Morbidity and severity of illness during interhospital transfer: impact of a specialised paediatric retrieval teamBMJ 1995; 311 doi: https://doi.org/10.1136/bmj.311.7009.836 (Published 30 September 1995) Cite this as: BMJ 1995;311:836
- Joseph Britto, paediatric intensive care fellowa,
- Simon Nadel, consultant in paediatric intensive carea,
- Ian Maconochie, clinical research fellowa,
- Michael Levin, professora,
- Parviz Habibi, senior lecturera
- Correspondence to: Dr Britto.
- Accepted 14 July 1995
Objective: To evaluate the morbidity and severity of illness during interhospital transfer of critically ill children by a specialised paediatric retrieval team.
Design: Prospective, descriptive study.
Setting: Hospitals without paediatric intensive care facilities in and around the London area, and a paediatric intensive care unit at a tertiary centre.
Subjects: 51 critically ill children transferred to the paediatric intensive care unit.
Main outcome measures: Adverse events related to equipment and physiological deterioration during transfer. Paediatric risk of mortality score before and after retrieval. Therapeutic intervention score before and after arrival of retrieval team.
Results: Two (4%) patients had preventable physiological deterioration during transport. There were no adverse events related to equipment. Severity of illness decreased during stabilisation and transport by the retrieval team, suggested by the difference between risk of mortality scores before and after retrieval (P<0.001). The median (range) difference between the two scores was 3.0 (−6 to 17). Interventions during stabilisation by the retrieval team increased, demonstrated by the difference between intervention scores before and after retrieval, median (range) difference between the two scores being 6 (−8 to 38) (P<0.001).
Conclusions: Our study indicates that a specialised paediatric retrieval team can rapidly deliver intensive care to critically ill children awaiting transfer. Such children can be transferred to a paediatric intensive care unit with minimal morbidity and mortality related to transport. There was no deterioration in the clinical condition of most patients during transfer.
The transfer of a critically ill child is inherently risky; up to 75% of patients transferred by non-specialised teams can suffer from serious clinical complications
These risks can be considerably reduced if a pecialised retrieval team stabilises the patient and establishes intensive care at the referring hospital before transfer
In this study physiological deterioration occurred in only two of 51 patients during transport, and there were no instances of adverse events related to equipment
There was no deterioration in the clinical condition of most patients during transfer
The improved outcome of critically ill children managed in paediatric intensive care units has increased the pressure to transfer such patients to a tertiary centre.1 The risk of deterioration from the primary illness, complications of treatment, and the transfer process itself make the interhospital transfer potentially hazardous.2 3 4 Unfortunately, despite recommendations, many critically ill children in the United Kingdom are still being transferred by non-specialised staff.5 6 7
There are no published data on the morbidity associated with transfer by a specialised paediatric retrieval team in the United Kingdom.
Patients and methods
This prospective, descriptive study at St Mary's Hospital, London, evaluated 78 consecutive patients transferred to the paediatric intensive care unit between October 1993 and May 1994. Fifty one patients were included in the study. Insufficient data from before retrieval prevented analysis of severity of illness in the 27 patients excluded from the study. In none of the patients excluded from the study was there any morbidity or mortality during transport.
The retrieval team consisted of a paediatric intensivist (senior registrar or consultant grade) and an experienced intensive care nurse.
ASSESSMENT OF MORBIDITY
Morbidity during transport was documented by using the criteria of Kanter and Tompkins (box).2 Transport was defined as the period between leaving the referring hospital and arrival of the patient at the paediatric intensive care unit.
QUANTIFYING SEVERITY OF ILLNESS AND THERAPEUTIC INTERVENTIONS
For each child we computed the score at three points in time:
The admission score by using values obtained on admission to the referring hospital
The score before retrieval by using values obtained before stabilisation by the retrieval team at the referring hospital
The score after retrieval by using values obtained after stabilisation and transport.
Morbidity during transport2
Respiratory arrest or cyanosis, or both Cardiac arrest
Systolic hypotension (child <65 mm Hg, infant <55 mm Hg)
Cardiac arrythmia including tachycardia (child >200/min, infant >220/min) or bradycardia (child <40/min, infant <50/min)
Loss of consciousness (Glasgow coma scale <7)
Loss of brainstem reflexes
Core temperature (<34°C)
Hypoglycaemia (<2.5 mmol/l)
Occluded endotracheal tube
Accidental tracheal extubation
Loss of intravenous access
Loss of monitoring
Malfunction of ventilator
Exhaustion of oxygen supply
The score on the therapeutic intervention scoring system (TISS) was used as a quantitative measure of interventions performed.12 The score before retrieval measured interventions performed before the arrival of the retrieval team, and the score after retrieval measured interventions performed by the retrieval team.
The Wilcoxon matched pairs signed ranks test was performed to compare the two types of score at different time points, and the comparisons were preceded by the Friedman two way analysis of variance by ranks test.
Two patients had preventable physiological deterioration during transport. One infant with bronchiolitis developed apnoea and cyanosis, and another child with meningococcal shock developed hypoglycaemia. There were no instances of adverse events related to equipment.
SEVERITY OF ILLNESS
There was a decrease in the severity of illness before arrival of the retrieval team, suggested by the difference between paediatric risk of mortality scores on admission and before retrieval (P<0.001). The median (range) difference between the scores on admission and before retrieval was 1.0 (0 to 24). The severity of illness worsened in none, remained unchanged in 23, and decreased in 28 patients before arrival of the retrieval team.
There was a further decrease in severity of illness during stabilisation and transport by the retrieval team, suggested by the difference between scores before and after retrieval (P<0.001). The median (range) difference between these scores was 3.0 (−6 to 17). The severity of illness worsened in six, remained unchanged in 11, and decreased in 34 patients.
There was an increase in therapeutic interventions during stabilisation by the retrieval team, demonstrated by the difference between therapeutic intervention scores before and after retrieval (P<0.001). The median (range) difference between these two scores was 6 (−8 to 38). Interventions decreased in two, remained the same in four, and increased in 45 patients.
INTERVENTIONS PERFORMED AND TREATMENT INITIATED (TABLE II)
Airway intervention--Forty three patients (84%) required endotracheal intubation. Fifteen underwent rapid sequence tracheal intubations by the retrieval team.13 Of the 28 patients already intubated at the referring hospital, the retrieval team had to reintubate 14 (50%). Airway intervention by the retrieval team was therefore required in 29 of the 51 patients (57%) transferred. There were no instances of accidental extubation or blockage of endotracheal tubes during transport.
Ventilatory intervention--All intubated patients were mechanically ventilated. In the 28 patients already intubated the retrieval team initiated mechanical ventilation in five, used controlled hyperventilation in three, and applied or withdrew positive end expiratory pressure in five. Ventilatory intervention was therefore modified in 13 (26%) patients by the retrieval team.
Vascular access--There were no complications associated with either central venous or arterial catheterisation.
Vasoactive drugs--Vasoactive drugs (dopamine, dobutamine, adrenaline, epoprostenol) were used singly or in combination in 22 (43%) patients. On 24 occasions a vasoactive drug was started by the retrieval team.
Colloid treatment and metabolic corrections during retrieval--The retrieval team gave >/= 20 ml/kg of colloid to 28 (55%) patients and corrected hypokalaemia in 16 (31%), hypocalcaemia in 15 (29%), metabolic acidosis in 15 (29%), and hypoglycaemia in nine (18%) patients.
OPERATIONAL DATA (TABLE III)
The median distance travelled to a referring hospital was 22.2 (range 5.0-218.0) km.
MORBIDITY DURING TRANSPORT
A recent study in the United Kingdom by Barry and Ralston highlighted the dangers of non-specialised paediatric transfers. Three quarters of patients transferred suffered serious clinical complications, nearly a quarter of which were defined as life threatening.7 These figures, although alarming, are not surprising considering that in most patients there was no monitoring of blood pressure, temperature, and blood sugar, and in 27% of patients oxygen saturation or electrocardiography were not monitored. On arrival at the paediatric intensive care unit 11% of their patients needed immediate endotracheal intubation and ventilation, and 9% were hypotensive.7 In a study in the United States of non-specialised transfer by Kanter and Tompkins the occurrence of both physiological deterioration and adverse events related to equipment were each 10.2%.2
By using the criteria of Kanter and Tompkins, a recent study in the United States by Edge et al comparing specialised (n=47) and non-specialised (n=92) paediatric retrieval demonstrated that the specialised team significantly reduced the occurrence of adverse events related to equipment (from 20% to 2%) but not the occurrence of physiological deterioration (12% to 11%) during transport.10 In our study physiological deterioration during transport occurred in 4% of patients, and there were no instances of adverse events related to equipment. The use of a control group in our study would have been unethical, but comparisons of our results with published data on non-specialised transfers demonstrates that a specialised paediatric retrieval team can transport patients with minimal physiological deterioration and morbidity related to equipment.
SEVERITY OF ILLNESS
Studies of non-specialised retrieval have not assessed the change in severity of illness during transport. In the absence of a control group in our study we cannot make comparisons nor reach any definite conclusions as to what effect a non-specialised team would have had on the severity of illness during transfer. The significant decrease in the severity of illness, however, measured by the paediatric risk of mortality score clearly demonstrates that there was no deterioration in the clinical condition of most of our patients during transfer.
There are several settings in which intervention by the retrieval team during stabilisation could lead to a decrease in morbidity during transport.
In the study by Fuller et al of transportation of critically ill children in Canada 43% of patients required some form of airway or respiratory intervention.14 In our study 57% (29) required airway intervention and 26% (13) needed ventilatory interventions.
Loss or lack of intravenous access during transport by non-specialised teams occurs in 3.4% to 10% of patients.2 3 7 10 Central venous catheters were inserted in 37 (73%) of our patients. There were no instances of loss of intravenous access during transport. Forty four (86%) of our patients underwent arterial catheterisation and direct monitoring of blood pressure.
The objectives of a retrieval team are to treat the underlying cause and to achieve physiologically normal haemodynamic and metabolic parameters.15 To this end, half of our patients received >/= 20 ml/kg of colloid from the retrieval team. On 24 occasions a vasoactive drug was started by the retrieval team. Hypokalaemia, hypocalcaemia, and metabolic acidosis were corrected in 29% of our patients by the retrieval team.
Our median response time of 100 minutes compares with the average response time of another specialised paediatric team in Glasgow, reported as 121 minutes.16 The immediate institution of therapeutic measures by the referring hospital while awaiting the arrival of the retrieval team is more crucial to patient outcome than the rapidity with which the retrieval team reaches the patient's bedside.17 It is essential therefore that the response time be used by the retrieval team to give advice regarding management of the patient.
The concept that a retrieval team must “swoop and scoop” is no longer considered appropriate.18 19 The level of treatment and monitoring by our retrieval team, measured by the therapeutic intervention scores, increased significantly during stabilisation. The median stabilisation time in our study was 162.5 minutes, which compares with the stabilisation time of other specialised paediatric retrieval teams reported as 74.4-156 minutes.4 7 20 The new ethos of mobile intensive care implies that intensive care is firmly established at the patient's bedside at the referring hospital.18 19 20
Our study indicates that a specialised paediatric retrieval team can rapidly deliver intensive care to critically ill children awaiting transfer. Such children can be transferred to a paediatric intensive care unit with minimal morbidity and mortality related to the transfer. Furthermore, there was no deterioration in the clinical condition of most patients during transfer.
We thank Dr Nigel Curtis for his assistance in setting up the database used in this study and Dr Sam Walters for his advice on the preparation of the manuscript.
Funding George John Livanos charitable trust.
Conflict of interest None.