a controlled trial of prehospital advanced life support in trauma

ORIGINAL CONTRIBUTION emergency medical services, trauma, advanced life support; trauma, prehospital, advanced life support

A Controlled Trial of Prehospital Advanced Life Support in Trauma

We compared the outcome of 472 trauma patients who required ambulance attention and who received prehospital advanced life support (ALS) with another similar 589 patients who received only basic life support (BLS). Nontrapped, critically injured ALS patients were treated for an average of 13 minutes at the scene of injury, compared with 17 minutes for BLS cases ~ < .05). Seventeen of 37 ALS deaths (36%) occurred within 24 hours of injury, compared with 24 of 33 BLS fatalities (73%) (P < .05). However, the overall case fatality rate was similar in the two groups, and regression analyses did not demonstrate an impact of ALS care on mortality, ALS resuscitation did not reduce the duration of hospital or intensive care unit stay, or the incidence of disability after head injury. However, the incidence of respiratory failure in the critically injured patients was 5% (ALS) and 19% (BLS) (P < .025). ALS care appeared to influence patient outcome during the first 24 hours after injury, but had little impact on the later clinical course. Our sample size was too small to rule out any effect of ALS on in-hospital mortality. However, the improved 24-hour survival associated with ALS care suggests some benefit of prehospital resuscitation in major trauma. [Potter D, Goldstein G, Fung SC, 8elig M: A controlled trial of prehospital advanced life support in trauma. Ann Emerg Med June 1988;17:582-588.]

INTRODUCTION Trauma is a major health problem in Australian urban areas. There were

370 fatalities from road accidents in the Sydney metropolitan area (popula- tion 3,166,000) in 1984.1 A number of cities including Sydney and Melbourne have introduced prehospital advanced life support (ALS}. ALS personnel assist regular ambulance officers (basic life support [BLS] officers) in the prehospital management and transport of serious emergencies, including trauma. Compared with BLS personnel, ALS personnel receive more extensive training in the recognition of symptoms and signs of major injury. They are able to initiate treatment with additional measures not available to regular ambulance officers: intubation, cannulation and administration of IV fluids, use of pneumat ic ant ishock garments (MAST), decompression of tension pneumothorax, treatment of the crush syndrome, and defibrillation and drug treatment for cardiac arrest.

There has been no definitive demonstration either in Australia or else- where of an improved outcome in trauma with ALS care. 2 Studies that sug- gest such a result have been based on small data sets,3, 4 or have been prelim- inary in nature, s Similarly, the finding that ALS resuscitation is "a failed concept" was based on small case numbers, wi thout the use of control groups for comparison. 6 Also, controversy exists about what procedures are useful in prehospital trauma care2, 6 and whether ALS intervention delays transport to hospitals, adversely affecting outcome.3,5, 6 Trunkey poses the question, "Is it justifiable for most communit ies to invest in ALS (advanced life support) training for prehospital trauma care?" and concludes, "I do not believe so. "2

ALS for trauma is only resuscitative and not a definitive therapy.5, 6 If any group is amenable to ALS, it is the intermediate severity of injury group (ie, neither overwhelmingly fatal, nor mild injuries);3, s injury severity,3-s, 7 age,3, 7 response time, treatment time on scene, transport time, t ime to definitive care, 3,5,6 type of trauma,8, 9 and quality of hospital care lo all affect outcome.

David Potter, BSc* Gregory Goldstein, MB, FRACP* SC Fung, CPH, MPhil* Sydney, Australia Murray Selig, MB, ChB (Otago),

FFARACS, FFARCS (Eng), FFARCS (I)t St Leonards, Australia

From the School of Public Health and Tropical Medicine, University of Sydney, Sydney, Australia;* and Royal North Shore Hospital, St Leonards, Australia.l-

This study was assisted by a Health Services Research and Development Grant awarded by the Australian Institute of Health.

Address for reprints: Gregory Goldstein, MB, FRACP, Environmental Health in Rural and Urban Development and Housing, Division of Environmental Health, 1211 Geneva, 27-Switzerland.

17:6 June 1988 Annals of Emergency Medicine 582/55

ADVANCED LIFE SUPPORT Potter et al

We tested the hypothesis that ALS had a favorable impact on mortal i ty and morbidity by comparing the outcomes in t r auma pat ients in a city where prehospital ALS is not available (Brisbane) and one where it is (Syd- ney).

METHODS Ambulance Systems

The S y d n e y m e t r o p o l i t a n area served as the s tudy c o m m u n i t y for cases receiving ALS care. The New South Wales Ambulance Service pro- vides a two-tiered response to emergency calls. Forty ambulance stations are staffed by BLS officers. Ten of these also have a mobile unit staffed by ALS officers on a 24-hour basis. BLS officers are trained in basic resuscitation, spl int ing of fractures, and movement of spinal injuries. ALS officers receive an additional 760 hours of training in the recognition and in- vasive treatment of emergencies.

Medical control is achieved by 58 protocols , each l is t ing an order of treatment for a given injury or set of observations, and the circumstances in which immediate transport is to be instituted. Ambulance controllers tri- age phone calls for assistance, dis- patching ALS units to cases likely to be severe. BLS officers at an accident scene also may reques t ALS assistance.

The Br i sbane /Gold Coas t Metro- poli tan areas (combined populat ion, 1,262,000) served as the study commu- nity for the control group receiving BLS care only Seven centers of the Queensland Ambulance Transport Bri- gade maintain 24 ambulance stations staffed by BLS officers with compara- ble training to those in Sydney. Doc- tors may assist with prehospital care in individual cases. Ambulance controllers dispatch the nearest available ambulance to an incident.

Selection of Cases Both mechanical and thermal inju-

ries (bums) were included. In the ALS group, consecutive ambulance report forms detailing ALS treatment of trauma patients in Sydney during a three- month period from February to May 1984 were prospectively reviewed. All cases admi t t ed to the hospi ta l for more than 24 hours and/or who died after ambulance treatment was com- menced were included, with the exception of 32 of 504 cases in which details were unavailable due to legal

TABLE 1. Distribution of injury severity

ISS ALS BLS No. (%) No. (%)

0-8 148 (31) 199 (34)

0-15 177 (37) 257 (44)

16-24 51 (11) 59 (10)

25-40 69 (15) 48 (8)

41-75 27 (6) 26 (4)

Total 472* (100) 589 (100)

*This figure includes three ALS cases who had no vital signs on arrival of the ambulance, but on whom resuscitation was attempted without success. These cases were excluded from the subsequent analysis. There were no such cases in the BLS group.

and ethical considerations. Complete data were recorded for the remaining 472 ALS patients.

In the BLS group, during another three-month period from July to Oc- tober 1984, consecutive ambulance report forms noting BLS management of traumatic conditions in the Brisbane/ Gold Coast area were prospect ively studied and similarly followed Up. Full information was recorded for all 589 cases a d m i t t e d to the hospi ta l for more than 24 hours and/or who died after ambulance management began.

Recording of Data Ambulance report forms and con-

trollers' radio logs were used to obtain the response t ime (the t ime interval from receipt of the emergency call to the arrival of the ambulance responsi- ble for treatment), the treatment time at the scene, and the transport time to hospi ta l . En t rapment , hazards, and mul t ip le v ic t im s i tuat ions delaying transport were recorded, as were effec- tive first aid and the cause of injuI3a. The total time from the initial emergency call to arrival at a hospital with appropriate specialist facilities for the injuries sustained was calculated and termed the time to definitive care. If a patient required acute transfer (within 48 hours) from the initial receiving hospital for further management, this time interval was included in the time to definitive care.

All 40 public hospitals in the study areas receiving tral~ma cases cooper- ated in making patient records available, and all relevant autopsy reports were reviewed. Patient age, sex, and details of injuries were recorded. Only clinically confirmed injuries noted in operation and autopsy reports or reli-

able d iagnos t i c tes t s ( radiographs, c o m p u t e d t o m o g r a p h y scans, an- giography, pyelograms, etc) were included, and discharge summaries were not used.

Injury scoring using the Abbreviated Injury Scale (AIS) and Injury Severity Score (ISS) systems was performed by one of the authors and was rescored blindly in a subset of cases by a senior surgeon experienced in t r auma and AIS and ISS scoring. Major injuries were defined as those of AIS score of 4 or 5 severity, and cases with an ISS of 25 or more were classified as critically injured.

Information on outcomes also was derived from hospital and autopsy records. Cases dead on arrival at hospital or who died during hospital stay were noted. The cause of death n and the t ime interval f rom the initial emergency call to death (time of survival before death) were recorded. Estimates of morbidity were made based on duration of continuous hospital and ICU stay. Resp i ra tory and renal fai lure were defined by blood gas and bio- c h e m i s t r y resul ts , u Res idua l disability of major head injury was as- sessed usin K the Glasgow Ot i tcome Scale12,13 thre 'e 'months after the accident.

Data analysis was performed using the SPSS-X package for general fre- q u e n c y d i s t r ibu t ion and mu l t i -way cross-tabulation table production, and the GENSTAT package for the multi: ple regression and logistic regression analyses.

RESULTS Description of Cases

The distribution of injury severity in the study groups is shown (Table 1).

56/583 Annals of Emergency Medicine 17:6 June 1988

Proportion of Critically Injured (•)

60 5g.4

50

40 3g.2

;O I

20

10

0 - - HeadiNec k Abdominal

47.3 ~1 Advanced Life Support n=96

Basic Life Support n=74

6.3 8.1

Thoracic Spinal Extremity Burns Facial

Region of Major Injuries

Cumulative Proportion of Deaths (g) too

75

50

25

100.0

D.O.A. ot hospitoI

less then less then less then Iess then less than less then less than 1 hour 2 hours 6 hours 24 hours 3 days 7 days 14 days

Time interval from Ambulance Notification to Death 2

The mean ISS of ALS cases was 14.5 (SD, _+ 13.3); that of BLS cases was 12.2 (SD, + 11.6). Using a t test on continuous values, ALS cases had a s igni f icant ly h igher ISS (t, 3.0; df, 1,059; P < .005). Ninety-six ALS and 74 BLS patients were critically injured (Iss > 25).

Crit ical ly injured pat ients in the ALS and BLS groups are compared (Table 2). No significant differences in the distributions of injury severity, patient age, and time to definitive care could be demonstrated using t tests on continuous values. ALS officers spent on average 13 minutes treating nontrapped critically injured at the inci-

dent scene compared with a mean of 17 minutes for BLS personnel manag- ing similar cases (P < .05).

A higher proportion of critically injured sustained major head/neck injuries in the ALS group (59%) compared with the BLS group (39%), as shown (Figure 1). The difference was significant (chi-square [1, df], 6; P < .025). Similarly, 47% of critically injured in the BLS group sustained major thoracic injuries, nearly double the proportion of ALS cases (chi-square [1, d)q, 8.2; P < .005).

Doctors assisted with 17 of 74 critically injured cases (23%) in the BLS group. They were either general practi-

FIGURE 1. Critically injured: Regions of major injuries (AIS, 4 or 5).

FIGURE 2. Fatal cases: Duration of survival.

tioners or junior hospital staff called in on an ad hoc basis. They carried a m i n i m u m of e q u i p m e n t and performed very few ALS-type interven- tions such as endotracheal intubation or application of MAST

Mortality A total of 73 patients in the study

died. Three of these had no vital signs at the commencement of ambulance treatment. All three were ALS cases with an ISS of 75 (unsurvivable). They were excluded from further analysis as resuscitation was only attempted because of the availability of ALS care. Such cases were not treated by the BLS officers.

Of the remaining 70 deaths, 37 occurred in the ALS group and 33 in the BLS group. In cri t ical ly injured patients there were 33 ALS deaths (case fatality, 35%) and 30 BLS deaths (case fatality, 41%) (difference not significant; chi-square [1, df], 0.26). Of the seven deaths in patients who did not sustain critical injuries, six occurred in persons 70 years of age or more.

Brain injury was the cause of death in 14 ALS (35%) and 11 BLS cases (33%). Another 14 ALS (35%) and ten BLS deaths (30%) occurred from hemorrhage. Seven ALS (18%) and five BLS cases (15%)died after organ failure/ sepsis. The remaining five ALS deaths (including the three who had no vital signs on arrival of the ambulance) and seven BLS fatalities had unsurvivable injuries (ISS of 75).

The t ime of survival before death for fatal cases is shown (Figure 2). Sev- enteen of 37 ALS cases (46%) died within 24 hours of injury compared with 24 of 33 BL8 cases (73%). This difference was significant (chi-square [1, df], 4.1; P < .05). Sixteen ALS cases (43%) died before being admitted to an ICU, a significantly lower proportion than that of the BLS group, where 26 fatalities (79%) occurred before ICU admission (chi-square [1, df], 7.7; P < .01).

The increased survival t ime before death of ALS fatalities also was demonstrated in the subsets of cases dying due to brain damage and hemorrhage. However, the proportion of deaths oc-



TABLE 2. Variables in critically injured patients (ISS > 25)

ISS SD

Mean age (yr) SD

Mean treatment time at the scene (min)*

SD

Mean time to definitive care (min)t

SD

Proportion of cases requiring interhospital transfer (%)

Proportion of cases with blunt trauma (%)

Proportion of cases road accident rebated (%)

*Excluding entrapped patients. tExcluding values > 360 minutes.

ALS BLS (n = 93) (n = 74) t Value df Significance

37 37 0.7 165 NS 1.3 1.6

32 34 0.003 165 NS 1.8 2.4

13 17 2.8 7.9 8.7

59 80 - 1.9 57.5 72.6

Chi-Square

32 27 0.5

96 96 0.1

86 72 7.34

127 P < .05

164 NS

df Significance

1 NS

1 NS

1 P < .01

curring before a definitive care hospital was reached was similar in both groups (38% of ALS and 36% of BLS fatalities). This suggests that the higher proportion of BLS fatalities within 24 hours occurred between arrival at a definitive care hospital and admission to the ICU.

Morbidity A higher incidence of respiratory

failure was noted in the BLS group of cr i t ica l ly injured (Tables 3 and 4). Fourteen of the 74 BLS cases (19%) de- veloped this complicat ion, as compared with five of 93 ALS cases (5%), a significant result (chi-square [1, df], 6.2; P < .025). The difference re- mained significant when only major thoracic injuries were included (chi- square [1, df], 6.8; P < .01). All other measures of morbidity were similar in the two groups (Table 4). Performance of t tests on continuous values showed no significant difference in dura- t ion of hospi ta l and ICU stay, nor could any reduct ion in head-injury disabil i ty be demons t ra ted (using a chi-square test). Only three cases of acute renal failure occurred.

Multiple and Logistic Regression Analyses

Analysis was undertaken using the

following independent variables: case type (ALS or BLS), injury severity, patient age, sex, total time to definitive care, and various severity factors re- lated to the accident (ie, degree of first aid, cause of injury, number of injured persons, entrapment, hazards, transport by helicopter). The dependent variables were death during hospital stay, duration of ICU stay, duration of hospital stay, and incidence of respiratory failure. For duration of ICU and hospital stay, multiple regression analysis was used, with fatalities excluded and case type treated as a contributory variable. A final best fit regression was ob ta ined by the forward se lec t ion method.

Logistic regression techniques were used for the d i cho tomous ou t come measures "death during hospital stay" and "incidence of respiratory failure" (yes/no outcomes). Results of the regressions for critically injured cases are presented (Table 5). Case type was not found to be a significant independent variable in duration of ICU and hospital stay or death during hospital stay. The BLS case type was associated w i t h the i n c i d e n c e of r e sp i r a to ry failure (P < .01).

Reliability of Severity Scoring The mean ISS score for the research

officer was 18.1 compared with 19.1 for the external consultant, wi th a cor- relation coefficient of 0.95. A paired t tes t of the difference be tween the scores showed the average difference of one ISS point was signif icant (t, 1.93; df, 98; P < .03).

DISCUSSION The hypothesis that ALS care might

reduce the case fatality rate was not confirmed in this study. The most significant finding of the study was that 54% of ALS deaths occurred more than 24 hours after injury compared with 27% of BLS fatalities. There are a number of possible explanations for this result. BLS cases may have been more severely injured or may have sustained injuries of a type associated with a more rapid deterioration; delays in reaching definitive care in BL8 cases may have adversely affected outcome; the ALS group may have received a higher standard of acute care on arrival at hospital ; and ALS re- susc i t a t ion m a y have reduced the mortal i ty within 24 hours of critical mjury

Both severity and age distributions in the two groups of critically injured were very similar, with the mean ISS scores being identical. In addition, the validity of severity classification was


TABLE 3. Respiratory failure in critically injured patients

Region of Major Injury (AIS, 4 or 5)

Thoracic Burns Other Total Number of cases respiratory failure

ALS 1 2 2 5 BLS 13 1 0 14

Total critically injured

ALS 24 3 66 93 BLS 35 3 36 74

Percentage of cases with respiratory failure (a/b x 100)

ALS 4 67 3 5 BLS 37 33 0 19

TABLE 4. Measures of morbidity in critica]ly hTjured patients

Mean length of stay Hospital (days)* SD

Mean length of stay ICU (days)* SD

ALS BLS

34 36 20.5 22.1

5 6 3.5 3.8

Head injuries with residual disabil i tyt 15%

Patients with respiratory failure 5%

Thoracic and burns injury cases with respiratory failure 11%

Thoracic injury cases with respiratory failure 4%

*Calculated on surviving patients only.

14

19

37

37

t Value df Significance

0.4 100 NS

- 0 . 4 81 NS

Chi-Square df Significance

0.02 1 NS

6.21 1 P < .025

4.16 1 P < .05

6.8 1 P < .01

tPersistent vegetative state, severe or moderate disability, as a proportion of critically injured head injury patients.

tested with a high degree of correla- t ion wi th an independent observer. There was a difference in injury pat = tern, with more major thoracic injuries in the BLS group and more major head injuries in ALS patients. It could be argued that severe chest injuries are more rapidly fatal than severe neu- rotrauma. However, the comparative proportions of fatalities due to hemorrhage were similar, suggesting that injuries causing rapid exsanguina t ion were not more c o m m o n in the BLS group.

An analysis of the specific injuries shows a similar frequency of tension pneumothorax and cardiac/great ves- sel trauma. Flail chest and myocardial

con tus ion were the injuries wi th a higher incidence in the BLS group. These are not rapidly fatal with ade- quate management. Furthermore, the improved survival rate of ALS cases in the first 24 hours was noted for each of the causes of death, including hemorrhage and brain injury..

The BLS group did appear to experi- ence a longer delay in reaching definitive hospital care, with a mean time of 80 minutes (SD, -+ 72.6) compared with 59 minutes (SD, + 57.5) for ALS cases. The difference was not statis- tically significant; however, the large standard deviation may explain this. We doubt that this delay was responsi- ble for the increased early mortality.

The logistic regression analysis did not show any rela t ionship between time to definitive care and mortality. Because only urban areas were studied, arrival at a hospital invariably occurred within two hours of the injury, and delays of a few minutes in this setting were probably of little significance for most cases.

As regards the subset of very rapidly deteriorating patients, it would have been expected that a prolonged time to definitive care would have resulted in more of such cases dying en route to the eventual treating hospital. This was not the case as similar propor- t ions of fatalities (38% of ALS and 36% of BLS deaths) occurred before ar-



TABLE 5. Regression coefficients of independent variables significantly influencing outcome in critically injured

Logistic Regressions Equations Death During Chi- Respiratory Chi- Hospital Stay Square Failure Square

Constant - 6.55 37.2 - 3.58 28.4

ISS 0,1263 28.3 - -

Age 0,0299 6.1 0,025 3.8

Case type - 0 . 0 9 6 (NS)* 0,044 1,443 6.7

Signi f icance level P < .001 P < .01

Multiple Regressions Equations Length of Length of ICU Stay F Hospital Stay

Constant - 3.76 1.04 68.4

ISS 0.278 7.29 - -

Age 0.0885 4,45 - -

Case type - 1.05 (NS) 0.51 - -

Adjusted R square (%) 8,6 R square (%) 11,9

*NS, not significant.

F 25.7

rival at such a hospital; the increased mortality in the BLS group occurred after arrival.

Definitive care for critical injuries almost always involved a teaching hospital. Management was performed by senior surgeons, anesthetists, and emergency care specialists with aca- demic appointments. We believe that there are no differences in management or the availability of advanced diagnostic or therapeutic technology be tween Brisbane hospi ta ls (BLS group) or the Sydney (ALS group) hospitals, although this possibility cannot be ruled out without additional study.

ha our view, the most likely expla- nation is that ALS care was responsi- ble for the reduced early mortality. This result must be accepted with caution, as many tests were performed causing the functional statistical signif icance levels to be lower than stated. Many more of the early BLS deaths occurred during resuscitation, diagnostic work-up, and surgery at teaching hospitals.

Researchers have found an associa- tion between ALS care and improved vital signs on hospital arrival.~, s Un- fortunately, such information was unavailable for BLS cases in this study, as neither BLS personnel nor Brisbane hospitals routinely collected the information necessary to derive a Trauma Score. 14 However, we believe that a

similar effect occurred here, with BLS cases arriving with a poorer respiratory and cardiovascular function than their ALS counterparts.

Most of the critically injured were young pat ients with some init ial physiological response to hypoxia and hypotension. However, in the absence of early resuscitation, compensatory mechanisms tended to fail during the time interval to definitive care. BLS fatalities often arrived at teaching hospitals in a premorbid state, with respiratory or cardiac arrest occurring within minutes of arrival. Shock ,usu- ally proved refractory to treatment.

No matter how skillful, the success rate of emergency surgery on patients with blunt trauma in such a condition is minimal.S,9 Little diagnostic information is available about the multi- plicity of injuries, and the patient presents a great anesthetic risk. In contrast, ALS cases responded to further hospital resuscitation and many more survived surgery to reach the ICU.

C O N C L U S I O N N o difference could be demon-

strated in the overall mortality and morbidity of the two groups of critically injured, with the exception of the higher rate of respiratory failure in BLS cases. This suggests that the impact of ALS care is mainly on the

clinical course during the first 24 hours. ALS care in trauma is not definitive, and its main role is to tem- porarily maintain homeostasis until surgical repair or other supportive treatment can take place. Once this has occurred, a new set of management problems and potential com- plications arises in which prehospital treatment is of.lit t le consequence. Does this mean that ALS care is not worthwhile? It could be argued that ALS treatment merely delayed a fatal outcome that was inevitable due to the severity of injury.

On the other hand, further advances and refinements in hospital management may ultimately result in the survival of such patients. One-third of cases died from hemorrhage and one- sixth from organ failure and sepsis, both poterttially manageable problems. Establishment in the future of dedicated trauma centers in Australia may influence overall outcome. How- ever, the overwhelming proportion sustaining blunt trauma suggests that the prognosis may remain poor, as most of the success with aggressive hospital management thus far has been achieved in victims with pen- etrating injuries.8,9

The power of our study was rela- tively weak in regard to mortality. Only small numbers of cases with critical injuries (96 ALS and 74 BLS)


were at r i sk of death , of w h o m on ly 40 ALS a n d 33 BLS p a t i e n t s died. A s t u d y sample size of a b o u t 500 cr i t ica l in juries in e a c h group w o u l d be n e e d e d to de tec t a 10% di f ference in i n -hosp i t a l m o r t a l i t y due to ALS care w i t h a 90% power. Therefore , we be l ieve t h a t t h e i m p a c t of ALS care n o t e d in t he f i rs t 24 h o u r s a f te r in ju ry jus t i f ies t he on- go ing p r o v i s i o n of ALS s e r v i c e s for t r auma , w i t h s o m e con f idence t h a t an overall i m p r o v e m e n t in m o r t a l i t y m a y even tua l ly be d e m o n s t r a t e d .

Finally, t w o o t h e r i m p o r t a n t f ind- ings s h o u l d be discussed. T h e r e was a r e d u c e d i n c i d e n c e of r e s p i r a t o r y fa i lure i n t h e ALS c r i t i c a l l y in ju red , even a l l o w i n g for t h e lower f r e q u e n c y of t h o r a c i c in ju r i e s . T h i s m a y be at- t r i b u t e d to a n i m p r o v e d r e c o g n i t i o n of s y m p t o m s a n d s igns of c h e s t t r a u m a and h y p o x i a b y ALS staff, a n d t h e rou- t ine a d m i n i s t r a t i o n of h igh - f low oxy- gen to al l se r ious ly i n ju red pa t i en t s .

S e c o n d l y , ALS o f f i c e r s a c t u a l l y

s p e n t l ess t i m e t r e a t i n g n o n - t r a p p e d cr i t ica l in ju r i e s a t t he s cene t h a n d id t h e i r BLS c o u n t e r p a r t s . T h e s t r i c t gu ide l ines o u t l i n i n g c o n d i t i o n s for i m - m e d i a t e t r a n s p o r t appear to h a v e b e e n f o l l o w e d . M a n y ALS i n t e r v e n t i o n s were p e r f o r m e d e n r o u t e to hosp i ta l , a n d o f t e n p r o c e d u r e s t h a t w e r e ind i - ca ted were n o t p e r f o r m e d due to t i m e cons t r a in t s .

REFERENCES 1. Traffic Authority of New South Wales: Road Traffic Crashes in New South Wales. Sydney, Traffic Authority of NSW, 1985.

2. Trunkey DD: Is ALS necessary for prehospital trauma care? (editorial). J Trauma 1984; 24:86-87,

3. Hedges JR, Sacco WJ, Champion HR: An analysis of pre-hospital care of blunt trauma. ] Trauma 1982;22:989-993.

4. Compton J, Little JM: Role of the intensive care ambulance in the transport of accident victims. Aust NZ J Surg 1983;53:435-438.

5. Jacobs LM, Sinclair A, Beiser A, et al: Pre- hospital advanced life support: Benefits in trauma. [ Trauma 1984;24:8-13.

6. Smith P, Bodai BI, Hill AS, et ah Pre-hospital stabilization of critically injured patients: A failed concept. J Trauma 1985;25:65-70.

7. Baker SP, O'Neill B, Haddon W, et al: The In- jury Severity Score: A method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187-196.

8. Bodai BI, Smith JP, Blaisdell W: The role of the emergency thoracotomy in blunt trauma. ] Trauma 1982;22:487-491.

9. Baker CC, Thomas AN, Trunkey DD: The role of emergency thoracotomy in trauma. J Trauma 1980;20:848-855.

10. West JG, Trunkey DD, Lira RC: Systems of trauma care: A study of two counties. Arch Surg 1979;114:455-459.

11. Goris RJA, Draaisma J: Causes of death after blunt trauma. J Trauma 1982;22:141-146.

12. Jennett B, Bond M: Assessment of outcome after severe brain damage. Lancet 1975;1: 480-484.

13. Teasdale G, Knill-Jones R, Van Der Sande J: Observer variability in assessing impaired con- sciousness and coma. J Neurol Neurosurg Psy- chiatry 1978;41:603-610.

14. Champion HR, Sacco WJ, Carnazzo AJ, et ah Trauma Score. Crit Care Med 1981~9:672-676.

17:6 June t988 Annals of Emergency Medicine 588/61

a controlled trial of prehospital advanced life support in trauma

Documents