Joeky Senders

26 Chapter 2 for 30 days after their surgery and includes data from over 600 hospitals across the United States. 16 This validated dataset is collected by trained surgical reviewers at each site using a standardized protocol and includes the most common postoperative complications, length of stay, occurrence of reoperations and readmissions together with associated reasons, and mortality. 17 The NSQIP registry has previously been used to study surgical outcomes after neurosurgical procedures. 6,8,17-26 Our institutional review board has exempted this study from review. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were used in this study. Inclusion criteria Patients were included who met the following criteria: 1) aged 18 years or older; 2) Current Procedural Terminology (CPT) code(s) indicating craniotomy for surgical resection of brain tumors (CPT: 61500, 61510, 61512, 61518, 61519, 61520, 61521, 61526, and 61530); 3) postoperative diagnosis indicative of primary malignant brain tumor according to International Classification of Diseases, Ninth Revision [ICD-9]: 191.x. CPT codes are medical codes maintained by the American Medical Association to communicate uniform information and billing on medical, surgical, and diagnostic procedures. Similarly, ICD-9 codes are also used for uniform communication; however, ICD-9 codes refer to diagnoses, whereas CPT codes refer to medical services. Covariates Covariates were extracted and analyzed as follows. Age, body mass index (BMI), and operative time were examined continuously. American Society of Anesthesiologists (ASA) classification was examined categorically (I-II, III, IV-V). Other evaluated preoperative patient characteristics included: gender, race, smoking status during the past year, dyspnea, chronic obstructive pulmonary disease (COPD), hypertension requiring medication, bleeding disorders, diabetes mellitus (insulin dependent vs. non-insulin dependent), steroid usage, recent congestive heart failure, preoperative functional status (dependent vs. independent), and systemic sepsis. Preoperative laboratory values were also extracted and categorized according to standard reference ranges and clinical significance. When ‘elevated’ or ‘decreased’ laboratory groups consisted of a very low number of cases or when its association with the independent variables did not significantly differ from the normal group, this category was deemed to be clinically non-significant and merged with the ‘normal’ group. This reduces the degrees of freedom and preserves statistical power, resulting in the following categories: creatinine (<1.4 mg/dL vs. ≥1.4 mg/dL), hematocrit (<36% vs. ≥36%), platelet count (100,000-450,000/µL, <100,000/µL, or >450,000/µL), sodium (135-145 mEq/L, <135 mEq/L, or >145 mEq/L), and white blood cell count (WBC, ≤12,000/µL or >12,000/