Improvements in trauma survival in Leeds

Proceedings

of the British

Trauma

Society

M. G. Chambers, C. M. Airey, S. Chell, A. S. Rigby, J. Connelly, A. Tennant (York, UK and Leeds, UK) A cosf analysis of major fraumafic inju y Individuals who suffer serious injury are significant consumers of health care and other services. Identifying and analysing the costs of care are increasingly important for establishing the financial burden of disease to individuals, society and the NHS, forming the basis for sound service planning, and acting as a baseline for cost-effectiveness studies of interventions in this area. The Yorkshire major trauma study provides a unique opportunity to analyse and compare the costs of initial hospitalization and subsequent care after discharge for a geographically defined cohort, survivors and non-survivors, for incidents occurring in 1 year. Victims of major injuries (Injury Severity Score (ISS)> 15) admitted initially to four hospitals in the region have been selected. Of a total of 328 cases, 221(67 per cent) arrived alive at hospital Accident and Emergency (A&E) resuscitation facilities. Of these, 191 (86 per cent) survived to be admitted to a ward or transferred to another specialist hospital. One hundred and sixteen individuals were eventually discharged alive from hospital, of whom 83 (72 per cent) have been interviewed. This cohort gnerated 9275 bed days of initial (acute) hospital care in the Region, with a median stay of 15 days (range 0-1594, IQR 4-44, mean 50.1 days). Removing one individual who stayed for 1594 days reduces the mean stay to 41.7 days. For survivors, the median stay was 32.0 days (range 3-1594, IQR 15-60, mean 77.9, excluding outlier 64.1 days). Of all bed days 47 per cent were in orthopaedic units, 20 per cent in neurosurgery units, 20 per cent in neurology/spinal injury units, and 5 per cent in intensive care units (ICU). For survivors, the mean stays were 26.9,64.3 and 236.7 days for ISS bands 16-19,20-29 and 30 + respectively. An illustrative analysis of hospital costs gives an estimate of f7092 per admission (El0816 for survivors and f1455 for non-survivors). Total ward treatment costs were El.34m, to which should be added the costs of resuscitation in A&E (221 cases), mortuary/pathology (105 hospital deaths, 107 deaths before arrival), emergency services (328 initial call-outs, 55 transfers and !S re-transfers). Mean hospital costs were f3923, xl0 562 and f27 057 for survivors in ISS bands 15-19, 20-29 and 30 + respectively. Orthopaedics contributed 30 per cent of all costs, neurosurgery 14 per cent, spinal injury 17 per cent and ICU 20 per cent. The most severely injured and disabled generated mean costs of E30 000 and above. Follow-up (hospital) costs averaged f2000 per survivor, with high costs concentrated among the severely disabled. Total direct hospital costs incurred by injuries in 1988-1989 in the Yorkshire region is at least iZ4.9 million, 80 per cent being the cost of initial care of the survivors. Further analysis will compare hospital costs with on-going costs of care, and distinguish between average and marginal costs of care.

V. Neumann, A. Bowen, M. Conner, M. A. Chamberlain (Leeds, UK) Lack of occupation after traumatic brain injury: who is affected? The first 100 consecutive patients recruited to a study conceming rehabilitation following traumatic brain injury (TBI) were assessed 6 months after injury to establish: (i) whether they were occupied pre-injury; (ii) whether they had returned to their occupation; (iii) what differences could be detected between those occupied and not occupied at 6 months. Occupation was defined as including full-time (FT) or parttime (PT) paid work, full-time study or full-time involvement in homecare. All patients had been hospital in-patients for at least 3

369 days following TBI and had no history of significant psychiatric disorder, alcohol or drug abuse. Before their injury 80 were occupied (57 FT, 10 PT, nine FT study, four homecare) and 20 were unoccupied. At 6 months only SO were occupied (36 FT, eight PT, six FT study). Analyses of the 80 patients showed that the occupied and unoccupied groups did not differ with respect to age, sex, duration of post-traumatic amnesia or presence of other injuries. The worst recorded Glasgow Coma Score was lower in the unoccupied group (P
P. Burdett-Smith, M. Airey, A. Franks (Leeds, UK) Improvements in trauma survival in Leeds All cases of major injury (Injury Severity Score (ISS)> 15) admitted to the two Leeds hospitals in 1988-1989 were compared to all cases admitted in 1991-1993, to ascertain if a system of trauma care which had evolved in the hospitals during this time had affected outcome of these patients. Significant improvements in survival were shown especially in less seriously injured patients. Data from an existing Regional study of serious injury were compared to current data collected as part of the MTOS(UK) study. One hundred and eighty-six patients admitted to the two Leeds hospitals in 1988-1989 were compared to 198 cases admitted in 1992-1993. ISS, age, and sex distribution were similar for the two groups. There were significantly fewer deaths in 1992-1993: particularly in those patients with ISS < 41. This shows that although the anatomical indicators of injury did not change, the deleterious physiological effects of trauma were better corrected in the later group. This would suggest that early treatment had improved over the study period. The changes that had occurred included regular Advanced Trauma Life Support (ATLS) courses being run in Leeds, increased senior medical staff presence in the Accident and Emergency departments, nurse training in ATNC, the development of trauma teams, regular multi-disciplinary trauma audit meetings and a general raising of awareness of the importance of rapid treatment of the seriously injured. The emphasis, since 1988, on timely and definitive manage-

370

Injury:

International

ment of patients with major injuries along the lines of the ATLS teaching had lead to improved survival of the seriously injured in Leeds. S. D. Deo, J. D. Knottenbelt (Cape Town, Republic of South Africa) The use of midazolam in trauma resuscitation This is the first paper describing the use of midazolam as an adjunct in the initial phase of resuscitation after injury. It is a retrospective study of the use of midazolam in resuscitations after injury over a 6-week period in 1994. In this period 45 patients required endotracheal intubation as part of resuscitation, of whom 33 (73 per cent) were given midazolam. The indications for its use, dose given, and physiological and side effects were noted. Midazolam was found to be a very safe and viable alternative to muscle relaxants and other anaesthetic agents in injuries allowing endotracheal intubation, as well as other invasive procedures to be carried out with minimal distress to the patient. Sixty-four per cent of patients had a low Glasgow Coma Score and required intubation for airway control and oxygenation. Adjuvant analgesia was required in less than 40 per cent of patients, amnesic properties were excellent and there were no significant physiological or other side effects. The use of midazolam is recommended in the appropriate environment and its use in the pre-hospital setting should be given consideration. M. J. Clancy, J. Alderman, C. Case, K. J. W. Taylor (Bristol, UK) The use of ultrasound in the non-invasive defection of changes in the renal circulation in response to blood loss using an animal model Using a continuous haemorrhage model, eight anaesthetized swine were bled I ml/kg/min for 30 min. The resistance index (RI) of the main renal artery, interlobar and arcuate vessels all significantly increased. Cortical Doppler signals were lost in four animals at a mean arterial pressure of 26 mmHg. After reinfusion of blood and normal saline only the RI of the interlobar vessels was significantly different from baseline readings. Ultrasound non-invasively demonstrated changes in regional blood flow within the kidney in response to hypovolaemic shock. T. Wardle, P. Driscoll, C. Oxbey, C. Dryer, F. Campbell, M. Woodford, F. Munsal (Chester, UK, and Salford, UK) The effect of pre-existing medical conditions on the outcome of injured trauma

patients

In a recent retrospective study we have shown that pre-morbid medical conditions (PMC) were present in 39 per cent of UK injured patients with complete records. In view of the incidence of incomplete documentation, it could be argued that this figure was artificially high. The aim of this study therefore was to carry out a detailed investigation in one UK centre to ascertain the incidence of PMC and determine its effect on patient outcome. Injured patients from Hope hospital included in the Major Trauma Outcome Study between 1988-1990 were investigated. All deaths from injuries (N= 121) and a random sample of survivors (N= 1350) were analysed for PMC (i.e. cardiovascular, respiratory, metabolic, neurological, and others), Injury Severity Score (ISS), Revised Trauma Score (RTS), age and outcome. PMC were found in all age groups but their occurrence increased directly with age (Mantel-Haenszel 1’ P= 0.0001). The most common PMC was cardiovascular disease which was found in 14.8 per cent of all patients and 19 per cent of those who died. Six hundred and sixty-four patients (45 per cent) had one or more PMCs. This group had a percentage mortality of 10.7 per cent which compared with 6.2 per cent in patients with no PMC (odds ratio = 1.8). Logistic regression showed that the presence, as well as the number, of PMCs were significant risk factors in the

Journal

of the Care of the Injured

Vol. 27, No. 5, 1996

prediction of survival after injury. After taking into account the effects of ISS, RTS and age, the risk of death increased by 2.25 in the presence of PMC (CI = 1.07-4.07, P= 0.032). Consequently traumatologists must be aware that PMCs are common and can have a profound effect on the outcome of injured victims. It is therefore essential that an accurate medical history is obtained from these patients and management carried out. C. Milroy, D. Warwick, M. Clancy (Bristol, UK) External pelvic fixation in the South West External pelvic fixation is recommended for the management of hypovolaemia in the presence of unstable pelvic fractures. Its success may depend on the immediate availability of experienced personnel and appropriate equipment. To assess this availability a survey of 30 middle-grade trauma surgeons representing 12 hospitals in the South West was undertaken. Ninety per cent of these surgeons understood the indications for the external pelvic fixator but only 70 per cent felt able to apply a frame without senior assistance. A further 23 per cent felt clear diagrams would be necessary. Of the 12 hospitals surveyed only one had an external fixator available in the Accident and Emergency (A&E) department. Only four had a dedicated pelvic fixator within the hospital. Five hospitals had no appropriate external pelvic fixator at all. Although the majority of middle-grade trauma surgeons understand the indications for pelvic fixation, a significant proportion are not confident to apply currently available models. Furthermore, in the majority of hospitals appropriate equipment was not readily available. These results suggest there is considerable scope for improvement. A protocol and dedicated pelvic fixator with clear diagrams in each A&E department may enhance the early management of these life-threatening injuries. D. A. Lloyd, H. Carty, D. A. Roe, M. Patterson (Liverpool, UK) The value of skull X-rays in head-injured children Trauma guidelines recommend a skull X-ray (SXR) in most children with a head injury. As a result, a large number of patients attending Accident and Emergency (A&E) departments with a head injury receive an SXR. We believe that most of these are unnecessary and contribute little to management, as a skull fracture is a poor predictor of intracranial injury in children with blunt head injury and does not influence management decisions. From February 1993 to January 1995 we prospectively documented all children admitted to Alder Hey Children’s Hospital with a head injury. During this period it was policy for all children with a skull fracture to have a computed tomography (CT) scan of the head. CT was also performed when indicated for neurological reasons. Patients with a skull fracture and/or neurological symptoms and signs were admitted. During the study period 6011 children had a skull X-ray of which 162 (2.7 per cent) demonstrated a skull fracture. There were 849 children admitted with a head injury. A CT scan was obtained for 106 (62 per cent) of patients with a skull fracture and 15 were found to have a brain injury or an intracranial collection. Eleven of these 15 had a CT scan for neurological indications; the four positive CT scans in neurologically normal children were a minor haemorrhage in three and a possible infarction beneath a clinically depressed fracture in one. There were 56 children with a fracture who did not have a CT, the significant reason being initial failure by the A&E or surgical staff to recognize the skull X-ray in 34 (21 per cent of total fractures). A total of 34 children without a fracture had a CT for neurological indications and seven of these showed a significant