Dale Ding, M.D., Daniel M. Raper, M.B.B.S., Gautam U. Mehta, M.D., Robert M. Starke, M.D., M.Sc.
University of Virginia, Department of Neurological Surgery, Charlottesville, Virginia, United States
Journal Club Article: Siddiq F, Chaudhry SA, Tummala RP, Suri MFK, Qureshi AI. Factors and outcomes associated with early and delayed aneurysm treatment in subarachnoid hemorrhage patients in the United States. Neurosurgery. 2012; 71(3): 670-678.
Significance/Context and Importance of the Study
Despite advances in medical and surgical technology, aneurysmal subarachnoid hemorrhage (SAH) remains a challenging cerebrovascular disease affecting relatively young patients with a high morbidity and mortality. The pathophysiology of aneurysm formation and rupture, and the sequelae of untreated aneurysmal SAH remain incompletely understood. Traditionally, the risk of increased surgical complications due to cerebral edema and acute SAH had to be weighed against the morbidity and mortality associated with aneurysm re-rupture, the most significant cause of morbidity and mortality in untreated patients. Following the publication of the International Subarachnoid Aneurysm Trial (ISAT) and the wide-spread incorporation of endovascular coil embolization, treatment of ruptured aneurysms through a minimally invasive approach may circumvent obstacles encountered during surgical treatment in the early period following SAH.5 In fact, a policy of early treatment (here defined as <48 hours) of ruptured aneurysms has become fairly standard in the modern literature. There is, however, a possibility that early endovascular treatment may be associated with higher incidence of complications. Although this is not definitively substantiated by the literature, the question provides impetus for this survey report based on the Nationwide Inpatient Sample. This interesting study provides additional evidence of the overall benefit of early over delayed obliteration of ruptured aneurysms, based on a real-world sample.
Originality of the Work
Since the seminal work by Kassell et al three decades ago, the overall national trend has been toward early treatment of aneurysmal SAH.3 There is little debate to the advantage of early treatment. Early aneurysm treatment significantly decreases the rate of rebleeding and allows for aggressive treatment of vasospasm.4 This study looks at the effect of a current practice whereby early treatment is the preferred management and few patients are left untreated or are treated in a delayed fashion. There is a small population of patients who may present a number of days after the onset of an aneurysmal headache, but this is clearly a minority. The strategy of analyzing the NIS to answer questions of outcome after SAH in a large population has been utilized in a number of previous studies.1 While the concept of early treatment is not original, the analysis of the benefit of early treatment is interesting and reinforces the impact of current standards.
Appropriateness of the Study Design or Experimental Approach
The authors screened the Nationwide Inpatient Sample (NIS) for diagnostic codes defined by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) to identify patients treated for aneurysmal SAH. Patients with head trauma, other vascular malformations, and SAH without procedure codes were excluded. The patients arbitrarily classify treatment as early if administered within 48 hours of hospital admission and delayed if administered greater than 48 hours after admission. A number of factors, including medical and treatment complications, need for additional procedures, vasospasm, degree of disability at discharge, and in-hospital mortality were identified by ICD-9-CM codes and their occurrences were compared between the early and delayed treatment groups. The NIS database does not have a record of the clinical or radiographic grade of SAH. Therefore the authors used the 3M Health Information Systems All Patient Refined Diagnosis Related Group (3M APR-DRG) mortality risk algorithm to assess the degree of patient debilitation after SAH and treatment. The outcome measures were length of stay, hospitalization charges, and discharge destination. The major strength of the NIS is the large number of patients representing approximately 20% of all U.S. hospital admissions. It is however, limited by the drawbacks of a retrospective analysis of a large nationwide database that is lacking in significant aneurysm-related details. Although the 3M APR-DRG 4-point system has been shown to be a valid marker for a patient’s clinical condition of the patient, it cannot supplant standard SAH grading systems, such as the Hunt and Hess and World Federation of Neurological Societies scales. Some papers have suggested possible inaccuracies in the NIS data including inaccurate physician and administrative reporting of ICD-9-CM codes.2 ICD-9-CM codes may be incomplete, and patients without a procedure code may, in fact, be part of the population of the ‘late treatment’ cohort but not included in the analysis. Therefore, it is hazardous to draw solid conclusions on specific questions unless the data is explicitly collected as part of the NIS methodology. For the question of timing of aneurysm intervention, the applicability of the available data is limited. Additionally, inference cannot be drawn about aneurysm rupture and deaths that occur prior to patient admittance or after hospital discharge.
Adequacy of Experimental Techniques
The study presents calculations from three separate logistic regression analyses. The first identifies demographic and clinical predictors of receiving early versus delayed treatment. The second shows the effect of early versus delayed treatment on in-hospital mortality. The third analysis determines the effect of early versus delayed treatment on degree of disability at discharge. A limitation of the analysis is those receiving early versus delayed therapy significantly differed in a number of important factors. Although the authors found that endovascular treatment and weekday admission (P<0.001) were associated with early treatment, these factors were not included in the overall analysis of outcome. Thus there is room for significant confounding and interaction that may over- or underestimate the effect of timing on outcome. Additionally, the authors find that early treatment is associated with an increased risk of in-hospital mortality, but an increased likelihood of discharge favorable outcome. It is possible that early treatment results in significant complications and in-hospital mortality that artificially inflates the beneficial role of early treatment on discharge outcome. Lastly, discharge destination was used to define the overall clinical condition with discharge to self-care classified as ‘none to minimal’ disability and discharge to extended care facility classified as ‘moderate to severe’ disability. In a large patient cohort discharge destination may be a suitable surrogate indicator of level of disability. These outcome measures may not account for patients whose care is attended to by family members, and patients with inadequate funds to afford extended care facilities. This may confound the results in socioeconomically disadvantaged patients.
Soundness of Conclusions and Interpretation
Two of the authors principal conclusions, namely that early treatment is associated with decreased disability at discharge (P=0.002) but increased in-hospital mortality (P<0.001), are difficult to reconcile and raise a significant concern of whether a benefit of early treatment even exists. It is possible that the superiority of the clinical outcomes in the early treatment group is offset by the mortality incurred by early treatment. Female sex (P=0.002), endovascular treatment (P<0.001), and weekday admission (P<0.001) were associated with early treatment. The conclusions have potentially far-reaching implications, especially for hospitals without endovascular capabilities. The authors recommend transfer to a teaching hospital, perhaps with the assumption that teaching hospitals are more likely to offer endovascular treatment and have more readily available physicians than non-teaching hospitals. While we agree with this notion, the authors’ recommendation is not supported by their data which shows that the difference in early treatment between teaching and non-teaching hospitals is not significant (P=0.06) While beyond the scope of the article, we emphasize that although the ability to treat ruptured aneurysms is important, the best therapeutic paradigm includes a hospital that is able to provide a multimodality approach to aneurysmal SAH that includes specialty-trained neurosurgeons, neurointerventionalists, and neurointensivists.
Relevance of Discussion
The authors propose four potential reasons for improved outcomes with early treatment. The first reason is prevention of rebleeding, the most dreaded consequence of an unsecured ruptured aneurysm. The second reason is decreased morbidity and mortality from vasospasm due to the ability to more aggressively manage vasospasm in patients with treated aneurysms. The occurrence of vasospasm is not different between the two groups (P=0.06) which is not surprising. However, one would perhaps expect the rate of angioplasty, which was also not different (P=0.91), to be higher in the delayed treatment group since most physicians would be hesitant to aggressively treat vasospasm with only medical therapy in a patient with an unsecured ruptured aneurysm. The third reason is a higher likelihood of delayed treatment in patients with more medical comorbidities. Only hypertension was significantly less common in the early treatment group (P=0.005). The prevalence of diabetes, congestive heart failure, renal failure, and the 3M APR-DRG disease severity score, which accounts for medical comorbidities, were not significantly different between the two treatment groups. The final reason is early treatment suggests a higher quality of care. This claim is refuted by the data presented by the authors. The rate of all reported in-hospital events and all procedures, except transfusion (P=0.004) and ventriculostomy (P=0.001), are not different between the two groups. Furthermore, the rate of post-procedure myocardial infarction and neurological complications is not decreased with early treatment. Of note, the hospital cost was significantly lower in the early treatment group (P<0.001) which may have significant implications in the near future as the government seeks to decrease healthcare spending. This discussion is well-reasoned and clearly expressed. However, the explanations for the observed trend only throw the inadequacies of the primary data into sharper relief. The strength of large samples for relatively rare clinical conditions is to identify large-scale trends and treatment effects. Identification of causative factors of a difference in treatment of 48 hours may be beyond the granularity obtainable from this data.
Clarity of Writing, Strength and Organization of the Paper
The strength of the paper lies in the large number of patients, approximately 32,000. In the final paragraph of the discussion, the authors provide an excellent description of potential shortcomings in using the NIS database, ICD-9-CM codes, and the 3M APR-DRG disease severity scale to assess the differences in early and delayed treatment. The paper is well-organized, allowing the reader to follow the authors’ methods and results, and the bottom line is clear: early treatment is superior to delayed treatment of ruptured aneurysms.
Economy of Words
Overall, the article is concise and clearly expressed. Instead of listing the ICD-9-CM codes in text under the ‘Patient Selection’ and ‘Variables Analyzed’ subsections, it may be better to list them in a table or figure for easier readability. Due to the large quantity of data, the wording of the results section is dense. The clarity of the results may be improved by subsections, each emphasizing an important statistical finding of which there were many.
Relevance, Accuracy, and Completeness of Bibliography
The authors provide an appropriate bibliography. Notably they cite the early work done on the timing of ruptured aneurysm treatment before the advent of endovascular coil embolization. The citations of the accuracy and reliability of the NIS database and of ICD-9-CM codes were particularly helpful in putting the analyzed data in context.
Number and Quality of Figures, Tables, and Illustrations
Table 1 is provides a wealth of information but its clarity might be improved through division into 2-4 smaller tables for easier readability. Tables 2 and 3 are short and emphasize important findings.
Future/Next Steps the Paper Logically Leads To
The treatment of aneurysmal SAH has advanced significantly since the International Cooperative Study on the Timing of Aneurysm Surgery.3 With the constant evolution of endovascular technology, our ability to treat ruptured aneurysms continues to improve, and with the outpouring of new basic and translation scientific studies, our understanding of the pathophysiology of SAH continues to evolve. This article stresses the importance of early treatment of ruptured aneurysms. In order to definitively cement the merits of early treatment, a large prospective trial is necessary comparing early to late treatment of aneurysmal SAH. Given current treatment practices, this is unlikely to be feasible in the United States. A cohort study comparing patients presenting in a delayed fashion after SAH, matched with equivalent controls who present early after the onset of headache, may be a more practical way to address the question. Knowledge of the patient’s neurological condition, which is a cardinal measure of outcomes following aneurysmal SAH treatment, is crucial in future studies.
1. Andaluz N and Zuccarello M: Recent trends in the treatment of cerebral aneurysms: analysis of a nationwide inpatient database. J Neurosurg 108: 1163-9, 2008.
2. Hertzer NR: Reasons why data from the Nationwide Inpatient Sample can be misleading for carotid endarterectomy and carotid stenting. Semin Vasc Surg 25: 13-7, 2012.
3. Kassell NF, Torner JC, Haley EC, Jr., Jane JA, Adams HP, and Kongable GL: The International Cooperative Study on the Timing of Aneurysm Surgery. Part 1: Overall management results. J Neurosurg 73: 18-36, 1990.
4. Macdonald RL, Higashida RT, Keller E, Mayer SA, Molyneux A, Raabe A, Vajkoczy P, Wanke I, Bach D, Frey A, Marr A, Roux S, and Kassell N: Clazosentan, an endothelin receptor antagonist, in patients with aneurysmal subarachnoid haemorrhage undergoing surgical clipping: a randomised, double-blind, placebo-controlled phase 3 trial (CONSCIOUS-2). Lancet Neurol 10: 618-25, 2011.
5. Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J, and Holman R: International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 360: 1267-74, 2002.