Results: randomised

Reproduced with permission from Hutchinson, et al.6 © 2023 Massachusetts Medical Society.

Study flow chart

Please see Appendix 4.

Recruitment

Trial enrolment started in September 2014 through to April 2019 in 40 centres across 11 countries. A total of 3566 patients were screened for eligibility, and 462 were randomised. Twelve patients were withdrawn due to lack of valid informed consent or withdrawal of consent. This resulted in a total of 228 in the craniotomy arm and 222 in the DC arm.

Numbers analysed

From the 228 patients in the craniotomy arm, 208 received the allocated intervention and 20 received a DC. Regarding the 222 patients in the DC arm, 210 received the allocated intervention and 12 received a craniotomy. The primary outcome was reported in 426 patients (215 in the craniotomy arm and 211 in the DC arm) (Table 1).

Baseline demographics

Injury and pre-hospital data

Almost all injuries were closed-head injuries (96%). The distribution of the mechanism of injury and presence of major extracranial injury was similar in the two treatment groups. The presence of pre-hospital hypoxia, hypotension and hypothermia and of cardiac arrest, seizures or clinical deterioration in the two patient groups was similar (Table 2).

Medical history

There were no major differences in medical history between the two treatment groups (Table 3).

Primary outcome: Glasgow Outcome Scale-Extended

Death occurred in 30.2% in the craniotomy group versus 32.2% in the craniectomy group; vegetative state occurred in 2.3% versus 2.8%, and good recovery occurred in 25.6% versus 19.9%, respectively. Non-statistically significant better outcomes were seen in patients undergoing craniotomy initially than in those undergoing DC, with a common OR across the GOSE scores of 0.85 [95% CI (0.6 to 1.18, p = 0.324)] (Figure 1 and Table 4).

FIGURE 1 Primary end point: stacked bar chart of GOSE results at 12 months. Reproduced with permission from Hutchinson et al.6 © 2023 Massachusetts Medical Society. (Figure not reproduced here; see the original article.)

Secondary outcomes: Glasgow Outcome Scale-Extended, efficacy and safety

FIGURE 2 Stacked bar chart of GOSE results at 6 months. Reproduced with permission from Hutchinson et al.6 © 2023 Massachusetts Medical Society. (Figure not reproduced here; see the original article.)

Additional data randomised arm

Computed tomography scan results

The characteristics of the CT scans were very similar in the two treatment groups.

The distribution of pre-intubation GCS and the pupil reactivity was similar in the two treatment groups (see Appendix 3).

Surgery details and postoperative destination

There was a small difference in the location of the ASDH between the treatment groups. Patients in the craniotomy group were more likely to have the ASDH located on the right, while those in the DC group were more likely to have the ASDH located on the left.

There were no differences between the treatment groups in terms of the type of skin incision, with 84% reverse question mark, or in the size of the bone flap [with a mean of 12.9 cm (± 1.7 cm)].

A drain was used in 85% of patients and was usually subgaleal 70% rather than subdural (15%). Intracranial monitoring devices were used in 51% of patients, with the majority being parenchymal and most (69%) being located ipsilateral to the ASDH.

For the patients who underwent DC, the bone flap was either stored inside the body (40%) or disposed (340).

There was no difference in the occurrence of intraoperative complications (χ² p = 0.82) or the destination from the operating theatre (χ² p = 0.89).

Treatment received

Most patients received the treatment they were allocated to. The main reason given for patients crossing over from the craniotomy group was due to intraoperative swelling. The main reason given for patients crossing over from the DC group was because the brain was relaxed (see Appendix 4).

Discharge from the intensive care unit

There was no difference in mortality in ICU (χ² p = 0.789), and survivors were no more likely go to anywhere other than the NSU (χ² p = 0.207).

For those that died in ICU, there was no difference by treatment group in the length of time they spent in ICU (Mann–Whitney U-test p = 0.823). For those that were discharged from ICU, there was no difference in their GCS score at discharge (Mann–Whitney U-test p = 0.884), in whether their pupil reactivity was positive or negative in either eye (right χ² p = 0.484, left χ² p = 0.492) or in their LOS (Mann–Whitney U-test p = 0.680).

Please see Appendix 5 for table.

Neurosurgical interventions within 14 days

These are recorded on the discharge from NSU case report form (CRF). The information is not recorded for the patient who did not go from the operating theatre into the ICU or NSU or for a further 69 patients (35 craniotomy and 34 DC) who initially went into ICU (including most of the patients who died in ICU or were discharged to a different hospital). Please see Appendix 6 for table.

Discharge from neurosurgical unit

Patients were admitted to the NSU either directly from operating theatre (56 craniotomy, 56 DC) or after a stay in the ICU (126 craniotomy, 128 DC) but excludes cases who died in the ICU or were discharged directly from the ICU (Table 8).

Length of stay in intensive care unit and neurosurgical unit

Figure 3 is a survival plot of LOS in ICU (deaths are censored). The mean LOS for craniotomy is 12.8 days (95% CI 10.8 to 14.9), and for craniectomy, 13.0 days (95% CI 11.2 to 14.9). The median LOS for craniotomy is 10 days (95% CI 8.6 to 11.4), and for DC, 10 days (95% CI 9.0 to 11.0). Log-rank p = 0.681.

FIGURE 3 Randomised data: survival plot of LOS in ICU. Reproduced with permission from Hutchinson et al.6 © 2023 Massachusetts Medical Society. (Figure not reproduced here; see the original article.)

Figure 4 demonstrates that the median LOS in the NSU is 18 days and does not differ by treatment group (craniotomy 17 days, DC 18 days). Log-rank p = 0.51.

FIGURE 4 Randomised data: survival plot of median LOS in NSU. Reproduced with permission from Hutchinson et al.6 © 2023 Massachusetts Medical Society, (Figure not reproduced here; see the original article.)

Patients who died in hospital are censored for this analysis.

Eighty-two patients died in hospital after a mean of 14.8 days (median 8.5 days). No difference in time to death in hospital (Mann–Whitney U-test p = 0.831) (note – this will be further described using Kaplan–Meier). The median LOS for patients who were discharged from this hospital alive (to other hospitals, rehab, home nursing homes, etc.) was 15 days. There was no difference in LOS (Mann–Whitney U-test p = 0.831).

Therapy intensity level

The median number of days of ICP-directed therapy was 8 days in the craniotomy group and 10 days in the DC group (p = 0.128, Mann–Whitney U-test).

Please see Appendix 7 for table.

Cranioplasty and shunts for hydrocephalus

Adverse events: randomised

Extracranial adverse events

Two hundred and seventy AEs are reported for 113 (50%) of the 228 patients randomised to craniotomy. The maximum number of events reported for a single patient was 9, and the average number of events reported by those randomised to craniotomy was 1.18. Two hundred and eighty-nine AEs were reported for 104 (47%) of the 222 patients randomised to DC. The maximum number of events reported for a single patient in this group was 6, and the average number of events reported by a patient randomised to DC was 1.30. The types of AEs experienced between the two treatment groups were similar.

Please see Appendix 8 for AEs and procedure-related complications tables.