This retrospective study was approved by the Institutional Review Board of our center (IRB No. 2021-0663). The study included patients who underwent pterional craniotomy for the treatment of cerebral aneurysms between January 2006 and December 2020 at our center. Patients who underwent a surgical approach other than pterional craniotomy, expired within 14 days after surgery, underwent postoperative craniotomy extension due to brain swelling, had underlying coagulopathy, or underwent extracranial-intracranial or intracranial-intracranial bypass were excluded from the study.

A subgaleal suction drain was utilized in all surgeries performed before May 2013, whereas its use was discontinued after this time period. Since May 2013, a subgaleal suction drain was selectively applied based on the surgeon’s discretion regarding whether antiplatelet drugs or anticoagulants were being taken, or if there was a tendency to bleed.

Demographic characteristics were collected retrospectively from electronic medical records, including information on sex, age, weight, height, and body mass index (BMI). The presence of underlying diseases, such as diabetes mellitus (DM), hypertension, antiplatelet or anticoagulation agent intake, immunosuppressant use, smoking, and time of operation were documented as variables to identify any significant confounding variables that affect EDH requiring evacuation and wound complications. Computed tomography (CT) of the brain was routinely performed immediately after surgery and on postoperative day 3. Supplementary brain imaging was conducted as required based on clinical indications, such as the presence of severe headache, neurological decline, or indications suggesting a potential surgical site infection.

Patients were instructed to stop taking antiplatelet agents for at least 7 days and anticoagulants (including warfarin and direct oral anticoagulants) for at least 48 hours before elective surgeries for non-emergent unruptured aneurysms. Platelet transfusions were done during emergency surgeries for individuals who were receiving antiplatelet medications. Antidotes and fresh frozen plasma were administered to patients who were using anticoagulants in order to regulate the prothrombin time international normalized ratio at a normal level.

The craniotomies were conducted in the operating room under controlled laminar air flow conditions. To reduce the risk of deep vein thrombosis, an intermittent pneumatic compression device was used for all patients, and low molecular weight heparin was not administered. Prophylactic antibiotics were given within one hour before making the incision on the scalp. Cefazolin, a first-generation cephalosporin, was injected repeatedly every four hours. Ciprofloxacin was injected every eight hours in patients with cephalosporin allergy.

If a patient had a drain, antibiotics followed a set schedule until the drain’s removal, as per local protocol. However, antibiotics were discontinued within 24 hours after surgery in patients without drain insertion. Prior to elective surgeries, patients performed a hair and scalp cleansing using betadine soap on the day preceding the operation. Hair and scalp washing was not allowed prior to emergent surgeries, or to patients with betadine allergy. To prepare the scalp incision based on the surgeon’s preference, the scalp was shaved in either of the following methods: complete shaving of the frontotemporal area or partial shaving with a width of 1-2 cm along the designated incision line. After the shaving process, the incision site was cleansed using a solution containing alcohol and iodine derivatives. Subsequently, an iodine-impregnated incision drape called Ioban® (3-M, St. Paul, MN, USA) was applied to cover the prepared area. After the intradural procedure was completed, the intradural space was filled with warm saline. The dura was closed watertightly, using a continuous suture of Prolene 4-0. Subsequently, the bone flap was secured with a screw and plate system. In cases that required drainage, the Jackson-Pratt drain was positioned subgaleally, filling the space between the bone flap and the underlying temporalis muscle fascia. The drain tube’s end was situated near the drilled portion of the lateral sphenoid wing, serving to facilitate communication with the epidural space. The remote end of the drain tube was linked to the trocar, then passed through the scalp and connected to a closed 200 mL bulb suction system. Temporalis muscle was directly reattached with Vicryl 2-0 sutures. Interrupted sutures made with 3-0 Vicryl were used for both the subgaleal and subcutaneous layers. The skin was then closed with 3-0 Nylon sutures or, in some cases, with staples. Compressive dressings were applied in most of cases. Saline irrigation of surgical field was frequently performed during surgery in all patients, particularly after dura closure, bone flap closure, and temporalis muscle suture.

The operated area was inspected on a daily basis, and the wound dressings were regularly replaced. The compression bandages were taken off a day following the surgical procedure. The amount of drained fluid was measured every 8 hours and the reservoir was emptied for the drain group. If the drainage volume was less than 100 mL in the prior 24 hours and below 30 mL in the preceding 8 hours, the drains were taken out in a sterile environment. All drains were withdrawn at least after 48 hours, and the exit site was closed with a single suture. The sutures or staples placed on the scalp were typically taken out between the seventh and tenth days following the operation.

Two outcome variables were recorded: first, the occurrence of EDH requiring evacuation; and second, the occurrence of wound complications. Complications related to the wound were categorized as any of the subsequent conditions: a surgical site infection, which is evidenced by positive cultures from surgical samples, pus or clear liquid discharge from the wound accompanied by local heat, redness, sensitivity, or subdural empyema, and wound dehiscence, characterized by either spontaneous or medically induced separation of the skin layer that necessitates re-closure. Revisional surgery was carried out to manage epidural fluid accumulation or hematoma in individuals experiencing severe headaches or focal neurologic dysfunction due to the mass effects.

All statistical analyses were performed using R Version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Independent-sample t-test was used for the baseline characteristics; Wilcoxon rank sum test was used for continuous variables, and the chi-squared test or Fisher’s exact test was used for categorical variables. Univariate logistic regression analyses were conducted to identify the risk factors associated with the two outcome variables. The factors analyzed included the drain insertion, sex, age, BMI, the presence of hypertension, DM, smoking, use of antithrombotic agent, use of immunosuppressant, craniotomy side, and the duration of operation. If the univariate analyses identified factors other than drain insertion was associated with the two outcome variables with p-value <0.10, multivariate logistic analyses were performed further to identify the association with drain insertion and the two outcome variables adjusted for those variables. A p-value <0.05 was considered statistically significant.

A total of 5,280 pterional craniotomies were performed in 5,139 patients for clipping of cerebral aneurysms. Of these, 1,637 cases (31%) were assigned to the subgaleal suction drain (SGSD) group and 3,643 cases (69%) were assigned to the no drain (ND) group.

Baseline characteristics of the SGSD group and the ND group were compared. The prevalence of subarachnoid hemorrhage (SAH) was significantly higher in the SGSD group compared to the ND group (19% vs 5.5%, p<0.001). The mean age was significantly lower in the SGSD group compared to the ND group (55.9±10.0 years vs 58.6±9.57 years, p<0.001), while the mean BMI was significantly lower in the SGSD group compared to the ND group (24.5±3.2 kg/m² vs 24.8±3.3 kg/m², p=0.003). The prevalence of hypertension was significantly lower in the SGSD group compared to the ND group (48.1% vs 53.8%, p<0.001). The sex, prevalence of diabetes, smoking and the proportion of patients who received immunosuppressant agents were similar in both groups. The proportion of patients who received antiplatelet agents was similar in both groups (p=0.154), while the proportion of patients who received anticoagulant agents was significantly lower in the SGSD group compared to the ND group (p=0.005). The mean time of operation was significantly longer in the SGSD group compared to the ND group (244.4±73.3 vs 219.9±60.4 minutes, p<0.001). There was a single case with catheter-related complication, which was while the drain tube was being removed, it was caught on the edge of the bone flap, leaving a part of the distal end, which was removed through revisional surgery. The baseline characteristics between SGSD and ND group are detailed in Table 1.

Among all patients, 14 patients (0.27%) had surgical complication of EDH requiring evacuation. Eight cases (0.5% of all SGSD cases) occurred in the SGSD group and 6 cases (0.2% of all ND cases) occurred in the ND group. Fig. 1 shows the brain CT of the representative patient.

The univariate analysis showed that the presence of SAH (OR=12.57, 95% CI=4.36-38.31, p<0.001), drain insertion (OR=2.98, 95% CI=1.03-9.06, p=0.044) and time of operation (OR=1.01, 95% CI=1.00-1.01, p<0.001) were significantly associated with EDH requiring evacuation. In the multivariate analysis, the presence of SAH (OR=10.72, 95% CI=3.50-34.58, p<0.001) the time of operation (OR=1.01, 95% CI=1.00-1.01, p<0.001) was significantly associated with EDH requiring evacuation. The association between drain insertion and EDH requiring evacuation was no longer significant in the multivariate analysis (OR=1.62, 95% CI=0.53-5.21, p=0.402) (Table 2).

A total of 30 (0.57%) patients had wound complications among the 5,280 craniotomies. Fifteen cases (0.9% of all SGSD cases) occurred in the SGSD group and 15 cases (0.4% of all ND cases) occurred in the ND group. Fig. 2 indicates the brain magnetic resonance imaging (MRI) of the representative case of surgical site infections (SSIs).

The univariate analysis showed that drain insertion (OR=2.24, 95% CI=1.08-4.62, p=0.028), the presence of SAH (OR=7.27, 95% CI=3.44-14.99, p<0.001), male sex (OR=2.33, 95% CI=1.13-4.82, p=0.021), age (OR=0.96, 95% CI=0.93-0.99, p=0.021) and time of operation (OR=1.004, 95% CI=1.00-1.01, p=0.033) were significantly associated with wound complications. In the multivariate analysis, the presence of SAH (OR=6.12, 95% CI=2.79-13.18, p<0.001), male sex (OR=2.12, 95% CI=1.02-4.40, p=0.042), were found to be significant risk factors for wound complications. The association between drain insertion and wound complications was no longer significant in the multivariate analysis (OR=1.45, 95% CI=0.67-3.11, p=0.342) (Table 3).

A subgroup analysis was performed on cases with SAH. Out of the total 5,280 cases, 513 (9.72%) were aneurysmal SAH cases, and 311 SAH cases (60.6% of total SAH cases) had a subgaleal suction drain. The EDH requiring evacuation event occurred in 8 cases (1.56% of total SAH cases), and wound complications occurred in 13 cases (2.53% of total SAH cases). There were no significant differences in baseline characteristics between the two groups (Table 4). Univariate analysis showed that the use of a subgaleal suction drain was not associated with EDH requiring evacuation (OR=4.63, 95% CI=0.82-86.85, p=0.153) or wound complications (OR=3.67, 95% CI=0.97-23.85, p=0.093) (Table 5).