There is a national trend for accountable care and

Transcription

1 J Neurosurg 120: , 2014 AANS, 2014 Predicting inpatient complications from cerebral aneurysm clipping: the Nationwide Inpatient Sample Clinical article *Kimon Bekelis, M.D., 1 Symeon Missios, M.D., 2 Todd A. MacKenzie, Ph.D., 3 Atman Desai, M.D., 1 Adina Fischer, Ph.D., 4 Nicos Labropoulos, Ph.D., D.I.C., 5 and David W. Roberts, M.D. 1,4,6 2 Appledore Neurosurgery Group, Portsmouth Hospital, Portsmouth; 4 Geisel School of Medicine at Dartmouth, Hanover; 1 Section of Neurosurgery, and Departments of 3 Medicine and 6 Neurology, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire; and 5 Section of Vascular Surgery, SUNY Stony Brook, New York Object. Precise delineation of individualized risks of morbidity and mortality is crucial in decision making in cerebrovascular neurosurgery. The authors attempted to create a predictive model of complications in patients undergoing cerebral aneurysm clipping (CAC). Methods. The authors performed a retrospective cohort study of patients who had undergone CAC in the period from 2005 to 2009 and were registered in the Nationwide Inpatient Sample (NIS) database. A model for outcome prediction based on preoperative individual patient characteristics was developed. Results. Of the 7651 patients in the NIS who underwent CAC, 3682 (48.1%) had presented with unruptured aneurysms and 3969 (51.9%) with subarachnoid hemorrhage. The respective inpatient postoperative risks for death, unfavorable discharge, stroke, treated hydrocephalus, cardiac complications, deep vein thrombosis, pulmonary embolism, and acute renal failure were 0.7%, 15.3%, 5.3%, 1.5%, 1.3%, 0.6%, 2.0%, and 0.1% for those with unruptured aneurysms and 11.5%, 52.8%, 5.5%, 39.2%, 1.7%, 2.8%, 2.7%, and 0.8% for those with ruptured aneurysms. Multivariate analysis identified risk factors independently associated with the above outcomes. A validated model for outcome prediction based on individual patient characteristics was developed. The accuracy of the model was estimated using the area under the receiver operating characteristic curve, and it was found to have good discrimination. Conclusions. The featured model can provide individualized estimates of the risks of postoperative complications based on preoperative conditions and can potentially be used as an adjunct in decision making in cerebrovascular neurosurgery. ( Key Words cerebral aneurysm aneurysm clipping stroke risk prediction Nationwide Inpatient Sample vascular disorders There is a national trend for accountable care and rationalization of the decision-making process in all areas of medicine. 6 Although large randomized trials 3,11 13 have focused on the appropriate treatment in patients with cerebral aneurysms, little attention has been paid to the short-term complications of such interventions, especially in real-world practice. Negative outcomes can result in a significant morbidity burden and should be taken into account when considering treatment options. The parallel trend for negative financial incentivization of complications necessitates benchmarking of cerebral aneurysm clipping (CAC) outcomes and identifying the modifiable patient-level risk factors associated with them. Abbreviations used in this paper: ARF = acute renal failure; AVM = arteriovenous malformation; CAC = cerebral aneurysm clipping; CAD = coronary artery disease; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; CRF = chronic renal failure; DVT = deep vein thrombosis; NIS = Nationwide Inpatient Sample; PE = pulmonary embolism; PVD = peripheral vascular disease; SAH = subarachnoid hemorrhage; SE = standard error. * Drs. Bekelis and Missios contributed equally to this work. J Neurosurg / Volume 120 / March 2014 Although some practicing surgeons recognize the risks of the procedures, justifying the decision-making process based on individual patient characteristics is frequently arbitrary. Several studies have attempted to identify such complications; however, they have mostly focused on comparing adverse events in patients undergoing clipping versus coiling. There has been no investigation of modifiable patient-level risk factors that can affect outcomes. Further, most of the studies have been retrospective analyses of single-institution experiences, 10 demonstrating results with limited generalization given their inherent selection bias. The interpretation of other multicenter studies is equally limited given their focus on specific subgroup data 1,2,4,5,7,8 or their consideration of all morbidity as one variable. 3,15 The Nationwide Inpatient Sample (NIS) is a hospital discharge database that represents approximately 20% of all inpatient admissions to nonfederal hospitals in the US. 17 It allowed unrestricted study of the patient popu- This article contains some figures that are displayed in color on line but in black-and-white in the print edition. 591

2 K. Bekelis et al. lation in question. Using this database, we identified the preoperative comorbidities associated with postoperative death, unfavorable discharge, stroke, treated hydrocephalus, cardiac complications, deep vein thrombosis (DVT), pulmonary embolism (PE), acute renal failure (ARF), and prolonged hospital stay in patients who underwent CAC. With these data, we developed a risk factor based predictive model for negative outcomes in CAC surgery. Methods Nationwide Inpatient Sample Database All patients in the NIS database 17 (Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality, Rockville, MD) who had undergone CAC between 2005 and 2009 were included in our analysis. For these years, the NIS contains discharge data for 100% of discharges from a stratified random sample of nonfederal hospitals in several states to approximate a representative 20% subsample of all nonfederal US hospital discharges. More information about the NIS is available at Cohort Definition To establish the study cohort, we used ICD-9-CM codes to identify patients in the registry who had undergone clipping (ICD-9-CM code 39.51) for ruptured (ICD- 9-CM code 430, excluding for ruptured syphilitic aneurysm; for cerebral arteritis; for arteriovenous malformation [AVM]; , , , and for traumatic hemorrhage; and or for treatment diagnosis for AVM repair or radiosurgery) and unruptured (ICD-9-CM code 437.3) cerebral aneurysms between 2005 and Independent and Outcome Variables The primary outcome variables (Table 1) were the inpatient postoperative risks of stroke, cardiac complications, death, treated hydrocephalus, DVT, PE, ARF, length of hospital stay more than 4 days for patients with unruptured aneurysms, length of hospital stay more than 14 days for patients with ruptured aneurysms, and unfavorable discharge (transfer to short-term hospital, skilled nursing facility, intermediate care, or rehabilitation center). The effect of pertinent exposure variables on outcomes was examined in a multivariate analysis. Age was the only continuous variable. Sex, diabetes mellitus, chronic obstructive pulmonary disease (COPD), hypertension, hypercholesterolemia, peripheral vascular disease (PVD), congestive heart failure (CHF), coronary artery disease (CAD), history of stroke, obesity, chronic renal failure (CRF), and coagulopathy were categorical variables. Statistical Analysis Logistic regression was used to assess the ability of each preoperative patient characteristic to predict inpatient postoperative complications. Based on these data, a multivariable logistic regression model was developed to predict each complication. The predictive ability of each of these models was assessed using the c-index (area under the receiver operating characteristic curve), which was corrected for overfitting bias by using leaveout cross-validation (for example, leave 10% out, repeated 1000 times, as well as bootstrap validation, repeated 1000 times). The Hosmer-Lemeshow test was used to assess the calibration of each model. Interactions were tested, but none were significant at the threshold (0.001) we set to correct for multiple testing (20 19/2 = 190 interactions in all). Moreover, none improved the c-index by more than Statistical analyses were performed using Stata version 12 (StataCorp LP), XLSTAT version (Adinsoft), and R version (R Project for Statistical Computing). The level of statistical significance was set at 0.05 based on 2-tailed tests. Results Demographics and Clinical Characteristics In the selected study period, 7651 patients registered in the NIS had undergone CAC (71.4% females, mean age 53.5 years), and 3682 (48.1%) of them had presented with unruptured aneurysms and 3969 (51.9%) with SAH (Fig. 1 and Table 2). Clinical Outcomes The inpatient postoperative incidence of death, unfavorable discharge, stroke, treated hydrocephalus, cardiac complications, DVT, PE, and ARF was 0.7%, 15.3%, 5.3%, 1.5%, 1.3%, 0.6%, 1.96%, and 0.1% for unruptured aneurysms and 11.5%, 52.8%, 5.5%, 39.2%, 1.7%, 2.8%, 2.6%, and 0.8% for ruptured aneurysms, respectively (Table 3). All of the outcomes, except for stroke, cardiac complications, and PE, were encountered significantly more often in cases of SAH. Multivariate Analysis of Inpatient Postoperative Outcomes A multivariate analysis of the effects of several preoperative factors on the risks of inpatient death (Fig. 2A), stroke (Fig. 2B), treated hydrocephalus (Fig. 2C), unfavorable discharge (Fig. 2D), length of hospital stay more than 4 days (Fig. 2E), cardiac complications (Fig. 3A), ARF (Fig. 3B), DVT (Fig. 3C), and PE (Fig. 3D) for patients undergoing CAC for unruptured aneurysms was performed. A similar analysis for patients undergoing CAC for ruptured aneurysms was also performed (Figs. 4 and 5). Increasing age and history of stroke were universally associated with higher mortality. Risk factors frequently associated with other postoperative complications were diabetes, PVD, hypertension, and coagulopathy. Predictive Model Application The predicted risk for an unfavorable discharge in patients with unruptured and ruptured aneurysms, as well as for death and prolonged hospital stay in patients with ruptured aneurysms, was plotted against age for the most significant individual variables and their combinations (Fig. 6). An additive effect of the variables was observed for the risk for all 4 outcomes. The presence of CHF in patients with a history of stroke increased the possibility of an unfavorable discharge by one-third in patients with 592 J Neurosurg / Volume 120 / March 2014

3 Nationwide Inpatient Sample: cerebral aneurysm clipping TABLE 1: Definitions using ICD-9-CM codes* Category Code Note/Coding Rule diagnosis unruptured aneurysm (exclude 430 & 431) ICD-9 diagnosis codes ruptured aneurysm 430 (SAH) exclude ICD-9 diagnostic code (ruptured syphilitic aneurysm already excluded in 430); (cerebral arteritis), (AVM); , , , & (traumatic hemorrhage); or (treatment diagnosis for AVM repair or radiosurgery) procedure, clipping ICD-9 procedure codes independent variable sex age stroke , , ICD-9 diagnosis codes MI 410, , 411, 411.1, , 412, 413, 413.1, 413.9, 414, 414.2, 414.3, 414.4, 414.8, COPD 491.0, 492.0, 494.0, 496.0, & 305.1, V15.82, ICD-9 diagnosis codes CHF , , , , , , , DM , , coagulopathy , 287.1, CRF , , , , 585, 586, V42.0, V45.1, V56.0, V56.8 hypertension hypercholesterolemia 272, 272.0, 272.1, 272.2, 272.3, obesity , ICD-9 diagnosis codes alcohol abuse 291.1, 291.2, 291.5, 291.8, 291.9, , , V113 PVD , 441.2, 441.4, 441.7, 441.9, , 447.1, 557.1, 557.9, V43.4 outcome (complication) death unfavorable discharge (DISPUniform) stroke ICD-9 diagnosis codes treated hydrocephalus 02.2, 02.32, 02.33, 02.34, ICD-9 diagnosis codes cardiac complications ICD-9 diagnosis codes DVT 451.1, , , 451.2, , 451.9, 453.2, 453.8, ICD-9 diagnosis codes PE 415.0, 415.1, , , , V12.51 ICD-9 diagnosis codes ARF 584, 584.5, 584.6, 584.7, 584.8, 584.9, 583 length of hospital stay * DISPUniform = disposition of patient, uniform coding; DM = diabetes mellitus; MI = myocardial infarction. unruptured aneurysms. The presence of CRF had a similar effect in patients with a history of stroke and ruptured cerebral aneurysms. In these same patients, that is, those with a history of stroke and ruptured aneurysms, the presence of coagulopathy and COPD doubled the risk of death. Finally, the presence of coagulopathy in patients with ruptured aneurysms and a history of CAD increased the chance for a hospitalization longer than 2 weeks by one-third. Accuracy of the Model The area under the receiver operating characteristic curve, or the c statistic, was used to estimate the accuracy of the model. In patients with unruptured aneurysms, the bootstrap-validated c-index was calculated as 0.87 ± (mean ± SE) for postoperative risk of death, 0.74 ± for unfavorable discharge, 0.77 ± for J Neurosurg / Volume 120 / March 2014 stroke, 0.69 ± for treated hydrocephalus, 0.71 ± 0.14 for cardiac complications, 0.77 ± for DVT, 0.68 ± 0.03 for PE, 0.96 ± for ARF, and 0.62 ± for prolonged hospital stay, demonstrating good discrimination. In patients with ruptured aneurysms, the bootstrapvalidated c-index was calculated to be 0.66 ± for postoperative risk of death, 0.73 ± for unfavorable discharge, 0.70 ± for stroke, 0.57 ± for treated hydrocephalus, 0.63 ± 0.16 for cardiac complications, 0.58 ± for DVT, 0.62 ± for PE, 0.77 ± for ARF, and 0.65 ± for prolonged hospital stay, again demonstrating good discrimination. The cross-validated c-index gave similar results for ruptured and unruptured aneurysms. Each model had good calibration as assessed by a Hosmer-Lemeshow statistic. The Hosmer-Lemeshow test for the inpatient postoperative risk of death (p = 593

4 K. Bekelis et al. Fig. 1. Cohort selection for the study. 0.62), unfavorable discharge (p = 0.06), stroke (p = 0.88), treated hydrocephalus (p = 0.78), cardiac complications (p = 0.19), DVT (p = 0.50), PE (p = 0.13), ARF (p = 0.99), and prolonged hospital stay (0.002) demonstrated good calibration in patients with unruptured aneurysms. In those with ruptured aneurysms, the Hosmer-Lemeshow test for death (p = 0.59), unfavorable discharge (p = 0.37), stroke (p = 0.24), treated hydrocephalus (p = 0.77), cardiac complications (p = 0.04), DVT (p = 0.86), PE (p = 0.52), ARF (p = 0.22), and prolonged hospital stay (0.02) also demonstrated good calibration. Discussion The identification of modifiable risk factors associated with complications and the development of predictive models for outcomes are the cornerstones of defining quality in surgical health care delivery. 6 Accountable care will be based on such benchmarks. 6 Especially in cerebrovascular neurosurgery, the focus on mortality and angiographic outcomes has minimized attention to these important metrics. Using the NIS, we identified modifiable risk factors for such complications and created predictive models for negative outcomes. The application of these models can assist decision making by patients who are amenable to both clipping and coiling. Although an increasing age and a history of stroke were associated with higher mortality in all patients, a history of coagulopathy and COPD specifically increased the risk of death in those with SAH. The latter constellation of risk factors may be associated with a higher Hunt and Hess grade at presentation, resulting in worse outcomes. The protective effect of hypercholesterolemia may be associated with the potentially beneficial role of statins in the survival of patients with SAH. 19 The observed inpatient mortality 0.7% for those with unruptured aneurysms and 11.5% for those with SAH is slightly less than that reported in the current literature. 1 5,10,12 16 In previous studies of the NIS, that is, those not focusing on other complications or patient-level predictive factors, an increasing age 1 3 was also associated with higher mortality. Morbidity directly related to the surgical intervention (postoperative stroke and treated hydrocephalus) was also investigated. The incidence of stroke due to operative intervention was no different among patients with or without SAH. It was most commonly seen in patients with a history of ischemic stroke, probably because of their higher atherosclerotic burden. Treated hydrocephalus was observed in 39.2% of the patients with SAH and only a minority (1.5%) of patients with unruptured aneurysms. These rates were, as expected, higher than those reported for shunt placement in similar patients registered in the NIS. 7 Patients with a higher atherosclerotic burden (prior ischemic stroke or cardiac disease) appeared to have a higher incidence of severe hydrocephalus, perhaps because these patients have a higher Hunt and Hess grade at presentation and more complex operations, thus increasing the risk of hydrocephalus. The observed protective effect of hypercholesterolemia could again be associated with a beneficial effect of statin use on this outcome. 19 The observed incidence of general medical morbidity measures was similar in all patients undergoing CAC, except for the higher rate of DVT in patients with SAH. The lack of anticoagulation and prolonged immobilization in this group of patients may explain the latter. Increasing age and previous heart disease were associated with cardiac complications. A higher atherosclerotic burden (PVD) and CRF were common factors associated with ARF. Patients with coagulation disorders, including alcohol consumption, who probably underwent reversal of their coagulopathy, and others who were more prone to longer immobilization (obesity and prior stroke) had a higher incidence of DVT and PE. These rates are in accordance with those previously reported. 1 Finally, we investigated some metrics currently under scrutiny by Medicare when it considers appropriateness of reimbursement. A prolonged hospital stay was associated with increasing age, patient immobility (obesity and prior stroke), general medical comorbidities (CHF and diabetes), and coagulopathy. Unfavorable discharge was, as expected, more commonly seen in patients with SAH and was associated with increasing age, general medical comorbidities, and other factors (prior stroke and coagulopathy) that also contributed to prolonged hospitalizations. The effect of increasing age on unfavorable discharge has also been noted in prior studies. 1 3,19 The proposed predictive model for all outcomes was validated and demonstrated good accuracy and discrimination. In the context of the results of prior randomized studies, the model may serve as a useful tool for patient counseling and informed decision making in the preoperative phase of CAC. This model can be further refined and validated in prospective studies of this population. If the risk of morbidity and mortality is predicted to be unacceptably high for clipping, even after optimizing any modifi- 594 J Neurosurg / Volume 120 / March 2014

5 Nationwide Inpatient Sample: cerebral aneurysm clipping TABLE 2: Summary of preoperative demographic details in 7651 patients in the NIS who had undergone CAC between 2005 and 2009* Variable No. (%) All Patients Patients w/ Ruptured Aneurysm Patients w/ Unruptured Aneurysm p Value mean age in yrs ± ± ± sex F 5465 (71.43) 2748 (69.24) 2717 (73.79) <0.001 M 2186 (28.57) 1221 (30.76) 965 (26.21) TIA (6.05) 432 (10.88) 31 (0.84) < stroke (6.26) 316 (7.96) 163 (4.43) < CAD (7.24) 367 (9.25) 187 (5.08) < COPD (14.34) 460 (11.59) 637 (17.30) < CHF (2.98) 167 (4.21) 61 (1.66) < DM (10.47) 412 (10.38) 389 (10.56) coagulopathy (2.60) 153 (3.85) 46 (1.25) < CRF (0.77) 26 (0.66) 33 (0.90) hypertension (44.11) 1784 (44.95) 1591 (43.21) hypercholesterolemia (17.27) 495 (12.47) 826 (22.43) < obesity (4.64) 177 (4.46) 178 (4.83) alcohol abuse (4.30) 227 (5.72) 102 (2.77) < PVD (2.10) 89 (2.24) 72 (1.96) * Comparisons were performed using the chi-square test. TIA = transient ischemic attack; = without; + = with. Values represent means ± SD. J Neurosurg / Volume 120 / March

6 K. Bekelis et al. Fig. 2. Multivariate analysis of the inpatient postoperative risk for death (A), stroke (B), treated hydrocephalus (C), unfavorable discharge (D), and hospitalization longer than 4 days (E) in patients with unruptured aneurysms treated with clipping. Odds ratios are presented in Forest plots to the right of each table. Fig. 3. Multivariate analysis of the inpatient postoperative risk of cardiac complications (A), ARF (B), DVT (C), and PE (D) in patients with unruptured aneurysms treated with clipping. Odds ratios are presented in Forest plots to the right of each table. Fig. 4. Multivariate analysis of the inpatient postoperative risk of death (A), stroke (B), treated hydrocephalus (C), unfavorable discharge (D), and hospitalization longer than 14 days (E) in patients with ruptured aneurysms treated with clipping. Odds ratios are presented in Forest plots to the right of each table. 596 J Neurosurg / Volume 120 / March 2014

7 Nationwide Inpatient Sample: cerebral aneurysm clipping Fig. 5. Multivariate analysis of the inpatient postoperative risk of cardiac complications (A), ARF (B), DVT (C), and PE (D) in patients with ruptured aneurysms treated with clipping. Odds ratios are presented in Forest plots to the right of each table. able risk factors, alternative options should be sought more aggressively (coiling and observation) when appropriate. The present study has several limitations common to administrative databases. Indication bias and residual confounding could account for some of the observed associations. In addition, some coding inaccuracies will undoubtedly occur and can affect estimates. This is no different from what happens in other studies involving the NIS. However, several reports have demonstrated that coding for SAH has a nearly perfect association with medical record review. 9,18 The NIS for the years studied did not include hospitals from all states; however, the hospitals included were still diverse with respect to size, region, and academic status, supporting the generalizability of our findings. The NIS does not provide any clinical information on the structure, size, or location of the aneurysms, which are important factors in cerebrovascular neurosurgery. Other significant predictors of outcome that have been validated, such as baseline neurological status and Hunt and Hess grade, are also not included. Neither are a number of complications specific to aneurysms contained in the NIS, for example, rebleeding and vasospasm. Additionally, we lacked posthospitalization and long-term data in these patients, as well as disease severity in those presenting with SAH. There is potential bias in the patient-level risk factors selected for investigation. Further, there is also bias in our assignment of the length of prolonged hospitalization (more than 14 and 4 days for ruptured and unruptured aneurysms, respectively). However, these time limits were chosen based on mean values presented in the literature 1 3 and only served the purpose of creating a predictive model. Finally, the retrospective nature of our study introduces a selection bias that can affect the outcomes of clipping. Conclusions The NIS is a prospective hospital discharge database that contains a representative sample of all inpatient admissions to nonfederal hospitals in the US. Using its data, we identified independent preoperative risk factors associated with death, unfavorable discharge, stroke, treated hydrocephalus, cardiac complications, DVT, PE, ARF, and prolonged hospitalization in patients who had undergone CAC. With the results of multivariate analysis, we devised and validated a risk factor based predictive model of all the outcomes. Although these predictions should be generalized with caution, the model may assist with risk stratification and tailoring of surgical decision making in individual patients. Fig. 6. Diagrams of the risk for an unfavorable discharge in patients with unruptured aneurysms (A) and ruptured aneurysms (B), as well as for death (C) and a prolonged hospital stay (D) in patients with ruptured aneurysms for the individual variables found to be statistically significant (highest odds ratio) on multivariate analysis, as well as their combination against patient age (x-axis). The left y-axis represents the absolute probability of developing an outcome for individual risk factors or their combination. The right y-axis represents the absolute probability of developing an outcome for all risk factors combined (including nonsignificant variables). J Neurosurg / Volume 120 / March

8 K. Bekelis et al. TABLE 3: Outcomes in patients with ruptured and unruptured aneurysms treated with CAC* Variable Patients w/ Ruptured Aneurysm Patients w/ Unruptured Aneurysm p Value death 455 (11.46%) 25 (0.68%) < unfavorable discharge 2097 (52.83%) 562 (15.26%) < stroke 220 (5.54%) 196 (5.32%) treated hydrocephalus 1556 (39.20%) 55 (1.49%) < cardiac complications 67 (1.69%) 47 (1.28%) DVT 109 (2.75%) 21 (0.57%) < PE 105 (2.65%) 72 (1.96%) ARF 33 (0.83%) 4 (0.11%) <0.001 * Outcomes for ruptured and unruptured aneurysms were compared using the chi-square test. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Bekelis, Roberts. Acquisition of data: Bekelis, Missios, MacKenzie, Fischer. Analysis and interpretation of data: Bekelis, MacKenzie, Desai, Labropoulos. Drafting the article: Bekelis. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Bekelis. Statistical analysis: Bekelis, Missios, MacKenzie, Labropoulos. Administrative/technical/material support: Bekelis, Desai, Roberts. Study supervision: Bekelis, Desai, Labropoulos, Roberts. References 1. Barker FG II, Amin-Hanjani S, Butler WE, Hoh BL, Rabinov JD, Pryor JC, et al: Age-dependent differences in short-term outcome after surgical or endovascular treatment of unruptured intracranial aneurysms in the United States, Neurosurgery 54:18 30, Brinjikji W, Lanzino G, Rabinstein AA, Kallmes DF, Cloft HJ: Age-related trends in the treatment and outcomes of ruptured cerebral aneurysms: a study of the Nationwide Inpatient Sample AJNR Am J Neuroradiol34: , Brinjikji W, Rabinstein AA, Lanzino G, Kallmes DF, Cloft HJ: Effect of age on outcomes of treatment of unruptured cerebral aneurysms: a study of the National Inpatient Sample Stroke 42: , Brinjikji W, Rabinstein AA, Lanzino G, Kallmes DF, Cloft HJ: Patient outcomes are better for unruptured cerebral aneurysms treated at centers that preferentially treat with endovascular coiling: a study of the national inpatient sample AJNR Am J Neuroradiol 32: , Brinjikji W, Rabinstein AA, Nasr DM, Lanzino G, Kallmes DF, Cloft HJ: Better outcomes with treatment by coiling relative to clipping of unruptured intracranial aneurysms in the United States, AJNR Am J Neuroradiol 32: , Fisher ES, McClellan MB, Safran DG: Building the path to accountable care. N Engl J Med 365: , Hoh BL, Kleinhenz DT, Chi YY, Mocco J, Barker FG II: Incidence of ventricular shunt placement for hydrocephalus with clipping versus coiling for ruptured and unruptured cerebral aneurysms in the Nationwide Inpatient Sample database: 2002 to World Neurosurg 76: , Hoh BL, Nathoo S, Chi YY, Mocco J, Barker FG II: Incidence of seizures or epilepsy after clipping or coiling of ruptured and unruptured cerebral aneurysms in the nationwide inpatient sample database: Neurosurgery 69: , Kokotailo RA, Hill MD: Coding of stroke and stroke risk factors using international classification of diseases, revisions 9 and 10. Stroke 36: , Kotowski M, Naggara O, Darsaut TE, Nolet S, Gevry G, Kouznetsov E, et al: Safety and occlusion rates of surgical treatment of unruptured intracranial aneurysms: a systematic review and meta-analysis of the literature from 1990 to J Neurol Neurosurg Psychiatry 84:42 48, McDougall CG, Spetzler RF, Zabramski JM, Partovi S, Hills NK, Nakaji P, et al: The Barrow Ruptured Aneurysm Trial. Clinical article. J Neurosurg 116: , Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J, et al: International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 360: , Molyneux AJ, Kerr RS, Birks J, Ramzi N, Yarnold J, Sneade M, et al: Risk of recurrent subarachnoid haemorrhage, death, or dependence and standardised mortality ratios after clipping or coiling of an intracranial aneurysm in the International Subarachnoid Aneurysm Trial (ISAT): long-term follow-up. Lancet Neurol 8: , Qureshi AI, Vazquez G, Tariq N, Suri MF, Lakshminarayan K, Lanzino G: Impact of International Subarachnoid Aneurysm Trial results on treatment of ruptured intracranial aneurysms in the United States. Clinical article. J Neurosurg 114: , Raaymakers TW, Rinkel GJ, Limburg M, Algra A: Mortality and morbidity of surgery for unruptured intracranial aneurysms: a meta-analysis. Stroke 29: , Shea AM, Reed SD, Curtis LH, Alexander MJ, Villani JJ, Schulman KA: Characteristics of nontraumatic subarachnoid hemorrhage in the United States in Neurosurgery 61: , Steiner C, Elixhauser A, Schnaier J: The healthcare cost and utilization project: an overview. Eff Clin Pract 5: , Tirschwell DL, Longstreth WT Jr: Validating administrative data in stroke research. Stroke 33: , Tseng MY: Summary of evidence on immediate statins therapy following aneurysmal subarachnoid hemorrhage. Neurocrit Care 15: , 2011 Manuscript submitted February 8, Accepted August 6, Please include this information when citing this paper: published online September 13, 2013; DOI: / JNS Address correspondence to: Kimon Bekelis, M.D., Dartmouth- Hitchcock Medical Center, One Medical Center Dr., Lebanon, NH kbekelis@gmail.com. 598 J Neurosurg / Volume 120 / March 2014