- Email: [email protected]
Outcome of head injuries in general surgical units with an off-site neurosurgical service
Injury, Int. J. Care Injured (2007) 38, 576—583
www.elsevier.com/locate/injury
Outcome of head injuries in general surgical units with an off-site neurosurgical service§ B.M.W. Zulu, T.V. Mulaudzi, T.E. Madiba *, D.J.J. Muckart Department of Surgery, University of KwaZulu-Natal and King Edward VIII Hospital, Durban, South Africa Accepted 2 January 2007
KEYWORDS Head injury; Neurosurgical unit; General surgical unit
Summary Introduction: In order to cope with bed shortages in the only neurosurgical unit (NSU) in KwaZulu-Natal, it has become necessary to manage head injured patients in a general surgical unit (GSU) at the referral hospitals in consultation with the NSU. This study was undertaken to assess the outcome of patients with head injuries managed in a GSU in consultation with a regional NSU. Patients and methods: A prospective pilot study was carried out in a single surgical unit over an 18-month period (July 1997—December 1998), followed by a larger prospective study over a 6-month period (January—June 2001). All patients with head injuries severe enough to warrant admission to hospital (GCS 15 or 15 with localising signs) were included. Results: The pilot study comprised 86 patients and the subsequent study comprised 230 patients, giving a total of 316 patients in the whole study. The mean age was 31 12 years. Following consultation with the NSU, 265 (84%) patients were managed in the GSU and 51 (16%) required transfer to the NSU. Forty-one patients died giving a mortality rate of 13%. Twenty eight of the 42 patients with GCS 8 died (67%) compared to 13 out of 274 with GCS >8 (5%) ( p < 0.0001). Eleven of the 51 patients destined to the NSU died (22%) compared to 30 of 265 in the GSU (11%) ( p = 0.046). The average hospital stay was 10 18 days. Forty surviving patients (17%) from the GSU and 20 from the NSU (51%) were discharged with neurological sequelae ( p < 0.001). Conclusion: Head trauma is associated with high morbidity and mortality. Nonsurgical treatment of traumatic brain injury at the referral hospital by the GSU is
§
Presented at the Congress of the Association of Surgeons of South Africa on 6—10 August 2004, Cape Town, South Africa. * Corresponding author at: Department of Surgery, University of KwaZulu-Natal, Private Bag 7, Congella 4013, South Africa. Tel.: +27 31 260 4257; fax: +27 31 260 4389. E-mail address: [email protected] (T.E. Madiba). 0020–1383/$ — see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2007.01.002
Outcome of head injuries in GSU with NSU consultation
577
acceptable practice. Outcome is determined primarily by the GCS on presentation. NSU patients had a significantly higher mortality rate. A delay before surgery did not seem to affect outcome. # 2007 Elsevier Ltd. All rights reserved.
Introduction Trauma is the second most common cause of death in South Africa and the commonest cause in children (age 5—14) and young adults (age 15—34).15 Head injuries account for the majority of deaths, being marginally more common than torso trauma.14,15 Consequently an overwhelming number of head injured patients are treated in South African hospitals. Ideally, head injured patients should be managed in a neurosurgical unit (NSU) where they can receive expert neurosurgical care. In South Africa, however, NSU’s are in short supply owing to budgetary constraints; consequently the majority of these patients are managed in peripheral hospitals in consultation with a NSU via CT Telemedicine.16 KwaZulu-Natal province occupies the east coast of South Africa with a population of about 8 million, which represents about 20% of the national population. The Medical Faculty of the University of KwaZulu-Natal is situated in Durban, the main city of KwaZulu-Natal. This province has only one public sector NSU, formerly situated at Wentworth Hospital (WWH) at the time of the study, and now situated at the new Inkosi Albert Luthuli Central Hospital. WWH is situated 8 km from King Edward VIII Hospital (KEH), the main teaching hospital of the University of KwaZulu-Natal Medical School, which houses the general surgical wards. This NSU also serves communities as far south as the province of the Eastern Cape. Due to bed shortage this NSU has had to establish guidelines for referral to the NSU as it is incapable of accommodating all patients with head injury. Although the early transfer of appropriate patients to a NSU substantially reduces morbidity and mortality, inappropriate referrals may overburden not only such specialised units but also the transport services. Two studies were undertaken to assess the outcome of patients with head injuries managed in the GSU of KEH in consultation with a regional off-site NSU situated in WWH, namely a pilot study in July 1997—December 1998 and the main study in January—June 2001. The results of the two studies have been combined and analysed as one study.
Patients and methods The surgical service at KEH consists of three wards, each with two surgical units, giving a total of six
surgical units. The pilot study was undertaken in a single surgical unit at King Edward VIII Hospital over an 18-month period from July 1997 to December 1998. The second study was carried out in all six units over a 6-month period from January to June 2001. King Edward VIII Hospital (KEH) is one of the main referring hospitals to the NSU in WWH. The large period that elapsed between the two studies was unavoidable owing to lack of human resources. All patients with head injury severe enough to warrant admission to hospital were enrolled in the study. Indications for admission to hospital in the Durban Metropolitan Hospital services are (i) a period of loss of consciousness, (ii) presence of neurological deficit, and (iii) all patients with Glasgow coma scale (GCS) of <15. Indications for urgent CT scan on admission are a GCS 14 or any level of consciousness associated with neurological deficit. Indications for performance of a semi-elective CT scan during hospitalisation include (i) any deterioration in the level of consciousness, (ii) development of neurological deficit, or (iii) failure of improvement of the GCS for more than 48 h even in the absence of a fracture or focal signs. Patients with a GCS 4 from the time of injury and those that were hypotensive or hypoxic were resuscitated and reevaluated before consultation. Severe head injury was defined as GCS 8. All trauma patients, including those with head injury are admitted to an acute ward where triage is carried out. Patients requiring referral to the NSU are transferred directly from this unit. The rest are kept in the acute ward for a period of 24 h prior to being transferred to the general surgical unit (GSU). The criteria for triage of patients were admission GCS and CT scan. The CT scan was sent to the NSU by telemedicine and results were discussed with the neurosurgeon prior to an appropriate decision being made. Transfer of the patient to the NSU occurred only after the neurosurgeon had agreed to receive the patient. Additional criteria were patients with a compound injury and/or brain oozing. These were transferred immediately to the NSU. All fractures with deeply in-driven bone fragments, contaminated wounds, brain or cerebrospinal fluid oozing, retained foreign body or heavy bleeding underwent emergency debridement. Patients with intracranial pathology underwent surgery when they fulfilled one or more of the following
578
B.M.W. Zulu et al.
Table 1 Patient profile Pilot study
Main study
Whole study
Total number of patients Male M:F
86 74 (86%) 6:1
230 192 (83%) 5:1
316 266 (84%) 5:1
Age (years) Range Mean
16—65 31 11
12—80 31 13
12—80 31 12
GSU NSU
64 (74%) 22 (26%)
201 (87%) 29 (13%)
265 (84%) 51 (16%)
GCS >8 8
76 (88%) 10 (12%)
198 (86%) 32 (14%)
274 (87%) 42 (13%)
criteria: (i) size of more that 30 cm2, (ii) midline shift of more than 10 mm, and (iii) effacement of the ipsilateral ventricle. Patients with a GCS of 15 with no localising signs received clinical observation. Patients with intracranial pathology but no mass effect were managed non-operatively. An intracranial pressure monitor was inserted under the following circumstances: (i) patients undergoing surgery for evacuation of space occupying lesion (except for a pure extradural haematoma) and (ii) patients with evidence of increased intracranial pressure on CTscan but no mass lesion. The external ventricular drain was used to relieve intracranial pressure or to drain ventricular sepsis. All neurosurgical as well as operations on the cranium (such as debridement) were performed at the NSU. Data were analysed using the Statistical Package for the Social Sciences (SPSS) version 11.5. The chisquare method was used to compare groups and, where numbers were very small, Fisher’s exact test was used. A p value of <0.05 was taken as statistically significant.
Results The patient profiles of both studies are shown in Table 1. Patients in both studies had similar characteristics. There were 86 patients in the pilot study of whom 10 had a GCS 8 and 22 required management in the NSU. There were 230 patients in the main study of whom 32 had a GCS 8 and 29 required management in the NSU. For ease of analysis the studies have been combined. There were thus a total of 316 patients in the whole study with a male to female ratio of 5:1. Their mean age was 31 12 years (range 12—80 years). Injury mechanisms were blunt trauma (289), fire-arm (14), stab (7) and bush-knife (6) injuries. Forty-two patients suffered severe head injuries (GCS 8).
CT scan was performed on 189 patients, 186 on the day of admission and three subsequently due to deterioration; 129 of these CT scans were abnormal and the rest were normal (Table 2 and Fig. 1). Findings unrelated to trauma were age appropriate atrophy (2), tumour (1), neurocysticercosis (1), mastoid air cell opacification (1) and granuloma (1). Following consultation with the NSU, 265 patients were managed in the GSU in KEH and 51 required transfer to the NSU at WWH. Of the latter 51 patients two died in transit to the NSU and one died on arrival at the NSU. Forty-two of the remaining 48 patients underwent various neurosurgical procedures (Table 3) and six patients were admitted to the NSU for neurosurgical observation. Some of the patients had more than one neurosurgical procedure. Eight patients died following neurosurgical procedures. In total therefore 41 patients died giving a mortality rate of 13% (Table 4 and Fig. 2). Twenty eight of the 42 patients with GCS 8 died compared to 13 out of 274 with GCS >8 ( p < 0.0001). None of the deaths were deemed preventable. Table 2
CT scan findings
Finding
n
%
Positive CT scan No CT scan Normal Contusion Subdural haematoma Intracerebral haematoma Diffuse cerebral injury Extradural haematoma Compound fracture in-driven bone fragments Findings not related to trauma Oedema Subarachnoid haemorrhage Bullet
129 127 60 43 27 27 11 10 7
41 40 19 14 8 8 4 3 2
6 2 2 1
2 0.6 0.6 0.3
Outcome of head injuries in GSU with NSU consultation
Figure 1
Flow diagram showing the management of 316 patients admitted with head injury.
Table 3 Neurosurgical procedures and associated deaths in 42 patients Procedure
579
Number of procedures
Evacuation of SDH 10 Debridement 8 Evacuation of EDH 7 Evacuation of contusion 4 Evacuation of ICH 4 Insertion of ICP monitor 2 Evacuation of SDH + insertion of 2 ICP monitor Insertion of EVD 1 Evacuation of contusion + insertion of EVD 1 Evacuation of contusion + ICH 1 Evacuation of SDH + insertion of EVD 1 Debridement, evacuation of 1 ICH + insertion ICP monitor EDH, exrtadural haematoma; SDH, subdural haematoma; ICH, intracerebral haematoma; EVD, external ventricular drain; ICP, intracranial pressure.
The median delay from time of injury to surgery was 5.1 h (range 2.5—8.5 h). The median delay for survivors was 5 h (range 1.5—96 h) compared to 6.5 h (1.5—504 h) for non-survivors ( p = 0.073). The mean age of survivors was 31 12 years compared to 32 13 years for non-survivors ( p = 0.478). The average hospital stay was 10 18 days. As shown in Table 5, 40 surviving patients from the GSU and 20 from the NSU were discharged with neurological sequelae ( p < 0.001). There were
Table 4 Mortality in patients treated for head injury n
Mortality
%
Overall
316
41
13
GCS >8 8
274 42
13 28
5 67
GSU NSU
265 51
30 11
11 22
580
B.M.W. Zulu et al.
Figure 2
Breakdown of mortality in patients with head injury.
11 patients with serious sequelae from the GSU namely hemiparesis (6) and monoparesis (5) and there were 12 patients with serious sequelae from the NSU, namely hemiplegia (3), hemiparesis (7) and monoparesis (2).
Discussion Injury to the brain has devastating consequences.17—19 Almost two-thirds of vehicular trauma deaths arise as a direct result of head injury and, in some who survive, permanent neurological dysfunction is incapacitating.6 South Africa is a Third World country with characteristic Third World problems, amongst which are budgetary constraints on the provision of health care services. Vehicular and interpersonal trauma is endemic,15 NSU’s are in short supply and in great demand, and hence the
majority of brain injured patients must by necessity be managed initially in peripheral hospitals. In the face of such logistical problems, it is a fundamental necessity that general surgeons triage head injured patients on the basis of the GCS and CT scan findings.9 The large clinical load makes it impractical to perform CT scans on all patients with head injuries and the performance of CT scans is therefore selective, being limited to patients with significant head injury. Young patients in the prime of their lives constitute the vast majority of head injured patients in most studies.17—19 It is imperative that they are offered the most appropriate management whereby those with a rectifiable lesion are referred early to a NSU, thus reducing unnecessary morbidity and mortality, and those who will not benefit from neurosurgical intervention do not overburden a scarce resource.
Outcome of head injuries in GSU with NSU consultation Table 5 Neurological sequelae in patients discharged following head injury Sequelae a
GSU, n (%)
Hemiparesis Hemiplegia Monoparesis Post-traumatic epilepsy Confusion Unsteady gait Cranial nerve palsy Hydrocephalus Aphasia Fixed dilated pupil Diabetes insipidus Deafness Post-concussion headache Dysphasia Amnesia
6 — 5 3 11 6 3 — — — — 2 1 2 2
Total
41 (16)
a
(2) (2) (1) (4) (2) (1)
(0.7) (0.4) (0.7) (0.7)
NSU, n (%) 7 3 2 5 3 2 3 1 1 1 1 — — — —
(14) (6) (4) (10) (6) (4) (6) (2) (2) (2) (2)
29 (56)
Some patients had more than one sequelae.
Following the landmark publication by Teasdale and Jennet20 over 30 years ago, the GCS has become the internationally accepted gold standard for head injury triage. This distinguishes those with severe head injury, facilitates patient selection for NSU referral, and allows a reasonable estimate of mortality prediction. The overall mortality rate in the present study of 13% compares favourably with the 8—30% reported in the literature.5,8,19 The higher mortality rate for patients in the NSU compared to those in the GSU in this study is not surprising considering that the NSU caters for severely head injured patients with space occupying lesions who require neurosurgical intervention. The mortality rate for severe head injuries was significantly higher than for mild to moderate injuries. The trend shown in this study has also been demonstrated by other reports with comparable survival rates in both severe and mild to moderate head injury based on the admission GCS. Gennarelli et al.8 published the results of the Major Trauma Outcome Study involving over 50,000 patients and found a good correlation between the GCS and mortality for head injuries. Cooke et al.5, in a study of 131 head injured patients, reported a mortality rate of 44% for those with severe injury compared to 11% in those with mild to moderate injury. A similar finding was made ´ lvarez et al.1 who documented mortality rates by A of 51% and 2% in those patients with severe and mild injuries, respectively. Demetriades et al.6 reported that more than three-quarters of patients with a GCS of three died and 16% were left severely
581
disabled. In a study of 150 children Feickert et al.7 showed that the GCS had a very high predictive value of outcome. Other authors have found additional factors as predictors of outcome. In a study of severe head injury, Boto et al.3 found the independent predictors of early mortality to be non-evacuated mass, diffuse injury, decerebration and shock. Hiler et al.10 demonstrated in 126 patients with severe head injury that patients with early loss of autoregulation had a worse prognosis. The use of sophisticated neurophysiologic technology does not appear to outperform the GCS. Rae-Grant et al.19 using electrocardiography and trans-cranial Doppler ultrasound, could find no benefit compared to the GCS. In addition to GCS, both age and delays in surgical management have been reported to adversely influence outcome.3,12,13,18 Neither was found to correlate with outcome in the present study. The reasons are to be found in the population we serve and the logistics of the health care system in the public service in South Africa. Firstly, the majority of our patients were young and more able to withstand physiological aberrations associated with trauma. Secondly, prolonged injury-to-hospital times are not uncommon in South Africa4 and those who require urgent neurosurgical intervention for a mass lesion would not survive to reach hospital. This is substantiated by the low prevalence of severe head injuries in our studies. The combination of the GCS and CT scanning has allowed more specific selection of patients for transfer to the NSU.11,16 Nadvi et al.16 assessed the effect of managing head injured patients at a peripheral hospital with a CT scanner and transferring only those with surgically rectifiable lesions. Admissions to the NSU were reduced by 43%. In the present series, the combination of GCS, CT scanning, and neurosurgical consultation resulted in only 41% of patients being transferred to the NSU. Pivotal within this system, however, is the ability to transmit the image to the NSU so that the scan may be interpreted by a neurosurgeon. Should that facility not exist, it would be crucial for the staff at the referring hospital to be capable of interpreting the CT scan. In the absence of a CT scanning facility, the GCS must serve as the sole triage tool. The obvious concern with an off-site neurosurgical service is that there may be a number of preventable deaths in those patients managed by a GSU. Avoidable deaths will still occur due to the need for secondary transfer from the general surgical service to the NSU. Hypoxia and hypercarbia
582 may develop insidiously, and the disturbance caused by the ambulance journey can adversely affect cardiovascular stability in a seriously injured patient.9 Transfer must therefore be undertaken rapidly and efficiently by trained personnel. None of the deaths were deemed to be preventable in this study. In the presence of multiple injuries, an all too common scenario in the context of vehicular trauma, it is crucial that unnecessary transportation be minimised. Even with comprehensive monitoring and trained attendants, the movement of critically ill patients is associated with adverse physiological changes.2,21 This group of patients, in whom the management involves multiple surgical disciplines, benefit immensely from the early confirmation or exclusion of a rectifiable neurosurgical lesion. The higher proportion of neurological sequelae in patients managed in the NSU is understandable because they either required neurosurgical intervention or were so severely injured that they needed observation in the NSU’s ICU. Head injured patients have significantly greater impairment than patients with extra-cranial injury at the time of discharge from hospital and, despite advances in healthcare, head injuries remain the most important cause of disability.8 Rehabilitation therefore remains crucial in these patients. Head trauma is associated with high morbidity and mortality and outcome is determined primarily by the GCS on presentation. NSU patients have a significantly higher mortality compared to the GSU patients. Delay before surgery in those patients requiring neurosurgical intervention did not seem to affect outcome. Non-surgical treatment of traumatic brain injury at the referral hospital by the GSU is acceptable practice. Triage using the GCS and CT scanning therefore is an effective, safe and economical method of managing head injured patients with an off-site neurosurgical service. We believe that the guidelines for CT scan and for referral to the NSU currently in use in our institution are adequate and do not need revision. Although limited resources may restrict the provision of health care in developing countries, this should not deter clinicians from seeking First World solutions to their problems. Advances in information technology have resulted in communication on a global scale and the transmission of images by telemedicine has obviated the need for unnecessary patient transfer. This benefits the patient, the clinician, and the health service as a whole. In situations where specialist services are scarce and unlikely to expand, the use of telemedicine
B.M.W. Zulu et al. may allow unlimited access to a limited area of expertise, thus ensuring optimal use of available resources. It is our recommendation that communities with limited resources such as occurs in South Africa should consider the management of head injuries in the GSU with neurosurgical consultation via telemedicine.
References ´lvarez M, Nava J-M, Rue 1. A ` M, Quintana S. Mortality prediction in head trauma patients: performance of Glasgow Coma Score and general severity systems. Crit Care Med 1998;26:142—8. 2. Beckmann V, Gillies DM, Berenholtz SM, et al. Incidents relating to the intra-hospital transfer of critically ill patients. Int Care Med 2004;30:1508—10. 3. Boto GR, Gomez PA, De La Cruz J, Lobato RD. Severe head injury and the risk of early death. J Neurol Neurosurg Psych 2006;1:1—11. 4. Campbell NC, Thomson SR, Muckart DJJ, et al. Review of 1198 cases of penetrating cardiac trauma. Br J Surg 1997;84:1737— 40. 5. Cooke RS, McNicholl BP, Byrnes DP. Early management of severe head injury in Northern Ireland. Injury 1995;26: 395—7. 6. Demetriades D, Kuncir E, Velmahos GC, et al. Outcome and prognostic factors in head injuries with an admission Glasgow Coma Scale score of 3. Arch Surg 2004;139:1066—8. 7. Feickert H-J, Drommer S, Heyer R. Severe head injury in children: impact of risk factors on outcome. J Trauma 1999;47:33—8. 8. Gennarelli TA, Champion HR, Copes WS, Sacco WJ. Comparison of mortality, morbidity and severity of 59,713 head injured patients with 114,447 patients with extracranial injuries. J Trauma 1994;37:962—8. 9. Gentleman D, Dearden M, Midgley S, Maclean D. Guidelines for resuscitation and transfer of patients with serious head injury. Br Med J 1993;307:547—52. 10. Hiler M, Czosnyka M, Hutchinson P, et al. Predictive value of initial computerised tomography scan, intracranial pressure, and state of autoregulation in patients with traumatic brain injury. J Neurosurg 2006;104:731—7. 11. Jithoo R, Govender PV, Corr P, Nathoo N. Telemedicine and neurosurgery: experience of a regional unit based in South Africa. J Telemed Telecare 2003;9:63—6. 12. Mamelak AN, Pitts LH, Damron S. Predicting survival from head trauma 24 h after injury: a practical method with therapeutic implications. J Trauma 1996;41:91—9. 13. Mendelow AD, Karmi MZ, Paul KS, et al. Extradural haematoma: effect of delayed treatment. Br Med J 1979;1: 1240—2. 14. Muckart DJJ. Trauma–—the malignant epidemic. S Afr Med J 1991;79:93—5. 15. Muckart DJJ, Meumann C, Botha JBC. The changing pattern of penetrating torso trauma in KwaZulu-Natal–—a clinical and pathological review. S Afr Med J 1995;85:1172—4. 16. Nadvi SS, Parboosing R, van Dellen JR. Benefits to a regional neurosurgical unit following the introduction of a decentralised imaging facility. S Afr Med J 1997;87:1669—71. 17. Ommaya AK, Dannenberg AL, Salazar AM. Causation, incidence and costs of traumatic brain injury in the U.S. Military Medical System. J Trauma 1996;40:211—7.
Outcome of head injuries in GSU with NSU consultation 18. Quigley MR, Vidovich D, Cantella D, et al. Defining the limits of survivorship after severe head injury. J Trauma 1997;42:7—10. 19. Rae-Grant AD, Eckert N, Barbour PJ, et al. Outcome of severe brain injury: a multimodality neurophysiologic study. J Trauma 1996;40:401—7.
583
20. Teasdale G, Jennet B. Assessment of coma and impaired consciousness. Lancet 1974;ii:81—4. 21. Warren J, Fromm RE, Om RA, et al. Guidelines for inter- and intra-hospital transport of critically ill patients. Crit Care Med 2004;32:305—6.
www.elsevier.com/locate/injury
Outcome of head injuries in general surgical units with an off-site neurosurgical service§ B.M.W. Zulu, T.V. Mulaudzi, T.E. Madiba *, D.J.J. Muckart Department of Surgery, University of KwaZulu-Natal and King Edward VIII Hospital, Durban, South Africa Accepted 2 January 2007
KEYWORDS Head injury; Neurosurgical unit; General surgical unit
Summary Introduction: In order to cope with bed shortages in the only neurosurgical unit (NSU) in KwaZulu-Natal, it has become necessary to manage head injured patients in a general surgical unit (GSU) at the referral hospitals in consultation with the NSU. This study was undertaken to assess the outcome of patients with head injuries managed in a GSU in consultation with a regional NSU. Patients and methods: A prospective pilot study was carried out in a single surgical unit over an 18-month period (July 1997—December 1998), followed by a larger prospective study over a 6-month period (January—June 2001). All patients with head injuries severe enough to warrant admission to hospital (GCS 15 or 15 with localising signs) were included. Results: The pilot study comprised 86 patients and the subsequent study comprised 230 patients, giving a total of 316 patients in the whole study. The mean age was 31 12 years. Following consultation with the NSU, 265 (84%) patients were managed in the GSU and 51 (16%) required transfer to the NSU. Forty-one patients died giving a mortality rate of 13%. Twenty eight of the 42 patients with GCS 8 died (67%) compared to 13 out of 274 with GCS >8 (5%) ( p < 0.0001). Eleven of the 51 patients destined to the NSU died (22%) compared to 30 of 265 in the GSU (11%) ( p = 0.046). The average hospital stay was 10 18 days. Forty surviving patients (17%) from the GSU and 20 from the NSU (51%) were discharged with neurological sequelae ( p < 0.001). Conclusion: Head trauma is associated with high morbidity and mortality. Nonsurgical treatment of traumatic brain injury at the referral hospital by the GSU is
§
Presented at the Congress of the Association of Surgeons of South Africa on 6—10 August 2004, Cape Town, South Africa. * Corresponding author at: Department of Surgery, University of KwaZulu-Natal, Private Bag 7, Congella 4013, South Africa. Tel.: +27 31 260 4257; fax: +27 31 260 4389. E-mail address: [email protected] (T.E. Madiba). 0020–1383/$ — see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2007.01.002
Outcome of head injuries in GSU with NSU consultation
577
acceptable practice. Outcome is determined primarily by the GCS on presentation. NSU patients had a significantly higher mortality rate. A delay before surgery did not seem to affect outcome. # 2007 Elsevier Ltd. All rights reserved.
Introduction Trauma is the second most common cause of death in South Africa and the commonest cause in children (age 5—14) and young adults (age 15—34).15 Head injuries account for the majority of deaths, being marginally more common than torso trauma.14,15 Consequently an overwhelming number of head injured patients are treated in South African hospitals. Ideally, head injured patients should be managed in a neurosurgical unit (NSU) where they can receive expert neurosurgical care. In South Africa, however, NSU’s are in short supply owing to budgetary constraints; consequently the majority of these patients are managed in peripheral hospitals in consultation with a NSU via CT Telemedicine.16 KwaZulu-Natal province occupies the east coast of South Africa with a population of about 8 million, which represents about 20% of the national population. The Medical Faculty of the University of KwaZulu-Natal is situated in Durban, the main city of KwaZulu-Natal. This province has only one public sector NSU, formerly situated at Wentworth Hospital (WWH) at the time of the study, and now situated at the new Inkosi Albert Luthuli Central Hospital. WWH is situated 8 km from King Edward VIII Hospital (KEH), the main teaching hospital of the University of KwaZulu-Natal Medical School, which houses the general surgical wards. This NSU also serves communities as far south as the province of the Eastern Cape. Due to bed shortage this NSU has had to establish guidelines for referral to the NSU as it is incapable of accommodating all patients with head injury. Although the early transfer of appropriate patients to a NSU substantially reduces morbidity and mortality, inappropriate referrals may overburden not only such specialised units but also the transport services. Two studies were undertaken to assess the outcome of patients with head injuries managed in the GSU of KEH in consultation with a regional off-site NSU situated in WWH, namely a pilot study in July 1997—December 1998 and the main study in January—June 2001. The results of the two studies have been combined and analysed as one study.
Patients and methods The surgical service at KEH consists of three wards, each with two surgical units, giving a total of six
surgical units. The pilot study was undertaken in a single surgical unit at King Edward VIII Hospital over an 18-month period from July 1997 to December 1998. The second study was carried out in all six units over a 6-month period from January to June 2001. King Edward VIII Hospital (KEH) is one of the main referring hospitals to the NSU in WWH. The large period that elapsed between the two studies was unavoidable owing to lack of human resources. All patients with head injury severe enough to warrant admission to hospital were enrolled in the study. Indications for admission to hospital in the Durban Metropolitan Hospital services are (i) a period of loss of consciousness, (ii) presence of neurological deficit, and (iii) all patients with Glasgow coma scale (GCS) of <15. Indications for urgent CT scan on admission are a GCS 14 or any level of consciousness associated with neurological deficit. Indications for performance of a semi-elective CT scan during hospitalisation include (i) any deterioration in the level of consciousness, (ii) development of neurological deficit, or (iii) failure of improvement of the GCS for more than 48 h even in the absence of a fracture or focal signs. Patients with a GCS 4 from the time of injury and those that were hypotensive or hypoxic were resuscitated and reevaluated before consultation. Severe head injury was defined as GCS 8. All trauma patients, including those with head injury are admitted to an acute ward where triage is carried out. Patients requiring referral to the NSU are transferred directly from this unit. The rest are kept in the acute ward for a period of 24 h prior to being transferred to the general surgical unit (GSU). The criteria for triage of patients were admission GCS and CT scan. The CT scan was sent to the NSU by telemedicine and results were discussed with the neurosurgeon prior to an appropriate decision being made. Transfer of the patient to the NSU occurred only after the neurosurgeon had agreed to receive the patient. Additional criteria were patients with a compound injury and/or brain oozing. These were transferred immediately to the NSU. All fractures with deeply in-driven bone fragments, contaminated wounds, brain or cerebrospinal fluid oozing, retained foreign body or heavy bleeding underwent emergency debridement. Patients with intracranial pathology underwent surgery when they fulfilled one or more of the following
578
B.M.W. Zulu et al.
Table 1 Patient profile Pilot study
Main study
Whole study
Total number of patients Male M:F
86 74 (86%) 6:1
230 192 (83%) 5:1
316 266 (84%) 5:1
Age (years) Range Mean
16—65 31 11
12—80 31 13
12—80 31 12
GSU NSU
64 (74%) 22 (26%)
201 (87%) 29 (13%)
265 (84%) 51 (16%)
GCS >8 8
76 (88%) 10 (12%)
198 (86%) 32 (14%)
274 (87%) 42 (13%)
criteria: (i) size of more that 30 cm2, (ii) midline shift of more than 10 mm, and (iii) effacement of the ipsilateral ventricle. Patients with a GCS of 15 with no localising signs received clinical observation. Patients with intracranial pathology but no mass effect were managed non-operatively. An intracranial pressure monitor was inserted under the following circumstances: (i) patients undergoing surgery for evacuation of space occupying lesion (except for a pure extradural haematoma) and (ii) patients with evidence of increased intracranial pressure on CTscan but no mass lesion. The external ventricular drain was used to relieve intracranial pressure or to drain ventricular sepsis. All neurosurgical as well as operations on the cranium (such as debridement) were performed at the NSU. Data were analysed using the Statistical Package for the Social Sciences (SPSS) version 11.5. The chisquare method was used to compare groups and, where numbers were very small, Fisher’s exact test was used. A p value of <0.05 was taken as statistically significant.
Results The patient profiles of both studies are shown in Table 1. Patients in both studies had similar characteristics. There were 86 patients in the pilot study of whom 10 had a GCS 8 and 22 required management in the NSU. There were 230 patients in the main study of whom 32 had a GCS 8 and 29 required management in the NSU. For ease of analysis the studies have been combined. There were thus a total of 316 patients in the whole study with a male to female ratio of 5:1. Their mean age was 31 12 years (range 12—80 years). Injury mechanisms were blunt trauma (289), fire-arm (14), stab (7) and bush-knife (6) injuries. Forty-two patients suffered severe head injuries (GCS 8).
CT scan was performed on 189 patients, 186 on the day of admission and three subsequently due to deterioration; 129 of these CT scans were abnormal and the rest were normal (Table 2 and Fig. 1). Findings unrelated to trauma were age appropriate atrophy (2), tumour (1), neurocysticercosis (1), mastoid air cell opacification (1) and granuloma (1). Following consultation with the NSU, 265 patients were managed in the GSU in KEH and 51 required transfer to the NSU at WWH. Of the latter 51 patients two died in transit to the NSU and one died on arrival at the NSU. Forty-two of the remaining 48 patients underwent various neurosurgical procedures (Table 3) and six patients were admitted to the NSU for neurosurgical observation. Some of the patients had more than one neurosurgical procedure. Eight patients died following neurosurgical procedures. In total therefore 41 patients died giving a mortality rate of 13% (Table 4 and Fig. 2). Twenty eight of the 42 patients with GCS 8 died compared to 13 out of 274 with GCS >8 ( p < 0.0001). None of the deaths were deemed preventable. Table 2
CT scan findings
Finding
n
%
Positive CT scan No CT scan Normal Contusion Subdural haematoma Intracerebral haematoma Diffuse cerebral injury Extradural haematoma Compound fracture in-driven bone fragments Findings not related to trauma Oedema Subarachnoid haemorrhage Bullet
129 127 60 43 27 27 11 10 7
41 40 19 14 8 8 4 3 2
6 2 2 1
2 0.6 0.6 0.3
Outcome of head injuries in GSU with NSU consultation
Figure 1
Flow diagram showing the management of 316 patients admitted with head injury.
Table 3 Neurosurgical procedures and associated deaths in 42 patients Procedure
579
Number of procedures
Evacuation of SDH 10 Debridement 8 Evacuation of EDH 7 Evacuation of contusion 4 Evacuation of ICH 4 Insertion of ICP monitor 2 Evacuation of SDH + insertion of 2 ICP monitor Insertion of EVD 1 Evacuation of contusion + insertion of EVD 1 Evacuation of contusion + ICH 1 Evacuation of SDH + insertion of EVD 1 Debridement, evacuation of 1 ICH + insertion ICP monitor EDH, exrtadural haematoma; SDH, subdural haematoma; ICH, intracerebral haematoma; EVD, external ventricular drain; ICP, intracranial pressure.
The median delay from time of injury to surgery was 5.1 h (range 2.5—8.5 h). The median delay for survivors was 5 h (range 1.5—96 h) compared to 6.5 h (1.5—504 h) for non-survivors ( p = 0.073). The mean age of survivors was 31 12 years compared to 32 13 years for non-survivors ( p = 0.478). The average hospital stay was 10 18 days. As shown in Table 5, 40 surviving patients from the GSU and 20 from the NSU were discharged with neurological sequelae ( p < 0.001). There were
Table 4 Mortality in patients treated for head injury n
Mortality
%
Overall
316
41
13
GCS >8 8
274 42
13 28
5 67
GSU NSU
265 51
30 11
11 22
580
B.M.W. Zulu et al.
Figure 2
Breakdown of mortality in patients with head injury.
11 patients with serious sequelae from the GSU namely hemiparesis (6) and monoparesis (5) and there were 12 patients with serious sequelae from the NSU, namely hemiplegia (3), hemiparesis (7) and monoparesis (2).
Discussion Injury to the brain has devastating consequences.17—19 Almost two-thirds of vehicular trauma deaths arise as a direct result of head injury and, in some who survive, permanent neurological dysfunction is incapacitating.6 South Africa is a Third World country with characteristic Third World problems, amongst which are budgetary constraints on the provision of health care services. Vehicular and interpersonal trauma is endemic,15 NSU’s are in short supply and in great demand, and hence the
majority of brain injured patients must by necessity be managed initially in peripheral hospitals. In the face of such logistical problems, it is a fundamental necessity that general surgeons triage head injured patients on the basis of the GCS and CT scan findings.9 The large clinical load makes it impractical to perform CT scans on all patients with head injuries and the performance of CT scans is therefore selective, being limited to patients with significant head injury. Young patients in the prime of their lives constitute the vast majority of head injured patients in most studies.17—19 It is imperative that they are offered the most appropriate management whereby those with a rectifiable lesion are referred early to a NSU, thus reducing unnecessary morbidity and mortality, and those who will not benefit from neurosurgical intervention do not overburden a scarce resource.
Outcome of head injuries in GSU with NSU consultation Table 5 Neurological sequelae in patients discharged following head injury Sequelae a
GSU, n (%)
Hemiparesis Hemiplegia Monoparesis Post-traumatic epilepsy Confusion Unsteady gait Cranial nerve palsy Hydrocephalus Aphasia Fixed dilated pupil Diabetes insipidus Deafness Post-concussion headache Dysphasia Amnesia
6 — 5 3 11 6 3 — — — — 2 1 2 2
Total
41 (16)
a
(2) (2) (1) (4) (2) (1)
(0.7) (0.4) (0.7) (0.7)
NSU, n (%) 7 3 2 5 3 2 3 1 1 1 1 — — — —
(14) (6) (4) (10) (6) (4) (6) (2) (2) (2) (2)
29 (56)
Some patients had more than one sequelae.
Following the landmark publication by Teasdale and Jennet20 over 30 years ago, the GCS has become the internationally accepted gold standard for head injury triage. This distinguishes those with severe head injury, facilitates patient selection for NSU referral, and allows a reasonable estimate of mortality prediction. The overall mortality rate in the present study of 13% compares favourably with the 8—30% reported in the literature.5,8,19 The higher mortality rate for patients in the NSU compared to those in the GSU in this study is not surprising considering that the NSU caters for severely head injured patients with space occupying lesions who require neurosurgical intervention. The mortality rate for severe head injuries was significantly higher than for mild to moderate injuries. The trend shown in this study has also been demonstrated by other reports with comparable survival rates in both severe and mild to moderate head injury based on the admission GCS. Gennarelli et al.8 published the results of the Major Trauma Outcome Study involving over 50,000 patients and found a good correlation between the GCS and mortality for head injuries. Cooke et al.5, in a study of 131 head injured patients, reported a mortality rate of 44% for those with severe injury compared to 11% in those with mild to moderate injury. A similar finding was made ´ lvarez et al.1 who documented mortality rates by A of 51% and 2% in those patients with severe and mild injuries, respectively. Demetriades et al.6 reported that more than three-quarters of patients with a GCS of three died and 16% were left severely
581
disabled. In a study of 150 children Feickert et al.7 showed that the GCS had a very high predictive value of outcome. Other authors have found additional factors as predictors of outcome. In a study of severe head injury, Boto et al.3 found the independent predictors of early mortality to be non-evacuated mass, diffuse injury, decerebration and shock. Hiler et al.10 demonstrated in 126 patients with severe head injury that patients with early loss of autoregulation had a worse prognosis. The use of sophisticated neurophysiologic technology does not appear to outperform the GCS. Rae-Grant et al.19 using electrocardiography and trans-cranial Doppler ultrasound, could find no benefit compared to the GCS. In addition to GCS, both age and delays in surgical management have been reported to adversely influence outcome.3,12,13,18 Neither was found to correlate with outcome in the present study. The reasons are to be found in the population we serve and the logistics of the health care system in the public service in South Africa. Firstly, the majority of our patients were young and more able to withstand physiological aberrations associated with trauma. Secondly, prolonged injury-to-hospital times are not uncommon in South Africa4 and those who require urgent neurosurgical intervention for a mass lesion would not survive to reach hospital. This is substantiated by the low prevalence of severe head injuries in our studies. The combination of the GCS and CT scanning has allowed more specific selection of patients for transfer to the NSU.11,16 Nadvi et al.16 assessed the effect of managing head injured patients at a peripheral hospital with a CT scanner and transferring only those with surgically rectifiable lesions. Admissions to the NSU were reduced by 43%. In the present series, the combination of GCS, CT scanning, and neurosurgical consultation resulted in only 41% of patients being transferred to the NSU. Pivotal within this system, however, is the ability to transmit the image to the NSU so that the scan may be interpreted by a neurosurgeon. Should that facility not exist, it would be crucial for the staff at the referring hospital to be capable of interpreting the CT scan. In the absence of a CT scanning facility, the GCS must serve as the sole triage tool. The obvious concern with an off-site neurosurgical service is that there may be a number of preventable deaths in those patients managed by a GSU. Avoidable deaths will still occur due to the need for secondary transfer from the general surgical service to the NSU. Hypoxia and hypercarbia
582 may develop insidiously, and the disturbance caused by the ambulance journey can adversely affect cardiovascular stability in a seriously injured patient.9 Transfer must therefore be undertaken rapidly and efficiently by trained personnel. None of the deaths were deemed to be preventable in this study. In the presence of multiple injuries, an all too common scenario in the context of vehicular trauma, it is crucial that unnecessary transportation be minimised. Even with comprehensive monitoring and trained attendants, the movement of critically ill patients is associated with adverse physiological changes.2,21 This group of patients, in whom the management involves multiple surgical disciplines, benefit immensely from the early confirmation or exclusion of a rectifiable neurosurgical lesion. The higher proportion of neurological sequelae in patients managed in the NSU is understandable because they either required neurosurgical intervention or were so severely injured that they needed observation in the NSU’s ICU. Head injured patients have significantly greater impairment than patients with extra-cranial injury at the time of discharge from hospital and, despite advances in healthcare, head injuries remain the most important cause of disability.8 Rehabilitation therefore remains crucial in these patients. Head trauma is associated with high morbidity and mortality and outcome is determined primarily by the GCS on presentation. NSU patients have a significantly higher mortality compared to the GSU patients. Delay before surgery in those patients requiring neurosurgical intervention did not seem to affect outcome. Non-surgical treatment of traumatic brain injury at the referral hospital by the GSU is acceptable practice. Triage using the GCS and CT scanning therefore is an effective, safe and economical method of managing head injured patients with an off-site neurosurgical service. We believe that the guidelines for CT scan and for referral to the NSU currently in use in our institution are adequate and do not need revision. Although limited resources may restrict the provision of health care in developing countries, this should not deter clinicians from seeking First World solutions to their problems. Advances in information technology have resulted in communication on a global scale and the transmission of images by telemedicine has obviated the need for unnecessary patient transfer. This benefits the patient, the clinician, and the health service as a whole. In situations where specialist services are scarce and unlikely to expand, the use of telemedicine
B.M.W. Zulu et al. may allow unlimited access to a limited area of expertise, thus ensuring optimal use of available resources. It is our recommendation that communities with limited resources such as occurs in South Africa should consider the management of head injuries in the GSU with neurosurgical consultation via telemedicine.
References ´lvarez M, Nava J-M, Rue 1. A ` M, Quintana S. Mortality prediction in head trauma patients: performance of Glasgow Coma Score and general severity systems. Crit Care Med 1998;26:142—8. 2. Beckmann V, Gillies DM, Berenholtz SM, et al. Incidents relating to the intra-hospital transfer of critically ill patients. Int Care Med 2004;30:1508—10. 3. Boto GR, Gomez PA, De La Cruz J, Lobato RD. Severe head injury and the risk of early death. J Neurol Neurosurg Psych 2006;1:1—11. 4. Campbell NC, Thomson SR, Muckart DJJ, et al. Review of 1198 cases of penetrating cardiac trauma. Br J Surg 1997;84:1737— 40. 5. Cooke RS, McNicholl BP, Byrnes DP. Early management of severe head injury in Northern Ireland. Injury 1995;26: 395—7. 6. Demetriades D, Kuncir E, Velmahos GC, et al. Outcome and prognostic factors in head injuries with an admission Glasgow Coma Scale score of 3. Arch Surg 2004;139:1066—8. 7. Feickert H-J, Drommer S, Heyer R. Severe head injury in children: impact of risk factors on outcome. J Trauma 1999;47:33—8. 8. Gennarelli TA, Champion HR, Copes WS, Sacco WJ. Comparison of mortality, morbidity and severity of 59,713 head injured patients with 114,447 patients with extracranial injuries. J Trauma 1994;37:962—8. 9. Gentleman D, Dearden M, Midgley S, Maclean D. Guidelines for resuscitation and transfer of patients with serious head injury. Br Med J 1993;307:547—52. 10. Hiler M, Czosnyka M, Hutchinson P, et al. Predictive value of initial computerised tomography scan, intracranial pressure, and state of autoregulation in patients with traumatic brain injury. J Neurosurg 2006;104:731—7. 11. Jithoo R, Govender PV, Corr P, Nathoo N. Telemedicine and neurosurgery: experience of a regional unit based in South Africa. J Telemed Telecare 2003;9:63—6. 12. Mamelak AN, Pitts LH, Damron S. Predicting survival from head trauma 24 h after injury: a practical method with therapeutic implications. J Trauma 1996;41:91—9. 13. Mendelow AD, Karmi MZ, Paul KS, et al. Extradural haematoma: effect of delayed treatment. Br Med J 1979;1: 1240—2. 14. Muckart DJJ. Trauma–—the malignant epidemic. S Afr Med J 1991;79:93—5. 15. Muckart DJJ, Meumann C, Botha JBC. The changing pattern of penetrating torso trauma in KwaZulu-Natal–—a clinical and pathological review. S Afr Med J 1995;85:1172—4. 16. Nadvi SS, Parboosing R, van Dellen JR. Benefits to a regional neurosurgical unit following the introduction of a decentralised imaging facility. S Afr Med J 1997;87:1669—71. 17. Ommaya AK, Dannenberg AL, Salazar AM. Causation, incidence and costs of traumatic brain injury in the U.S. Military Medical System. J Trauma 1996;40:211—7.
Outcome of head injuries in GSU with NSU consultation 18. Quigley MR, Vidovich D, Cantella D, et al. Defining the limits of survivorship after severe head injury. J Trauma 1997;42:7—10. 19. Rae-Grant AD, Eckert N, Barbour PJ, et al. Outcome of severe brain injury: a multimodality neurophysiologic study. J Trauma 1996;40:401—7.
583
20. Teasdale G, Jennet B. Assessment of coma and impaired consciousness. Lancet 1974;ii:81—4. 21. Warren J, Fromm RE, Om RA, et al. Guidelines for inter- and intra-hospital transport of critically ill patients. Crit Care Med 2004;32:305—6.