y, Rinkel GJ, et al. Decision analysis to complete diagnostic research by closing the gap between test characteristics and cost-effectiveness. J Clin Epidemiol 2009; 62: 1248–1252. [DOI] [PubMed] [Google Scholar]

PERMALINK

A cost-effectiveness analysis of screening for intracranial aneurysms in persons with one first-degree relative with subarachnoid haemorrhage

Esther M Hopmans

Ynte M Ruigrok

Anne SE Bor

Gabriel JE Rinkel

Hendrik Koffijberg

Abstract

Introduction

Patients and methods

Results

Discussion

Conclusion

Introduction

Methods

Markov model

Figure 1.

Screening strategies

Transition probabilities

Table 1.

Development, growth and rupture of UIA

Costs

Utilities

Cost-effectiveness analysis

Sensitivity analysis

Results

Table 2.

Figure 2.

Discussion

Supplementary Material

Acknowledgements

Declaration of Conflicting Interests

Funding

Ethical approval

Informed consent

Guarantor

Contributorship

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases