|
The costs and benefits of paramedic skills in pre-hospital trauma care |
Nicholl J, Hughes S, Dixon S, Turner J, Yates D |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology The use of paramedic skills, that is, advanced life support (ALS) in pre-hospital trauma care. The authors stated that the training in the UK consisted of 8 weeks of instruction and practice (4-week training period and 4-week clinical placement) in endotracheal intubation, cannulation, and the administration of intravenous fluids and a limited range of drugs.
Type of intervention Other: pre-hospital care after a serious trauma injury.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised patients with a serious trauma injury who need to be picked up by ambulance at the scene after an incident.
Setting The settings were three different ambulance service areas and their corresponding 42 hospitals. The study was carried out in the UK.
Dates to which data relate The effectiveness data were collected between July 1994 and March 1996. Most of the cost data appear to have been collected during the same period. The price year was 1996 to 1997.
Source of effectiveness data The effectiveness data were derived from a single study and an expert panel.
Link between effectiveness and cost data The costing was undertaken on the same patient sample as that used in the effectiveness analysis, but it was unclear whether the costing data were collected prospectively or retrospectively.
Study sample The authors seem to have performed some power calculations in the planning phase of the study to determine the sample size, although the methods used were not reported. Patients were considered for the effectiveness analysis if:
they were admitted to the hospital because of trauma, with a hospital stay of 3 or more nights, or if they were transferred to another hospital for further emergency care with a total length of stay of at least 3 nights;
they were admitted either to an intensive care unit (ICU) or to a high dependency area because of a trauma injury;
they died before arrival at hospital or in hospital because of a trauma injury, but did not die before the ambulance arrived at the scene; or
they were admitted or re-admitted for treatment of their original trauma injuries after more than 2 days from the date of the incident.
Patients were excluded if:
the cause of the incident was poison;
they were transported to hospital by helicopter;
a doctor attended them on scene;
they died at the scene before the ambulance arrived;
they had superficial skin injuries, including simple penetrating injuries;
they had superficial burns, including partial thickness burns of less than 10%;
they were over 65 years old and had an isolated fracture neck or femur or single pubic rami fracture because of a fall of less than one metre, independently of whether or not they died;
if they presented an isolated simple facial injury, including simple eye injuries;
if they presented a simple spinal train (i.e. acute cervical, thoracic or lumbar sprain with no fracture of dislocation); or
if they were involved in 'major incidents', as defined by each respective ambulance service.
Patients from 3 different ambulance service areas were considered in the effectiveness analysis. A total of 2,076 patients met the inclusion criteria. There were 1,112 in area 1, 174 in area 2, and 790 in area 3. In area 1, an EMT attended 32.5% of the cases. In area 2, a paramedic attended 151 cases and an EMT attended 23 cases. In area 3, an EMT attended 213 cases, a paramedic attended 530 cases, and an unknown crew type attended 47 cases. The authors stated that the three areas were chosen in order to represent typical English environments, but they did not provide any evidence that the study sample was representative of the study population.
In total, 1,211 crewmembers were identified in the effectiveness analysis. There were 601 in area 1, 152 in area 2, and 458 in area 3.
The authors reported that they attempted to randomise the dispatch of EMT or paramedic crews to the trauma calls. Patients from area 3 were randomised during the 6 months from September 1995 to February 1996, whenever it was possible. For example, if there were no known restrictions such as the patient being trapped, if there were both EMT and paramedic vehicles available at the time of the call, and if both vehicles were expected to arrive on scene at the same time. The inclusion criteria for this additional study were different from those proposed for the initial study. The authors reported that, otherwise, only 16 cases from the 185 randomised calls would have met the main cohort inclusion criteria. Patients were included if they stayed in hospital at least overnight. Only 30 of the 185 randomised calls met the modified inclusion criteria.
Study design This was a cohort study, which was retrospective for area 1, and prospective for areas 2 and 3. The study was multi-centred as three different ambulance service areas and their corresponding 42 hospitals were included. The period of follow-up was 6 months (unless the patient died before this) in areas 2 and 3, and 5 months in area 1. There were 94 patients who were first attended by EMTs and later by paramedics, and they were considered as EMT cases.
One researcher in each of the areas under study classified the severity of the injuries, and an additional fourth researcher participated for the avoidable death study. All attended a training course in classifying the severity of the injuries.
Analysis of effectiveness The basis of the clinical study was intention to treat. The primary health outcomes assessed in the single study were:
the processes of care given by the EMT and paramedic crews;
the number of patients, among those with a patient report form (PRF), receiving an intervention that could only be administered by a paramedic (cannulation, administration of intravenous fluids, intubation and administration of drugs such as adrenaline, atropine, nubain, heparin sodium in saline or tramadol);
the mean time (and standard deviation, SD) spent on scene by the paramedics and EMTs for both the main cohort and the randomised cases;
the odds of admission to the ICU for paramedic patients relative to EMT patients;
the length of stay;
the crude and adjusted mortality rates, relative risk of dying and odds ratios for the paramedic and EMT patients, both overall and for specific groups; and
morbidity, in terms of the mean number of days off from usual activities, the percentage of patients still restricted after 6 months, and the response rates and SF-36 scores for patients attended by paramedics or EMTs, for patients first attended on-scene by paramedics or EMTs, and by type of injuries. The SF-36 scores were measured for physical functioning, extended physical functioning, social functioning, role physical, role emotional, mental health index, energy/vitality, pain and general health perceptions.
The analysis of processes included 2,059 cases. There were 2,029 patients from the main cohort for whom the pre-hospital care provided was known (there were 47 cases in area 3 with unknown type of pre-hospital care), and 30 cases from area 3 that were randomised to either EMT or paramedic care.
The analysis of outcomes included 2,045 cases, based on the inclusion criteria for the main cohort of patients. Of these, 2,029 cases from the main cohort and 16 randomised cases met the main cohort inclusion criteria.
For the analysis of morbidity, 273 questionnaires were sent out after 6 months to patients first attended on scene by paramedics, and 177 to those first attended by EMTs. The patients were over 16 years old. Questions about restrictions on work, usual activities and general health perception (using the SF-36) were asked.
The authors reported the number and percentage of paramedic and EMT cases that received wound care, fracture splintage, pulse monitor, cardiac monitor, oxygen, spinal immobiliser, ventilation, defibrillation, cardiopulmonary resuscitation, limb traction, burns dressing, nebuliser, pulse oximeter, pupils check, cervical collar, entonox, blood pressure monitor and other treatments. The authors reported differences in the experience and ages of the EMTs and paramedics. Area 1 had younger and less experienced personnel in comparison with areas 2 or 3. There were few differences between the areas in terms of the type of patient or incident attended. The authors also reported that the types of incidents, patients and injuries attended were very similar for both EMT and paramedic cases, although the EMTs attended a smaller proportion of trapped patients than the paramedics did.
Effectiveness results Processes of care.
The number of paramedic cases were wound care 91 (10.35), fracture splintage 167 (18.7%), pulse monitor 744 (84.7%), cardiac monitor 98 (11.1%), oxygen 243 (27.5%), spinal immobiliser 72 (8.1%), ventilation 17 (1.9%), defibrillation 3 (0.3%), cardiopulmonary resuscitation 11 (1.2%), limb traction 6 (0.7%), burns dressing 2 (0.2%), nebuliser 3 (0.3%), pulse oximeter 533 (60.7%), pupils check 562 (63.9%), cervical collar 208 (22.9%), entonox 113 (12.8%), blood pressure monitor 397 (46.2%), and other treatments 109 (12.3%).
The number of EMT cases by procedure were wound care 23 (6.9%), fracture splintage 51 (15.0%), pulse monitor 283 (85.5%), cardiac monitor 11 (3.3%), oxygen 71 (21.3%), spinal immobiliser 25 (7.5%), ventilation 1 (0.3%), defibrillation 0 (0%), cardiopulmonary resuscitation 1 (0.3%), limb traction 1 (0.3%), burns dressing 1 (0.3%), nebuliser 2 (0.6%), pulse oximeter 209 (63.7%), pupils check 182 (55.2%), cervical collar 62 (17.7%), entonox 45 (13.6%), blood pressure monitor 134 (41.0%), and other treatments 9 (2.7%).
Among 868 patients with a PRF, 288 received an intervention that could only be administered by a paramedic.
Time spent on scene.
For the main cohort of patients (non randomised cases), the paramedics spent an average of 15.2 minutes (SD=0.25) on scene, while EMTs spent 13.5 minutes (SD=0.30).
Among the randomised cases, the paramedics spent 14.1 minutes (SD=2.6) on scene, while EMTs spent 8.6 minutes (SD=1.0).
After adjusting for injury severity, head injury severity, type and place of incident, age and ambulance service area, the paramedics presented significantly longer times on scene than the EMTs. For the non randomised cohort of patients, this difference in time was +1.8 minutes (SD=0.46) before adjustment, and +2.0 minutes (SD=0.44) after adjustment, (p<0.01). For the randomised cohort of patients, paramedics spent +5.5 minutes more (SD=2.4) before adjustment and +7.2 minutes (SD=3.1) after adjustment, (p=0.04). The authors reported that 10.3% of the variance in the time spent on scene was explained by the adjustment performed in the non randomised cohort of patients. For the randomised cohort, the adjustment explained 24.1% of the variance in time spent on scene. Untrapped patients receiving any paramedic intervention had a mean time on scene of 25.4 minutes (SD=9.2), while those who did not receive a paramedic intervention spent 13.4 minutes (SD=7.1) on scene. There were no statistically significant differences between the non randomised and the randomised cohorts, and the different areas, for the time spent on scene.
There was no statistically significant evidence of differences between the paramedic and EMT patients in rates of admission to the ICU.
The mean length of stay was 15.2 days (SD=24.2) for patients attended by paramedics, and 15.0 days (SD=28.0) for patients attended by EMTs. This difference was not significant, even after adjusting for age, injury severity, type of incident and area.
Mortality.
The crude mortality rate was 6.0% (86 in 1,440) for paramedic patients and 4.6% (28 in 605) for EMT patients. There was a strong relationship between mortality and total travel time from the scene to the hospital, which, curiously, was higher for shorter travel time cases. The observed mortality was 9% for those cases with total travel times shorter than 10 minutes, and 2% for incidents with more than 25 minutes' travel time. After adjusting mortality for injury severity, age of patients, mechanism of injury, whether the patient was trapped or not, the type of incident and area, the authors reported that this relationship almost disappeared. It may have been mainly due to differences in case-mix. Even after adjustment, the differences in mortality rates between the crews remained, although they were non significant. The adjusted odds ratio of dying when attended by a paramedic, relative to a technician, was 1.86 (95% confidence interval, CI: 0.93 - 3.72). When the first crew attending the patient was considered as the basis for the effectiveness analysis, it was shown that there was a mortality rate of 6.2% (84 in 1,346) among patients first attended by paramedics, and 4.3% (30 in 699) among patients first attended by EMTs. Patients first attended by paramedics had a relative risk of death of 1.45, compared with 1.29 for those patients ever attended by paramedics, (p=0.05).
Sub-group analyses were performed to ascertain whether the increased mortality among paramedic patients was related to some characteristics of the patients or the incidents. For those patients presenting injuries in which blood loss may have been an important problem, such as deaths resulting from exsanguinations (hypovolaemic shock/recurrent haemorrhage) or other causes that may have resulted in exsanguinations (multiple fractures and injuries to the thorax and abdomen), the estimated crude relative risk of death from these causes in paramedic patients, compared with EMT patients, was 2.52 (95% CI: 098 - 6.61). The relative risk of death from other causes was nearly identical, 1.02 (95% CI: 0.62 - 1.68). When patients with no head injury were considered, those presenting bleeding injuries and attended first by paramedics had a relative risk of death of 4.60 (95% CI: 1.07 - 20.0), compared with those attended first by EMTs.
Morbidity.
The response rate to the questionnaire sent out 6 months after the incident was 65% for patients attended by paramedics, and 72% for patients attended by EMTs. The time off from work of the survivors was 30 days less for patients treated by paramedics, (p=0.05). After having adjusted for case-mix (area, severity of injuries, type of incident, whether trapped or not, age and gender), patients attended by paramedics had 14.4 fewer days off work (SD=9.5) than those attended by EMTs, (p=0.04).
When the type of crew that first arrived on scene was considered as the basis for the analysis, the difference was not statistically significant. After adjusting for case-mix, patients first attended by a paramedic had 9.1 (SD=9.2) fewer days off work, (p>0.1).
After adjusting for case-mix, the differences in the mean SF-6 scores for patients attended by paramedics, compared with EMT patients, were:
6.8 (SD=4.1) for physical functioning, (p=0.09);
6.5 (SD=4.0) for extended physical functioning, (p=0.08);
10.5 (SD=4.7) for social functioning, (p=0.02);
13.4 (SD=6.2) for role physical, (p=0.03);
12.8 (SD=7.0) for role emotional, (p=0.06);
1.2 (SD=3.4) for mental health index, (p=0.70);
5.4 (SD=3.4) for energy/vitality, (p=0.10);
0.1 (SD=3.9) for pain, (p=0.97); and
2.5 (SD=3.6) for general health perceptions, (p=0.46).
When the type of crew first attending the patient was considered as the basis for the analysis, these differences were still greater and more statistically significant.
The SF-36 scores for patients first attended by paramedics were 49.7 for physical functioning, 60.2 for social functioning, 33.8 for role physical, 61.9 for role emotional, 64.3 for mental health index, 48.5 for energy/vitality, 53.8 for pain, and 56.0 for general health perceptions.
The SF-36 scores for patients first attended by EMTs were 41.7 for physical functioning, 49.6 for social functioning, 24.3 for role physical, 47.5 for role emotional, 61.6 for mental health index, 42.3 for energy/vitality, 51.2 for pain, and 55.6 for general health perceptions.
The adjusted differences were:
+9.1 (SD=4.1) for physical functioning, (p=0.02);
+13.5 (SD=4.6) for social functioning, (p=0.005);
+13.6 (SD=6.2) for role physical, (p=0.03);
+17.2 (SD=6.8) for role emotional, (p=0.01);
+3.8 (SD=3.3) for mental health index, (p=0.20);
+6.4 (SD=3.4) for energy/vitality, (p=0.04);
+2.4 (SD=3.9) for pain, (p=0.55); and
+3.4 (SD=3.6) for general health perceptions, (p=0.30).
The authors reported that the differences in the mean SF-36 scores obtained for patients attended by paramedics were similar, independent of the type of injury they presented (head injury, no head injury with bleeding injuries or other kind of injuries).
Clinical conclusions Paramedics administered the most common treatments to a greater proportion of patients than EMTs, although the difference may have been due to the lack of reporting by the EMTs. The time spent on scene was longer for paramedics, and the difference depended on whether or not the patients received any paramedic intervention. The mortality rates were higher for paramedic patients even after adjusting for case-mix, although the difference was non significant. Patients attended by paramedics had a better residual health than those attended by EMTs. The authors potential explanation for the higher rates of mortality, besides better residual health for those surviving patients attended by a paramedic, was that the paramedic strategy, when successful, may achieve earlier stabilisation and better recovery. However, the extra time taken may involve a delay in some cases, leading to deaths that may have been avoided by the "scoop-and-run" approach adopted by EMTs.
Methods used to derive estimates of effectiveness An expert panel compared avoidable death rates in patients attended by paramedics and EMTs, that is, the number of trauma deaths that could have been prevented in the case of an optimal pre-hospital care. Two sub-studies were performed. The first study (study 1) clearly specified inclusion and exclusion criteria in order to exclude those inevitable deaths. Two different sub-groups were considered. These were death on arrival (if patients died shortly after the incident and before reaching the hospital), and hospital death. Hospital death included both patients for whom there were an attempt to resuscitate on arrival, and those dying within 3 days. The sample in study 1 consisted of 266 cases, from which 179 cases were selected and presented to the panel for assessment. The second sub-study (study 2) analysed 100 consecutive deaths, without any exclusion, in order to analyse whether these deaths were preventable.
Deaths were classified as unavoidable or potentially avoidable, and those potentially avoidable were further classified as possibly not avoidable and probably avoidable. The authors reported comparisons of the rates of potentially and probably avoidable deaths (in cases in which optimal pre-hospital care would have been given) in all patients who died and who were attended either by paramedics or by EMTs.
A panel of four experts was formed to assess the avoidable deaths. The experts were blinded to when, where and by whom each patient was treated.
Avoidable death was defined as death, which could have been prevented, and the outcome reversed, if the patient had been managed with all the necessary skills and resources appropriate to the severity of the injury.
Estimates of effectiveness and key assumptions The authors presented the rates of potentially and probably avoidable deaths in all patients who died for cases attended by paramedics and by EMTs.
For study 1, the potentially avoidable mortality rate was 14.2 % (17 in 120) for paramedic cases and 6.8% (4 in 59) for EMT cases. The difference was not statistically significant, (p=0.15). The probably avoidable mortality rate was 6.7% (8 in 120) in the paramedic group and 0% in the EMT group. The difference was statistically significant, (p=0.04).
Exsanguination was the most common contributory cause of death among the potentially avoidable deaths, with 71% (15 in 21) of deaths due to this cause. This was followed by airway obstruction, with 38% of deaths (8 in 21), and neurological injury with 19% (4 in 21).
A similar pattern was found in the probably avoidable deaths, with 75% (6 in 8) being caused by exsanguination.
Paramedics spent approximately double the length of time on scene in cases of serious injury leading to death (20.4 minutes), compared with EMTs (11.0 minutes).
The results of study 2 showed that the potentially avoidable mortality rate among the paramedic deaths was 7.1% (5 in 70), and the probably avoidable mortality rate was 4.5% (4 in 70). The potentially avoidable mortality rate among EMT cases was 11.1% (2 in 18), whilst the probably avoidable mortality rate was 0% (0 in 18).
The mean scene time for all the deaths studied by the expert panel was relatively longer (17.4 minutes) than the mean scene times found in the cohort study of mainly survivors (14.6 minutes).
Measure of benefits used in the economic analysis A cost-consequences analysis was performed. Therefore, no summary measure of health benefit was used in the economic analysis.
Direct costs The resource quantities and the costs were reported separately. The direct costs included in the analysis were those of the NHS. These costs included the ambulance service for ALS training, the associated service costs, hospitalisation costs and ambulatory care. The ambulance service costs for ALS training covered instructors, equipment, facilities, consumables, time off work for training and associated costs of skills maintenance. The associated service costs were for staff, consumables, support requested on scene and time on scene. The hospitalisation costs covered ward stay and ICU stay, including readmissions. Ambulatory care costs included outpatient care, general practitioner contacts and nurse visits. The costs related to the paramedic training were adjusted in order to represent only the trauma related part of the additional skills. The cost data on ALS training was obtained from the ambulance services, and were valued at local unit costs. The service costs were obtained from the PRFs and other ambulance service records, and were valued using local cost data, except for consumables, which were valued at national averages. Hospitalisation costs were obtained from the patient records and were valued at regional average costs. The ambulatory care costs were derived from the patient questionnaire, and were valued at national and regional average unit costs. Therefore, the unit costs were estimated on the basis of actual data and on modelling. The incremental costs of ALS training were estimated using a bottom-up approach, whereas the service costs were estimated using a top-down methodology. Discounting was not performed, but was irrelevant due to the short period of follow-up (up to 6 months). The study reported the average costs. The price year was 1996 to 1997.
Statistical analysis of costs An analysis of variance was performed to estimate the independent effect of the crew type on the resource use and costs.
Indirect Costs The authors reported that the time off work for survivors was lower for patients treated by paramedics (30 days less, p=0.05), but the precise value for the indirect costs was not estimated because of the problems that this estimation presented.
Sensitivity analysis The authors performed sensitivity analyses to investigate variability in the data. They considered alternative scenarios where the proportion of total paramedic and salary costs attributable to trauma skills varied from 0%, 20%, and 30% to 100%.
Estimated benefits used in the economic analysis See the 'Effectiveness Results' section.
Cost results The average cost per trauma patient was 2,231 for patients first attended by paramedics, and 2,209 for patients first attended by EMTs. The adjusted difference was 22 more for the paramedic patients. The authors reported that the costs were insensitive to changes in the assumptions made about the percentage of total paramedic and salary costs attributable to trauma skills.
Synthesis of costs and benefits Not applicable due to the cost-consequences approach.
Authors' conclusions The authors reported in their final conclusions that seriously injured patients who were attended by paramedics had a higher risk of death than those attended by emergency medical teams (EMTs). However, the results of the effectiveness analysis showed that those patients with less severe injuries who were treated by paramedics had higher mortality rates than those treated by EMTs. The authors also reported that all eight dimensions of health outcome on the SF-36 questionnaire were better in survivors attended by paramedics than by EMTs. They concluded that the study did not show evidence that the current use of paramedic skills was preventing possible avoidable deaths. The mean total cost of treatment was slightly higher for patients treated by paramedics than by EMTs.
CRD COMMENTARY - Selection of comparators The comparator was justified on the grounds that basic life support training is the alternative training to ALS in the authors' setting. You should decide if this is a widely used health technology in your own setting, and consider that paramedic training varies considerably from country to country. Training in the UK is 8 weeks in duration, whereas in the USA it consists of a 3-year period of professional training.
Validity of estimate of measure of effectiveness The main analysis used a cohort study. This seems to have been appropriate for the study question, although there were some difficulties in interpreting the results because not all the possible bias and confounding factors could be controlled for. An intention to randomise occurred, although the authors also reported some limitations with this sub-study. The authors did not show that the main study sample was representative of the study population. However, the fact that three different ambulance service areas were included in the study, to represent the types of environment typically encountered in England, may have made the study sample more representative of the study population.
The authors reported the methods and results of the clinical study in great detail, but there were some limitations that were beyond the authors' control. These were related to how the crews reported the data. The authors stated that differences in the treatments given by paramedics and EMTs should be viewed with caution, as they may be due to a lack in the reporting of treatments actually given by the EMTs. Moreover, the authors acknowledged that is was difficult to interpret some of the effectiveness results because there was no simple pattern in the way that paramedics and EMTs attended the incidents. There was a problem with missing cases in each one of the different analyses, and this introduced uncertainty into the reliability of the effectiveness results. Further, the authors reported that the effectiveness results should be treated with caution, as there is the possibility of selection bias and confounding factors. There was also uncertainty regarding the morbidity outcomes because the response rate to the questionnaire was not high, and because some data were collected retrospectively, which may have introduced inaccuracies in the analyses. However, the authors relied on the results obtained for higher mortality among the paramedic patients, because this finding was consistent with the results obtained from other studies.
A panel of experts was used to assess the number of avoidable deaths. The authors reported clearly the methods used for this assessment, and the process by which the physicians were selected. A study of the reliability of panel assessment showed high reliability of the assessments reported. Some statistical analyses were performed to compare the results obtained in the potential and probable mortality rates among paramedic and EMT deaths. The authors reported that caution should be exercised because the areas included in the effectiveness study might differ from other English areas. Moreover, the exclusion criteria used for the analysis of the expert panel have to be considered, and the length of time considered (death up to three days after the incident) may have influenced the results obtained. The analysis of avoidable deaths, as reported by the authors, presented some difficulties. First, higher avoidable mortality rates in the paramedic group that may have been balanced by unexpected survivors in this group. Second, the analysis was based on the experts' opinion. Third, there may be confounding factors affecting the results obtained. Finally, the authors tried to ensure that the experts on the panel were blinded to the type of crew attending the incidents that ended in death, but the experts could accurately infer the type of crew because of the different kind of care given by paramedics and EMTs.
Validity of estimate of measure of benefit The authors did not derive a summary measure of health benefit. The analysis was therefore categorised as a cost-consequences study.
Validity of estimate of costs All the categories of costs relevant to the perspective adopted seem to have been included in the analysis. The costs and the quantities were reported separately, and the price year was given. Discounting was not performed but was irrelevant because the costs were incurred over less than 2 years.
The authors reported that missing data (because of the lack of PRF) may have introduced uncertainty into the validity of the estimation of the pre-hospital treatment costs. However, they reported that these costs represented a small proportion of the total costs. Therefore, it was unlikely that the overall cost results would have been different because of the missing data. They also reported that the top-down method used to estimate some of the costs may have produced some imprecision in the unit costs estimated, although they stated that this imprecision would have been relatively unimportant.
Other issues The authors made extensive and appropriate comparisons of their findings with those from other studies. They reported similarities in terms of the on-scene time spent by paramedics and EMTs. The mortality rates were lower than those obtained in an earlier study, because the injuries involved were less severe. After a review of the literature about the benefits of ALS training, the authors concluded that the mortality rates were similar to those obtained in other studies. The differences between paramedic and EMT costs were larger in another study, but these may have arisen on account of the different services investigated in both studies.
The authors addressed the issue of generalisability of the results. They reported that, although it was difficult to determine to what extent the results were generalisable to other settings, the fact that three different ambulance areas were included in the effectiveness study, with similar on-scene times and SF-36 scores, and mortality rates in the same direction, led them to think that the results may be generalisable to the whole of England.
Implications of the study The results of the study did not show that the paramedics' skills were preventing avoidable deaths. Thus, the authors suggest that paramedics may require different training courses based around different protocols, to be able to make better judgements about when skills should be used, on whom, and in which circumstances.
The authors recommend further research in several areas. First, the analysis of alternative paramedic fluid resuscitation protocols in blunt trauma. Second, a comparison of the effectiveness of different pre-hospital time protocols in untrapped patients with bleeding injuries. Finally, a comparison of paramedic training programmes, using similar protocols, to examine whether the skills developed with longer paramedic training make a difference to the implementation and effectiveness of the protocols.
The results of this study should be treated with some caution because, although it was performed with great care and detail, there are several limitations that make the interpretation of the results difficult.
Bibliographic details Nicholl J, Hughes S, Dixon S, Turner J, Yates D. The costs and benefits of paramedic skills in pre-hospital trauma care. Health Technology Assessment 1998; 2(17): 1-67 Other publications of related interest Anderson IW, Black RJ, Ledingham I McA, Little K, Robertson CE, Urquhart JD. Early emergency care study: the potential and benefits of advanced prehospital care. BMJ 1987;294:228.
Bissell R, Eslinger D. Advanced life support literature review. The American Ambulance Association, National Study Centre for Trauma and EMS. Baltimore (MD): University of Maryland at Baltimore; 1995.
Nguyen-van-Tam JS, Dore AF, Bradley MP, Pearson JC, et al. Effectiveness of ambulance paramedics versus ambulance technicians in managing out of hospital cardiac arrest. Journal of Accident and Emergency Medicine 1997;14:142-8.
Nicholl JP, Turner J, Dixon S. The cost-effectiveness of the regional trauma system in the North West Midlands. Sheffield: University of Sheffield, Medical Care Research Unit; 1995.
Underhill TJ, Finlayson BJ. A review of trauma deaths in an accident and emergency department. Archives of Emergency Medicine 1989;6:90-6.
Indexing Status Subject indexing assigned by NLM MeSH Adolescent; Adult; Age Distribution; Aged; Child; Child, Preschool; Clinical Competence; Cost-Benefit Analysis; Emergency Medical Services /economics; Emergency Medical Technicians /economics /standards; Female; Great Britain /epidemiology; Hospital Mortality; Humans; Infant; Length of Stay; Male; Middle Aged; Quality of Life; Research Support, Non-U.S. Gov't; Sex Distribution; Surveys and Questionnaires; Wounds and Injuries /mortality AccessionNumber 21999008002 Date bibliographic record published 31/05/2003 Date abstract record published 31/05/2003 |
|
|
|