There is no established, objective, quantitative method and evaluation guide for surgical drainage in the literature. In neurosurgery, limited research has been conducted on using a drainage tube [3–9]. Su et al. [4] believe that the prophylactic negative pressure closed drainage under a myocutaneous flap cannot reduce the subcutaneous and epidural bleeding after a pterion craniotomy. Swollen muscles seriously affect the CT data for measuring haematomas. It is challenging to drain the blood out oozing from the muscle and fascia by drainage placed under the myocutaneous flap.
According to Zhang et al. [3], “ED cannot reduce the occurrence of epidural haematoma and subcutaneous hydrops in patients with supratentorial epilepsy”. The aim of epileptic surgery is mainly to remove the epileptic focus, which assists with internal decompression. It is different from the mechanism of meningioma and glioma. Less exudate and swelling can be observed. Therefore, epileptic surgery is also inappropriate to evaluate the efficacy of drainage.
Low intracranial pressure caused by negative pressure suction can cause a series of serious problems, including brain swelling, multiple epidural haematomas, sinus bradycardia, and cardiac arrest, all of which are mentioned in the literature [5–7, 9]. Therefore, in our study, the ED bag was placed at the level of the external auditory canal, without negative pressure suction. CSF is naturally induced by brain beat and pressure gradient. Therefore, the intracranial pressure of patients in DG is stable. No complications as mentioned above occurred in our study.
On analysing the above literature, it was clear that the designs had particular patient selection biases, or the selection of SD was not appropriate, especially regarding negative pressure suction. Therefore, our study was designed specifically for the problems of SFC, STH, and secondary infection. In our study, the site of the craniotomy was chosen as our focus. The whole cranium is supposed to be divided into two parts by a line depending on how much muscle is attached and whether it can be compression bandaged (Fig. 1). Near the parietal site, it is easier to form SFC and not STH. But near the “Sports hair band site”, it is easier to form STH than SFC.
The conditions and possible mechanisms for the formation of STH are as follows: deep and large tumour, large exudation of the cavity, the small opening of the cortex surface, no communication between the tumour cavity or ventricle, and a suture of the dura mater that is too tight. These lead to the failure of hypertonic fluid to enter into the CSF circulation through the subarachnoid space on the cortex surface or ventricle. Further, it increases the local cerebral pressure and aggravates the brain oedema. As a result, the outlet of the cavity or ventricles is narrower, and hypertonic fluid cannot flow out, leading to vicious circulation and the formation of STH. In our study, the hypertonic fluid will likely flow through the non-watertight dura and drain before the STH happens. This is preferable than removing more cortex tissue [10] for a larger CSF opening.
The results of our study have shown that the incidence of SFC in NDG (54.1%) was significantly higher than that in DG (11.5%). Further treatment, such as repeated scalp puncture, aspiration, and compressed bandage, may lead to the risk of a secondary wound infection and/or intracranial infection, or even debridement and osteotomy. It affects wound healing and brings unnecessary economic and pain burden to patients, especially in children. On the contrary, appropriate drainage could drain the haematogenous CSF. It can make the scalp and skull more attached, help the wound heal, and reduce the patients’ pain. On this basis, considering the local relative high pressure caused by exudation of the tumour cavity and swelling of the surrounding brain tissue, our study innovatively proposes the method of non-watertight dura suturing. This can make the CSF slowly transfer from the intracranial relative high-pressure site to the extradural low-pressure site by using extradural drainage through the small gap of the dura. Therefore, the occurrence of STH is low in our study.
This may be due to the craniotomy location being near the parietal site. Due to the influence of the large veins in the midline, most patients cannot achieve a watertight suture. The CSF flows out along the pressure gradient. Then, SFC occurs. In addition, repeated puncture and aspiration plays a role in the subcutaneous drainage to a certain extent. Therefore, SFC is common in NDG, and STH is rare in both groups. The craniotomy near the “Sports hair band site” was excluded from our study, making it easier to evaluate the role of the drainage tube.
In our study, STH occurred in a special patient in DG, which proved the mechanism from the opposite side (Fig. 2). This procedure confirmed the importance of effective liquid draining. The non-watertight suture of the dura mater is as important as the drainage tube as a part of the CSF drainage system and plays a key role in reducing the occurrence of STH, which is often ignored. This precise drainage system includes the pressure gradient from the inside to the outside of dura. Relying on watertight dura suturing and bandage pressure [3], sometimes cannot effectively prevent SFC but can also lead to STH and scalp necrosis, especially in children.
The selection of ED or SD needs to be further considered. The SD was easier to lead the haemorrhage out in the cavity, avoiding the formation of a haematoma. So most articles focused on SD. Many complications were reported, especially with negative pressure suction, including pseudo hypoxic brain swelling [6], subdural haematoma along with drainage tube which was removed [8], bradycardia [5], cardiac arrest [7], and extradural haematoma [5, 11]. The cases in these articles were analysed one by one, and the reasons are summarised as follows: (1) Excessive drainage of CSF leads to low intracranial pressure. Tumour cavity drainage with negative pressure suction is widely used in general surgery, orthopaedics, and plastic surgery [12–14]. However, the intracranial cavity is different from a soft tissue cavity. It is enclosed and connected with the whole CSF circulation. It characteristically has significant fluidity. Therefore, with the outflow of CSF, the pressure will be decreased. The fluidity means that the intracranial pressure tends to average with the pressure gradient. Then, if negative pressure continues, low intracranial pressure symptoms will occur. (2) The mechanism of a subdural haematoma during extubation may be as follows: (a) rupture of a vein in the cavity. With the slow CSF extraction, the haemostatic materials such as gauze enter into the micro-hole at the head of the drainage tube. Small vessels are winded with the haemostatic materials, rupture, and haemorrhage during extubation. (a) The vein on the brain surface near the drainage tube is ruptured by friction. (b) The bleeding of the scalp artery flows into the epidural or subdural areas due to inexperience. Incomplete haemostasis of the small scalp artery during craniotomy is the reason.
Based on the above analysis, the SD is inappropriate, especially with negative pressure suction. We promote ED and a non-watertight dura suture, which is safe and effective. Of course, this is based on satisfied haemostasis of the tumour cavity. The details are as follows: (1) The dura is non-watertight sutured. (a) Dural reconstruction is necessary to avoid the low intracranial pressure caused by the over drainage of CSF and to prevent a small amount of extradural blood from flowing into the cranial. (b) Non-watertight dura suturing is important, for an unobstructed drainage system. Specifically, the gap between the sutured dura help to let the fluid flow out slowly, which ensures that the subdural hypertonic fluid flows with the pulsation of the brain. Intracavity pressure can also be regulated by the outflowing of the CSF. Therefore, it is similar to a one-way external drainage system. (2) Place ED tube. The extradural drainage tube is separated from the brain tissue by the dura. It can avoid direct contact with the tumour cavity and the vein on the brain surface, to avoid subdural haemorrhage while the tube is being removed. The ED tube can not only drain the extradural exudation but also outflow the subdual haematogenous CSF slowly to avoid STH and SFC. Of course, it is important to standardise the manipulation of placing and removing the drainage tube, to prevent iatrogenic bleeding of the scalp arterial from flowing into the epidural or subdural space.
Infection is an important concern during placing a drainage tube. Previous literature has shown that it is not caused by the haematoma, but the internalisation of skin bacteria. Long-term drainage increases morbidity and even mortality [15–18]. However, in neurosurgery, only extra ventricular drainage may cause an intracranial infection [19]. In our study, the drainage tube was located in the epidural space and removed on the 2nd POD. As long as attention is paid to aseptic operation, local infection rarely occurs. It was concluded that the incidence of intracranial, wound, and lung infection was low in DG. One abscess in the cavity and skull occurred in a case in NDG, which was resolved with debridement and external drainage. The infection rate in DG was less than in NDG, due to the following reasons: (1) It is related to the haematogenous CSF transferred to the extracranial region through a single outward drainage system, which is caused by the haematogenous exudation of the tumour cavity and the degradation of haemostatic materials. The average exudation of patients in DG was 300–400 ml. The effect of this process was similar to that of an intermittent lumbar puncture. (2) The occurrence of SFC in DG was significantly reduced, which meant that the chances of needing another scalp puncture to aspirate the fluid were also reduced, thereby reducing the risk of scalp infection and fever. (3) The decrease of haematogenous CSF stimulation, steady body temperature, the stability of the intracranial pressure, led to the reduction of postoperative symptoms, and infusion treatment. The overall state of the patients improved faster. Removing ED on the 2nd POD in DG decreased the likelihood of complications from long-term bed rest, such as lung infection and lower extremity thrombosis.
The occurrence of epidural haematoma is closely related to the thoroughness of haemostasis because haematoma solidifies quickly and is difficult to drain. Therefore, drainage has little effect, which is consistent with previously reported findings [3, 4].
The prevalence of infection, patients’ body temperature, prolongation of hospital stay, and postoperative medicine fee were studied. It was found that those in DG were all better than those in NDG. However, there was no statistical significance. Our results initially suggested that the placement of extradural drainage perhaps reduces the length of stay and expenses.
The patients selected had a craniotomy performed near the parietal site, which is close to the cerebral functioning area. We, therefore, also statistically analysed the improvement of muscle strength and the control of their epileptic seizures. Muscle strength improvement in DG was better than that in NDG, but there was no statistical difference. The control of epileptic seizures in DG was significantly better than that in NDG. The improvement of epilepsy is perhaps not only related to the decompression of the local lesion but also the reduction of haematogenous CSF, the intracranial pressure, and brain oedema.