SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2322041667> ?p ?o ?g. }

Showing items 1 to 95 of 95 with 100 items per page.

W2322041667 endingPage "144" @default.
W2322041667 startingPage "137" @default.
W2322041667 abstract "Subarachnoid hemorrhage (SAH) is a significant cause of morbidity and mortality in neurosurgical patients. The most common cause of spontaneous SAH is a ruptured intracranial aneurysm, accounting for roughly 75% to 85% of cases.1-6 The gold standard test for diagnosing and characterizing an aneurysm is a digital subtraction angiogram (DSA) with 3-dimensional rotational angiography of the cerebral vessels. Although DSA is a safe procedure, there are associated risks, including puncture site hematoma, contrast-induced nephropathy, and stroke.7-9 As a result, noninvasive tests have been developed to visualize and characterize intracranial aneurysms. Computed tomographic angiography (CTA) has gained popularity as an initial imaging test in patients with SAH because of its reportedly high sensitivity and specificity for detecting ruptured aneurysms.10-12 The noninvasive nature of CTA and the ease and rapidity with which the study can be performed are additional reasons for its acceptance in the workup of spontaneous SAH. Another advantage of CTA is that treatment decisions regarding endovascular coiling or microsurgical clipping can be made without a formal DSA on every patient. However, with the rise in the number of aneurysms being treated with coiling, some have argued that the practice of performing a CTA is an unnecessary use of time and resources because a diagnostic angiogram is performed as part of the endovascular treatment.13,14 The purpose of this study is to perform a cost-effectiveness analysis of 2 strategies for treating patients found to have spontaneous SAH using a decision tree model. In 1 strategy, a spontaneous SAH was followed up with a CTA, whereas in the other, it was followed up with a DSA. The parameters of the decision tree were then allowed to vary over plausible distributions to understand which strategy would be most cost-effective in different scenarios. METHODS Decision Tree A decision tree was created in TreeAge Pro Suite 2012 with the initial condition of a spontaneous SAH. CTA and DSA were the potential options at the initial decision node. On the CTA arm, a chance node was created with 1 arm, labeled “aneurysm found,” representing the proportion of patients in whom the aneurysm was identified by CTA. The probability of this outcome corresponded to the sensitivity of CTA. The other arm of this chance node, labeled “aneurysm not found,” represented the proportion of patients with a false-negative CTA. If an aneurysm was not found on CTA, it was assumed that a follow-up DSA was used to identify the aneurysm. Chance nodes were then created corresponding to the treatment options, namely coiling or clipping. From that point, chance nodes were created corresponding to outcomes in each arm. Terminal nodes were divided on the basis of modified Rankin Scale (mRS) scores. An mRS score of 0 to 2 corresponded to a good outcome; an mRS score of 3 to 5 corresponded to a poor outcome; and an mRS score of 6 corresponded to death. Each terminal node had a cost and utility assigned to it. The second branch of the decision node represented the strategy of following up a spontaneous SAH with a DSA. This branch led to a chance node corresponding to the probability of developing a complication from the DSA itself. From there, chance nodes were made corresponding to the treatment options, as seen above for the CTA arm. As in the CTA arm, terminal nodes were based on mRS outcome groups, with costs and utilities assigned to each terminus. A diagram of the decision tree is given in Figure 1.FIGURE 1: Structure of decision tree model used in this study. The Table provides an explanation of abbreviations.Input Parameters Each probability in this model was based on values gathered from the published literature. The sensitivity of CTA was taken from a systematic review of several studies investigating the accuracy of CTA for detecting a ruptured aneurysm in the setting of SAH.11,15-35 Only studies that were based on a 16-slice or higher CTA resolution were considered. The complication rate of DSA was based on several studies investigating the rates of complications.7-9,36-38 Given the wide variability of reported complication rates, we assumed the mean value of 1% for the base-case scenario (This was reconciled in the sensitivity analysis by allowing each parameter to vary over a plausible range of values). The probabilities of each outcome group were based on the path of the decision tree. Paths that led to clipping had outcome braches constructed to model the results of the surgical arm of the International Subarachnoid Hemorrhage Trial (ISAT).39 Paths that lead to coiling had outcome branches constructed to model the endovascular arm of the ISAT study. Complication rates of both endovascular and surgical arms were also taken from the ISAT data.39 Costs Costs were approximated on the basis of whether they were associated with a procedure/diagnostic test, a complication, or hospitalization for a ruptured aneurysm. All costs were standardized to 2012 US dollars. Costs associated with procedures/diagnostic tests were taken from the Centers for Medicare and Medicaid Services Web site based on Current Procedural Terminology codes associated with each test.40 DSA was assumed to consist of a 6-vessel angiogram (bilateral internal carotid arteries, bilateral external carotid arteries, and bilateral vertebral arteries) when costs were computed. Costs associated with complications were estimated from published data in the cerebral angiography literature.7-9,36-38 Because cost associated with complications varied depending on reference cited, we assumed a large range (± 20%) when performing sensitivity analyses and Monte Carlo simulations. Costs associated with increased hospitalization were assigned to paths that had associated complications (whether from DSA or from endovascular/surgical treatment). The value of this increased hospitalization was based on estimates from the National Inpatient Survey for International Classification of Diseases, Ninth Modification code 430 (SAH).41 Because the National Inpatient Survey values represent only a sample of hospitals in the United States, we allowed this value to vary over a wide range (± 20%) when performing sensitivity analyses and Monte Carlo simulations. Utilities Each terminal node was associated with a utility value based on the commonly accepted concept of a quality-adjusted life-year (QALY). The utility weights for each mRS grouping were obtained from a study by Samsa et al.42 Utilities were allowed to vary over a plausible range for the Monte Carlo simulation using a uniform distribution. The simulation in this analysis was performed using a time threshold of 1 year, which is the time period after which many SAH patients, in our experience, have reached their new neurological baseline. Statistical Analysis All analyses were performed with TreeAge Pro Suite 2012 (TreeAge, Inc). The base-case analysis was performed using the mean value for each variable. Because there is high regional and institutional variability with respect to the proportion of aneurysms treated with clipping vs coiling, we assumed 25% of aneurysms to be treated with clipping for the base-case scenario. This value is derived from our institutional experience over the past year and seems to reflect the overall trend throughout the country. The willingness-to-pay (WTP) threshold for this analysis was set at $50 000/QALY, representing the dollar value of a year in perfect health. One-way sensitivity/threshold analysis was then performed for the variables corresponding to the proportion of aneurysms treated with clipping, sensitivity of CTA, and DSA complication rate because these were of most interest in this study. A 3-way sensitivity/threshold analysis also was performed, combining the effects of all 3 of these variables to determine the parameters of cost-effectiveness in the model. A Monte Carlo simulation (probabilistic sampling) was performed by giving each variable in the model a plausible distribution and varying all parameters simultaneously. Cost and utilities were approximated using a normal distribution. Probabilities that were based on strong evidence (eg, CTA sensitivity) were also approximated using a normal distribution. If the probability had high variability (eg, outcomes or complication rates), a β distribution was selected. For nodes with >2 branches (eg, outcome nodes), a Dirichlet distribution was used. The simulation was run using 10 000 iterations to capture stability in the model. RESULTS Base-Case Scenario The Table shows the parameters of the distributions used in this model. For the base-case scenario, the mean value from each distribution was used to assess the more cost-effective strategy. For these calculations, a WTP threshold of $50 000/QALY was used. CTA was found to have an expected cost of $40 197.13 with an expected utility of 0.63 QALY. DSA was found to have an expected cost of $40 248.58 with an expected utility of 0.63 QALY. The CTA branch was seen to be more cost-effective than DSA (cost/effectiveness ratio, $63 805/QALY vs $63 887/QALY, respectively).TABLE: Distributions Associated With Parameters Used in Decision TreeOne-Way Sensitivity Analysis One-way sensitivity analysis was performed on the proportion of aneurysms treated with surgery as it related to the model. As expected, the greater the proportion of aneurysms treated with surgery is, the more often CTA would be the most cost-effective option. With the use of a WTP threshold of $50 000/QALY, the proportion of cases treated by surgery would need to fall to <22.4% for DSA to become the more cost-effective strategy (Figure 2). This threshold assumes that the other variables are held at the mean values.FIGURE 2: One-way sensitivity analysis for the variable pClipping (proportion of aneurysms treated with surgical clipping). The threshold for cost-effectiveness was 22.4%. QALY, quality-adjusted life-year.A second key variable in the model was the sensitivity of CTA in detecting a ruptured aneurysm. As expected, the higher the sensitivity of CTA is, the more cost-effective CTA becomes as an initial screening test in this scenario. A one-way sensitivity analysis was performed on CTA sensitivity (again holding all other parameters at the mean value of their respective distributions), revealing that CTA was a more cost-effective screening test provided that the sensitivity was >87.0% (Figure 3).FIGURE 3: One-way sensitivity analysis for the variable pSensitivity_CTA (sensitivity of computed tomographic angiography for detecting a ruptured aneurysm). The threshold for cost-effectiveness was 87.0%. QALY, quality-adjusted life-year.Another variable that contributed strongly to the cost-effectiveness analysis was the complication rate related to DSA. As the complication rate of DSA decreases, the cost-effectiveness of DSA as the initial screening test increases. The results of the 1-way sensitivity analysis on this variable are demonstrated in Figure 4. The threshold for which DSA became the more cost-effective study, with all other variables kept constant, was a complication rate <0.58%.FIGURE 4: One-way sensitivity analysis for the variable pDSA_Complication (complication rate of digital subtraction angiography). The threshold for cost-effectiveness was 0.58%. QALY, quality-adjusted life-year.Three-Way Sensitivity Analysis A 3-way sensitivity analysis was performed using the aforementioned variables. This type of analysis is difficult to represent graphically because the threshold analysis corresponds to a 3-dimensional plane of values; however, the results can be represented mathematically. CTA would be the more cost-effective study if the equation below were satisfied: where CTA Sens corresponds to the sensitivity of CTA, Clip corresponds to the proportion of aneurysms treated with surgical clipping, and DSA Comp corresponds to the complication rate of DSA. For example, if the CTA sensitivity were 0.92, the proportion of aneurysm clipped were 0.20, and the DSA complication rate were 0.005, the left side of the equation would return a value of −1.41, implying that DSA would be the more cost-effective screening method based on those values. However, if the sensitivity of CTA were much higher (eg, 0.98) and the proportion of aneurysms treated with clipping were higher (0.3), then the left side of the equation would return a value of 1.05, implying that CTA would be the more cost-effective screening test. Monte Carlo Simulation Results of the 10 000-iteration probabilistic sampling analysis (or Monte Carlo simulation) demonstrated that DSA was the more cost-effective strategy in 54.3% of cases when a WTP threshold of $50 000/QALY was used. The incremental cost-effectiveness ratio (the difference in average cost between the 2 strategies divided by the difference in average effectiveness between the 2 strategies) was $14 110/QALY. DISCUSSION The advent of CTA has changed the ease and promptness with which treating physicians can identify and characterize a ruptured intracranial aneurysm. As a result, the use of CTA as the initial screening test for an aneurysm in the setting of a spontaneous SAH has been increasing. Theoretically, the increased efficiency in identifying an aneurysm can lead to faster decision making with respect to how the aneurysm should be treated; however, as the CTA technology has improved, there has been a concomitant improvement and use of endovascular treatments of ruptured aneurysms. As the number of aneurysms treated with endovascular therapies increases, screening for a ruptured aneurysm with CTA becomes less necessary because a diagnostic angiogram is performed as part of the treatment. In the present study, we sought to explore this situation from a cost-effectiveness perspective using modern decision analysis techniques. Decision-tree modeling allows a complex decision to be broken down into its simpler components so that each major factor leading to the decision can be analyzed. In the present study, the 3 most important factors for assessing the cost-effectiveness of CTA as the initial screening test for a ruptured aneurysm are the proportion of aneurysms treated by clipping, the sensitivity of CTA, and the complication rate from DSA. When the base-case scenario (each parameter of the model assigned the mean value of the distribution) was examined, CTA was found to be the more cost-effective study. This base-case scenario assumed a CTA sensitivity of 95%, a rate of 25% for surgical clipping as the modality of aneurysm treatment, and a DSA complication rate of 1%. The power of decision analysis techniques stems from the ability to adjust the parameters and to identify the thresholds at which the decision would change. In an examination of the parameter of surgical clipping rate alone, if the rate were <22.4%, DSA would become the more cost-effective screening test. This threshold makes intuitive sense, supporting the validity of the model. As the rate of aneurysms treated with endovascular methods increases, the more unnecessary CTA becomes as a screening test for the reasons mentioned previously. The same type of analysis was performed for the other main variables in the model. When CTA sensitivity alone is examined, the threshold for which DSA became the more cost-effective screening test was value below 87.0%. Again, the results here support intuition; if CTA were missing ≥13% of the ruptured aneurysms, it would no longer be cost-effective as a screening test because each of those negative CTA studies would need to be followed up with a DSA to establish the presence of an aneurysm. A similar analysis was performed for DSA complication rate, demonstrating that a complication rate from DSA <0.58% resulted in DSA becoming the more cost-effective screening study. One of the major advantages of CTA is its noninvasive nature, and the use of CTA helps avoid the potential risks of DSA, including access site hematoma, contrast-induced nephropathy, and stroke. On the other hand, if the rate of DSA complications is found to be acceptably low (<0.58% in this case), the theoretical advantage of avoiding those risks is not supported on a cost-effectiveness basis. The results of the 3-way sensitivity analysis allow us to incorporate all 3 variables described above into a single analysis. The 1 disadvantage of this type of analysis is that it is difficult to demonstrate the results graphically. As seen in the 1-way sensitivity analysis, the threshold value corresponds to the point where 2 lines intersect. Similarly, a 2-way sensitivity analysis would correspond to a threshold “line” where 2 planes would intersect, and a 3-way sensitivity analysis would correspond to a threshold “plane” where two 3-dimensional objects would intersect. Fortunately, the results of a 3-way sensitivity analysis can easily be represented and interpreted mathematically, as shown in the equation above. Using the equation, practitioners or institutions can take their own values for CTA sensitivity, proportion of aneurysms treated with surgical clipping, and DSA complication rate and use those values to determine which study is the most appropriate screening test in their particular clinical setting, as demonstrated above. Of course, further variables in the model could be incorporated into a higher-level sensitivity analysis (cost of a DSA, cost of surgical clipping, etc); however, the utility of the analysis would quickly diminish as the equations become more cumbersome. The final analysis performed in this study was a Monte Carlo simulation in which all the variables of the model were assigned a probable distribution and simulations were run with values chosen at random for each variable. The Monte Carlo simulation was run 10 000 times to ensure stability in the model and results. The results show that in 54.3% of cases (using a WTP of $50 000/QALY), DSA was the more cost-effective screening test to identify a ruptured aneurysm. The incremental cost-effectiveness ratio for the Monte Carlo simulation was $14 110/QALY, representing the WTP threshold at which there is equipoise between the 2 strategies. As with any decision analysis model, we recognize there are several limitations. Chiefly, the quality of the results is only as good as the quality of the input parameters (ie, the “garbage in, garbage out” phenomenon). Although we feel that the distributions and values used in this model all are based on realistic and valid data, we recognize that there is no way to model each variable perfectly. These distributions are all estimations and can be a source of error when interpreting the model. Additionally, cost data regarding complications are notoriously difficult to quantify. Although we have tried to identify appropriate sources for the values used in this study, we accept that they may not represent real-world values. We attempted to reconcile this uncertainty by assigning a wide distribution when performing the Monte Carlo simulation. An assumption used in the model was that any patient who had a CTA revealing an aneurysm that was treated with endovascular coiling had a full angiogram performed as part of the treatment. Although this practice may be common, we recognize that some physicians will use the CTA data to perform a targeted angiogram in the vessel harboring the aneurysm, thereby decreasing the cost of the study. Although we anticipate that this fact would not have a significant impact on the overall model, it may lower the threshold value for which CTA would be the more cost-effective study. Additionally, this study applies to aneurysms that can be treated by either endovascular or microsurgical methods. Obviously, there are aneurysms that are not amenable to 1 treatment modality or the other. This study does not apply to such cases; however, in our experience, these are much rarer situations and are unlikely to affect the overall results and conclusions significantly. CONCLUSION CTA has increasingly been used as an initial screening study to identify a ruptured aneurysm; however, with a similarly increasing trend toward endovascular aneurysm repair, performing a CTA on every spontaneous SAH may not be cost-effective. Myriad factors enter into the decision regarding the choice of screening test to be used for these patients. In our model, we identified CTA sensitivity, proportion of aneurysms treated with surgical clipping, and DSA complication rate as the most important variables affecting this decision from a cost-effectiveness standpoint. Our decision tree model and the results of the cost-effectiveness analysis can stand as a platform for providers to assess whether their practices regarding the screening of ruptured aneurysms are cost-effective. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article." @default.
W2322041667 created "2016-06-24" @default.
W2322041667 creator A5002202383 @default.
W2322041667 creator A5020822582 @default.
W2322041667 creator A5041173198 @default.
W2322041667 date "2014-08-01" @default.
W2322041667 modified "2023-09-24" @default.
W2322041667 title "Cost-Effectiveness of Computed Tomographic Angiography in Screening for Aneurysm in Spontaneous Subarachnoid Hemorrhage" @default.
W2322041667 cites W1589207099 @default.
W2322041667 cites W179951671 @default.
W2322041667 cites W1838439429 @default.
W2322041667 cites W1970647439 @default.
W2322041667 cites W1991798393 @default.
W2322041667 cites W1993897307 @default.
W2322041667 cites W2012097477 @default.
W2322041667 cites W2015416581 @default.
W2322041667 cites W2028531084 @default.
W2322041667 cites W2031187286 @default.
W2322041667 cites W2048997813 @default.
W2322041667 cites W2056227825 @default.
W2322041667 cites W2060209723 @default.
W2322041667 cites W2083182358 @default.
W2322041667 cites W2085746413 @default.
W2322041667 cites W2087938242 @default.
W2322041667 cites W2088738256 @default.
W2322041667 cites W2099026177 @default.
W2322041667 cites W2107655025 @default.
W2322041667 cites W2114389781 @default.
W2322041667 cites W2115815342 @default.
W2322041667 cites W2120793783 @default.
W2322041667 cites W2126826311 @default.
W2322041667 cites W2129177890 @default.
W2322041667 cites W2129957678 @default.
W2322041667 cites W2133408969 @default.
W2322041667 cites W2155157499 @default.
W2322041667 cites W2157312318 @default.
W2322041667 cites W2169436924 @default.
W2322041667 cites W262661252 @default.
W2322041667 cites W2729107688 @default.
W2322041667 doi "https://doi.org/10.1227/neu.0000000000000373" @default.
W2322041667 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/25036045" @default.
W2322041667 hasPublicationYear "2014" @default.
W2322041667 type Work @default.
W2322041667 sameAs 2322041667 @default.
W2322041667 citedByCount "6" @default.
W2322041667 countsByYear W23220416672015 @default.
W2322041667 countsByYear W23220416672016 @default.
W2322041667 countsByYear W23220416672022 @default.
W2322041667 countsByYear W23220416672023 @default.
W2322041667 crossrefType "journal-article" @default.
W2322041667 hasAuthorship W2322041667A5002202383 @default.
W2322041667 hasAuthorship W2322041667A5020822582 @default.
W2322041667 hasAuthorship W2322041667A5041173198 @default.
W2322041667 hasConcept C126838900 @default.
W2322041667 hasConcept C141071460 @default.
W2322041667 hasConcept C2776098176 @default.
W2322041667 hasConcept C2777736543 @default.
W2322041667 hasConcept C2778749279 @default.
W2322041667 hasConcept C2780643987 @default.
W2322041667 hasConcept C2983858294 @default.
W2322041667 hasConcept C2985771010 @default.
W2322041667 hasConcept C544519230 @default.
W2322041667 hasConcept C71924100 @default.
W2322041667 hasConceptScore W2322041667C126838900 @default.
W2322041667 hasConceptScore W2322041667C141071460 @default.
W2322041667 hasConceptScore W2322041667C2776098176 @default.
W2322041667 hasConceptScore W2322041667C2777736543 @default.
W2322041667 hasConceptScore W2322041667C2778749279 @default.
W2322041667 hasConceptScore W2322041667C2780643987 @default.
W2322041667 hasConceptScore W2322041667C2983858294 @default.
W2322041667 hasConceptScore W2322041667C2985771010 @default.
W2322041667 hasConceptScore W2322041667C544519230 @default.
W2322041667 hasConceptScore W2322041667C71924100 @default.
W2322041667 hasIssue "Supplement 1" @default.
W2322041667 hasLocation W23220416671 @default.
W2322041667 hasLocation W23220416672 @default.
W2322041667 hasOpenAccess W2322041667 @default.
W2322041667 hasPrimaryLocation W23220416671 @default.
W2322041667 hasRelatedWork W101948563 @default.
W2322041667 hasRelatedWork W1841115594 @default.
W2322041667 hasRelatedWork W1993578873 @default.
W2322041667 hasRelatedWork W2007073505 @default.
W2322041667 hasRelatedWork W2023684400 @default.
W2322041667 hasRelatedWork W2024110249 @default.
W2322041667 hasRelatedWork W2402771989 @default.
W2322041667 hasRelatedWork W4233227558 @default.
W2322041667 hasRelatedWork W4242019499 @default.
W2322041667 hasRelatedWork W4243347386 @default.
W2322041667 hasVolume "61" @default.
W2322041667 isParatext "false" @default.
W2322041667 isRetracted "false" @default.
W2322041667 magId "2322041667" @default.
W2322041667 workType "article" @default.