Better Outcomes and Reduced Hospitalization Cost are Associated with Ultra-Early Treatment of Ruptured Intracranial Aneurysms: A US Nationwide Data Sample Study

Better Outcomes and Reduced Hospitalization Cost are Associated with Ultra-Early Treatment of... Abstract BACKGROUND The benefit of surgical treatment of ruptured aneurysms is well established. OBJECTIVE To determine whether ultra-early ruptured aneurysm treatment leads to not only improved outcomes but also reduced hospitalization cost. METHODS Using 2008-2011 Nationwide Inpatient Sample data, we analyzed demographic, clinical, and hospital factors for nontraumatic subarachnoid hemorrhage (SAH) patients who were “directly” admitted to the treating hospital where they underwent intervention (clipping/coiling). Patients treated on the day of admission (day 0) formed the ultra-early cohort; others formed the deferred treatment cohort. All Patient Refined Diagnosis-Related Groups were also included in regression analyses. RESULTS A total of 17 412 patients were directly admitted to a hospital following nontraumatic SAH where they underwent intervention (clipping/coiling). Mean patient age was 53.87 yr (median 53.00, standard deviation 14.247); 68.3% were women (n = 11 893). A total of 6338 (36.4%) patients underwent treatment on the day of admission (ultra-early). Patients who underwent treatment on day 0 had significantly more routine discharge dispositions than those treated >admission day 0 (P < .0001). In regression analysis, treatment on day 0 was protective against other than routine discharge disposition outcome (P < .0001; odds ratio 0.657; 95% confidence interval 0.614-0.838). Total cost incurred by hospitals was $4.36 billion. Mean cost of hospital charges in the ultra-early cohort was $239 126.05, which was significantly lower than that for the cohort treated >day 0 ($272 989.56, P < .001), Mann-Whitney U-test). Performance of an intervention on admission day 0 was protective against higher hospitalization cost (P < .0001; odds ratio 0.811; 95% confidence interval 0.732-0.899). CONCLUSION Ultra-early treatment of ruptured aneurysms is significantly associated with better discharge disposition and decreased hospitalization cost. Aneurysmal subarachnoid hemorrhage, Clipping, Coiling, Discharge disposition, Hospitalization cost, Nationwide Inpatient Sample, Ruptured intracranial aneurysm ABBREVIATIONS ABBREVIATIONS APR-DRGs All Patient Refined Diagnosis Related Groups CAD coronary artery disease CCS Clinical Classifications Software CI confidence interval h hours HCUP Healthcare Cost and Utilization Project ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification ICH intracranial hemorrhage LOS length of stay MI myocardial infarction NIH National Institutes of Health NIS Nationwide Inpatient Sample NIS-SSS Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score OR odds ratio OTR other than routine SAH subarachnoid hemorrhage SD standard deviation The benefit of surgical treatment of ruptured aneurysms is well established. Traditionally, surgery by either endovascular or open approaches is considered early when performed within 48 to 72 hours (h) of presentation. Several studies1-5 have shown the benefit of early surgery, whereas some have advocated delayed treatment of poor-grade patients.6 To date, very few studies have discussed the benefits of ultra-early treatment (≤24 h of presentation) vs delayed treatment (>24 h of presentation) of aneurysmal subarachnoid hemorrhage (SAH).7-12 No study has compared the outcomes and cost-effectiveness of ultra-early surgery for ruptured aneurysms across the US. We analyzed Nationwide Inpatient Sample (NIS) data for our study on this subject.13,14 The NIS is the largest database of inpatient care. It comprises 20% of admissions to US nonfederal hospitals. The NIS database is managed by the Healthcare Cost and Utilization Project (HCUP). A new variable, “transfer status,” was incorporated in the database in 2008. This variable indicates whether a patient was a direct hospital admission or transferred from another facility. Additionally, the database included a coded variable: “number of days from admission to principal procedure.” In this study, we utilized the additional variables provided in the NIS database to study the timing of aneurysm surgery in the cohort of patients who were directly admitted to hospitals at the time of symptom onset and whether they underwent treatment on admission day 0 (ultra-early treatment; ie, on the day of admission as well as within 24 h of admission). Additionally, we analyzed the impact of ultra-early aneurysm treatment on the cost of hospitalization. Because the extent of a patient's initial disability can influence outcome, we also analyzed severity of disease using All Patient Refined Diagnosis-Related Groups (APR-DRGs).15 We posit that ultra-early aneurysm surgery not only improves outcome but also decreases hospitalization cost. METHODS The NIS data are deidentified by HCUP, and thus, institutional review board approval and patient consent were not needed for the present study. The NIS is the largest inpatient database available to the public. We extracted NIS data from 2008 to 2011 using codes from International Classification of Diseases, Clinical Modification (ICD-9-CM) and Clinical Classifications Software (CCS; http://www.hcup-us.ahrq.gov/). ICD-9-CM diagnosis codes of 430 (SAH); codes 3951 (clipping) and 3952, 3972, and 3979 (coiling) were then used to extract cases of SAH treated by primary clipping or coiling. We used the NIS-coded variable “transfer status” (direct admission to hospital or transfer-in from another facility) to extract only those patients who were directly admitted to the hospital. The NIS defines the day on which the procedure is performed (PRDAYn: procedure day n). PRDAYn is calculated from the procedure and admission dates. For a procedure performed on the day of admission, PRDAYn = 0; for the day after admission, PRDAYn = 1. Time of day is not a factor (Email from hcup-us@truvenhealth.com dated June 10, 2016). The following cohorts were formed based on the information on timing of aneurysm treatment. Ultra-early (procedure on admission day 0): the NIS-coded variable “procedure on admission day 0” was utilized to extract those patients who underwent treatment (clip/coil) on the day of admission, day 0. This cohort included patients who were treated <24 h from admission. This cohort was compared with the group who had the procedure on subsequent days (after day 0). Early (procedure on admission day 0 or 1): we created this cohort because a small subset of patients underwent their procedure the next day, after midnight (within 24 h, but on the next admission day, ie, admission day 1). Deferred (procedure on admission day 2 or more): the remaining patients formed the deferred treatment cohort. Patients who were not direct admits to a hospital were excluded from the study. The following additional NIS database variables were analyzed: gender; race; insurance status; hospital charges; APR-DRG severity, which was subclassified as minor function loss (including cases with no complications or comorbidity), moderate function loss, major function loss, and extreme function loss; hospital region (the US is divided into Midwest, West, Northeast, and Southern, regions); and teaching or nonteaching hospitals. An additional, new variable was created, Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score (NIS-SSS).16 This variable took into account several factors that are associated with poor presentation grade, including the following: ICD-9-CM diagnosis codes associated with altered mental status: coma (780.01, 780.03) and stupor (780.02, 780.09); ICD-9-CM codes associated with hydrocephalus (331.3, 331.4) and shunt/ventriculostomy (procedure codes 02.2 02.31-02.3); neurological deficit: aphasia (438.1-438.89), cranial nerve deficits (378.5-378.56, 379.4-379.43), and paresis or plegia (438.2-438.53, 781.4); and the CCS code for mechanical ventilation, 216. Several other factors that were generated secondarily from the database were also analyzed. These included age (dichotomized to ≥65 yr and <65 yr). The following CCS codes were used to generate these factors from the NIS database: 76 (meningitis); 100 (acute myocardial infarction), 101 (coronary atherosclerosis and other heart disease), 107 (cardiac arrest and ventricular fibrillation), 108 (congestive heart failure; non-hypertensive), 125 (acute bronchitis), 126 (other upper respiratory infections), 127 (chronic obstructive pulmonary disease and bronchiectasis), 128 (asthma), 129 (aspiration pneumonia), 130 (pleurisy; pneumothorax, pulmonary collapse), 131 (respiratory failure; insufficiency; arrest [adult]), 132 (lung disease due to external agents), 133 (other lower respiratory disease), 197 (wound infection), and ICD-9-CM codes for secondary intracranial hemorrhage (ICH; 430, 431, and 432). Patients who underwent intervention on day 1 of admission formed the “ultra-early treatment” cohort. Two additional variables were created based on existing information in the NIS database and percentile analysis: comorbidity index and higher hospitalization cost. The NIS database includes the following comorbidities: acquired immune deficiency syndrome, alcohol abuse, chronic blood loss anemia, chronic pulmonary disease, congestive heart failure, coagulopathy, deficiency anemias, diabetes with chronic complications, drug abuse, fluid and electrolyte disorders, hypertension, hypothyroidism, liver disease, lymphoma, metastatic cancer, obesity, other neurological disorders, paralysis, peptic ulcer disease excluding bleeding, peripheral vascular disorders, pulmonary circulation disorders, renal failure, rheumatoid arthritis/collagen vascular diseases, solid tumor without metastasis, uncomplicated diabetes, valvular disease, and weight loss. The number of comorbidities a patient had was calculated and summed. It ranged from 0 to 13. Patients with ≥2 comorbidities formed the “above the 75th percentile” high-comorbidity group. Patients with <2 comorbidities formed the low-comorbidity group. Similarly, the higher hospitalization cost variable was created based on existing “total charges” information in the NIS database. “Total charges” gives the cost incurred by the hospital while treating the patient. All the charges under the variable “total charges” were analyzed by descriptive statistics and segregation according to percentiles. Hospitalization cost of more than $511 451.00 formed the greater than 90th percentile cohort (higher hospitalization cost). For outcome analysis, a dichotomized discharge disposition was utilized: “routine” and “other than routine (OTR).”17 Any transfer from a hospital to acute rehabilitation, home health care, intermediate care, a short-term hospital facility, or a skilled nursing facility; discharge against medical advice; or death was considered an OTR discharge disposition. A direct disposition to home without home health care was considered a routine discharge disposition. Statistical Analysis Univariate analysis was performed using chi square or Fisher exact test. A multivariate binary logistic regression model was used to analyze variables with a P-value <.1, with hospitalization cost and discharge disposition as dependent variables. Means were compared with the Mann-Whitney U-test. All statistical analyses were performed using IBM SPSS Statistics 20 (IBM Corporation, Armonk, New York). RESULTS Data for 276 902 patients with nontraumatic SAH were analyzed. A total of 33 602 patients underwent intervention (clipping or coiling); and of those, 17 412 were directly admitted to a hospital following SAH. Among the patients who were directly admitted, the mean age was 53.87 yr (median 53.00 yr, standard deviation 14.24); 68.3% were women (n = 11 893); and 78.2% (n = 13 615) were <65 yr. The interventions of coiling (n = 8702, 49.9%) and clipping (n = 8710, 50.1%) were nearly equal in proportion. A total of 6338 (36.4%) patients underwent treatment on the day of admission (ultra-early), and a total of 12 159 (69.8%) of the 17 412 patients underwent treatment on admission day 0 or 1 (early). Coiling was the preferred modality on admission day 0, with 59.14% (n = 3748) of patients undergoing treatment with primary coiling (Table 1). Surgical clipping was the preferred modality when treatment was carried out beyond day 0 of admission (in 55.26% patients; n = 6120). When overall outcome was considered, 41.6% (n = 7245) of the patients had routine discharge disposition. Among these, patients who had treatment within admission day 0 had significantly more routine discharge dispositions (P < .001; Table 1). In the regression analysis, treatment within admission day 0 was protective against worse outcome (P = .001; odds ratio [OR] 0.657; confidence interval [CI] 0.614-0.838; Table 2). Similarly, early treatment (day 0/1) was protective against OTR outcome; however, the value was bordering on significance (P = .043; OR 0.907; CI 0.826-0.997). TABLE 1. Comparison of Treatment Timings (Ultra-Early, Early, and Deferred)     Procedure on admission day ≥1 (total n = 11 074)  Procedure on admission day 0 (ultra-early; total n = 6338)  P value  Procedure on admission day ≥2 (total n = 5253)  Procedure on admission day 0/1 (early; total n = 12 159)  P value      n  %  n  %    n  %  n  %    Died during hospitalization  1329  12.00  925  14.59  <.001  592  11.27  1662  13.67  <.001  Final outcome routine discharge disposition  4417  39.89  2828  44.62  <.001  2065  39.31  5180  42.60  <.001  >90 percentile cost (Higher cost)  1393  12.58  672  10.60  <.001  791  15.06  1274  10.48  <.001  Age ≥65 y  2493  22.51  1304  20.57  .003  1266  24.10  2531  20.82  <.001  Female sex  7615  68.76  4278  67.50  .086  3597  68.48  8296  68.23  .743  All patient refined DRG: severity of illness subclass  Minor loss of function (includes cases with no comorbidity or complications)  378  3.41  458  7.23  <0.001  143  2.72  693  5.70  <0.001    Moderate loss of function  1751  15.81  1027  16.20    875  16.66  1903  15.65      Major loss of function  4929  44.51  2649  41.80    2335  44.45  5243  43.12      Extreme loss of function  4016  36.27  2204  34.77    1900  36.17  4320  35.53    Comorbidity status (high comorbidity)  4471  40.37  2415  38.10  0.003  2097  39.92  4789  39.39  0.522  Coiling/clipping  Coiling  4954  44.74  3748  59.14  <0.001  2152  40.97  6550  53.87  <0.001    Clipping  6120  55.26  2590  40.86    3101  59.03  5609  46.13    Meningitis  Yes  318  2.87  242  3.82  0.001  148  2.82  412  3.39  0.04  CAD  Yes  898  8.11  422  6.66  0.001  475  9.04  845  6.95  <0.001  Intracerebral hemorrhage  Yes  598  5.40  305  4.81  0.092  297  5.65  606  4.98  0.067  Respiratory failure  Yes  3886  35.09  2328  36.73  0.03  1777  33.83  4437  36.49  0.001  Wound infection  Yes  84  0.76  40  0.63  0.336  34  0.65  90  0.74  0.419  Teaching status of hospital (#)  Teaching  9511  88.12  5022  80.84  <0.001  4517  88.24  10016  84.26  <0.001      Procedure on admission day ≥1 (total n = 11 074)  Procedure on admission day 0 (ultra-early; total n = 6338)  P value  Procedure on admission day ≥2 (total n = 5253)  Procedure on admission day 0/1 (early; total n = 12 159)  P value      n  %  n  %    n  %  n  %    Died during hospitalization  1329  12.00  925  14.59  <.001  592  11.27  1662  13.67  <.001  Final outcome routine discharge disposition  4417  39.89  2828  44.62  <.001  2065  39.31  5180  42.60  <.001  >90 percentile cost (Higher cost)  1393  12.58  672  10.60  <.001  791  15.06  1274  10.48  <.001  Age ≥65 y  2493  22.51  1304  20.57  .003  1266  24.10  2531  20.82  <.001  Female sex  7615  68.76  4278  67.50  .086  3597  68.48  8296  68.23  .743  All patient refined DRG: severity of illness subclass  Minor loss of function (includes cases with no comorbidity or complications)  378  3.41  458  7.23  <0.001  143  2.72  693  5.70  <0.001    Moderate loss of function  1751  15.81  1027  16.20    875  16.66  1903  15.65      Major loss of function  4929  44.51  2649  41.80    2335  44.45  5243  43.12      Extreme loss of function  4016  36.27  2204  34.77    1900  36.17  4320  35.53    Comorbidity status (high comorbidity)  4471  40.37  2415  38.10  0.003  2097  39.92  4789  39.39  0.522  Coiling/clipping  Coiling  4954  44.74  3748  59.14  <0.001  2152  40.97  6550  53.87  <0.001    Clipping  6120  55.26  2590  40.86    3101  59.03  5609  46.13    Meningitis  Yes  318  2.87  242  3.82  0.001  148  2.82  412  3.39  0.04  CAD  Yes  898  8.11  422  6.66  0.001  475  9.04  845  6.95  <0.001  Intracerebral hemorrhage  Yes  598  5.40  305  4.81  0.092  297  5.65  606  4.98  0.067  Respiratory failure  Yes  3886  35.09  2328  36.73  0.03  1777  33.83  4437  36.49  0.001  Wound infection  Yes  84  0.76  40  0.63  0.336  34  0.65  90  0.74  0.419  Teaching status of hospital (#)  Teaching  9511  88.12  5022  80.84  <0.001  4517  88.24  10016  84.26  <0.001  Abbreviations: CAD, coronary artery disease; DRG, diagnosis-related groups #, nonsignificant missing values; y, years View Large TABLE 2. Results of Multivariate Regression Analysis for Discharge Disposition       95% CI for OR  Factors  Significance  OR  Lower  Upper  Elderly population (>65 y)  <.0001  4.596  4.130  5.116  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  26.552  20.257  34.802  Major loss of function  <.0001  8.084  7.018  9.313  Moderate loss of function  <.0001  3.244  2.914  3.611  Comorbidity index  <.0001  1.185  1.093  1.285  Treatment type (coiling vs clipping)  .055  1.081  .998  1.170  Procedure on admission day 0/1 (early) vs >2 (deferred)  .043  0.907  0.826  0.997  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.657  0.614  0.838  Meningitis  .549  1.070  0.858  1.335  CAD and other heart disease  .391  1.069  0.918  1.243  ICH  <.0001  1.579  1.316  1.896  Respiratory failure  <.0001  2.760  2.489  3.060  Wound infection  <.0001  2.498  1.569  3.977  Hospital teaching status (teaching vs nonteaching)  <.0001  0.719  0.640  0.808  Region of hospital (ref: Northeast)  .000        West  .832  0.589  0.521  1.664  South  .553  0.965  0.857  1.086  Midwest  .731  1.422  0.292  1.566        95% CI for OR  Factors  Significance  OR  Lower  Upper  Elderly population (>65 y)  <.0001  4.596  4.130  5.116  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  26.552  20.257  34.802  Major loss of function  <.0001  8.084  7.018  9.313  Moderate loss of function  <.0001  3.244  2.914  3.611  Comorbidity index  <.0001  1.185  1.093  1.285  Treatment type (coiling vs clipping)  .055  1.081  .998  1.170  Procedure on admission day 0/1 (early) vs >2 (deferred)  .043  0.907  0.826  0.997  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.657  0.614  0.838  Meningitis  .549  1.070  0.858  1.335  CAD and other heart disease  .391  1.069  0.918  1.243  ICH  <.0001  1.579  1.316  1.896  Respiratory failure  <.0001  2.760  2.489  3.060  Wound infection  <.0001  2.498  1.569  3.977  Hospital teaching status (teaching vs nonteaching)  <.0001  0.719  0.640  0.808  Region of hospital (ref: Northeast)  .000        West  .832  0.589  0.521  1.664  South  .553  0.965  0.857  1.086  Midwest  .731  1.422  0.292  1.566  Abbreviations: APR-DRG, All Patient Refined Diagnosis Related Group; CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; ref=reference. aThe dependent variable was discharge disposition (other than routine vs routine); demographic and patient factors were the covariates. View Large The rates of mortality and OTR discharge disposition (worse outcome) were higher in patients >65 yr and in women, and associated with significant differences (Table, Supplemental Digital Content 1). This significance persisted even when these variables were put in the multivariate regression model (Table 2). Women had 1.248 times higher odds for a worse outcome than men (P < .0001, CI 1.144-1.362). Similarly, age ≥ 65 yr had 4.130 times higher odds for a worse outcome (P < .0001, CI 4.130-5.116). The predominant race was white (56.5%; n = 8535). Race had a high percentage (29.8%) of missing values, hence is not included in the univariate analysis. On the basis of the APR-DRG mortality subclass, 38.3% (n = 6677) had “minor likelihood of dying” with 1.1% in-hospital mortality and 29.7% (n = 5181) had “extreme likelihood of dying” with 31.9% mortality, the difference being significant (P < .001) (Table, Supplemental Digital Content 1). Similarly, on the basis of the APR-DRG severity subclass, 4.8% (n = 836) had “minor function loss” with 9.7% OTR outcomes, and 35.7% (n = 6220) had “extreme function loss” with 86.6% OTR outcomes (P < .001; Table, Supplemental Digital Content 1). Patients who were treated on the day of admission had lesser severity of APR-DRG (Table 1). In regression analysis, APR-DRG severity subclass “extreme function loss” had 26.55 times higher odds for a worse outcome than “minor function loss” (P < .0001; CI 20.26-34.80; Table 2). Similarly, we analyzed the relationship between poor NIS-SSS score and outcome. A total of 8034 (46.1%) patients had a poor NIS-SSS score and, of these, only 19.5% (1563 of 8034) had a routine discharge disposition. However, 60.5% (5682 of 9378) of good-grade patients had a routine discharge disposition. Regression analysis with discharge disposition as the outcome variable was performed by incorporating the NIS-SSS score (Table 3). A poor NIS-SSS score had an OR of 3.422 (P < .0001, CI 3.142-3.728) for OTR outcome, and patients who were treated on the day of admission (ultra-early) had protection against OTR discharge disposition (P = .001, OR 0.809, CI 0.741-0.884). TABLE 3. Results of Multivariate Regression Analysis Using the Variable NIS-SSS for Discharge Disposition       95% CI for OR  Factors  Significance  OR  Lower  Upper  +Elderly population (>65 y)  <.0001  4.953  4.465  5.493  NIS-SSS  <.0001  3.422  3.142  3.728  Comorbidity index  <.0001  1.572  1.455  1.699  Treatment type (coiling vs clipping)  .001  1.451  1.346  1.565  Procedure on admission day 0/1 (early) vs >2 (deferred)  .029  0.901  0.821  0.989  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.809  0.741  0.884  Meningitis  <.0001  1.680  1.350  2.091  CAD and other heart disease  .001  1.871  1.412  2.479  ICH  <.0001  1.687  1.411  2.016  Respiratory failure  <.0001  3.522  3.195  3.883  Wound infection  <.0001  3.111  1.982  4.884  Hospital teaching status (teaching vs non-teaching)  <.0001  0.710  0.636  0.792  Region of hospital (ref: Northeast)          West  .488  0.961  0.858  1.076  South  .327  0.945  0.845  1.058  Midwest  .743  1.335  0.217  1.466        95% CI for OR  Factors  Significance  OR  Lower  Upper  +Elderly population (>65 y)  <.0001  4.953  4.465  5.493  NIS-SSS  <.0001  3.422  3.142  3.728  Comorbidity index  <.0001  1.572  1.455  1.699  Treatment type (coiling vs clipping)  .001  1.451  1.346  1.565  Procedure on admission day 0/1 (early) vs >2 (deferred)  .029  0.901  0.821  0.989  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.809  0.741  0.884  Meningitis  <.0001  1.680  1.350  2.091  CAD and other heart disease  .001  1.871  1.412  2.479  ICH  <.0001  1.687  1.411  2.016  Respiratory failure  <.0001  3.522  3.195  3.883  Wound infection  <.0001  3.111  1.982  4.884  Hospital teaching status (teaching vs non-teaching)  <.0001  0.710  0.636  0.792  Region of hospital (ref: Northeast)          West  .488  0.961  0.858  1.076  South  .327  0.945  0.845  1.058  Midwest  .743  1.335  0.217  1.466  Abbreviations: CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; NIS-SSS, Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score; ref, reference, yrs, years aThe dependent variable was discharge disposition (other than routine vs routine); demographic and patient factors were the covariates. View Large Patients with high comorbidity status or coiling as the procedure, coronary artery disease, ICH, and respiratory failure had significantly higher mortality and OTR discharge disposition (Table, Supplemental Digital Content 1). However, in the regression model, only ICH, respiratory failure, and presence of wound infection had significantly higher odds for worse outcome (Table 2 and 3). Among the hospital factors, 80.84% of day 0 treatments were carried out in teaching hospitals (Table 1), and regression analysis showed that teaching hospitals (P < .0001; OR 0.719; CI 0.640-0.808; Table 2) were protective against a worse outcome. Hospitalization Cost The total cost incurred by hospitals was $4.36 billion. The mean cost of hospital charges in the ultra-early cohort was $239 126.05, which was significantly lower than that for the cohort treated >day 0 ($272 989.56; P < .001, Mann-Whitney U-Test). The “higher hospitalization” cost cohort was comprised of 11.9% of all patients (n = 2065). Age >65 yr, higher comorbidity status, APR-DRG subset “extreme loss of function,” procedure performed >24 h, meningitis, coronary atherosclerosis and other heart disease, ICH, respiratory failure, and wound infection were associated with higher cost of hospitalization in univariate analysis (Table, Supplemental Digital Content 2). These variables were tested in the regression analysis (Table 4). Procedure performed on admission day 0 was protective against higher cost of hospitalization (P < .0001; OR 0.811; CI 0.732-0.899). TABLE 4. Results of Multivariate Regression Analysis       95% CI for OR    Significance  Odds ratio  Lower  Upper  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  20.696  12.429  34.463  Major loss of function  <.0001  9.494  7.593  11.872  Moderate loss of function  <.0001  4.254  3.739  4.840  Elderly population (>65 y)  .014  0.867  0.773  0.972  Comorbidity severity  .001  1.178  1.067  1.301  Treatment type (coiling vs clipping)  <.0001  0.670  0.607  0.740  Meningitis  .007  1.343  1.086  1.661  CAD and other heart disease  .214  0.895  0.752  1.066  ICH  <.0001  1.641  1.379  1.952  Respiratory failure  <.0001  1.668  1.593  1.752  Wound infection  <.0001  2.557  1.650  3.961  Procedure on admission day 0 (ultra-early) vs >1 admission day  <.0001  0.811  0.732  0.899        95% CI for OR    Significance  Odds ratio  Lower  Upper  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  20.696  12.429  34.463  Major loss of function  <.0001  9.494  7.593  11.872  Moderate loss of function  <.0001  4.254  3.739  4.840  Elderly population (>65 y)  .014  0.867  0.773  0.972  Comorbidity severity  .001  1.178  1.067  1.301  Treatment type (coiling vs clipping)  <.0001  0.670  0.607  0.740  Meningitis  .007  1.343  1.086  1.661  CAD and other heart disease  .214  0.895  0.752  1.066  ICH  <.0001  1.641  1.379  1.952  Respiratory failure  <.0001  1.668  1.593  1.752  Wound infection  <.0001  2.557  1.650  3.961  Procedure on admission day 0 (ultra-early) vs >1 admission day  <.0001  0.811  0.732  0.899  APR-DRG, All Patient Refined Diagnosis-Related Group; CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; ref, reference; y, years. View Large DISCUSSION Earlier studies have segregated the timing of ruptured aneurysm treatment in relation to the time interval between symptom onset and treatment. The term early surgery refers to <3 d18 or <2 d19 of the ictus, and the term “ultra-early surgery” is applied to intervention/surgery within 24 h of presentation.6,9,12 Our study is based on NIS data provided by HCUP. The database divides patients into mainly 2 categories, patients who are admitted directly to a hospital and those who are transferred to a hospital from another medical facility. As this study focused on the timing of surgery/intervention, we selected only those patients who were directly admitted to a facility where they underwent treatment. We excluded those patients who were transferred from another hospital because in the latter scenario, true timing of surgery/intervention is difficult to assess owing to the time elapsed between admission to the first and subsequent hospitals. Patients who were admitted to the hospital directly were divided into cohorts based on the timing of surgery. The coded variable of NIS Procedure on day 0 of admission was used to form the ultra-early cohort, and patients who underwent procedures on day 0 and day 1 formed the early cohort. A similar analysis of the impact of timing of intervention/surgery on the outcome of aneurysmal SAH across the United States was performed by Siddiq et al.19 However, in that study, the admission status (direct admission or transferred from another facility) was not considered and the impact of early treatment (<48 h from admission), rather than ultra-early treatment, was analyzed. Additionally, the study analyzed both directly admitted and transferred patients. The time elapsed in transfer was not taken into account. In the present study, univariate analysis showed that clipping was significantly associated with less mortality and OTR discharge disposition than coiling. However, the regression analysis did not show significance (P = .055; OR 1.081; CI 0.998-1.170). The aim of our study was not to compare endovascular and surgical treatment techniques because the 2 techniques have been compared earlier in an evaluation of NIS data.20 Instead, we were interested in the timing of treatment and its relationship with outcome and cost. In the present study, 6338 (36.4%) patients underwent treatment on the day of admission (day 0, ultra-early), and coiling was the preferred modality for that cohort (59.14%; n = 3748). The percentage of coiling procedures decreased in the early (day 0/1) cohort (53.9%, n = 6550). Surgical clipping was the preferred modality when treatment was carried out beyond day 0 of admission (in 55.26% patients; n = 6120; Table 1). Several factors determine the modality of treatment, including presence of ICH, mass effect, morphology of the aneurysm, and predilections of the treatment team. One possible reason for the preference for endovascular treatment within day 0 of admission was that the endovascular intervention could become a part of the diagnostic angiogram, thus saving time and hospital resources. Another factor could be the predilection of the treating team and the relative ease of delivering endovascular treatment compared with surgical clipping. At some centers, including ours, endovascular treatment is successfully performed under conscious sedation, and it can be extended to SAH patients who are in good grade.21 Further analysis on the timing of intervention (clipping/coiling) demonstrated that despite a higher rate of mortality on admission days 0 and 1, procedures performed on the day of admission (day 0) were associated with significantly less occurrence of OTR discharge disposition than those performed >day 0 of admission (P = .001; OR 0.657). This difference persisted in the regression analysis despite adjusting for several factors, including clipping/coiling (Table 2). Significance was maintained (P = .043) on procedure day 0/1, but the OR was only 0.907 compared to that for the deferred intervention cohort. The higher mortality on days 0 and 1 may be secondary to poor general condition of the patient upon admission. We adjusted for this factor in regression analysis, and mortality was included in OTR discharge disposition. A possible explanation could be that ultra-early surgery confers protection against rebleed and future neurological complications. The occurrence of rebleeding is known to peak within the first 24 h of rupture.22 In a systemic review, Starke and Connolly23 found the frequency of rebleeding to be as high as 9% to 17% in the first 24 h of presentation. Also, ultra-early aneurysm treatment allows for an early start of aggressive medical therapy to counter cerebral vasospasm and to tackle delayed neurological deficits.24,25 In our study, patients who received treatment within day 0 of admission had lower odds (0.657; Table 2) of worse outcome. In a similar study on the benefits of ultra-early treatment, Wong et al12 reported a significant association between better SF-36 mental scores and ultra-early aneurysm treatment (P = .041) and favorable outcome (P = .086). This benefit persisted, even in poor-grade patients (P = .062). Studies have shown that outcomes are better when surgeries/interventions are performed at teaching hospitals.19,26 In our study, 85.5% (n = 14 533) patients had surgery/intervention at teaching hospitals. Mortality and morbidity were significantly lower at teaching hospitals than nonteaching hospitals. Regression analysis showed that the odds of worse outcomes are lower at teaching hospitals (P < .001; OR 0.719; CI 0.640-0.808; Table 2). In our study, large hospitals had a protective effect against worse outcome. This could be due to readily available resources at larger hospitals. A similar association was seen in earlier studies.27,28 When we analyzed clinical factors, higher comorbidity status was associated with increased mortality and OTR discharge disposition. In multivariate regression analysis (Table 2), several other patient factors—ICH, respiratory infection, wound infection—and APR-DRG severity of illness were associated with OTR discharge disposition. The odds for a worse outcome were 26.552 times greater when the patient had extreme function loss, compared with minimal function loss, during the hospital stay (Table 2). APR-DRGs are based on clinical events at admission or occurring during the hospital stay.15 We created another variable, NIS-SSS, as described by Washington et al.16 In their study of the validation of SAH severity based on NIS data, they found a significant correlation between NIS-SSS score and Hunt and Hess scores. When we performed regression analysis by incorporating the NIS-SSS variable, ultra-early timing of surgery still had the best preventive rate against OTR discharge disposition (Table 3). In the current scenario of ever-increasing healthcare cost, the economics of cost utilization become paramount to the discussion of timing of aneurysm surgery. Several studies have compared the cost utilization of clipping vs coiling at a single institution29 or nationwide;30,31 but to our knowledge, none has looked at the impact of ultra-early intervention on cost of hospitalization. Costs are affected with increasing comorbidities and complications, length of stay (LOS), clinical outcome, and hospitalization.17 We analyzed hospitalization cost and various associated factors. The mean hospitalization cost was significantly less for patients who received treatment within admission day 1 for those who received treatment after day 1 of admission (P < .0001). This association persisted when hospitalization cost was adjusted for comorbidities, complications, and APR-DRG severity in regression analysis. The odds were 0.8 for higher hospitalization cost when treatment was performed within admission day 0 (Table 4); thus, ultra-early intervention was protective against higher hospitalization cost. This finding is important as it suggests that ultra-early treatment not only provides better outcomes but also is cost-effective. LOS had a significantly high correlation with hospital cost incurred (P < .001: bivariate Spearman nonparametric correlation analysis). Therefore, LOS was not included in the regression analysis to assess factors associated with higher hospitalization cost. It is often assumed that patients are transferred to another facility if their case is complicated, the grade is poor, or when treatment options are not available at the initial facility. To avoid this bias, we analyzed only those patients who were directly admitted. However, there are limitations to our study. Limitations Our study is a retrospective analysis, and there may be inherent errors in coding or under-reporting of events in the database. It is not possible to distinguish preoperative neurological deficit from postoperative deficit, including rebleed. The database does not provide information on radiological outcomes associated with aneurysm treatment, so it is not possible to extrapolate the findings of this study in that respect. APR-DRG severity and NIS-SSS scores cannot be used as surrogate markers for World Federation of Neurosurgical Societies or Hunt and Hess scores at admission. However, the APR-DRGs and NIS-SSS scores do provide valuable data regarding the disability status of the patient; and by including these scores in our regression analysis, the testing of other covariates was more robust. Although not addressed in the present analysis, hospital volumes can also affect outcomes. Hospitals with a high volume of SAH patients tend to have better outcomes.32 Further, the NIS defines “day 0” as time prior to the first midnight of admission, thus patients admitted late in the day might be treated the next day, technically within 24 h, but not on “day 0.” Finally, for the analysis of cost, we looked at single hospitalization, not taking into account readmissions or cost incurred for treatment at rehabilitation facilities. Such data on long-term rehabilitation services and readmissions are not provided in the NIS database. CONCLUSION Several factors can affect clinical outcome and hospitalization cost following treatment for ruptured aneurysms. Our study has shown that treatment of ruptured aneurysm within day 0 of admission following SAH is associated with better discharge disposition and decreased hospitalization cost. Disclosures The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Financial relationships/potential conflicts of interest: Sonig, Shallwani, Natarajan, Shakir: none. Hopkins: grant/research support-Toshiba; consultant-Abbott, Boston Scientific, Cordis, Covidien, Medtronic; financial interests-Boston Scientific, Valor Medical, Claret Medical Inc., Augmenix, Endomation, Silk Road, Ostial, Apama, StimSox, Photolitec, ValenTx, Ellipse, Axtria, NextPlain, Ocular; board/trustee/officer position-Claret Medical, Inc.; honoraria-Complete Conference Management, Covidien, Memorial Healthcare System. Levy: shareholder/ownership interests–Intratech Medical Ltd., Blockade Medical LLC, NeXtGen Biologics. Principal investigator: Covidien US SWIFT PRIME Trials. Honoraria–Covidien. Consultant–Pulsar, Blockade Medical. Advisory Board-Stryker, NeXtGen Biologics, MEDX. Other financial support–Abbott for carotid training sessions. Siddiqui: Research grants: The National Institutes of Health (co-investigator: NINDS 1R01NS064592-01A1, Hemodynamic induction of pathologic remodeling leading to intracranial aneurysms), The National Institutes of Health (co-investigator: NIBIB 5 R01 EB002873-07, Micro-Radiographic Image for Neurovascular Interventions), The National Institutes of Health (co-investigator: NIH/NINDS 1R01NS091075 Virtual Intervention of Intracranial Aneurysms). None of these grants are relevant to this paper. Financial interests:Hotspur, Intratech Medical, StimSox, Valor Medical, Blockade Medical, Lazarus Effect, Pulsar Vascular, Medina Medical; Consultant: Codman & Shurtleff, Covidien Vascular Therapies, GuidePoint Global Consulting, Penumbra, Stryker, Pulsar Vascular, MicroVention, Lazarus Effect, Blockade Medical, Reverse Medical, W.L. Gore & Associates; National Steering Committees:Penumbra-3D Separator Trial, Covidien-SWIFT PRIME Trial, MicroVention-FRED Trial; Speakers’ bureau: Codman & Shurtleff, Inc.; Advisory Board:Codman & Shurtleff, Covidien Neurovascular, ICAVL, Medina Medical; Honoraria:Penumbra, Toshiba Medical Systems. Snyder: Speaker's Bureau: Toshiba and Jacobs Institute REFERENCES 1. de Gans K, Nieuwkamp DJ, Rinkel GJ, Algra A. Timing of aneurysm surgery in subarachnoid hemorrhage: a systematic review of the literature. Neurosurgery . 2002; 50( 2): 336- 340; discussion 340-332. Google Scholar PubMed  2. Deruty R, Mottolese C, Pelissou-Guyotat I, Soustiel JF. Management of the ruptured intracranial aneurysm–early surgery, late surgery, or modulated surgery? Personal experience based upon 468 patients admitted in two periods (1972-1984 and 1985-1989). Acta Neurochir (Wien) . 1991; 113( 1-2): 1- 10. Google Scholar CrossRef Search ADS PubMed  3. Nieuwkamp DJ, de Gans K, Algra A et al.   Timing of aneurysm surgery in subarachnoid haemorrhage–an observational study in The Netherlands. Acta Neurochir (Wien) . 2005; 147( 8): 815- 821. Google Scholar CrossRef Search ADS PubMed  4. Ross N, Hutchinson PJ, Seeley H, Kirkpatrick PJ. Timing of surgery for supratentorial aneurysmal subarachnoid haemorrhage: report of a prospective study. J Neurol Neurosurg Psychiatry . 2002; 72( 4): 480- 484. Google Scholar PubMed  5. Whitfield PC, Moss H, O’Hare D, Smielewski P, Pickard JD, Kirkpatrick PJ. An audit of aneurysmal subarachnoid haemorrhage: earlier resuscitation and surgery reduces inpatient stay and deaths from rebleeding. J Neurol Neurosurg Psychiatry . 1996; 60( 3): 301- 306. Google Scholar CrossRef Search ADS PubMed  6. Egashira Y, Yoshimura S, Enomoto Y, Ishiguro M, Asano T, Iwama T. Ultra-early endovascular embolization of ruptured cerebral aneurysm and the increased risk of hematoma growth unrelated to aneurysmal rebleeding. J Neurosurg . 2013; 118( 5): 1003- 1008. Google Scholar CrossRef Search ADS PubMed  7. Baltsavias GS, Byrne JV, Halsey J, Coley SC, Sohn MJ, Molyneux AJ. Effects of timing of coil embolization after aneurysmal subarachnoid hemorrhage on procedural morbidity and outcomes. Neurosurgery . 2000; 47( 6): 1320- 1329; discussion 1329-1331. Google Scholar CrossRef Search ADS PubMed  8. Gu DQ, Zhang X, Luo B, Long XA, Duan CZ. Impact of ultra-early coiling on clinical outcome after aneurysmal subarachnoid hemorrhage in elderly patients. Acad Radiol . 2012; 19( 1): 3- 7. Google Scholar CrossRef Search ADS PubMed  9. Laidlaw JD, Siu KH. Ultra-early surgery for aneurysmal subarachnoid hemorrhage: outcomes for a consecutive series of 391 patients not selected by grade or age. J Neurosurg . 2002; 97( 2): 250- 258; discussion 247-259. Google Scholar CrossRef Search ADS PubMed  10. Phillips TJ, Dowling RJ, Yan B, Laidlaw JD, Mitchell PJ. Does treatment of ruptured intracranial aneurysms within 24 hours improve clinical outcome? Stroke . 2011; 42( 7): 1936- 1945. Google Scholar CrossRef Search ADS PubMed  11. Weil AG, Zhao JZ. Treatment of ruptured aneurysms: earlier is better. World Neurosurg . 2012; 77( 2): 263- 265. Google Scholar CrossRef Search ADS PubMed  12. Wong GK, Boet R, Ng SC et al.   Ultra-early (within 24 hours) aneurysm treatment after subarachnoid hemorrhage. World Neurosurg . 2012; 77( 2): 311- 315. Google Scholar CrossRef Search ADS PubMed  13. 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 2001–2009. AJNR Am J Neuroradiol . 2013; 34( 5): 1022- 1027. Google Scholar CrossRef Search ADS PubMed  14. Leake CB, Brinjikji W, Kallmes DF, Cloft HJ. Increasing treatment of ruptured cerebral aneurysms at high-volume centers in the United States. J Neurosurg . 2011; 115( 6): 1179- 1183. Google Scholar CrossRef Search ADS PubMed  15. Reed SD, Cramer SC, Blough DK, Meyer K, Jarvik JG. Treatment with tissue plasminogen activator and inpatient mortality rates for patients with ischemic stroke treated in community hospitals. Stroke . 2001; 32( 8): 1832- 1840. Google Scholar CrossRef Search ADS PubMed  16. Washington CW, Derdeyn CP, Dacey RG Jr, Dhar R, Zipfel GJ. Analysis of subarachnoid hemorrhage using the nationwide inpatient sample: the NIS-SAH severity score and outcome measure. J Neurosurg . 2014; 121( 2): 482- 489. Google Scholar CrossRef Search ADS PubMed  17. Sonig A, Khan IS, Wadhwa R, Thakur JD, Nanda A. The impact of comorbidities, regional trends, and hospital factors on discharge dispositions and hospital costs after acoustic neuroma microsurgery: a United States nationwide inpatient data sample study (2005-2009). Neurosurg Focus . 2012; 33( 3): E3. Google Scholar CrossRef Search ADS   18. Chyatte D, Fode NC, Sundt TM Jr. Early versus late intracranial aneurysm surgery in subarachnoid hemorrhage. J Neurosurg . 1988; 69( 3): 326- 331. Google Scholar CrossRef Search ADS PubMed  19. Siddiq F, Chaudhry SA, Tummala RP, Suri MF, 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- 677; discussion 677-678. Google Scholar CrossRef Search ADS PubMed  20. 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, 2001-2008. AJNR Am J Neuroradiol . 2011; 32( 6): 1071- 1075. Google Scholar CrossRef Search ADS PubMed  21. Kan P, Jahshan S, Yashar P et al.   Feasibility, safety, and periprocedural complications associated with endovascular treatment of selected ruptured aneurysms under conscious sedation and local anesthesia. Neurosurgery . 2013; 72( 2): 216- 220; discussion 220. Google Scholar CrossRef Search ADS PubMed  22. Cho YD, Lee JY, Seo JH et al.   Early recurrent hemorrhage after coil embolization in ruptured intracranial aneurysms. Neuroradiology . 2012; 54( 7): 719- 726. Google Scholar CrossRef Search ADS PubMed  23. Starke RM, Connolly ES Jr, Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid, Hemorrhage. Rebleeding after aneurysmal subarachnoid hemorrhage. Neurocrit Care . 2011; 15( 2): 241- 246. Google Scholar CrossRef Search ADS PubMed  24. Eddleman CS, Hurley MC, Naidech AM, Batjer HH, Bendok BR. Endovascular options in the treatment of delayed ischemic neurological deficits due to cerebral vasospasm. Neurosurg Focus . 2009; 26( 3): E6. Google Scholar CrossRef Search ADS PubMed  25. Steiner T, Juvela S, Unterberg A et al.   European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage. Cerebrovasc Dis . 2013; 35( 2): 93- 112. Google Scholar CrossRef Search ADS PubMed  26. Yuan Z, Cooper GS, Einstadter D, Cebul RD, Rimm AA. The association between hospital type and mortality and length of stay: a study of 16.9 million hospitalized Medicare beneficiaries. Med Care . 2000; 38( 2): 231- 245. Google Scholar CrossRef Search ADS PubMed  27. Khuri SF, Najjar SF, Daley J et al.   Comparison of surgical outcomes between teaching and nonteaching hospitals in the Department of Veterans Affairs. Ann Surg . 2001; 234( 3): 370- 382; discussion 382-373. Google Scholar CrossRef Search ADS PubMed  28. Rosenthal GE, Harper DL, Quinn LM, Cooper GS. Severity-adjusted mortality and length of stay in teaching and nonteaching hospitals. Results of a regional study. JAMA . 1997; 278( 6): 485- 490. Google Scholar CrossRef Search ADS PubMed  29. Hoh BL, Chi YY, Dermott MA, Lipori PJ, Lewis SB. The effect of coiling versus clipping of ruptured and unruptured cerebral aneurysms on length of stay, hospital cost, hospital reimbursement, and surgeon reimbursement at the university of Florida. Neurosurgery . 2009; 64( 4): 614- 619; discussion 619-621. Google Scholar CrossRef Search ADS PubMed  30. Brinjikji W, Kallmes DF, Lanzino G, Cloft HJ. Hospitalization costs for endovascular and surgical treatment of unruptured cerebral aneurysms in the United States are substantially higher than medicare payments. AJNR Am J Neuroradiol . 2012; 33( 1): 49- 51. Google Scholar CrossRef Search ADS PubMed  31. Hoh BL, Chi YY, Lawson MF, Mocco J, Barker FG 2nd. Length of stay and total hospital charges of clipping versus coiling for ruptured and unruptured adult cerebral aneurysms in the Nationwide Inpatient Sample database 2002 to 2006. Stroke . 2010; 41( 2): 337- 342. Google Scholar CrossRef Search ADS PubMed  32. Prabhakaran S, Fonarow GC, Smith EE et al.   Hospital case volume is associated with mortality in patients hospitalized with subarachnoid hemorrhage. Neurosurgery . 2014; 75( 5): 500- 508. Google Scholar CrossRef Search ADS PubMed  Acknowledgments We thank Paul H. Dressel, BFA, for preparation of the illustrations and Debra J. Zimmer for editorial assistance. Supplemental digital content is available for this article at www.neurosurgery-online.com. COMMENTS The authors report 17 412, directly admitted SAH patients from the NIS database. Ultra-early (<24 hours) vs >24 hours treatment for securing the aneurysm (clipping vs coiling was compared). This manuscript is important in helping to solidify what most high-volume vascular neurosurgeons “know” anecdotally and supports the practice of early treatment. In most centers, stable patients with aneurysmal SAH will be treated less than 24 hours from hospital admission. These results must be understood and interpreted carefully, however. Due to the method of data collection and reporting, a patient admitted at 11:59 pm on day 0 and treated at 7 am on day 1 (7 hours from admission), will be looked at as if treated in a more delayed fashion than the patient admitted at midnight and treated at noon the same day (12 hours from admission), based solely on the way hospital days are counted. Therefore, it is expected that some of the patients evaluated in the “early” group (hospital 0 to 1), were actually treated “ultra-early” or less than 24 hours from admission. This does muddy the results and can lead to misinterpretation. In addition, more patients treated in the ultra-early group underwent coiling procedures. We already know from ISAT (International Subarachnoid Aneurysm Trial) and BRAT (Barrow Ruptured Aneurysm Trial) that patients in coiling cohorts have shorter hospital stays and better immediate outcomes. Therefore, this data can be confounded additionally simply by the mode of treatment in the ultra-early cohort. Finally, volume data for each hospital was not available. In most cases, hospitals that have the resources to offer ultra-early or early coiling or clipping will also be high-volume centers with better through-put, more experienced surgeons and staff and better infrastructure. This too will confound this data. Overall, this was a nice manuscript with a large cohort of retrospectively reviewed patient data that suggests that treatment for ruptured intracranial aneurysms within 24 hours of presentation portends a better outcome. We agree with this in general, but also think this should be reviewed with objectivity. Mandy Binning Erol Veznedaroglu Pennington, New Jersey The authors present a retrospective review of 17 412 patients in the Nationwide Inpatient Sample with non-traumatic SAH in an effort to determine the effect of ultra-early treatment on outcome. They found that ultra-early aneurysm treatment (defined as treatment on hospital day 0) is associated with better outcome (defined as discharge home) and decreased hospital cost. Although it has been shown that early aneurysm treatment is associated with decreased incidence of rebleeding and improved outcomes, the effect of ultra-early treatment is less well described. The present study is a valuable contribution to this body of knowledge. The use of nationwide data samples is appropriate for answering simple clinical questions, such as this. Additionally, the very large number of patients adds significant power to the results. On the other hand, nuances can be lost in the data and frequently cannot be recovered. Misdiagnosis can be a problem and an example in this study would be differentiating non-traumatic and traumatic SAH. Additionally, the day of treatment may be less important than the actual number of hours between admission and treatment as a patient admitted at 11 PM and treated the following day at 1 AM was actually treated incredibly quickly yet was grouped as being treated after admission day 0. Similarly, judging good outcome as only the patients who went home neglects a large number of patients who will have excellent clinical outcomes after a short stay in acute rehabilitation. The limitations of nationwide samples should be kept in mind when interpreting studies such as this. Justin Mascitelli J. Mocco New York, New York Copyright © 2017 by the Congress of Neurological Surgeons http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Neurosurgery Oxford University Press

Better Outcomes and Reduced Hospitalization Cost are Associated with Ultra-Early Treatment of Ruptured Intracranial Aneurysms: A US Nationwide Data Sample Study

Loading next page...
 
/lp/ou_press/better-outcomes-and-reduced-hospitalization-cost-are-associated-with-jcyVbsRRZk
Publisher
Congress of Neurological Surgeons
Copyright
Copyright © 2017 by the Congress of Neurological Surgeons
ISSN
0148-396X
eISSN
1524-4040
D.O.I.
10.1093/neuros/nyx241
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND The benefit of surgical treatment of ruptured aneurysms is well established. OBJECTIVE To determine whether ultra-early ruptured aneurysm treatment leads to not only improved outcomes but also reduced hospitalization cost. METHODS Using 2008-2011 Nationwide Inpatient Sample data, we analyzed demographic, clinical, and hospital factors for nontraumatic subarachnoid hemorrhage (SAH) patients who were “directly” admitted to the treating hospital where they underwent intervention (clipping/coiling). Patients treated on the day of admission (day 0) formed the ultra-early cohort; others formed the deferred treatment cohort. All Patient Refined Diagnosis-Related Groups were also included in regression analyses. RESULTS A total of 17 412 patients were directly admitted to a hospital following nontraumatic SAH where they underwent intervention (clipping/coiling). Mean patient age was 53.87 yr (median 53.00, standard deviation 14.247); 68.3% were women (n = 11 893). A total of 6338 (36.4%) patients underwent treatment on the day of admission (ultra-early). Patients who underwent treatment on day 0 had significantly more routine discharge dispositions than those treated >admission day 0 (P < .0001). In regression analysis, treatment on day 0 was protective against other than routine discharge disposition outcome (P < .0001; odds ratio 0.657; 95% confidence interval 0.614-0.838). Total cost incurred by hospitals was $4.36 billion. Mean cost of hospital charges in the ultra-early cohort was $239 126.05, which was significantly lower than that for the cohort treated >day 0 ($272 989.56, P < .001), Mann-Whitney U-test). Performance of an intervention on admission day 0 was protective against higher hospitalization cost (P < .0001; odds ratio 0.811; 95% confidence interval 0.732-0.899). CONCLUSION Ultra-early treatment of ruptured aneurysms is significantly associated with better discharge disposition and decreased hospitalization cost. Aneurysmal subarachnoid hemorrhage, Clipping, Coiling, Discharge disposition, Hospitalization cost, Nationwide Inpatient Sample, Ruptured intracranial aneurysm ABBREVIATIONS ABBREVIATIONS APR-DRGs All Patient Refined Diagnosis Related Groups CAD coronary artery disease CCS Clinical Classifications Software CI confidence interval h hours HCUP Healthcare Cost and Utilization Project ICD-9-CM International Classification of Diseases, Ninth Revision, Clinical Modification ICH intracranial hemorrhage LOS length of stay MI myocardial infarction NIH National Institutes of Health NIS Nationwide Inpatient Sample NIS-SSS Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score OR odds ratio OTR other than routine SAH subarachnoid hemorrhage SD standard deviation The benefit of surgical treatment of ruptured aneurysms is well established. Traditionally, surgery by either endovascular or open approaches is considered early when performed within 48 to 72 hours (h) of presentation. Several studies1-5 have shown the benefit of early surgery, whereas some have advocated delayed treatment of poor-grade patients.6 To date, very few studies have discussed the benefits of ultra-early treatment (≤24 h of presentation) vs delayed treatment (>24 h of presentation) of aneurysmal subarachnoid hemorrhage (SAH).7-12 No study has compared the outcomes and cost-effectiveness of ultra-early surgery for ruptured aneurysms across the US. We analyzed Nationwide Inpatient Sample (NIS) data for our study on this subject.13,14 The NIS is the largest database of inpatient care. It comprises 20% of admissions to US nonfederal hospitals. The NIS database is managed by the Healthcare Cost and Utilization Project (HCUP). A new variable, “transfer status,” was incorporated in the database in 2008. This variable indicates whether a patient was a direct hospital admission or transferred from another facility. Additionally, the database included a coded variable: “number of days from admission to principal procedure.” In this study, we utilized the additional variables provided in the NIS database to study the timing of aneurysm surgery in the cohort of patients who were directly admitted to hospitals at the time of symptom onset and whether they underwent treatment on admission day 0 (ultra-early treatment; ie, on the day of admission as well as within 24 h of admission). Additionally, we analyzed the impact of ultra-early aneurysm treatment on the cost of hospitalization. Because the extent of a patient's initial disability can influence outcome, we also analyzed severity of disease using All Patient Refined Diagnosis-Related Groups (APR-DRGs).15 We posit that ultra-early aneurysm surgery not only improves outcome but also decreases hospitalization cost. METHODS The NIS data are deidentified by HCUP, and thus, institutional review board approval and patient consent were not needed for the present study. The NIS is the largest inpatient database available to the public. We extracted NIS data from 2008 to 2011 using codes from International Classification of Diseases, Clinical Modification (ICD-9-CM) and Clinical Classifications Software (CCS; http://www.hcup-us.ahrq.gov/). ICD-9-CM diagnosis codes of 430 (SAH); codes 3951 (clipping) and 3952, 3972, and 3979 (coiling) were then used to extract cases of SAH treated by primary clipping or coiling. We used the NIS-coded variable “transfer status” (direct admission to hospital or transfer-in from another facility) to extract only those patients who were directly admitted to the hospital. The NIS defines the day on which the procedure is performed (PRDAYn: procedure day n). PRDAYn is calculated from the procedure and admission dates. For a procedure performed on the day of admission, PRDAYn = 0; for the day after admission, PRDAYn = 1. Time of day is not a factor (Email from hcup-us@truvenhealth.com dated June 10, 2016). The following cohorts were formed based on the information on timing of aneurysm treatment. Ultra-early (procedure on admission day 0): the NIS-coded variable “procedure on admission day 0” was utilized to extract those patients who underwent treatment (clip/coil) on the day of admission, day 0. This cohort included patients who were treated <24 h from admission. This cohort was compared with the group who had the procedure on subsequent days (after day 0). Early (procedure on admission day 0 or 1): we created this cohort because a small subset of patients underwent their procedure the next day, after midnight (within 24 h, but on the next admission day, ie, admission day 1). Deferred (procedure on admission day 2 or more): the remaining patients formed the deferred treatment cohort. Patients who were not direct admits to a hospital were excluded from the study. The following additional NIS database variables were analyzed: gender; race; insurance status; hospital charges; APR-DRG severity, which was subclassified as minor function loss (including cases with no complications or comorbidity), moderate function loss, major function loss, and extreme function loss; hospital region (the US is divided into Midwest, West, Northeast, and Southern, regions); and teaching or nonteaching hospitals. An additional, new variable was created, Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score (NIS-SSS).16 This variable took into account several factors that are associated with poor presentation grade, including the following: ICD-9-CM diagnosis codes associated with altered mental status: coma (780.01, 780.03) and stupor (780.02, 780.09); ICD-9-CM codes associated with hydrocephalus (331.3, 331.4) and shunt/ventriculostomy (procedure codes 02.2 02.31-02.3); neurological deficit: aphasia (438.1-438.89), cranial nerve deficits (378.5-378.56, 379.4-379.43), and paresis or plegia (438.2-438.53, 781.4); and the CCS code for mechanical ventilation, 216. Several other factors that were generated secondarily from the database were also analyzed. These included age (dichotomized to ≥65 yr and <65 yr). The following CCS codes were used to generate these factors from the NIS database: 76 (meningitis); 100 (acute myocardial infarction), 101 (coronary atherosclerosis and other heart disease), 107 (cardiac arrest and ventricular fibrillation), 108 (congestive heart failure; non-hypertensive), 125 (acute bronchitis), 126 (other upper respiratory infections), 127 (chronic obstructive pulmonary disease and bronchiectasis), 128 (asthma), 129 (aspiration pneumonia), 130 (pleurisy; pneumothorax, pulmonary collapse), 131 (respiratory failure; insufficiency; arrest [adult]), 132 (lung disease due to external agents), 133 (other lower respiratory disease), 197 (wound infection), and ICD-9-CM codes for secondary intracranial hemorrhage (ICH; 430, 431, and 432). Patients who underwent intervention on day 1 of admission formed the “ultra-early treatment” cohort. Two additional variables were created based on existing information in the NIS database and percentile analysis: comorbidity index and higher hospitalization cost. The NIS database includes the following comorbidities: acquired immune deficiency syndrome, alcohol abuse, chronic blood loss anemia, chronic pulmonary disease, congestive heart failure, coagulopathy, deficiency anemias, diabetes with chronic complications, drug abuse, fluid and electrolyte disorders, hypertension, hypothyroidism, liver disease, lymphoma, metastatic cancer, obesity, other neurological disorders, paralysis, peptic ulcer disease excluding bleeding, peripheral vascular disorders, pulmonary circulation disorders, renal failure, rheumatoid arthritis/collagen vascular diseases, solid tumor without metastasis, uncomplicated diabetes, valvular disease, and weight loss. The number of comorbidities a patient had was calculated and summed. It ranged from 0 to 13. Patients with ≥2 comorbidities formed the “above the 75th percentile” high-comorbidity group. Patients with <2 comorbidities formed the low-comorbidity group. Similarly, the higher hospitalization cost variable was created based on existing “total charges” information in the NIS database. “Total charges” gives the cost incurred by the hospital while treating the patient. All the charges under the variable “total charges” were analyzed by descriptive statistics and segregation according to percentiles. Hospitalization cost of more than $511 451.00 formed the greater than 90th percentile cohort (higher hospitalization cost). For outcome analysis, a dichotomized discharge disposition was utilized: “routine” and “other than routine (OTR).”17 Any transfer from a hospital to acute rehabilitation, home health care, intermediate care, a short-term hospital facility, or a skilled nursing facility; discharge against medical advice; or death was considered an OTR discharge disposition. A direct disposition to home without home health care was considered a routine discharge disposition. Statistical Analysis Univariate analysis was performed using chi square or Fisher exact test. A multivariate binary logistic regression model was used to analyze variables with a P-value <.1, with hospitalization cost and discharge disposition as dependent variables. Means were compared with the Mann-Whitney U-test. All statistical analyses were performed using IBM SPSS Statistics 20 (IBM Corporation, Armonk, New York). RESULTS Data for 276 902 patients with nontraumatic SAH were analyzed. A total of 33 602 patients underwent intervention (clipping or coiling); and of those, 17 412 were directly admitted to a hospital following SAH. Among the patients who were directly admitted, the mean age was 53.87 yr (median 53.00 yr, standard deviation 14.24); 68.3% were women (n = 11 893); and 78.2% (n = 13 615) were <65 yr. The interventions of coiling (n = 8702, 49.9%) and clipping (n = 8710, 50.1%) were nearly equal in proportion. A total of 6338 (36.4%) patients underwent treatment on the day of admission (ultra-early), and a total of 12 159 (69.8%) of the 17 412 patients underwent treatment on admission day 0 or 1 (early). Coiling was the preferred modality on admission day 0, with 59.14% (n = 3748) of patients undergoing treatment with primary coiling (Table 1). Surgical clipping was the preferred modality when treatment was carried out beyond day 0 of admission (in 55.26% patients; n = 6120). When overall outcome was considered, 41.6% (n = 7245) of the patients had routine discharge disposition. Among these, patients who had treatment within admission day 0 had significantly more routine discharge dispositions (P < .001; Table 1). In the regression analysis, treatment within admission day 0 was protective against worse outcome (P = .001; odds ratio [OR] 0.657; confidence interval [CI] 0.614-0.838; Table 2). Similarly, early treatment (day 0/1) was protective against OTR outcome; however, the value was bordering on significance (P = .043; OR 0.907; CI 0.826-0.997). TABLE 1. Comparison of Treatment Timings (Ultra-Early, Early, and Deferred)     Procedure on admission day ≥1 (total n = 11 074)  Procedure on admission day 0 (ultra-early; total n = 6338)  P value  Procedure on admission day ≥2 (total n = 5253)  Procedure on admission day 0/1 (early; total n = 12 159)  P value      n  %  n  %    n  %  n  %    Died during hospitalization  1329  12.00  925  14.59  <.001  592  11.27  1662  13.67  <.001  Final outcome routine discharge disposition  4417  39.89  2828  44.62  <.001  2065  39.31  5180  42.60  <.001  >90 percentile cost (Higher cost)  1393  12.58  672  10.60  <.001  791  15.06  1274  10.48  <.001  Age ≥65 y  2493  22.51  1304  20.57  .003  1266  24.10  2531  20.82  <.001  Female sex  7615  68.76  4278  67.50  .086  3597  68.48  8296  68.23  .743  All patient refined DRG: severity of illness subclass  Minor loss of function (includes cases with no comorbidity or complications)  378  3.41  458  7.23  <0.001  143  2.72  693  5.70  <0.001    Moderate loss of function  1751  15.81  1027  16.20    875  16.66  1903  15.65      Major loss of function  4929  44.51  2649  41.80    2335  44.45  5243  43.12      Extreme loss of function  4016  36.27  2204  34.77    1900  36.17  4320  35.53    Comorbidity status (high comorbidity)  4471  40.37  2415  38.10  0.003  2097  39.92  4789  39.39  0.522  Coiling/clipping  Coiling  4954  44.74  3748  59.14  <0.001  2152  40.97  6550  53.87  <0.001    Clipping  6120  55.26  2590  40.86    3101  59.03  5609  46.13    Meningitis  Yes  318  2.87  242  3.82  0.001  148  2.82  412  3.39  0.04  CAD  Yes  898  8.11  422  6.66  0.001  475  9.04  845  6.95  <0.001  Intracerebral hemorrhage  Yes  598  5.40  305  4.81  0.092  297  5.65  606  4.98  0.067  Respiratory failure  Yes  3886  35.09  2328  36.73  0.03  1777  33.83  4437  36.49  0.001  Wound infection  Yes  84  0.76  40  0.63  0.336  34  0.65  90  0.74  0.419  Teaching status of hospital (#)  Teaching  9511  88.12  5022  80.84  <0.001  4517  88.24  10016  84.26  <0.001      Procedure on admission day ≥1 (total n = 11 074)  Procedure on admission day 0 (ultra-early; total n = 6338)  P value  Procedure on admission day ≥2 (total n = 5253)  Procedure on admission day 0/1 (early; total n = 12 159)  P value      n  %  n  %    n  %  n  %    Died during hospitalization  1329  12.00  925  14.59  <.001  592  11.27  1662  13.67  <.001  Final outcome routine discharge disposition  4417  39.89  2828  44.62  <.001  2065  39.31  5180  42.60  <.001  >90 percentile cost (Higher cost)  1393  12.58  672  10.60  <.001  791  15.06  1274  10.48  <.001  Age ≥65 y  2493  22.51  1304  20.57  .003  1266  24.10  2531  20.82  <.001  Female sex  7615  68.76  4278  67.50  .086  3597  68.48  8296  68.23  .743  All patient refined DRG: severity of illness subclass  Minor loss of function (includes cases with no comorbidity or complications)  378  3.41  458  7.23  <0.001  143  2.72  693  5.70  <0.001    Moderate loss of function  1751  15.81  1027  16.20    875  16.66  1903  15.65      Major loss of function  4929  44.51  2649  41.80    2335  44.45  5243  43.12      Extreme loss of function  4016  36.27  2204  34.77    1900  36.17  4320  35.53    Comorbidity status (high comorbidity)  4471  40.37  2415  38.10  0.003  2097  39.92  4789  39.39  0.522  Coiling/clipping  Coiling  4954  44.74  3748  59.14  <0.001  2152  40.97  6550  53.87  <0.001    Clipping  6120  55.26  2590  40.86    3101  59.03  5609  46.13    Meningitis  Yes  318  2.87  242  3.82  0.001  148  2.82  412  3.39  0.04  CAD  Yes  898  8.11  422  6.66  0.001  475  9.04  845  6.95  <0.001  Intracerebral hemorrhage  Yes  598  5.40  305  4.81  0.092  297  5.65  606  4.98  0.067  Respiratory failure  Yes  3886  35.09  2328  36.73  0.03  1777  33.83  4437  36.49  0.001  Wound infection  Yes  84  0.76  40  0.63  0.336  34  0.65  90  0.74  0.419  Teaching status of hospital (#)  Teaching  9511  88.12  5022  80.84  <0.001  4517  88.24  10016  84.26  <0.001  Abbreviations: CAD, coronary artery disease; DRG, diagnosis-related groups #, nonsignificant missing values; y, years View Large TABLE 2. Results of Multivariate Regression Analysis for Discharge Disposition       95% CI for OR  Factors  Significance  OR  Lower  Upper  Elderly population (>65 y)  <.0001  4.596  4.130  5.116  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  26.552  20.257  34.802  Major loss of function  <.0001  8.084  7.018  9.313  Moderate loss of function  <.0001  3.244  2.914  3.611  Comorbidity index  <.0001  1.185  1.093  1.285  Treatment type (coiling vs clipping)  .055  1.081  .998  1.170  Procedure on admission day 0/1 (early) vs >2 (deferred)  .043  0.907  0.826  0.997  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.657  0.614  0.838  Meningitis  .549  1.070  0.858  1.335  CAD and other heart disease  .391  1.069  0.918  1.243  ICH  <.0001  1.579  1.316  1.896  Respiratory failure  <.0001  2.760  2.489  3.060  Wound infection  <.0001  2.498  1.569  3.977  Hospital teaching status (teaching vs nonteaching)  <.0001  0.719  0.640  0.808  Region of hospital (ref: Northeast)  .000        West  .832  0.589  0.521  1.664  South  .553  0.965  0.857  1.086  Midwest  .731  1.422  0.292  1.566        95% CI for OR  Factors  Significance  OR  Lower  Upper  Elderly population (>65 y)  <.0001  4.596  4.130  5.116  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  26.552  20.257  34.802  Major loss of function  <.0001  8.084  7.018  9.313  Moderate loss of function  <.0001  3.244  2.914  3.611  Comorbidity index  <.0001  1.185  1.093  1.285  Treatment type (coiling vs clipping)  .055  1.081  .998  1.170  Procedure on admission day 0/1 (early) vs >2 (deferred)  .043  0.907  0.826  0.997  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.657  0.614  0.838  Meningitis  .549  1.070  0.858  1.335  CAD and other heart disease  .391  1.069  0.918  1.243  ICH  <.0001  1.579  1.316  1.896  Respiratory failure  <.0001  2.760  2.489  3.060  Wound infection  <.0001  2.498  1.569  3.977  Hospital teaching status (teaching vs nonteaching)  <.0001  0.719  0.640  0.808  Region of hospital (ref: Northeast)  .000        West  .832  0.589  0.521  1.664  South  .553  0.965  0.857  1.086  Midwest  .731  1.422  0.292  1.566  Abbreviations: APR-DRG, All Patient Refined Diagnosis Related Group; CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; ref=reference. aThe dependent variable was discharge disposition (other than routine vs routine); demographic and patient factors were the covariates. View Large The rates of mortality and OTR discharge disposition (worse outcome) were higher in patients >65 yr and in women, and associated with significant differences (Table, Supplemental Digital Content 1). This significance persisted even when these variables were put in the multivariate regression model (Table 2). Women had 1.248 times higher odds for a worse outcome than men (P < .0001, CI 1.144-1.362). Similarly, age ≥ 65 yr had 4.130 times higher odds for a worse outcome (P < .0001, CI 4.130-5.116). The predominant race was white (56.5%; n = 8535). Race had a high percentage (29.8%) of missing values, hence is not included in the univariate analysis. On the basis of the APR-DRG mortality subclass, 38.3% (n = 6677) had “minor likelihood of dying” with 1.1% in-hospital mortality and 29.7% (n = 5181) had “extreme likelihood of dying” with 31.9% mortality, the difference being significant (P < .001) (Table, Supplemental Digital Content 1). Similarly, on the basis of the APR-DRG severity subclass, 4.8% (n = 836) had “minor function loss” with 9.7% OTR outcomes, and 35.7% (n = 6220) had “extreme function loss” with 86.6% OTR outcomes (P < .001; Table, Supplemental Digital Content 1). Patients who were treated on the day of admission had lesser severity of APR-DRG (Table 1). In regression analysis, APR-DRG severity subclass “extreme function loss” had 26.55 times higher odds for a worse outcome than “minor function loss” (P < .0001; CI 20.26-34.80; Table 2). Similarly, we analyzed the relationship between poor NIS-SSS score and outcome. A total of 8034 (46.1%) patients had a poor NIS-SSS score and, of these, only 19.5% (1563 of 8034) had a routine discharge disposition. However, 60.5% (5682 of 9378) of good-grade patients had a routine discharge disposition. Regression analysis with discharge disposition as the outcome variable was performed by incorporating the NIS-SSS score (Table 3). A poor NIS-SSS score had an OR of 3.422 (P < .0001, CI 3.142-3.728) for OTR outcome, and patients who were treated on the day of admission (ultra-early) had protection against OTR discharge disposition (P = .001, OR 0.809, CI 0.741-0.884). TABLE 3. Results of Multivariate Regression Analysis Using the Variable NIS-SSS for Discharge Disposition       95% CI for OR  Factors  Significance  OR  Lower  Upper  +Elderly population (>65 y)  <.0001  4.953  4.465  5.493  NIS-SSS  <.0001  3.422  3.142  3.728  Comorbidity index  <.0001  1.572  1.455  1.699  Treatment type (coiling vs clipping)  .001  1.451  1.346  1.565  Procedure on admission day 0/1 (early) vs >2 (deferred)  .029  0.901  0.821  0.989  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.809  0.741  0.884  Meningitis  <.0001  1.680  1.350  2.091  CAD and other heart disease  .001  1.871  1.412  2.479  ICH  <.0001  1.687  1.411  2.016  Respiratory failure  <.0001  3.522  3.195  3.883  Wound infection  <.0001  3.111  1.982  4.884  Hospital teaching status (teaching vs non-teaching)  <.0001  0.710  0.636  0.792  Region of hospital (ref: Northeast)          West  .488  0.961  0.858  1.076  South  .327  0.945  0.845  1.058  Midwest  .743  1.335  0.217  1.466        95% CI for OR  Factors  Significance  OR  Lower  Upper  +Elderly population (>65 y)  <.0001  4.953  4.465  5.493  NIS-SSS  <.0001  3.422  3.142  3.728  Comorbidity index  <.0001  1.572  1.455  1.699  Treatment type (coiling vs clipping)  .001  1.451  1.346  1.565  Procedure on admission day 0/1 (early) vs >2 (deferred)  .029  0.901  0.821  0.989  Procedure on admission day 0 (ultra-early) vs >1 admission day  .001  0.809  0.741  0.884  Meningitis  <.0001  1.680  1.350  2.091  CAD and other heart disease  .001  1.871  1.412  2.479  ICH  <.0001  1.687  1.411  2.016  Respiratory failure  <.0001  3.522  3.195  3.883  Wound infection  <.0001  3.111  1.982  4.884  Hospital teaching status (teaching vs non-teaching)  <.0001  0.710  0.636  0.792  Region of hospital (ref: Northeast)          West  .488  0.961  0.858  1.076  South  .327  0.945  0.845  1.058  Midwest  .743  1.335  0.217  1.466  Abbreviations: CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; NIS-SSS, Nationwide Inpatient Sample-Subarachnoid Hemorrhage Severity Score; ref, reference, yrs, years aThe dependent variable was discharge disposition (other than routine vs routine); demographic and patient factors were the covariates. View Large Patients with high comorbidity status or coiling as the procedure, coronary artery disease, ICH, and respiratory failure had significantly higher mortality and OTR discharge disposition (Table, Supplemental Digital Content 1). However, in the regression model, only ICH, respiratory failure, and presence of wound infection had significantly higher odds for worse outcome (Table 2 and 3). Among the hospital factors, 80.84% of day 0 treatments were carried out in teaching hospitals (Table 1), and regression analysis showed that teaching hospitals (P < .0001; OR 0.719; CI 0.640-0.808; Table 2) were protective against a worse outcome. Hospitalization Cost The total cost incurred by hospitals was $4.36 billion. The mean cost of hospital charges in the ultra-early cohort was $239 126.05, which was significantly lower than that for the cohort treated >day 0 ($272 989.56; P < .001, Mann-Whitney U-Test). The “higher hospitalization” cost cohort was comprised of 11.9% of all patients (n = 2065). Age >65 yr, higher comorbidity status, APR-DRG subset “extreme loss of function,” procedure performed >24 h, meningitis, coronary atherosclerosis and other heart disease, ICH, respiratory failure, and wound infection were associated with higher cost of hospitalization in univariate analysis (Table, Supplemental Digital Content 2). These variables were tested in the regression analysis (Table 4). Procedure performed on admission day 0 was protective against higher cost of hospitalization (P < .0001; OR 0.811; CI 0.732-0.899). TABLE 4. Results of Multivariate Regression Analysis       95% CI for OR    Significance  Odds ratio  Lower  Upper  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  20.696  12.429  34.463  Major loss of function  <.0001  9.494  7.593  11.872  Moderate loss of function  <.0001  4.254  3.739  4.840  Elderly population (>65 y)  .014  0.867  0.773  0.972  Comorbidity severity  .001  1.178  1.067  1.301  Treatment type (coiling vs clipping)  <.0001  0.670  0.607  0.740  Meningitis  .007  1.343  1.086  1.661  CAD and other heart disease  .214  0.895  0.752  1.066  ICH  <.0001  1.641  1.379  1.952  Respiratory failure  <.0001  1.668  1.593  1.752  Wound infection  <.0001  2.557  1.650  3.961  Procedure on admission day 0 (ultra-early) vs >1 admission day  <.0001  0.811  0.732  0.899        95% CI for OR    Significance  Odds ratio  Lower  Upper  APR-DRG (ref: minor loss of function)          Extreme loss of function  <.0001  20.696  12.429  34.463  Major loss of function  <.0001  9.494  7.593  11.872  Moderate loss of function  <.0001  4.254  3.739  4.840  Elderly population (>65 y)  .014  0.867  0.773  0.972  Comorbidity severity  .001  1.178  1.067  1.301  Treatment type (coiling vs clipping)  <.0001  0.670  0.607  0.740  Meningitis  .007  1.343  1.086  1.661  CAD and other heart disease  .214  0.895  0.752  1.066  ICH  <.0001  1.641  1.379  1.952  Respiratory failure  <.0001  1.668  1.593  1.752  Wound infection  <.0001  2.557  1.650  3.961  Procedure on admission day 0 (ultra-early) vs >1 admission day  <.0001  0.811  0.732  0.899  APR-DRG, All Patient Refined Diagnosis-Related Group; CAD, coronary artery disease; CI, confidence interval; ICH, intracranial hemorrhage; ref, reference; y, years. View Large DISCUSSION Earlier studies have segregated the timing of ruptured aneurysm treatment in relation to the time interval between symptom onset and treatment. The term early surgery refers to <3 d18 or <2 d19 of the ictus, and the term “ultra-early surgery” is applied to intervention/surgery within 24 h of presentation.6,9,12 Our study is based on NIS data provided by HCUP. The database divides patients into mainly 2 categories, patients who are admitted directly to a hospital and those who are transferred to a hospital from another medical facility. As this study focused on the timing of surgery/intervention, we selected only those patients who were directly admitted to a facility where they underwent treatment. We excluded those patients who were transferred from another hospital because in the latter scenario, true timing of surgery/intervention is difficult to assess owing to the time elapsed between admission to the first and subsequent hospitals. Patients who were admitted to the hospital directly were divided into cohorts based on the timing of surgery. The coded variable of NIS Procedure on day 0 of admission was used to form the ultra-early cohort, and patients who underwent procedures on day 0 and day 1 formed the early cohort. A similar analysis of the impact of timing of intervention/surgery on the outcome of aneurysmal SAH across the United States was performed by Siddiq et al.19 However, in that study, the admission status (direct admission or transferred from another facility) was not considered and the impact of early treatment (<48 h from admission), rather than ultra-early treatment, was analyzed. Additionally, the study analyzed both directly admitted and transferred patients. The time elapsed in transfer was not taken into account. In the present study, univariate analysis showed that clipping was significantly associated with less mortality and OTR discharge disposition than coiling. However, the regression analysis did not show significance (P = .055; OR 1.081; CI 0.998-1.170). The aim of our study was not to compare endovascular and surgical treatment techniques because the 2 techniques have been compared earlier in an evaluation of NIS data.20 Instead, we were interested in the timing of treatment and its relationship with outcome and cost. In the present study, 6338 (36.4%) patients underwent treatment on the day of admission (day 0, ultra-early), and coiling was the preferred modality for that cohort (59.14%; n = 3748). The percentage of coiling procedures decreased in the early (day 0/1) cohort (53.9%, n = 6550). Surgical clipping was the preferred modality when treatment was carried out beyond day 0 of admission (in 55.26% patients; n = 6120; Table 1). Several factors determine the modality of treatment, including presence of ICH, mass effect, morphology of the aneurysm, and predilections of the treatment team. One possible reason for the preference for endovascular treatment within day 0 of admission was that the endovascular intervention could become a part of the diagnostic angiogram, thus saving time and hospital resources. Another factor could be the predilection of the treating team and the relative ease of delivering endovascular treatment compared with surgical clipping. At some centers, including ours, endovascular treatment is successfully performed under conscious sedation, and it can be extended to SAH patients who are in good grade.21 Further analysis on the timing of intervention (clipping/coiling) demonstrated that despite a higher rate of mortality on admission days 0 and 1, procedures performed on the day of admission (day 0) were associated with significantly less occurrence of OTR discharge disposition than those performed >day 0 of admission (P = .001; OR 0.657). This difference persisted in the regression analysis despite adjusting for several factors, including clipping/coiling (Table 2). Significance was maintained (P = .043) on procedure day 0/1, but the OR was only 0.907 compared to that for the deferred intervention cohort. The higher mortality on days 0 and 1 may be secondary to poor general condition of the patient upon admission. We adjusted for this factor in regression analysis, and mortality was included in OTR discharge disposition. A possible explanation could be that ultra-early surgery confers protection against rebleed and future neurological complications. The occurrence of rebleeding is known to peak within the first 24 h of rupture.22 In a systemic review, Starke and Connolly23 found the frequency of rebleeding to be as high as 9% to 17% in the first 24 h of presentation. Also, ultra-early aneurysm treatment allows for an early start of aggressive medical therapy to counter cerebral vasospasm and to tackle delayed neurological deficits.24,25 In our study, patients who received treatment within day 0 of admission had lower odds (0.657; Table 2) of worse outcome. In a similar study on the benefits of ultra-early treatment, Wong et al12 reported a significant association between better SF-36 mental scores and ultra-early aneurysm treatment (P = .041) and favorable outcome (P = .086). This benefit persisted, even in poor-grade patients (P = .062). Studies have shown that outcomes are better when surgeries/interventions are performed at teaching hospitals.19,26 In our study, 85.5% (n = 14 533) patients had surgery/intervention at teaching hospitals. Mortality and morbidity were significantly lower at teaching hospitals than nonteaching hospitals. Regression analysis showed that the odds of worse outcomes are lower at teaching hospitals (P < .001; OR 0.719; CI 0.640-0.808; Table 2). In our study, large hospitals had a protective effect against worse outcome. This could be due to readily available resources at larger hospitals. A similar association was seen in earlier studies.27,28 When we analyzed clinical factors, higher comorbidity status was associated with increased mortality and OTR discharge disposition. In multivariate regression analysis (Table 2), several other patient factors—ICH, respiratory infection, wound infection—and APR-DRG severity of illness were associated with OTR discharge disposition. The odds for a worse outcome were 26.552 times greater when the patient had extreme function loss, compared with minimal function loss, during the hospital stay (Table 2). APR-DRGs are based on clinical events at admission or occurring during the hospital stay.15 We created another variable, NIS-SSS, as described by Washington et al.16 In their study of the validation of SAH severity based on NIS data, they found a significant correlation between NIS-SSS score and Hunt and Hess scores. When we performed regression analysis by incorporating the NIS-SSS variable, ultra-early timing of surgery still had the best preventive rate against OTR discharge disposition (Table 3). In the current scenario of ever-increasing healthcare cost, the economics of cost utilization become paramount to the discussion of timing of aneurysm surgery. Several studies have compared the cost utilization of clipping vs coiling at a single institution29 or nationwide;30,31 but to our knowledge, none has looked at the impact of ultra-early intervention on cost of hospitalization. Costs are affected with increasing comorbidities and complications, length of stay (LOS), clinical outcome, and hospitalization.17 We analyzed hospitalization cost and various associated factors. The mean hospitalization cost was significantly less for patients who received treatment within admission day 1 for those who received treatment after day 1 of admission (P < .0001). This association persisted when hospitalization cost was adjusted for comorbidities, complications, and APR-DRG severity in regression analysis. The odds were 0.8 for higher hospitalization cost when treatment was performed within admission day 0 (Table 4); thus, ultra-early intervention was protective against higher hospitalization cost. This finding is important as it suggests that ultra-early treatment not only provides better outcomes but also is cost-effective. LOS had a significantly high correlation with hospital cost incurred (P < .001: bivariate Spearman nonparametric correlation analysis). Therefore, LOS was not included in the regression analysis to assess factors associated with higher hospitalization cost. It is often assumed that patients are transferred to another facility if their case is complicated, the grade is poor, or when treatment options are not available at the initial facility. To avoid this bias, we analyzed only those patients who were directly admitted. However, there are limitations to our study. Limitations Our study is a retrospective analysis, and there may be inherent errors in coding or under-reporting of events in the database. It is not possible to distinguish preoperative neurological deficit from postoperative deficit, including rebleed. The database does not provide information on radiological outcomes associated with aneurysm treatment, so it is not possible to extrapolate the findings of this study in that respect. APR-DRG severity and NIS-SSS scores cannot be used as surrogate markers for World Federation of Neurosurgical Societies or Hunt and Hess scores at admission. However, the APR-DRGs and NIS-SSS scores do provide valuable data regarding the disability status of the patient; and by including these scores in our regression analysis, the testing of other covariates was more robust. Although not addressed in the present analysis, hospital volumes can also affect outcomes. Hospitals with a high volume of SAH patients tend to have better outcomes.32 Further, the NIS defines “day 0” as time prior to the first midnight of admission, thus patients admitted late in the day might be treated the next day, technically within 24 h, but not on “day 0.” Finally, for the analysis of cost, we looked at single hospitalization, not taking into account readmissions or cost incurred for treatment at rehabilitation facilities. Such data on long-term rehabilitation services and readmissions are not provided in the NIS database. CONCLUSION Several factors can affect clinical outcome and hospitalization cost following treatment for ruptured aneurysms. Our study has shown that treatment of ruptured aneurysm within day 0 of admission following SAH is associated with better discharge disposition and decreased hospitalization cost. Disclosures The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Financial relationships/potential conflicts of interest: Sonig, Shallwani, Natarajan, Shakir: none. Hopkins: grant/research support-Toshiba; consultant-Abbott, Boston Scientific, Cordis, Covidien, Medtronic; financial interests-Boston Scientific, Valor Medical, Claret Medical Inc., Augmenix, Endomation, Silk Road, Ostial, Apama, StimSox, Photolitec, ValenTx, Ellipse, Axtria, NextPlain, Ocular; board/trustee/officer position-Claret Medical, Inc.; honoraria-Complete Conference Management, Covidien, Memorial Healthcare System. Levy: shareholder/ownership interests–Intratech Medical Ltd., Blockade Medical LLC, NeXtGen Biologics. Principal investigator: Covidien US SWIFT PRIME Trials. Honoraria–Covidien. Consultant–Pulsar, Blockade Medical. Advisory Board-Stryker, NeXtGen Biologics, MEDX. Other financial support–Abbott for carotid training sessions. Siddiqui: Research grants: The National Institutes of Health (co-investigator: NINDS 1R01NS064592-01A1, Hemodynamic induction of pathologic remodeling leading to intracranial aneurysms), The National Institutes of Health (co-investigator: NIBIB 5 R01 EB002873-07, Micro-Radiographic Image for Neurovascular Interventions), The National Institutes of Health (co-investigator: NIH/NINDS 1R01NS091075 Virtual Intervention of Intracranial Aneurysms). None of these grants are relevant to this paper. Financial interests:Hotspur, Intratech Medical, StimSox, Valor Medical, Blockade Medical, Lazarus Effect, Pulsar Vascular, Medina Medical; Consultant: Codman & Shurtleff, Covidien Vascular Therapies, GuidePoint Global Consulting, Penumbra, Stryker, Pulsar Vascular, MicroVention, Lazarus Effect, Blockade Medical, Reverse Medical, W.L. Gore & Associates; National Steering Committees:Penumbra-3D Separator Trial, Covidien-SWIFT PRIME Trial, MicroVention-FRED Trial; Speakers’ bureau: Codman & Shurtleff, Inc.; Advisory Board:Codman & Shurtleff, Covidien Neurovascular, ICAVL, Medina Medical; Honoraria:Penumbra, Toshiba Medical Systems. Snyder: Speaker's Bureau: Toshiba and Jacobs Institute REFERENCES 1. de Gans K, Nieuwkamp DJ, Rinkel GJ, Algra A. Timing of aneurysm surgery in subarachnoid hemorrhage: a systematic review of the literature. Neurosurgery . 2002; 50( 2): 336- 340; discussion 340-332. Google Scholar PubMed  2. Deruty R, Mottolese C, Pelissou-Guyotat I, Soustiel JF. Management of the ruptured intracranial aneurysm–early surgery, late surgery, or modulated surgery? Personal experience based upon 468 patients admitted in two periods (1972-1984 and 1985-1989). Acta Neurochir (Wien) . 1991; 113( 1-2): 1- 10. Google Scholar CrossRef Search ADS PubMed  3. Nieuwkamp DJ, de Gans K, Algra A et al.   Timing of aneurysm surgery in subarachnoid haemorrhage–an observational study in The Netherlands. Acta Neurochir (Wien) . 2005; 147( 8): 815- 821. Google Scholar CrossRef Search ADS PubMed  4. Ross N, Hutchinson PJ, Seeley H, Kirkpatrick PJ. Timing of surgery for supratentorial aneurysmal subarachnoid haemorrhage: report of a prospective study. J Neurol Neurosurg Psychiatry . 2002; 72( 4): 480- 484. Google Scholar PubMed  5. Whitfield PC, Moss H, O’Hare D, Smielewski P, Pickard JD, Kirkpatrick PJ. An audit of aneurysmal subarachnoid haemorrhage: earlier resuscitation and surgery reduces inpatient stay and deaths from rebleeding. J Neurol Neurosurg Psychiatry . 1996; 60( 3): 301- 306. Google Scholar CrossRef Search ADS PubMed  6. Egashira Y, Yoshimura S, Enomoto Y, Ishiguro M, Asano T, Iwama T. Ultra-early endovascular embolization of ruptured cerebral aneurysm and the increased risk of hematoma growth unrelated to aneurysmal rebleeding. J Neurosurg . 2013; 118( 5): 1003- 1008. Google Scholar CrossRef Search ADS PubMed  7. Baltsavias GS, Byrne JV, Halsey J, Coley SC, Sohn MJ, Molyneux AJ. Effects of timing of coil embolization after aneurysmal subarachnoid hemorrhage on procedural morbidity and outcomes. Neurosurgery . 2000; 47( 6): 1320- 1329; discussion 1329-1331. Google Scholar CrossRef Search ADS PubMed  8. Gu DQ, Zhang X, Luo B, Long XA, Duan CZ. Impact of ultra-early coiling on clinical outcome after aneurysmal subarachnoid hemorrhage in elderly patients. Acad Radiol . 2012; 19( 1): 3- 7. Google Scholar CrossRef Search ADS PubMed  9. Laidlaw JD, Siu KH. Ultra-early surgery for aneurysmal subarachnoid hemorrhage: outcomes for a consecutive series of 391 patients not selected by grade or age. J Neurosurg . 2002; 97( 2): 250- 258; discussion 247-259. Google Scholar CrossRef Search ADS PubMed  10. Phillips TJ, Dowling RJ, Yan B, Laidlaw JD, Mitchell PJ. Does treatment of ruptured intracranial aneurysms within 24 hours improve clinical outcome? Stroke . 2011; 42( 7): 1936- 1945. Google Scholar CrossRef Search ADS PubMed  11. Weil AG, Zhao JZ. Treatment of ruptured aneurysms: earlier is better. World Neurosurg . 2012; 77( 2): 263- 265. Google Scholar CrossRef Search ADS PubMed  12. Wong GK, Boet R, Ng SC et al.   Ultra-early (within 24 hours) aneurysm treatment after subarachnoid hemorrhage. World Neurosurg . 2012; 77( 2): 311- 315. Google Scholar CrossRef Search ADS PubMed  13. 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 2001–2009. AJNR Am J Neuroradiol . 2013; 34( 5): 1022- 1027. Google Scholar CrossRef Search ADS PubMed  14. Leake CB, Brinjikji W, Kallmes DF, Cloft HJ. Increasing treatment of ruptured cerebral aneurysms at high-volume centers in the United States. J Neurosurg . 2011; 115( 6): 1179- 1183. Google Scholar CrossRef Search ADS PubMed  15. Reed SD, Cramer SC, Blough DK, Meyer K, Jarvik JG. Treatment with tissue plasminogen activator and inpatient mortality rates for patients with ischemic stroke treated in community hospitals. Stroke . 2001; 32( 8): 1832- 1840. Google Scholar CrossRef Search ADS PubMed  16. Washington CW, Derdeyn CP, Dacey RG Jr, Dhar R, Zipfel GJ. Analysis of subarachnoid hemorrhage using the nationwide inpatient sample: the NIS-SAH severity score and outcome measure. J Neurosurg . 2014; 121( 2): 482- 489. Google Scholar CrossRef Search ADS PubMed  17. Sonig A, Khan IS, Wadhwa R, Thakur JD, Nanda A. The impact of comorbidities, regional trends, and hospital factors on discharge dispositions and hospital costs after acoustic neuroma microsurgery: a United States nationwide inpatient data sample study (2005-2009). Neurosurg Focus . 2012; 33( 3): E3. Google Scholar CrossRef Search ADS   18. Chyatte D, Fode NC, Sundt TM Jr. Early versus late intracranial aneurysm surgery in subarachnoid hemorrhage. J Neurosurg . 1988; 69( 3): 326- 331. Google Scholar CrossRef Search ADS PubMed  19. Siddiq F, Chaudhry SA, Tummala RP, Suri MF, 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- 677; discussion 677-678. Google Scholar CrossRef Search ADS PubMed  20. 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, 2001-2008. AJNR Am J Neuroradiol . 2011; 32( 6): 1071- 1075. Google Scholar CrossRef Search ADS PubMed  21. Kan P, Jahshan S, Yashar P et al.   Feasibility, safety, and periprocedural complications associated with endovascular treatment of selected ruptured aneurysms under conscious sedation and local anesthesia. Neurosurgery . 2013; 72( 2): 216- 220; discussion 220. Google Scholar CrossRef Search ADS PubMed  22. Cho YD, Lee JY, Seo JH et al.   Early recurrent hemorrhage after coil embolization in ruptured intracranial aneurysms. Neuroradiology . 2012; 54( 7): 719- 726. Google Scholar CrossRef Search ADS PubMed  23. Starke RM, Connolly ES Jr, Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid, Hemorrhage. Rebleeding after aneurysmal subarachnoid hemorrhage. Neurocrit Care . 2011; 15( 2): 241- 246. Google Scholar CrossRef Search ADS PubMed  24. Eddleman CS, Hurley MC, Naidech AM, Batjer HH, Bendok BR. Endovascular options in the treatment of delayed ischemic neurological deficits due to cerebral vasospasm. Neurosurg Focus . 2009; 26( 3): E6. Google Scholar CrossRef Search ADS PubMed  25. Steiner T, Juvela S, Unterberg A et al.   European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage. Cerebrovasc Dis . 2013; 35( 2): 93- 112. Google Scholar CrossRef Search ADS PubMed  26. Yuan Z, Cooper GS, Einstadter D, Cebul RD, Rimm AA. The association between hospital type and mortality and length of stay: a study of 16.9 million hospitalized Medicare beneficiaries. Med Care . 2000; 38( 2): 231- 245. Google Scholar CrossRef Search ADS PubMed  27. Khuri SF, Najjar SF, Daley J et al.   Comparison of surgical outcomes between teaching and nonteaching hospitals in the Department of Veterans Affairs. Ann Surg . 2001; 234( 3): 370- 382; discussion 382-373. Google Scholar CrossRef Search ADS PubMed  28. Rosenthal GE, Harper DL, Quinn LM, Cooper GS. Severity-adjusted mortality and length of stay in teaching and nonteaching hospitals. Results of a regional study. JAMA . 1997; 278( 6): 485- 490. Google Scholar CrossRef Search ADS PubMed  29. Hoh BL, Chi YY, Dermott MA, Lipori PJ, Lewis SB. The effect of coiling versus clipping of ruptured and unruptured cerebral aneurysms on length of stay, hospital cost, hospital reimbursement, and surgeon reimbursement at the university of Florida. Neurosurgery . 2009; 64( 4): 614- 619; discussion 619-621. Google Scholar CrossRef Search ADS PubMed  30. Brinjikji W, Kallmes DF, Lanzino G, Cloft HJ. Hospitalization costs for endovascular and surgical treatment of unruptured cerebral aneurysms in the United States are substantially higher than medicare payments. AJNR Am J Neuroradiol . 2012; 33( 1): 49- 51. Google Scholar CrossRef Search ADS PubMed  31. Hoh BL, Chi YY, Lawson MF, Mocco J, Barker FG 2nd. Length of stay and total hospital charges of clipping versus coiling for ruptured and unruptured adult cerebral aneurysms in the Nationwide Inpatient Sample database 2002 to 2006. Stroke . 2010; 41( 2): 337- 342. Google Scholar CrossRef Search ADS PubMed  32. Prabhakaran S, Fonarow GC, Smith EE et al.   Hospital case volume is associated with mortality in patients hospitalized with subarachnoid hemorrhage. Neurosurgery . 2014; 75( 5): 500- 508. Google Scholar CrossRef Search ADS PubMed  Acknowledgments We thank Paul H. Dressel, BFA, for preparation of the illustrations and Debra J. Zimmer for editorial assistance. Supplemental digital content is available for this article at www.neurosurgery-online.com. COMMENTS The authors report 17 412, directly admitted SAH patients from the NIS database. Ultra-early (<24 hours) vs >24 hours treatment for securing the aneurysm (clipping vs coiling was compared). This manuscript is important in helping to solidify what most high-volume vascular neurosurgeons “know” anecdotally and supports the practice of early treatment. In most centers, stable patients with aneurysmal SAH will be treated less than 24 hours from hospital admission. These results must be understood and interpreted carefully, however. Due to the method of data collection and reporting, a patient admitted at 11:59 pm on day 0 and treated at 7 am on day 1 (7 hours from admission), will be looked at as if treated in a more delayed fashion than the patient admitted at midnight and treated at noon the same day (12 hours from admission), based solely on the way hospital days are counted. Therefore, it is expected that some of the patients evaluated in the “early” group (hospital 0 to 1), were actually treated “ultra-early” or less than 24 hours from admission. This does muddy the results and can lead to misinterpretation. In addition, more patients treated in the ultra-early group underwent coiling procedures. We already know from ISAT (International Subarachnoid Aneurysm Trial) and BRAT (Barrow Ruptured Aneurysm Trial) that patients in coiling cohorts have shorter hospital stays and better immediate outcomes. Therefore, this data can be confounded additionally simply by the mode of treatment in the ultra-early cohort. Finally, volume data for each hospital was not available. In most cases, hospitals that have the resources to offer ultra-early or early coiling or clipping will also be high-volume centers with better through-put, more experienced surgeons and staff and better infrastructure. This too will confound this data. Overall, this was a nice manuscript with a large cohort of retrospectively reviewed patient data that suggests that treatment for ruptured intracranial aneurysms within 24 hours of presentation portends a better outcome. We agree with this in general, but also think this should be reviewed with objectivity. Mandy Binning Erol Veznedaroglu Pennington, New Jersey The authors present a retrospective review of 17 412 patients in the Nationwide Inpatient Sample with non-traumatic SAH in an effort to determine the effect of ultra-early treatment on outcome. They found that ultra-early aneurysm treatment (defined as treatment on hospital day 0) is associated with better outcome (defined as discharge home) and decreased hospital cost. Although it has been shown that early aneurysm treatment is associated with decreased incidence of rebleeding and improved outcomes, the effect of ultra-early treatment is less well described. The present study is a valuable contribution to this body of knowledge. The use of nationwide data samples is appropriate for answering simple clinical questions, such as this. Additionally, the very large number of patients adds significant power to the results. On the other hand, nuances can be lost in the data and frequently cannot be recovered. Misdiagnosis can be a problem and an example in this study would be differentiating non-traumatic and traumatic SAH. Additionally, the day of treatment may be less important than the actual number of hours between admission and treatment as a patient admitted at 11 PM and treated the following day at 1 AM was actually treated incredibly quickly yet was grouped as being treated after admission day 0. Similarly, judging good outcome as only the patients who went home neglects a large number of patients who will have excellent clinical outcomes after a short stay in acute rehabilitation. The limitations of nationwide samples should be kept in mind when interpreting studies such as this. Justin Mascitelli J. Mocco New York, New York Copyright © 2017 by the Congress of Neurological Surgeons

Journal

NeurosurgeryOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off