The unruptured intracranial aneurysm treatment score: a multidisciplinary consensus

Abstract

Objective: We endeavored to develop an unruptured intracranial aneurysm (UIA) treatment score (UIATS) model that includes and quantifies key factors involved in clinical decision-making in the management of UIAs and to assess agreement for this model among specialists in UIA management and research.

Methods: An international multidisciplinary (neurosurgery, neuroradiology, neurology, clinical epidemiology) group of 69 specialists was convened to develop and validate the UIATS model using a Delphi consensus. For internal (39 panel members involved in identification of relevant features) and external validation (30 independent external reviewers), 30 selected UIA cases were used to analyze agreement with UIATS management recommendations based on a 5-point Likert scale (5 indicating strong agreement). Interrater agreement (IRA) was assessed with standardized coefficients of dispersion (vr*) (vr* = 0 indicating excellent agreement and vr* = 1 indicating poor agreement).

Results: The UIATS accounts for 29 key factors in UIA management. Agreement with UIATS (mean Likert scores) was 4.2 (95% confidence interval [CI] 4.1-4.3) per reviewer for both reviewer cohorts; agreement per case was 4.3 (95% CI 4.1-4.4) for panel members and 4.5 (95% CI 4.3-4.6) for external reviewers (p = 0.017). Mean Likert scores were 4.2 (95% CI 4.1-4.3) for interventional reviewers (n = 56) and 4.1 (95% CI 3.9-4.4) for noninterventional reviewers (n = 12) (p = 0.290). Overall IRA (vr*) for both cohorts was 0.026 (95% CI 0.019-0.033).

Conclusions: This novel UIA decision guidance study captures an excellent consensus among highly informed individuals on UIA management, irrespective of their underlying specialty. Clinicians can use the UIATS as a comprehensive mechanism for indicating how a large group of specialists might manage an individual patient with a UIA.

Figure 1. Study flow of the Delphi consensus process

Participant frequencies for each round are given in parentheses. The panel member group consisted of 28 neurosurgeons (5 of whom were dually trained in endovascular and microsurgical aneurysm repair), 7 interventional neuroradiologists, 3 neurologists, and 1 clinical epidemiologist. The external reviewer group consisted of 15 neurosurgeons (7 of whom were dually trained in endovascular and microsurgical aneurysm repair), 7 interventional neuroradiologists, and 8 neurologists. CV = cerebrovascular; UIATS = unruptured intracranial aneurysm treatment score.

Figure 2. The unruptured intracranial aneurysm treatment score

The unruptured intracranial aneurysm treatment score (UIATS) model includes and quantifies the key factors for clinical decision-making in the management of unruptured intracranial aneurysms (UIAs), developed based on relevance rating data from Delphi consensus rounds 1–4. To calculate a management recommendation for a UIA, the number of points corresponding to each patient-, aneurysm-, or treatment-related feature on both management columns of the scoring form (“in favor of UIA repair” and “in favor of UIA conservative management”) are added up. This will lead to 2 numerical values, 1 favoring aneurysm repair (surgical or endovascular), and 1 favoring conservative management. The definitions for each category and factor are found in the Methods section. For cases with a score difference of 3 points or more, the direction, i.e., the difference between the calculated numerical values on each side of the recommendation columns, will suggest an individual management recommendation (i.e., aneurysm repair or conservative management). For cases that have similar aneurysm treatment and conservative management scores (±2 point difference or less), the recommendation is “not definitive” and either management approach could be supported, as additional factors apart from those used in the development of UIATS may be considered in making a final decision regarding the management recommendation and long-term follow-up. For cases with multiple aneurysms, every aneurysm must be evaluated separately, which will then also result in separate recommendations for each aneurysm. *The minimal intervention-related risk is always added as a constant factor (5 points). AComA = anterior communicating artery; BasA = basilar artery; BP = blood pressure; multiple = multiple selection category; PComA = posterior communicating artery; SAH = subarachnoid hemorrhage; single = single selection category.

Figure 3. Validation of the UIATS

(A) Agreement with unruptured intracranial aneurysm treatment score (UIATS)-derived recommendations per case and per rater. Means for Likert scores (y-axis) are illustrated for each case (dots) and for each rater (circles) among panel members and the blinded external reviewers (x-axis). A Likert score of 5 indicates strong agreement; 4 indicates agreement; 3 indicates neutrality; 2 indicates disagreement; and 1 indicates strong disagreement. Since means for Likert scores did not fall below a score of 3, the y-axis scale does not show disagreement and strong disagreement. Compared to agreement of panel members with UIATS-derived treatment recommendations for each case, the mean Likert score (y-axis) was significantly higher among the external reviewers, who were completely blinded to the score raw data and the survey development and design (*indicates p = 0.017). (B) Correlation of Likert scores and UIATS characteristics. The level of agreement (y-axis) between panel members and external reviewers was correlated (Pearson product-moment) with the UIATS differences between aneurysm repair and conservative management for every UIATS treatment recommendation (x-axis). The score magnitude correlated significantly with agreement among the reviewers, independently for panel members (r² = 0.323, p = 0.002; solid line) and external reviewers (r² = 0.399, p < 0.001; dotted line).

Figure 4. UIATS case recommendation resulting in the highest agreement among the internal and external reviewer cohorts

Corresponding scores in accordance with unruptured intracranial aneurysm treatment score (UIATS) features are given in parentheses. A catheter angiogram of a 47-year-old woman (3 points favoring treatment for patient age 41–60 years) who previously underwent cranial MRI for chronic headaches with (A) posterior-anterior and (B) lateral projection as well as (C) posterior-anterior and (D) lateral 3-dimensional reconstructions is shown. This incidental irregularly shaped (3 points favoring treatment for irregular morphology) anterior communicating artery aneurysm (arrow, 2 points favoring treatment for aneurysm location) had a maximum diameter of 7.6 mm (2 points favoring treatment for aneurysm diameter) with a neck diameter of 3.5 mm. Aspect and size ratios were calculated to be 2.1 and 3.8, respectively (1 point favoring treatment for aspect or size ratio greater than 1.6 or 3.0, respectively). Her medical history included arterial hypertension (2 points favoring treatment for risk factor hypertension) but no other chronic comorbidities. The resulting scoring based on the UIATS was 13 points in favor of aneurysm repair and 7 points in favor of conservative management (1 point for patient age 41–60 years, 1 point for aneurysm size 6–10 mm, and 5 points for the constant intervention-related risk). The resulting UIATS recommendation was “aneurysm repair.” Overall agreement with this UIATS recommendation was 4.73 (95% confidence interval 4.62–4.85) for both reviewer cohorts.

Figure 5. UIATS case recommendation resulting in the lowest agreement among the internal and external reviewer cohorts

Corresponding scores in accordance with unruptured intracranial aneurysm treatment score (UIATS) features are given in parentheses. A catheter angiogram of a 25-year-old woman (4 points favoring treatment for patient age younger than 40 years) who previously underwent MRI for persistent headaches, vertigo, and occasional bitemporal vision disturbances (due to a migraine aura) with (A) posterior-anterior and (B) lateral projection as well as (C) posterior-anterior and (D) lateral 3-dimensional reconstructions is shown. This left-sided wide-necked paraophthalmic internal carotid artery aneurysm had a maximum diameter of 3 mm with a neck diameter of 2.7 mm (3 points favoring conservative management for aneurysm complexity due to wide neck). The patient did not report any additional risk factors or comorbidities in her past medical history. The resulting scores based on the UIATS were 4 points in favor of aneurysm repair and 8 points (including 5 points favoring conservative management for the constant intervention-related risk) in favor of conservative management. The resulting UIATS recommendation was “conservative management.” Overall agreement with this UIATS recommendation was 3.7 (95% confidence interval 3.44–3.96) for both reviewer cohorts.