Patients and methods

Between 1995 and 1998, 122 consecutive patients with chronic subdural haematoma were treated in our institution. Patients with an acute subdural haematoma—that is, patients presenting with symptoms within three days of trauma—and patients with subdural hygroma containing xanthochromic or thin bloody fluid were excluded from the evaluation. Four patients were treated by a different therapy regimen and were, therefore, also excluded. The remaining group of 118 patients consisted of 71 men (60%) and 47 women (40%), ranging in age from 19 to 95 years (mean 67 years).

To facilitate comparison with previously reported studies, patients were assessed with the most commonly used neurological grading system for CSDH devised by Markwalder et al 19 as follows:

Grade 0: patient neurologically normal

Grade 1: patient alert and oriented; mild symptoms, such as headache; absent or mild symptoms or neurological deficit, such as reflex asymmetry

Grade 2: patient drowsy or disoriented with variable neurological deficit, such as hemiparesis

Grade 3: patient stuporous but responding appropriately to noxious stimuli; severe focal signs, such as hemiplegia

Grade 4: patient comatose with absent motor response to painful stimuli; decerebrate or decorticate posturing.

The initial diagnosis was confirmed by CT, and treatment was initiated if the thickness of the haematoma exceeded that of the skull and if the patient was symptomatic—that is, at least grade 1.

All patients were treated at the bedside by percutaneous subdural tapping under local anaesthesia. After partial shaving and disinfection, a stab incision of the skin was performed at the site of maximum thickness of the CSDH, preferably anterior to the coronal suture. Twist drill trephination was performed using a 3 mm hand driven drill. After perpendicular penetration of the skull, the dura and the outer neomembrane of the haematoma were perforated with a needle. Thereafter, a 14G teflon cannula was perpendicularly introduced, exceeding the tabula interna of the skull only slightly. The haematoma was evacuated only by spontaneous efflux and Valsalva's manoeuvre with the patient positioned in a 30° Trendelenburg position with the head placed so that the cannula was situated at the deepest point of the haematoma cavity (fig 1). After cessation of CSDH efflux, the cannula was removed and the skin was closed by a single suture with the patient still in the 30° Trendelenburg position. The amount of haematoma efflux was quantified and immediately postoperatively the patient was examined clinically. A control by CT was performed within the next 48 hours. If the thickness of the residual haematoma still exceeded that of the skull and if the patient was still symptomatic, the procedure was repeated. If there was no septation visible on CT, the repeated tapping was performed through the same burr hole. If there was septation visible or if the amount of haematoma efflux was below 10 ml, repeated tapping was performed through a new burr hole. If repeated subdural tappings did not improve neurological grade to at least grade 1, secondary treatment by continuous drainage after burr hole trephination and/or membranectomy after craniotomy was initiated. The patients were discharged the day after the last subdural tapping if their signs and symptoms were reduced at least to grade 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Positioning of the head for evacuation of chronic subdural haematoma by spontaneous efflux. (A) Overview. (B) Partial magnification. The patient is placed in the 30° Trendelenburg position and the head is placed such that the teflon cannula is situated at the deepest point of the haematoma cavity. To prevent injury to the brain the cannula is introduced so that it exceeds the tabula interna of the skull only slightly. 1=haematoma cavity; 2=outer neomembrane; 3=dura; 4=scull; 5=skin.

After discharge, the patients were controlled clinically and by CT examination within 6 weeks, and whenever there was clinical suspicion of haematoma reaccumulation. An indication for readmission to hospital and reoperation was made if the patients showed typical signs and symptoms and reaccumulation of the haematoma exceeding the thickness of the skull. Before completion of this study, in all patients the referring physicians were asked for a re-evaluation.

Results

Table 1 shows the patients of the present series, classified according to their preoperative neurological grade secondary to the grading system by Markwalder et al.19

The aetiology of CSDH and the distribution of concomitant diseases are shown in table 2. In the group of patients with known head trauma in the past, the mean duration from trauma to therapy was 6 weeks, ranging from 0.5 to 20 weeks.

The CSDH was right sided in 40 patients (34%), left sided in 59 patients (50%), and bilateral in 19 patients (16%) (fig 2). Septation was visible on the first CT in 25 patients (21%) (fig 3). On initial CT the mean thickness of the CSDH was 2 cm, ranging from 1 to 3.5 cm. The density of the CSDH on first CT is shown in table3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Serial CT of a 66 year old man with bilateral chronic subdural haematoma who presented with left sided hemiparesis and headache. (A) Scan obtained before treatment, demonstrating bilateral hyperdense haematoma. (B) Scan obtained 1 month after first subdural tapping,—that is, after three right sided and three left sided subdural tappings, demonstrating reduction of the—now hypodense—haematoma. After the sixth subdural tapping the patient was free of signs and symptoms. (C) Scan obtained at follow up 4 months after the last subdural tapping, demonstrating complete re-expansion of the brain.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Serial CT of an 83 year old man with a left sided chronic subdural haematoma who presented with right sided hemiparesis and headache. The patient was treated by eight subdural tappings without sufficient haematoma evacuation and without permanent improvement of neurological status. Therefore, he was finally treated by craniotomy, evacuation of the haematoma, and continuous catheter drainage. (A) Scan obtained before treatment, showing septation of the isodense haematoma (arrows). (B) Scan obtained 1 day after first subdural tapping through a left frontal twist drill craniostomy, demonstrating slight reduction of haematoma size and postoperative intracranial air. Note air bubbles at the septa. (C) Scan obtained after eight subdural tappings (17 days after first subdural tapping), still demonstrating the—now hypodense— haematoma, although it is somewhat reduced in size.

The number of subdural tappings for each patient is shown in fig 4. The mean number of subdural tappings in the 99 patients with unilateral CSDH was 2.8, ranging from 1 to 10; the mean number in the 19 patients with bilateral CSDH was 5.0, ranging from two to 11. Altogether, the mean number of subdural tappings in all 118 patients was 3.3. Ninety one out of the 99 patients with unilateral CSDH (92%) were treated by up to five tappings and 18 out of the 19 patients with bilateral CSDH (95%) were treated by up to 10 subdural tappings. Altogether, 377 subdural tappings through 225 twist drill craniostomies were performed. In 66 out of the 377 procedures (18%) no or only up to 10 ml haematoma efflux was achieved. Immediately after every subdural tapping the neurological status ameliorated in 253 procedures (67%), worsened in two procedures (1%), and remained stable in 115 procedures (31%).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Number of treated patients (99 patients with unilateral and 19 patients with bilateral CSDH) in relation to the number of subdural tappings in each patient. The figure includes two patients with unilateral CSDH, acutely operated on via craniotomy and haematoma evacuation because of procedure related acute subdural bleeding after the first subdural tapping. Further, two patients with unilateral CSDH are included, operated on via craniotomy and evacuation of empyema and CSDH. Finally, seven patients are included, operated on via continuous drainage and/or membranectomy because of insufficient haematoma evacuation after 1, 2, 3, 4, 4, 8, and 8 subdural tappings, (see also table 4).

In five out of the 118 patients (4%)—that is, in five out of 377 subdural tappings (1.3%)—acute subdural bleeding and in one patient (1%) intracerebral bleeding was seen on early postoperative CT. No correlation between these haemorrhagic complications and the surgeon‘s experience could be demonstrated. Two patients (2%) had to be operated on acutely due to a procedure related acute subdural bleeding with acute worsening of the neurological status. Seven patients (6%) were finally operated on using continuous drainage after burr hole trephination and/or membranectomy after craniotomy because of insufficient haematoma evacuation and failure to improve in neurological status (fig 3). Subdural empyema occurred in three patients (3%) after five, six, and 10 subdural tappings, resulting in operative revision in two patients (2%). Thus 11 patients in total (9%) were finally treated by a different protocol (table4).

Symptomatic postoperative pneumocephalus was not seen in any patient. No patient died as a consequence of the operative procedure, one patient (1%) died during her hospital stay from stroke due to cardiogenic embolism.

The mean duration of inpatient treatment was 12 days with a range from 1 to 35 days. Apart from the one patient (1%) who died, 75% of the patients were free of signs and symptoms after the last subdural tapping, the remaining 24% presented with only mild signs and symptoms—that is, grade 1 according to the grading system of Markwalder et al.19Postoperative neurological grading after the last subdural tapping and at the time of follow up evaluation, mean 22 (range 1–46) months after the last subdural tapping is shown in table 1. Fifteen patients (13%) died within the follow up period. However, death was related in all cases to concomitant diseases such as heart failure or malignant tumour and not to a recurrence of CSDH.

Statistical analysis showed a highly significant correlation (χ² 18.556; Fisher’s exact test, p<0.001) between septation visible on preoperative CT and patients who had finally to be treated by a changed treatment protocol because of insufficient haematoma evacuation. No other significant predisposing factors for unsuccessful treatment were found.

Discussion

Even though there is general agreement that the treatment of CSDH should be surgical, the extent of surgical therapy is still under debate.5-31 However, it has been demonstrated that increased surgical invasiveness does not guarantee better results.14 28 29 For example, the worst surgical results were reported after extensive craniotomy and membranectomy.28 29

The current theory about persistence, enlargement, and healing of CSDH is that there is local hyperfibrinolysis with a large concentration of fibrinolytic enzymes in the outer neomembrane and in the subdural fluid.33-36 Further, endothelial gap junctions of macrocapillaries in the outer neomembrane potentially lead to leakage of blood substances into the membrane and the haematoma cavity.37-39 Finally, it is the ratio between rebleeding and reabsorption of the subdural collection which determines whether the CSDH will resolve, persist, or increase.32 36 The improvement of local hyperfibrinolysis by reducing the amount of fibrinolytic enzymes through surgical removal of the CSDH leads to a tightening of the outer neomembrane and to a decrease in repetitive haemorrhage from the macrocapillaries, in the case of healing resulting in progressive increase of collagen and concurrent reduction of macrocapillaries.32 38

Thus Aoki and Drapkin advocated surgical removal of CSDH as completely as possible.5 32 By contrast, other authors reported that complete drainage of the haematoma seemed to be unnecessary.9 12 19 20 The probable pathophysiological basis is that incomplete haematoma evacuation may change the ratio between rebleeding and reabsorption towards reabsorption and, thus initiate a self healing process.

Consequently, there have been reports on various surgical attempts to treat CSDH radically on the one hand, but as minimally invasive as possible on the other hand. In 1968, Burton reported the management of CSDH by twist drill craniostomy and aspiration of the subdural haematoma.8 In 1975, Negrón et al reported on seven patients with subdural haematoma, five of whom had been successfully treated by bedside needle trephination.22 Hubschmann reported 22 cases of CSDH treated by twist drill craniostomy and closed system drainage.17 In his series, one patient required repeated twist drill evacuation of the haematoma, another patient required craniotomy to achieve total haematoma removal, and in two patients the procedure failed.17 Aoki presented a study of 24 patients with CSDH treated by subdural tapping and spontaneous evacuation of the haematoma.5 However, because of a recurrence rate of 29% he changed to subdural tapping and irrigation, resulting in a recurrence rate of 7%.5 Finally, Burchiel and Taylor treated 131 patients by serial twist drill craniostomy and haematoma aspiration.7 In their study, 73% of the patients achieved significant neurological improvement, 23% required an additional operative procedure.7

The therapy protocol used here is similar to that already described.5 7 8 22 However, to our knowledge, no prospective study on the efficacy of this form of treatment has been documented.

The epidemiological data of the present series are similar to other published series dealing with the treatment of CSDH,1-17 19-32 40 so that the results are comparable. To evaluate the described therapy regimen according to the duration from trauma to therapy, we subsumed CSDH and so called subacute subdural haematoma—that is, haematoma more than 3 weeks old and haematoma between 3 days and 3 weeks old.40

In accordance with other authors,9 12 19 20 we found that it is unnecessary to remove the subdural collection completely. A partial reduction of the CSDH may suffice to change the imbalance of rebleeding and reabsorption towards reabsorption and, thus, healing, may be due to an improvement of local hyperfibrinolysis or due to a reduction of tensile forces on the macrocapillaries. However, a considerable reduction in volume is necessary to initiate this self healing process, as Ito et al have shown that between 6 and 24 hours after infusion of labelled erythrocytes, haemorrhage into the haematoma cavity accounted for a mean of 6.7% of the haematoma content.34 Correspondingly, in the present series an indication for reoperation was made if the remaining or reaccumulating haematoma exceeded the thickness of the skull and the patient was symptomatic.

The rationale for treating CSDH by bedside percutaneous tapping after twist drill craniostomy under local anaesthesia without continuous drainage or irrigation is that it offers a less invasive treatment alternative to other presently available surgical treatment protocols. In the treatment protocols with drainage by placing the subdural drain obliquely and deeply into the haematoma cavity through one or more burr holes, there is always the danger of creating acute subdural bleeding or of damaging the brain by uncontrolled placement of the drain.26 27 31 Further, by placing a drain into the haematoma cavity and by irrigating the subdural space, there is the risk of not only damaging superficial cerebral vessels or bridging veins, but also of damaging macrocapillaries of the outer neomembrane of the CSDH, potentially resulting in acute subdural bleeding or reaccumulation.26 27 31 There is also the risk of damage to fragile vessels and the neomembrane by creating negative subdural pressure due to evacuation of the CSDH by suction and even by gravity forces in continuous subdural drains. Thus, in the present series evacuation of the CSDH was achieved only by spontaneous efflux and Valsalva's manoeuvre with the patient placed in the 30° Trendelenburg position. Further, continuous catheter drainage and irrigation carry the risk of infection.20 23 26 27 31Wound infection or meningitis occurred in 18% of the patients reported by Smely et al.26 We tried to reduce the risk of infection by immediate removal of the subdural cannula after cessation of spontaneous efflux of the haematoma in every subdural tapping. Finally, in all surgical techniques for treating CSDH there is the risk of postoperative pneumocephalus, potentially resulting in increased intracranial pressure.6 23 41 42The risk of pneumocephalus is reduced by the described treatment protocol, as the skin is closed immediately after cessation of spontaneous blood efflux with the patient still in the 30° Trendelenburg position.

However, here also the therapy regimen does not overcome some of the shortcomings and risks of conventional surgical therapy of CSDH:

By repeated subdural tapping after twist drill craniostomy there is the risk of damaging the cerebral surface with each tapping, resulting in five acute subdural bleedings and one intracerebral bleeding in our series, requiring acute craniotomy and haematoma evacuation in two patients (table 4). However, this rate is comparable with other series, which also reported procedure related acute bleedings.26 27 31 This risk can be reduced by the use of a twist drill that does not penetrate the skull43 44and by careful placement of the perpendicularly introduced teflon cannula only slightly exceeding the tabula interna of the skull.

By repeated subdural tapping, the risk of wound infection and subdural empyema increases, resulting in three subdural empyemas after 10, six, and five subdural tappings in the present series, requiring reoperation via craniotomy in two patients (table 4). However, this rate of wound infection and subdural empyema is similar to other published series, ranging between 1% and 18%.12 14-16 23 24 26 28 31 The risk of infection should be reducible by defining a maximal number of subdural tappings before changing to another therapy protocol, as in the present series in the two patients requiring craniotomy and evacuation of the subdural empyema, infection occurred after 10 and six subdural tappings for unilateral CSDH (table 4). Thus, in our opinion, the maximal number of subdural tappings should not exceed five in unilateral CSDH and 10 in bilateral CSDH, as in the present series 92% and 95% of the patients, respectively, were successfully treated by five and10 subdural tappings.

The risk of pneumocephalus is reduced by the treatment protocol, resulting in no symptomatic postoperative pneumocephalus in the present series. Therefore, subdural tapping should be performed in the 30° Trendelenburg position with Valsalva's manoeuvre to prevent air from penetrating into the subdural space.

The 6% rate of patients who had to be treated finally by continuous drainage and/or membranectomy because of insufficient haematoma evacuation and failure to improve in neurological status is similar to the reported rate of reoperation in other series, ranging between 2% and 37% of recurrence and reoperation.5 6 10-12 14-16 19 21 24-28 31 32However, these patients can potentially be identified by extensive septation visible on preoperative CT before the procedure, as septation was the only significant predictor for unsuccessful treatment in the present series. There were no other significant correlations found, especially not between the need for changing the therapy regimen and the thickness of the CSDH on preoperative CT, by contrast with the report of Ram and colleagues and in accordance with Stoobandt and colleagues.23 27

The mean duration of inpatient treatment of 12 days in the present series is comparable with other recently reported surgical treatment regimens which range between 4.9 and 23 days of mean hospital stay.7 9 11 12 15 21 24 26 29 32 However, it has to be considered that in the present series many patients were in a poor general condition, so called subacute subdural haematoma were included, and that 19 patients with bilateral CSDH required bilateral therapy. Using the described treatment protocol, the duration of hospital stay should also be reducible by defining the above mentioned maximal number of subdural tappings as five in unilateral and 10 in bilateral CSDH, and to change to another therapeutic regimen after that number is reached without significant improvement of the neurological status. This may be the compromise between less invasiveness on the one hand and maximum effectiveness on the other.

Finally, the clinical outcome of our patients with (apart from one patient who had died during her hospital stay from stroke due to cardiogenic embolism)75% of the patients being free of signs and symptoms and the remaining 24% of the patients presenting with only mild signs and symptoms after the last subdural tapping is at least as good as in other, more invasive, modern surgical therapy protocols.6 9-12 14-17 19 21 24-27 29 31 32

Conclusions

The ideal surgical treatment of CSDH has not been definitely defined. However, in the present prospective series the less invasive surgical technique of bedside percutaneous subdural tapping and spontaneous haematoma efflux after twist drill craniostomy under local anaesthesia has resulted in satisfactory results. It can be performed on the ward with a complication rate as low as other modern surgical techniques. We recommend it in all patients as a first and minimally invasive attempt, especially in patients in poor general condition. Patients with extensive septation visible on preoperative CT should be excluded. If the procedure fails after five subdural tappings in unilateral CSDH and after 10 subdural tappings in bilateral CSDH (five subdural tappings on each side), the regimen should be changed to prevent a rise in surgical morbidity. However, secondary treatment by continuous drainage after burr hole trephination and/or membranectomy after craniotomy—which are much more invasive and carry a higher procedure related morbidity should be performed only if the described initial treatment protocol fails.