Original Article

Craniotomy for Resection of Meningioma: An Age-Stratified Analysis of the MarketScan

Longitudinal Database
Ian D. Connolly1, Tyler Cole1, Anand Veeravagu1, Rita Popat2, John Ratliff1, Gordon Li1

- OBJECTIVE: We sought to describe complications after - CONCLUSIONS: In this study, we report complication
resection for meningioma with the use of longitudinal rates in patients undergoing resection for meningioma.
administrative data, which our group has shown recently to Because of the longitudinal nature of the MarketScan
be superior to nonlongitudinal administrative data. database, we were able to capture a wide array of specific
postoperative complications associated with meningioma
- METHODS: We identified patients who underwent
resection procedures. Care should be taken in the selec-
resection for meningioma between 2010 and 2012 in the
tion of candidates for meningioma resection.
Thomson Reuters MarketScan database. Current Proce-
dural Terminology coding at inpatient visit was used to
select for meningioma resection procedure. Comorbidities
and complications were obtained by use of the Interna-
tional Classification of Diseases, Ninth Revision or Current
Procedural Terminology coding. Associations between INTRODUCTION
complications and demographic and clinical factors were
evaluated with logistic regression.
- RESULTS: We identified a total of 2216 patients.
Approximately 41% developed 1 or more perioperative
M eningiomas are benign neoplasms that arise from the
dural layers covering the brain and spinal cord. They
account for up to 30% of all primary intracranial tumors
and typically are slow-growing.1 Many are asymptomatic and
frequently are discovered incidentally on magnetic resonance
complications. Approximately 15% were readmitted within imaging.2 Resection is indicated when the tumor becomes
30 days of their procedure. The most frequent complica- symptomatic or grows significantly in size over a limited time
tions that occurred in our cohort were new postoperative period. It is expected that the meningiomas will be diagnosed
seizures (11.8%), postoperative dysrhythmia (7.9%), intra- more frequently as the average human life expectancy continues
cranial hemorrhage (5.9%), and cerebral artery occlusion to increase and as diagnostic modalities continue to improve.
(5.4%). General neurosurgical complications and general This, coupled with the inherent risks associated with resection,
neurologic complications occurred in 4.4% and 16.1% of presents issues with regard to choice of operative therapy.
Several studies have reported widely varying results on operative
patients, respectively. Nearly 55% of elderly patients (‡70
outcomes.3-12
years) developed 1 or more perioperative complication (vs. We sought to undertake a national, longitudinal database study
39% of nonelderly patients). After we adjusted for comor- investigating operative outcomes of meningioma resection to
bidities, elderly status and male sex were found to be further describe the risks and benefits associated with this pro-
significantly associated with increased odds for a variety cedure. We used the Thomson Reuters MarketScan database to
of complications. investigate demographics, complications, and postoperative

Key words From the 1Department of Neurosurgery, Stanford University School of Medicine, Stanford,
- Adverse events California, USA; and 2Division of Epidemiology, Department of Health Research and Policy,
- Craniotomy Stanford University, Stanford, California, USA
- MarketScan To whom correspondence should be addressed: Gordon Li, M.D.
- Meningioma [E-mail: [email protected]]
Citation: World Neurosurg. (2015) 84, 6:1864-1870.
Abbreviations and Acronyms http://dx.doi.org/10.1016/j.wneu.2015.08.018
CPT: Current Procedural Terminology
Journal homepage: www.WORLDNEUROSURGERY.org
ICD-9: International Classification of Diseases, Ninth Revision
NIS: Nationwide Inpatient Sample Available online: www.sciencedirect.com
1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.

1864 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.08.018

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

governmental organizations, and public organizations. It captures

Table 1. CPT and ICD-9 Codes Used to Select Cohort, patient-specific longitudinal data independent of treatment loca-
Comorbidities, and Complications tion through claims from hospitals, outpatient centers, surgeons,
CPT or ICD-9 Code Number physicians, and independent laboratories.

Benign neoplasm of the cerebral 225.2 Cohort Selection

meninges (ICD-9) Supratentorial and infratentorial meningioma resection procedures
Deep vein thrombosis (ICD-9) 451, 453.4, 453.8, 453.2, 453.1, 453.9 were selected by the use of Current Procedural Terminology (CPT)
code 61512 or 61519 with concurrent International Classification of
Delirium (ICD-9) 293
Diseases, Ninth Revision (ICD-9) code 225.2 (Table 1). Only patients
Dysphagia (ICD-9) 787.2 having 1 of CPT codes 61512 (supratentorial meningioma) or 61519
Dysrhythmia (ICD-9) 427, 426.1, 426, 426.3, 426.4, 426.5, (infratentorial meningioma) with a concurrent ICD-9 code of 225.2
426.6, 426.7, 426.8 were included in the study. A total of 127 patients did not contain
General neurological 430, 431, 432, 433, 434, 435, both of the CPT and ICD-9 codes mentioned and were excluded
complication (ICD-9) 436, 438.2, 438.3, 438.4, 438.5 from the analysis. Altogether, a total of 2216 patients met our final
inclusion criteria. Patients of age greater or equal to 70 years were
General neurosurgical 997.0
considered elderly and those of age younger than 70 years were
complication (ICD-9)
considered nonelderly. Overall, we identified 239 elderly patients
Iatrogenic stoke (ICD-9) 997.02 and 1977 nonelderly patients who met our final inclusion criteria.
MI (ICD-9) 410, 412, 998.0, 997.1, 411, 429.7
Osteoporosis (ICD-9) 733, V17.81, 731.3, V82.81
Outcomes Analyzed
The primary outcome analyzed was the overall, aggregated (“any”)
Other wound complication (ICD-9) 998.3, 998.81, 998.83, 998.4 complication rate. Secondary outcomes analyzed were mortality,
Post-op infection (CPT) 10060, 10140, 10180, 12020, total payment, general neurologic complications, general neuro-
12021, 20005, 21501, 22010, 22015 surgical complications, and the additional complications listed in
Post-op infection (ICD-9) 998.0, 998.5, 999.3 Table 2. All the aforementioned outcomes were within 30 days
of the initial meningioma resection procedure. To avoid
Pulmonary complication (ICD-9) 997.3, 518.4, 518.5, 518.7, 518.81,
attributing preexisting conditions as new onset, comorbidities
518.82, 518.83, 518.84, 518.89,
519.1
and complications were only considered new if the patient did
not have a history of the corresponding outcome at any time in
Pulmonary embolism (ICD-9) 415.1 the past. Patient demographic characteristics assessed in our
Seizure (ICD-9) 345, 780.3 study included sex, age, and comorbid conditions. ICD-9-
Supratentorial meningioma 61512, 61519 Clinical Modification codes for determining comorbidities were
resection (CPT) based on those used in the Elixhauser and Deyo-Charlson co-
morbidity indices (Table 2).13,14 ICD-9-Clinical Modification codes
Thromboembolism (ICD-9) 453.0, 453.4, 453.8,453.2, 453.3
were used to identify drug use, hypertension, congestive heart
Tobacco use (ICD-9) 305.1, V15.82, 989.84, 649.0 failure, chronic obstructive pulmonary disease, myocardial
Wound dehiscence (ICD-9) 998.3 infarction, neurologic dysfunction, diabetes, and obesity. Addi-
tional comorbid conditions used in this study and not included in
Wound hematoma (ICD-9) 998.1
these indices were tobacco use and osteoporosis. The codes used
CPT, Current Procedural Terminology; ICD-9, International Classification of Diseases, to identify these conditions are listed in Table 1.
Ninth Revision.
Statistical Analysis
Data preparation and analysis were performed with SAS software
mortality rates in patients undergoing resection for meningioma (version 9.3; SAS Institute, Inc., Cary, North Carolina, USA). Odds
with a minimum 30-day follow up. ratios were calculated by use of the nonelderly cohort as the
reference group. A Fisher exact test was used to investigate the
METHODS differences between elderly and nonelderly patients. A P value of
0.05 was considered significant. For analysis of predictor variables
Data Source of any complications, a stepwise forward logistic regression model
We performed a retrospective, observational administrative data- was used.
base study of patients in the United States who underwent
resection for meningioma. The data were obtained through the RESULTS
Thomson Reuters MarketScan Commercial Claims and Encoun-
ters and Medicare Supplemental and Coordination of Benefits Patient Demographics and Comorbidities
databases from 2010 through 2012. This database contains data Results are summarized in Table 2. The cohort identified was
from more than 100 payers and includes inpatient, outpatient, and composed primarily of women (79.2%), supratentorial tumors
pharmacy services from a range of large employers, health plans, (88.2%), and nonelderly patients younger than 70 years of age

WORLD NEUROSURGERY 84 [6]: 1864-1870, DECEMBER 2015 www.WORLDNEUROSURGERY.org 1865

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

Table 2. Patient Demographics and Comorbidities

Odds Ratio (95%
Demographics Total N [ 2216 Age ‡70 years N [ 239 Age < 0 years N [ 1977 confidence interval) P value

Age, mean (SD) 54.61 (12.74) 77.3 (4.7) 51.9 (10.5) < 0.0001
Follow-up, mean, (SD) 337 (222) 316 (212) 340 (223) 0.1110
Length of stay (nights), mean (SD) 4.87 (6.17) 7.1 (7) 4.6 (6) < 0.0001
Medicare, n (%) 405 (18.28) 239 (100) 166 (8.40) < 0.0001
Male, n (%) 639 (28.84) 94 (39.33) 545 (27.57) 0.20 (0.15e0.27) < 0.0001
Supratentorial (%) 1954 (88.18) 224 (93.72) 1730 (87.51) 2.13 (1.24e3.66) 0.0039
Discharge home, n (%) 1812 (81.77) 128 (53.55) 1684 (85.18) 0.20 (0.15e0.27) < 0.0001
Comorbidities, n (%)
Tobacco use 120 (5.49) 8 (3.4) 112 (5.7) 0.57 (0.27e1.18) 0.1313
Osteoporosis 353 (16.14) 65 (27.3) 288 (14.8) 2.17 (1.59e2.96) < 0.0001
Hypertension (Elixhauser) 1026 (46.91) 179 (75.2) 847 (43.5) 3.95 (2.9e5.37) < 0.0001
CHF (Charlson) 79 (3.61) 24 (10.1) 55 (2.8) 3.86 (2.34e6.37) < 0.0001
MI (Charlson) 41 (1.87) 17 (7.1) 24 (1.2) 6.17 (3.26 e11.66) < 0.0001
COPD (Charlson) 285 (13.03) 46 (19.3) 239 (12.3) 1.71 (1.21e2.43) 0.0041
COPD (Elixhauser) 280 (12.8) 45 (18.9) 235 (12.1) 1.70 (1.2e2.42) 0.004
Diabetes (Elixhauser) 352 (16.1) 76 (31.9) 276 (14.2) 2.84 (2.11e3.84) < 0.0001
Obesity (Elixhauser) 177 (8.09) 15 (6.3) 162 (8.3) 0.74 (0.43 e1.28) 0.3158
Region 0.0227
Northeast 445 (20.08) 46 (19.20) 399 (20.20)
North Central 553 (24.95) 71 (29.70) 482 (24.40)
South 757 (34.16) 69 (28.90) 688 (34.80)
West 404 (18.23) 52 (21.80) 352 (17.80)
Unknown 57 (2.57) 1 (0.40) 56 (2.80)

CHF, congestive heart failure; MI, myocardial infarction; COPD, chronic obstructive pulmonary disease.

(89.2%). In total, the average patient age was 54.6 years (SD, 12.7). (IQR, $28,390e$57,904), whereas the median 90-day total post-
Patients had a mean duration of follow-up of 11.2 months and a discharge payment was $3506 (IQR, $1128e$16,207) (Table 3). The
mean duration of stay of 4.87 days. The mean duration of stay was majority of patients in the cohort were discharged home (81.8%).
significantly greater in elderly patients compared with nonelderly The most frequent comorbid conditions were hypertension
patients (7.1 vs. 4.6, P < 0.0001). All elderly patients were enrolled (Elixhauser) (46.9%), osteoporosis (16.1%), diabetes (Elixhauser)
under Medicare (vs. 8.4% of nonelderly patients). The median (16.1%), and chronic obstructive pulmonary disease (Charlson)
total payment for the initial admission for resection was $39,297 (13.0%).

Table 3. Procedure Payments (USD)

Characteristic Total Age ‡70 years Age <70 years Difference (%) P value

Hospital payments, median (IQR) 30,770 (20,398e45,672) 26,170 (15,036e40,584) 31,157 (21,220e46,528)
4987 (
0.19) 0.0463
Physician payments, median (IQR) 4727 (3241e6723) 3047 (2550e3791) 5013 (3505e7087)
1967 (
0.65) < 0.0001
Total payments, median (IQR) 39,297 (28,390e57,904) 33,145 (19,125e48,667) 40,405 (29,539e58,624)
7259 (
0.22) 0.0014
90 day postdischarge payments, median (IQR) 3506 (1128e16207) 4498 (1460e20,005) 3402 (1091e15,852) 1096 (0.24) 0.9772

IQR, interquartile range.

1866 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.08.018

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

Table 4. 30-Day Postoperative Complications

Odds Ratio (95%
Complication Total Age ‡70 years Age <70 years confidence interval) P Value

Any complication 898 (41.1) 130 (54.6) 768 (39.4) 1.85 (1.41e2.43) < 0.0001
All-cause readmission within 30 days 338 (15.45) 30 (12.6) 308 (15.8) 0.77 (0.51e1.15) 0.2174
Mortality, n (%) 19 (0.86) 4 (1.43) 15 (0.63) 2.29 (0.75e6.94) 0.158
Wound infection 48 (2.19) 3 (1.3) 45 (2.3) 0.54 (0.17e1.75) 0.4781
Wound dehiscence 14 (0.64) 2 (0.8) 12 (0.6) 1.37 (0.3e6.15) 0.6588
Wound hematoma 94 (4.3) 7 (2.9) 87 (4.5) 0.65 (0.3e1.42) 0.395
Other wound complication 23 (1.05) 2 (0.8) 21 (1.1) 0.78 (0.18e3.34) 1.00
General neurosurgical complication 97 (4.44) 20 (8.4) 77 (4) 2.23 (1.34e3.72) 0.0039
Iatrogenic stroke 63 (2.88) 14 (5.9) 49 (2.5) 2.42 (1.32e4.46) 0.0068
General neurological complication 353 (16.14) 48 (20.2) 305 (15.6) 1.36 (0.97e1.91) 0.0765
Intracranial hemorrhage NOS 128 (5.85) 32 (13.4) 96 (4.9) 3 (1.96e4.59) < 0.0001
Precerebral arterial occlusion 23 (1.05) 3 (1.3) 20 (1) 1.23 (0.36e4.17) 0.7324
Cerebral artery occlusion 118 (5.4) 24 (10.1) 94 (4.8) 2.21 (1.38e3.54) 0.002
Transient ischemia attack 25 (1.14) 2 (0.8) 23 (1.2) 0.71 (0.17e3.03) 1.00
Acute complication NOS 50 (2.29) 11 (4.6) 39 (2) 2.37 (1.2e4.7) 0.0188
Hemiplegia/hemiparalysis 39 (1.78) 11 (4.6) 28 (1.4) 3.32 (1.63e6.77) 0.0021
Subarachnoid hemorrhage 46 (2.1) 9 (3.8) 37 (1.9) 2.03 (0.97e4.26) 0.0871
Delirium 35 (1.6) 11 (4.6) 24 (1.2) 3.89 (1.88e8.04) 0.0008
Postoperative seizure 259 (11.84) 38 (16.0) 221 (11.3) 1.49 (1.02e2.16) 0.0431
Postoperative dysphagia 57 (2.61) 13 (5.5) 44 (2.3) 2.5 (1.33e4.72) 0.0078
Postoperative dysrhythmia 172 (7.86) 33 (13.9) 139 (7.1) 2.1 (1.4e3.15) 0.0008
Postoperative MI 64 (2.93) 15 (6.3) 49 (2.5) 2.61 (1.44e4.73) 0.0033
DVT 95 (4.34) 16 (6.7) 79 (4.1) 1.71 (0.98e2.97) 0.0635
Any thromboembolism 61 (2.79) 15 (6.3) 46 (2.4) 2.78 (1.53e5.07) 0.0024
Pulmonary complication 184 (8.41) 39 (16.4) 145 (7.4) 2.44 (1.66e3.57) < 0.0001
Pulmonary embolism 20 (0.91) 2 (0.8) 18 (0.9) 0.91 (0.21e3.94) 1.00

NOS, not otherwise specified; MI, myocardial infarction; DVT, deep-vein thrombosis.

Postoperative Outcomes hemorrhage (5.9%), cerebral artery occlusion (5.4%), and deep-
Postoperative outcomes are summarized in Table 4. Overall, the vein thrombosis (4.3%). Elderly patients generally displayed
30-day mortality rate was 0.86% (n ¼ 19). No significance differ- increased frequencies of complications compared with nonelderly
ence was detected in this rate for elderly versus nonelderly pa- patients. The risk of developing intracranial hemorrhage or
tients. Approximately 41.1% of all patients developed 1 or more hemiplegia/hemiparalysis was nearly 3 times as large in elderly
complications. Elderly patients had a complication rate of nearly patients. Likewise, the risk of developing postoperative delirium
55%, which was statistically significant compared with nonelderly was nearly 4 times as large.
patients (39%, P  0.0001). The overall rate of all cause 30-day
readmission was 15.5%, which did not differ significantly be- Multivariate Analysis
tween elderly and nonelderly patients. General neurosurgical, Using a stepwise logistic regression model, we examined age
neurologic, and pulmonary complications occurred in 4.4%, group, sex, and all previously identified comorbidities included in
16.1%, and 8.4% of patients, respectively. The most frequent in- Table 2 as predictor variables for the development of any
dividual complication measured was postoperative seizure, which complication, 30-day readmittance, general neurosurgical
occurred in 11.8% of patients. Other frequent complications complication, general neurologic complication, intracerebral
included postoperative dysrhythmia (7.9%), intracranial hemorrhage, cerebral artery occlusion, pulmonary complication,

WORLD NEUROSURGERY 84 [6]: 1864-1870, DECEMBER 2015 www.WORLDNEUROSURGERY.org 1867

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

adults, Bartek et al.3 showed that patient age, duration of surgery

Table 5. Stepwise Logistic Regression greater than 4 hours, and Karnofsky Performance Status score of
Odds Ratio (95% less than 70 to be significant independent predictors of serious
confidence interval) complications. They reported a 7% overall rate of development
of severe complications, which was defined as requiring
Any complication admission to the intensive care unit or resulting in death.
Age 70 years 1.69 (1.28e2.23) However, the study was limited to a Scandinavian population
seen at 3 neurosurgical centers, and no specific complications
Male 1.46 (1.21e1.76)
were reported.
Diabetes (Elixhauser) 1.34 (1.06e1.70) To our knowledge, only 3 studies thus far have used large-scale
30-day readmittance national databases to address the dilemma of surgical interven-
tion for intracranial meningiomas in elderly patients. Patil et al.8
Age 70 years 0.65 (0.43e0.98)
analyzed a total of 1281 patients by using data from the National
Male 1.48 (1.16e1.90) Surgical Quality Improvement Program, a prospectively collected
Hypertension (Elixhauser) 1.47 (1.16e1.86) database from 123 Veterans Affairs hospitals across the United
General neurosurgical complication
States. Not surprisingly, their cohort was predominantly
composed of male patients (98% in the elderly cohort and 94% in
Age 70 years 1.94 (1.13e3.31) the nonelderly cohort). They reported an increased rate of 1 or
Male 2.13 (1.40e3.24) more complications in the elderly cohort (29.8% vs. 13.1%). This
Infratentorial 3.05 (1.86e5.00) finding is less than our reported rates of 55% in elderly patients
and 39% in nonelderly patients. One possible source of this
Hypertension (Elixhauser) 1.74 (1.13e2.70) discrepancy could be attributed to the sex discrepancies between
General neurological complication these 2 studies. Our study population, which consisted of
Male 1.34 (1.05e1.70) approximately 70% women, more accurately reflects the sex
distribution of meningioma, which affects women more than men.
COPD (Charlson) 0.66 (0.45e0.97)
The remaining 2 studies both used data from the Nationwide
Intracerebral hemorrhage Inpatient Sample (NIS) database. Bateman et al.4 identified 2304
Age 70 years 2.84 (1.85e4.36) patients 70 years of age and older and 6557 patients younger than
the age of 70 years who underwent resection for meningioma.
Male 1.59 (1.10e2.30)
Similar to our study, their population consisted of more women
Cerebral artery occlusion than men. Their reported duration of stay for elderly (7.2 days)
Age 70 years 1.79 (1.10e2.90) and nonelderly groups (4.5 days) was similar to our findings.
Interestingly, their reported rate of adverse outcome (death or
Male 1.59 (1.08e2.34)
discharge to a facility other than home) in elderly patients (53.2%)
Pulmonary complication was similar to our rate of 1 or more complications whereas the
Age 70 years 1.94 (1.30e2.90) rate in nonelderly patients was approximately half our rate of one
or more complications. In our study, we believe that analyzing a
Diabetes (Elixhauser) 1.88 (1.32e2.68)
wider array of complications provides more informative and
Postoperative seizure granular insights into perioperative outcomes because various
Male 1.37 (1.04e1.80) other factors outside of the variables assessed also may influence
discharge to a facility other than home, such as age and family
Infratentorial 0.31 (0.17e0.57)
social structure.
COPD, chronic obstructive pulmonary disease. Finally, the study by Grossman et al.,6 who also used the NIS
database, excluded patients younger than 65 years of age. They
did, however, conclude that each year beyond 65 years of age was
and postoperative seizure. Adjusted results are summarized in predictive of increased rates of inpatient mortality, duration of
Table 5. Elderly status, sex, and diabetes were found to be stay, and complications. In this study, the authors limited their
significantly associated with developing any complication. analysis to neurologic deficits, stroke, respiratory complications,
Elderly status and male sex were the most consistent statistically and cardiac complications. Their reported rates of neurologic
significant variables and were included in 6 and 7 of the 8 deficits (5.7%), stroke (5.3%), and respiratory complications
stepwise logistic regression models respectively. (4.8%) were markedly lower than our corresponding outcomes for
elderly patients.
Although the aforementioned studies that used the NIS data-
DISCUSSION base had larger patient numbers, recent evidence sheds light on
Several studies have reported outcomes of meningioma resection an important limitation of nonlongitudinal retrospective databases
in the elderly with varying results. Most of these studies are such as the NIS database. Because the NIS database contains no
retrospective and tend to support an increased surgical risk in an patient-level identifiers, patients are unable to be followed over
elderly population. Most recently in a retrospective study of 979 time. Complications that arise after initial hospitalization for a

1868 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.08.018

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

procedure are thus unable to be captured. To replicate this limi- Limitations

tation, Veeravagu et al.15 investigated the ability of the MarketScan There are several inherent limitations of our study. Because our
database to capture the rate of spine surgery complications at the data were derived from a retrospective database, we were limited
initial date of the procedure as well as at 5, 10, 20, and 30 days to the predetermined variables that were captured as well as the
after operation. They observed increasing rates of complications information defined by each clinical code. Thus, we were unable
over time, especially among wound-related complications, sug- to assess several other important clinical variables, such as tumor
gesting that important outcomes are missed in nonlogitudinal size, location, and preoperative performance status. Another
retrospective databases. important piece of clinical information that we were unable to
Although our reported complication rates for elderly patients assess because of the limitations of diagnostic coding was the
are greater than the corresponding outcomes in the study of severity of complication. Although some complications such as
Grossman et al.,6 it is difficult to directly compare outcomes postoperative myocardial infarctions or iatrogenic strokes are
because of our limited numbers of elderly patients; however, we inherently severe and emergent, other complications that we
believe that our reported outcomes for nonelderly patients are captured may fall under a wider range of severity. This is partic-
perhaps more accurate than the corresponding outcomes ularly important to consider for complications such as dysphagia,
derived from the NIS. This finding is consistent with our dysrythmia, and pulmonary complications, which may range in
markedly greater overall complication rate in nonelderly patients severity from mild to significant.
compared with the study of Bateman et al.4 (39.4% vs. 16.6%). Demographic characteristics such as patient ethnicity were
It has been estimated that 15%e20% of patients develop epi- similarly unavailable to us. It is also important to note that mor-
lepsy after a supratentorial craniotomy.16 Although reports of tality after discharge may not have been adequately captured as
incidence rates of postoperative seizures in patients without this is only captured through inpatient MarketScan records.
preoperative seizures varies, evidence does suggest that a non- Further, MarketScan is primarily a database of privately insured
negligible portion of this population is affected.17,18 The subject patients. Because the majority of Medicare patients do not contain
of anticonvulsant prophylaxis for these patients still remains a additional private coverage, elderly and nonelderly patients were
widely debated topic. In 2005, more than 70% of polled neuro- distributed unevenly in our cohort. Similarly, because uninsured
surgeons reported routine use of antiepileptic drugs despite patients are not included in the database our study population may
studies showing limited support for its effectiveness.19 Our not be truly reflective of the actual population.
detected postoperative seizure incidence of 11.8% highlights that
this complication is an important concern for patients
undergoing meningioma resection. However, recent meta- CONCLUSIONS
analyses do not support the effectiveness of perioperative anti- Patients who undergo resection for meningioma are at risk of
seizure prophylaxis and advise against its use.20,21 developing a variety of complications. As the incidence of me-
Interestingly, our study shows one of the lowest rates of mor- ningiomas continues to increase, the practicing neurosurgeon
tality (0.86%) in the literature. However, this is potentially the must adequately weigh the risks and benefits of intervention. This
result of the MarketScan database not adequately capturing mor- is especially important for patients who display minimal symp-
tality after discharge because this is only captured in inpatient toms. Previous studies based on nonlongitudinal data may have
MarketScan records. This discrepancy also may be the result of the underestimated operative risks associated with meningioma
lack of uninsured patients in our cohort because these patients are resection. Despite our study’s limitations, we believe that our re-
not included in the MarketScan database. As previously mentioned ported results may more accurately reflect the true population due
mortality tends to vary widely between studies, likely because of to the longitudinal nature of the MarketScan database. We believe
varying study populations and study limitations. this is especially true for the nonelderly cohort that we identified.

5. Boviatsis EJ, Bouras TI, Kouyialis AT, Improvement Program. J Neurol Neurosurg Psy-
REFERENCES Themistocleous MS, Sakas DE. Impact of age on chiatr. 2010;81:502e505.
complications and outcome in meningioma sur-
1. Wiemels J, Wrensch M, Claus EB. Epidemiology
gery. Surg Neurol. 2007;68:407e411. 9. Poon MT, Fung LH, Pu JK, Leung GK. Outcome
and etiology of meningioma. J Neurooncol. 2010;9:
comparison between younger and older patients
307e314.
6. Grossman R, Mukherjee D, Chang DC, et al. undergoing intracranial meningioma resections.
Preoperative Charlson comorbidity score pre- J Neurooncol. 2013;114:219e227.
2. Vernooij MW, Ikram MA, Tanghe HL, et al. dicts postoperative outcomes among older intra-
Incidental findings on brain MRI in the general cranial meningioma patients. World Neurosurg.
10. Rogne SG, Konglund A, Meling TR, et al. Intra-
population. N Engl J Med. 2007;357:1821e1828. 2011;75:279e285. cranial tumor surgery in patients >70 years of age:
is clinical practice worthwhile or futile? Acta Neurol
3. Bartek J, Sjåvik K, Förander P, et al. Predictors of Scand. 2009;120:288e294.
7. Konglund A, Rogne SG, Lund-Johansen M,
severe complications in intracranial meningioma Scheie D, Helseth E, Meling TR. Outcome 11. Roser F, Ebner FH, Ritz R, Samii M, Tatagiba MS,
surgery: a population-based multicenter study. following surgery for intracranial meningiomas in Nakamura M. Management of skull based me-
World Neurosurg. 2015;83:673e678. the aging. Acta Neurol Scand. 2013;127:161e169. ningiomas in the elderly patient. J Clin Neurosci.
2007;14:224e228.
4. Bateman BT, Pile-Spellman J, Gutin PH, 8. Patil CG, Veeravagu A, Lad SP, Boakye M.
Berman MF. Meningioma resection in the elderly: Craniotomy for resection of meningioma in 12. Schul DB, Wolf S, Krammer MJ, Landscheidt JF,
nationwide inpatient sample, 1998-2002. Neuro- the elderly: a multicentre, prospective analysis Tomasino A, Lumenta CB. Meningioma surgery in
surgery. 2005;57:866e871. from the National Surgical Quality the elderly: outcome and validation of two

WORLD NEUROSURGERY 84 [6]: 1864-1870, DECEMBER 2015 www.WORLDNEUROSURGERY.org 1869

 ORIGINAL ARTICLE
IAN D. CONNOLLY ET AL. CRANIOTOMY FOR RESECTION OF MENINGIOMA

proposed grading scores systems. Neurosurgery. 17. Mahaley MS, Dudka L. The role of anticonvulsant 21. Sayegh ET, Fakurnejad S, Oh T, Bloch O, Parsa AT.
2011;70:555e565. medications in the management of patients with Anticonvulsant prophylaxis for brain tumor sur-
anaplastic gliomas. Surg Neurol. 1981;16:399e401. gery: determining the current best available evi-
13. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical dence. J Neurosurg. 2014;121:1139e1147.
comorbidity index for use with ICD-9-CM adminis- 18. Ramamurthi B, Ravi B, Ramachandran V. Con-
trative databases. J Clin Epidemiol. 1992;45:613e619. vulsions with meningiomas: incidence and sig-
nificance. Surg Neurol. 1980;14:415e416.
14. Elixhauser A, Steiner C, Harris DR, Coffey RM.
Comorbidity measures for use with administrative 19. Siomin V, Angelov L, Li L, Vogelbaum MA. Re-
data. Med Care. 1998;36:8e27. sults of a survey of neurosurgical practice patterns
regarding the prophylactic use of anti-epilepsy Received 10 July 2015; accepted 19 August 2015
15. Veeravagu A, Cole TS, Azad TD, Ratliff JK. drugs in patients with brain tumors. J Neurooncol.
Citation: World Neurosurg. (2015) 84, 6:1864-1870.
Improved capture of adverse events after spinal 2005;74:211e215.
surgery procedures with a longitudinal adminis- http://dx.doi.org/10.1016/j.wneu.2015.08.018
trative database. J Neurosurg Spine. 2015;23:374e382. 20. Kong X, Guan J, Yang Y, Li Y, Ma W, Wang R. Journal homepage: www.WORLDNEUROSURGERY.org
A meta-analysis: do prophylactic antiepileptic
Available online: www.sciencedirect.com
16. Foy PM, Copeland GP, Shaw MD. The incidence drugs in patients with brain tumors decrease the
of postoperative seizures. Acta Neurochir (Wien). incidence of seizures? Clin Neurol Neurosurg. 2015; 1878-8750/$ - see front matter ª 2015 Elsevier Inc.
1981;55:253e264. 134:98e103. All rights reserved.

1870 www.SCIENCEDIRECT.com WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.08.018