INJURY SURVEILLANCE AT A LEVEL I TRAUMA CENTRE IN JOHANNESBURG, SOUTH AFRICA

An analysis of 16 357 trauma patients seen over a one year period at the trauma casualty of an academic hospital in Johannesburg was carried out to determine the profile of injuries sustained by victims in the Johannesburg region. A retrospective survey was conducted between January and December 2001 to compile a composite trauma morbidity and mortality profile, and to create baseline data for future comparison. The objectives of the survey were: to describe the frequencies, distribution and categories of injuries; to assess, on admission to the trauma casualty, the severity of injuries according to the TRISS method; and the outcomes and/or placement of patients after initial treatment in the trauma casualty. Guided by a structured checklist, data were collected by reviewing trauma registers and patients’ documents. The TRISS method was used to determine injury severity and descriptive statistics were used to present and describe the results. A preview of the survey results indicates that males are a high-risk category for trauma, particularly over weekends, during their nocturnal activities. More than two thirds of all patients sorted in the 16-35 year age group. Injuries to the limbs and head and neck regions accounted for the highest percentage of cases with assault or interpersonal violence a major cause in an estimated 70% of cases. More than 60% of a random sample of 163 patients had sustained serious injuries with an ISS between 16 and 75; the majority however had a survival probability (Ps) of > 50%. This paper describes the methodology and results of the survey in relation to a proposed long-term injury surveillance project.


INTRODUCTION AND BACKGROUND
Trauma is ranked as the second most important cause of all deaths in South Africa, with violence related deaths as a central feature. A homicide rate of 61 per 100 000, compared to that of 10 per 100 000 in the United States of America (USA), places the South African society among the most violent in the world (Marais, 1998:1). The collective impact this has on the economy, the health services and societal stability is incalculable. The majority of trauma patients are young and productive participants in the country's economy, and who, as a result of their injuries, incur high personal, social and economic costs. In the USA trauma has been described as the most serious public health problem (Cardona, Hurn, Mason, Scanlon & Veise-Berry, 1994:28), a position which the South African trauma scenario is rapidly surpassing.
Locally, trauma is a significant community health and primary health care problem, yet trauma care services and the Ambulance and Emergency Medical Services (AEMS) are generally not recognised as important links in the chain of primary health care (Fuhri, 1998:36). The infancy stage in which the Primary Health Care (PHC) services find themselves increases the strain on hospital services mainly due to geographic inaccessibility, to unavailability (in service hours) and the irregular supply, and sometimes the lack of essential drugs, equipment and appropriately skilled personnel at PHC clinics. Continued budgetary cuts to hospital services have made it virtually impossible to provide primary emergency care to the most needy.
As to be expected there has been no proportionate

LITERATURE REVIEW
By definition a trauma centre operates in a designated hospital and provides resuscitation and other forms of surgical and non-surgical emergency treatment (Cardona et al. 1994:29). Trauma centres are further stratified into three levels according to specific criteria, which relate to the functions/services provided, physical facilities and clinical expertise available. Criteria for level II and level III trauma centres are variable but a common feature is the referral system, which exists between these levels and a level I trauma centre. Level I trauma centres provide 24-hour emergency services by qualified trauma surgeons, physicians and nurses, and are active in teaching, research and systems development. The development of efficient and cost effective trauma systems is thus a major function and can be effected and sustained by appropriate surveillance mechanisms to determine comprehensive trauma patterns and profiles. In surveying injuries, Van der Spuy (1996:10) is of the opinion that a simple head count of trauma cases would not suffice. Important considerations include injury distribution, categories and outcomes, and the severity of injury, most of which are determined by the mechanism of injury, and necessitate a cursory overview.

Mechanism of injury
Relating to the injuring force and subsequent tissue response, the mechanism of injury is useful to explain the nature of the injury, identify common combinations of injury, and predict eventual outcome (Cardona et al. 1994:94) according to the severity of injury. How the body responds to the injuring force is determined by the interplay between biophysical factors, for instance, force and velocity, the shape of the object and tissue rigidity. The force inflicting injury may be blunt or penetrating, causing a varying amount of energy to be delivered to the area of contact, at varying velocity.
Blunt or non-penetrating trauma is caused by direct or indirect forces from traffic accidents, falls, assaults and contact sports. Direct impact delivers a concentration of energy to the point of contact, causing greatest injury at the site. An indirect force on the other hand, causes energy to be dissipated internally over a larger area. The extent of the damage in both instances however, is dependent on the force velocity; the greater the velocity, the greater the damage to tissue. Dependent on the cause, blunt trauma usually involves a combination of forces, notably acceleration, deceleration, shearing and crushing forces. Hence multiple injuries are common (Cardona et al. 1994:95).
In penetrating trauma the mechanism of injury is the energy generated and dissipated by the object such as a knife, into the surrounding tissues (Cardona et al. 1994:104). The extent of the damage is directly proportional to the amount of energy created (kinetic energy) of which velocity in an important factor. Whilst several factors, such as distance from the force, tissue density and site of penetration, are considered in estimating the extent of tissue damage, it is generally agreed that high-velocity forces have the potential of inducing injuries with a high index of severity. Common types of penetrating injuries include stab wounds, gunshot wounds and impalements from falls and forceful collisions with an object.
Apart from biomechanics, associated risk factors such as age, sex, geographic location, alcohol consumption and many more, play a significant role in trauma and injury severity. Within the context of this study, the age and sex of trauma patients were established.
The question of age is found to be particularly significant in trauma, based on two assumptions around demography and morbidity and mortality. Firstly, older people are less inclined to sustain injuries than their counterparts younger than 50 years, and secondly, injuries in the older patient are likely to be more severe and carry a higher mortality than in younger victims. Age as an important determinant of human physiological reserve and hence patient survival, is discussed with the TRISS method of analysing injury severity.

Injury severity: The TRISS method
The TRISS method gives a physiological and anatomical index of injury severity based on the Injury Severity Score (ISS), the Revised Trauma Score (RTS), the patient's age and whether the injury was blunt or penetrating (Woodford, 2001:5). The ISS assigns numerical scores to body regions injured, in order to determine the extent of multiple injuries and correlating these with mortality risk. The ISS ranges from 1 to 75.
When the ISS is 25, the mortality risk is minimal (approximately 25%), but thereafter the mortality risk increases almost linearly with the injury score; a score of above 70 being close to 100% mortality.
The second parameter, the Revised Trauma Score . The sum of these products consti-tutes the RTS, which comprises a range of scores from 0 to 8. The lower the RTS, the lesser the probability to survive. If the RTS is used to triage in the pre-hospital environment, it is suggested by several authors that a trauma patient with a value of less than 4 in any parameter, be taken to a trauma centre (Boyd et al. 1987:373).
As stated earlier, age is an important risk factor in trauma. Age of the trauma patient is considered a significant determinant of physiological reserve, which in turn is related to patient survival (Bion, 1993:11). Studies have shown that cardiovascular compromise associated with increasing age, particularly greater than 55 years, relates significantly to probability to survive.
In addition to age, the mechanism of injury (blunt or penetrating) produces different physiological derangements and injury characteristics (Woodford, 2001:6) with resultant influence on the patient's probability of survival.
In combining these four parameters, the TRISS method is useful to quantify the probability of survival (Ps) and to evaluate the outcomes of trauma care. In this study the TRISS method was applied to determine trauma patients' Ps. Woodford (2001:5) however, cautions that Ps is merely a mathematical calculation, which indicates a patient's probability of survival and should not be regarded as an absolute measure of mortality.

PROBLEM STATEMENT
Several factors which complicate trauma systems development and trauma practice have been identified; the most pertinent being the low priority given to trauma care in the public sector (Van der Spuy, 1996:7). This in turn results in unrepresentative profiles of injury mortality, which may serve as a skewed basis for decision-making. A surveillance project was begun to provide a comprehensive description of the nature and severity of injuries, including mortality data, at a Level 1 Trauma Centre in Johannesburg. The use of the trauma casualty as opposed to hospital wards as the first point of entry into the hospital, was anticipated to provide a more representative profile of trauma. The provision of a severity profile as opposed to a head count of trauma cases will, in turn, enhance the planning of trauma services. According to Van der Spuy (1996:10), severity profiles are best done by using objective criteria such the Glascow Coma Scale for head injuries, the RTS and the Injury Severity Score (ISS) or an equivalent for assessing trauma. As part of a surveillance project, the study set out to provide comprehensive data for profiling trauma patients and determining their injury severity.

PURPOSE AND OBJECTIVES OF THE STUDY
The purpose of the study was to describe the morbidity and mortality profile of patients seen, over a one year period, at the Trauma Casualty of an academic hospital in Johannesburg.
The main objectives of the study were to: • describe the frequencies, distribution and categories of injuries seen in the trauma casualty over a one year period; • assess the severity of injuries, according to the TRISS method, on admission to the trauma casualty; and • determine the patient outcomes and/or placement after initial treatment in the trauma casualty.

POPULATION AND SAMPLE
The study population comprised all patients seen at the trauma casualty of an academic hospital, as determined by entries in the trauma register over a one

METHODS AND PROCEDURES
A descriptive survey was used to determine the morbidity and mortality profile of trauma patients between January to December 2001. Data were collected by means of a retrospective record review of the trauma register, hospital statistics and patient records. A check-list with items derived from the trauma register and the trauma resuscitation form was used for data recording. A pilot study was done to determine whether the checklist would capture the data needed to meet the study objectives. Twenty patients' records, randomly selected from the trauma register entries over a period of one month, were used to conduct the pilot study.
Data analysis involved the use of descriptive statistics to analyse numerical and categorical data. Indices of injury severity were calculated based on anatomical, physiological, injury and age characteristics to quantify probability of survival. Coefficients in the mathematical formula for these calculations were derived from regression analysis applied to data from the Major Trauma Outcomes Study (MTOS) as described by Boyd et al. (1987:372).

ETHICAL CONSIDERATIONS
Permission to conduct the study and to access hospital and patient records was obtained from the Chief Executive Officer of the Hospital, the Head of the Trauma Unit and the Registered Nurse in charge of Trauma Casualty. Ethical clearance was obtained from the University Committee for Research on Human Subjects to ensure compliance with ethical standards.

Patients' names and other identifying characteristics
were not documented and records were encoded to ensure anonymity and confidentiality during data collection and reporting. The identity of trauma personnel treating patients and recording such treatment, remained confidential throughout the study.

DISCUSSION OF FINDINGS
In discussing the findings it is important to note that years. Refer to Figure 1 for the age distribution according to four categories.

Frequencies and distribution of cases
An analysis of the annual caseload showed a distinct difference between weekdays and weekends in relation to patient numbers and time of day. On weekdays, from Monday to Thursday, a total of 7 961 patients (48.7%) were treated and 8 396 (51.3%) over weekends from Friday to Sunday. While it can be assumed that less patients are seen and treated during the week, the number of cases attended to during the day time period between 08h00 and 18h00, by far exceeds those seen during the same time period over a weekend. Figure 2 compares the daytime distribution of 59.2% and 44.2% between weekdays and weekends respectively. As anticipated, patient frequencies at night, after 18h00 until 08h00 the following morning, are reversed over weekends, when 55.8% of cases were treated as compared to 40.2% in the same time period during weekdays. This pattern is all too familiar to trauma facilities countrywide and is clearly reflected in their statistics. Although these findings have not been analysed on a day-to-day basis, the overall results nevertheless, are consistent with those of the Cape Metropolitan study reported by Van der Spuy and Steenkamp (1996:17). In their study, the weekend loading was considerably higher than weekdays, with Saturdays becoming particularly catastrophic.

Categories of injuries
Injuries were categorised according the body region/s afflicted (Figure 3).Where two or more body regions were involved, poly-trauma was documented. Approximately 1 in 20 patients seen at the trauma casualty sustained multiple injuries, with an annual count of 944 (5.8%). Trauma to the limbs, including hip and shoulder injuries, accounted for 45.2% of all injuries and involved mainly the disciplines of orthopaedic and plastic surgery. Gunshot wounds and injuries sustained during traffic and domestic accidents were the major causes of fractures and dislocations whereas occupational/industrial accidents accounted for a significant demand for plastic surgery to the hands and fingers.
Injuries to the head and neck region accounted for 30.1% of all trauma cases seen over a one year period. Consistent with popular opinion, the head (including the face) and neck are more prone to injury be-  Totals 08h00 -18h00 18h00 -08h00

Results of ISS and TRISS analysis
Serious injuries occurred mainly in the 25 to 34 year age group (n=86), followed by those in the 15 to 24 year age group (n=40). In total, these age groups ac-

Overall patient outcomes
To meet the third study objective, total patient outcomes were determined in relation to patient admis-