Background
A critical component of the emergency care of very sick infants and children is fluid management in the first 1–2 h after initial assessment. In children in whom adequate tissue perfusion appears to be compromised, additional fluids are commonly given as a means of improving circulating volume. Children with a history of diarrhoea with severe dehydration urgently need additional fluids to restore circulating volume. In other sick children, for example those with malaria, pneumonia or meningitis and those without overt fluid loss, giving additional fluids may be detrimental.
Generating the evidence base for such practice is challenging because it is complex both practically and ethically to conduct randomized comparative studies of fluid types, volumes and rates in critically ill children. Undertaking such studies in low-resource settings, where the disease profile of children is likely to be very different from that in high-resource settings and where staffing levels and competencies are less consistent, is even more difficult.
The second edition of the WHO Pocket book of hospital care for children (WHO, 2013b) did not include a revision of fluid management in triage and emergency conditions, as the document was revised before publication of a large RCT on use of fluid boluses in sick infants and children with febrile illness and signs of circulatory impairment: the trial of fluid expansion as supportive therapy trial (FEAST) (Maitland et al., 2011). No other new data were available at the time of the revision of the Pocket book. The FEAST study investigators anticipated that the study would confirm the value of fluid boluses in such children and would help to identify the best fluid strategy in terms of fluid type and rate; however, the study was stopped prematurely by the Data Safety Monitoring Board because fluid boluses were found to increase rather than decrease mortality.
The findings of the FEAST study should be understood in the context of the children who were recruited into the study and the populations addressed by the WHO ETAT guidelines. The study recruited infants and children with febrile illness and some signs of circulatory impairment, while WHO ETAT recommendations provide guidance on the management of children with a range of clinical presentations or combinations of clinical signs, including shock. The clinical definition of shock used by WHO, which can be assessed in resource-limited settings by non-specialist health workers, is the presence of three clinical signs at one time, i.e. cold extremities with capillary refill time > 3 s and a weak and fast pulse. The presence of one or two of these signs indicates nonspecific circulatory impairment that could be due to conditions other than circulatory shock. For example, cold extremities may be due to exposure; prolonged capillary refill may be due to exposure to cold; and a fast pulse may be due to pain or distress. Children in shock as defined by WHO, that is, who have all three signs, are at high risk for death.
There is no universally agreed definition of shock. International guidelines and training courses such as the Advanced Paediatric Life Support (Mackway-Jones et al., 2005) and Advanced Pediatric Life Support (Fuchs et al., 2007) provide lists of clinical signs that, if present, indicate that a child is in “shock” (); however, these guidelines and courses do not indicate how many signs must be present in order for shock to be diagnosed, nor is there a clear distinction between severe circulatory impairment, shock and severe shock. WHO refers only to “shock” with no other sub-classifications. If a child has only one or two of the three signs, the diagnosis is only circulatory impairment, whereas if all three signs are present the child is in “shock”.
Shock in children may be considered according to the underlying pathophysiology.
Hypovolaemic shock, due to a marked decrease in blood volume, may follow haemorrhage due to major trauma or large fluid losses from burns, severe dehydration from vomiting and diarrhoea, heat stroke or water deprivation. These are major causes of mortality among children in developing countries. In children, hypovolaemic shock may also be associated with sepsis (broad systemic response to infection).
Cardiogenic shock refers to failure of the heart to sustain an adequate output. The many causes include: excessive volume or pressure load on the heart chambers (e.g. congenital heart disease), dysfunction of the heart valves (e.g. valvular heart disease such as rheumatic or congenital valve stenosis or regurgitation), impaired cardiac muscle function (e.g. myocarditis), cardiomyopathy, acute arrhythmia (such as supraventricular tachycardia), obstruction of blood flow to or from the heart (e.g. tension pneumothorax, cardiac tamponade or pulmonary embolism); myocardial ischaemia (such as in Kawasaki disease); or severe acidosis.
Vasodilatory shock, sometimes called “distributive shock”, is associated with hypotension due to vasodilation of blood vessels and, sometimes, leaky capillaries; it may be caused by sepsis, anaphylaxis, dengue or spinal cord injury.
Septic shock in children can be a mixture of vasodilatory, hypovolaemic and cardiogenic shock. It may be due to bacteria, viruses (such as dengue), fungi or parasites (malaria). Sometimes, the specific pathogen is not identified.
The treatment required depends on the type of shock. The progression of shock is commonly divided into three phases: compensated, uncompensated and irreversible, describing the opportunities for intervention to improve outcome.
The fluid management strategy is decided on the basis of whether some or all signs of impaired circulation are present and characterization of the type and underlying shock, if present. Fluid bolus management may be entirely appropriate in some settings, such as hypovolaemic shock due to severe dehydration, but may be harmful if cardiogenic shock due to myocarditis is the underlying problem.
The management of critically ill infants and children depends on the skill and competence of health workers in identifying these children, the availability of basic equipment for detecting hypotension and hypoxaemia and the time available to adequately monitor and reassess the response of children to treatment.
At the scoping meeting in March 2013, members of the GDG prioritized PICO questions 5 and 6.
Question 5. In infants and children presenting with emergency signs (as described in WHO ETAT guidelines), which clinical signs or disease markers or combination of signs or other markers indicate severely impaired circulation and the need for intravenous fluids?
Question 6. In infants and children with severely impaired circulation, which intravenous fluids, at what rate and for how long are associated with the lowest risk for mortality?
During the guideline development meeting, the GDG reframed question 6 and the discussion as follows:
Question 6a. In infants and children who are not in shock but have signs of severely impaired circulation, which intravenous fluids, at what rate and for how long are associated with the lowest risk for mortality?
and
Question 6b. In infants and children in shock, which intravenous fluids, at what rate and for how long are associated with the lowest risk for mortality?
During the scoping meeting in March 2013, the GDG considered that there was no new evidence and no concern that would warrant a review of current WHO recommendations on fluid management of children with severe dehydration. The GDG noted that WHO recommendations on the management of children with dengue shock syndrome, including fluid management, had been updated in 2012 (WHO & Special Programme for Research and Training in Tropical Diseases, 2012) and there was no further evidence or concern that would warrant an additional review of these recommendations. The recommendations provide important evidence-based guidance to health care workers on the management of such children, and the group decided that the recommendations should be included in this guideline as they provide a context for other recommendations. Extracts of the recommendations are shown below.
Fluid resuscitation must be clearly separated from simple fluid administration. Fluid resuscitation is a strategy in which larger volumes of fluids (e.g. boluses of 10–20 mL/kg bw) are administered for a limited time under close supervision to evaluate the patient's response and to avoid the development of pulmonary oedema. These fluids should not contain glucose.
The plan for treating patients with compensated shock is as follows:
Obtain a reference haematocrit before starting IV fluid therapy. Start IV fluid resuscitation with isotonic crystalloid solutions at 5–10 mL/kg bw per h over 1 h in adults and 10–20 mL/kg bw per h over 1 h in infants and children. Then reassess the patient's condition (vital signs, capillary refill time, haematocrit, urine output).
If the condition of the infant or child improves, IV fluids should be reduced to 10 mL/kg bw per h for 1–2 h, then to 7 mL/kg bw per h for 2 h, 5 mL/kg bw per h for 4 h and then to 3 mL/kg bw per h for up to 24–48 h. Consider reducing IV fluid earlier if oral fluid intake improves. The total duration of IV fluid therapy should not exceed 48 h.
If vital signs are still unstable (i.e. shock persists), check the haematocrit after the first bolus.
In infants and children:
If the haematocrit increases or is still high, change to colloid solution at 10–20 mL/kg bw per h. After the initial dose, reduce the rate to 10 mL/kg bw per h for 1 h, then reduce to 7 mL/kg bw per h. As mentioned above, change to crystalloid solutions when the patient's condition improves.
If the haematocrit decreases from the initial reference value (especially if the repeat haematocrit is below the baseline, for example, < 35–40%) and the patient still has unstable vital signs, look for severe bleeding. Cross-match fresh whole blood or fresh packed red cells and transfuse if there is severe overt bleeding. If there is no bleeding, give a bolus of 10–20 mL/kg bw of colloid over 1 h, repeat clinical assessment and determine the haematocrit. A senior staff member should carry out a review to consider blood transfusion.
Further boluses of crystalloid or colloidal solutions may have to be given during the next 24–48 h.
Treatment of profound shock (hypotensive; undetectable pulse and blood pressure)
All patients (infants, children and adults) with hypotensive shock should be managed more vigorously. The plan for treating patients with hypotensive shock is outlined below. For all patients (infants, children and adults), initiate IV fluid resuscitation with crystalloid or colloid solution at 20 mL/kg bw as a bolus given over 15–30 min to bring the patient out of shock as quickly as possible. Colloids may be preferred if the blood pressure has to be restored urgently, i.e. in patients with a pulse pressure < 10 mm Hg. If the patient's condition improves:
In infants and children, give colloid infusion of 10 mL/kg bw per h for 1 h. Then, continue with crystalloid solution at 10 mL/kg bw per h for 1 h, then to 7.5 mL/kg bw per h for 2 h, to 5 mL/kg bw per h for 4 h and to 3 mL/kg bw per h for up to 24−48 h. Consider reducing IV fluid earlier if oral fluid intake and urine output improve. The total duration of IV fluid therapy should not exceed 48 h.
3.2.1. Children who are not in shock but have signs of circulatory impairment
Question 6a. In Infants and children who are not in shock but have signs of severely impaired circulation, which intravenous fluids, at what rate and for how long are associated with the lowest risk for mortality?
(Fluid resuscitation of critically ill children aged 2 to 59 months with impaired circulation)
Summary of evidence
The search initially identified 1600 references, including three RCTs, only one of which met the inclusion criteria. The “fluid expansion as supportive therapy” (FEAST) trial (Maitland et al., 2011) provided evidence for the population directly addressed by this recommendation.
In the FEAST trial, children were enrolled in two strata (A without and B with severe hypotension) in general paediatric hospital wards in six centres, one in Kenya, four in Uganda and one in the United Republic of Tanzania. The three arms of the study were: saline bolus, albumin bolus and no bolus. The intervention arms received either IV 0.9% saline solution (20 mL/kg bw over 1 h) or IV albumin solution (20 mL/kg bw over 1 h), and children in the comparator arm received no bolus. The saline versus no bolus comparison was the most relevant for the guideline. All three treatment arms received IV maintenance fluids (2.5–4.0 mL/kg per h); 90% received hypotonic maintenance fluids (5% dextrose). Participants also received antimalarial, antipyretic and anticonvulsant drugs. Treatment for hypoglycaemia and blood transfusions were provided if necessary. Additional boluses of 20 mL/kg bw over 1 h were given if impaired perfusion persisted. If severe hypotension (defined as systolic blood pressure < 50 mm Hg in children < 12 months, < 60 mm Hg in children aged 1–5 years and < 70 mm Hg in children > 5 years of age) developed in children in stratum A, 40 mL/kg boluses of study fluid or saline (no bolus group) were given. The initial boluses were increased to 40 mL/kg after a protocol amendment. Three reasons for increasing the initial bolus are provided in a commentary on the FEAST trial: (i) the original initial bolus fluid volume (20 mL/kg bw) might have been insufficient to answer the study question; (ii) if international guidelines on fluid bolus volume were not followed, policy-makers would not be convinced by the results; and (iii) only 1% of trial participants had severe hypotensive shock (whereas 5% had been predicted in the trial design), which might have affected the saline versus albumin comparison (secondary end-point).
Stratum B (children with severe hypotension) comprised two treatment arms: IV 0.9% saline solution (40 mL/kg bw over 1 h) and IV albumin (40 mL/kg bw over 1 h). Additional boluses of 20 mL/kg bw were given after 1 h if impaired perfusion persisted. Initial boluses were increased to 60 mL per kg bw after a protocol amendment in June 2010 (reasons as described for stratum A). Other treatments were the same as reported for stratum A.
The primary outcome of the trial was mortality 48 h after randomization, and the secondary outcomes were mortality at 4 weeks, neurological sequelae, episodes of hypotensive shock 48 h after randomization and adverse events. The children were followed up for 4 weeks, and any child with neurological impairment was followed up at 24 weeks.
Stratum A comprised 3141 children aged 60 days to 12 years, with a median age of 23–25 months (interquartile range, 13–40 months), and 46–48% were female. The median systolic blood pressure was 92–93 mm Hg; 58–60% had a positive temperature gradient, and 69–71% had severe tachycardia. The proportion of children with a capillary refill time ≥ 2 s was 64–69%, and 25–29% had a capillary refill time ≥ 3 s. One third of the children had severe anaemia (33% in both groups), and about 15% were in a coma; 58% had malaria parasitaemia. The trial excluded children with severe malnutrition, gastroenteritis, non-infectious causes of shock and conditions for which volume expansion was contraindicated
Stratum B comprised 29 children with severe hypotension. The median age was 21 months (interquartile range, 10–47 months) in the saline bolus group and 28 months (22–84 months) in the albumin bolus group; 50% of children given a saline bolus and 38% of those given the albumin bolus were girls. The median systolic pressure was 56–59 mmHg; 77–88% had a positive temperature gradient, and 42–43% had severe tachycardia. The capillary refill time was ≥ 2 s in 8–19% of children and ≥ 3 s in 69–77%. The proportion of children with severe anaemia was 38% in the group given saline bolus and 58% that given albumin bolus; 56% and 77% of children in the two groups, respectively, were in a coma. Malaria parasitaemia was found in 31% and 62% of the children, respectively.
Similar mortality rates in the first hour after randomization were observed among children in the two groups who had severe febrile illness complicated by impaired consciousness, respiratory distress or both and with impaired perfusion recognized by a capillary refill time of ≥ 3 s, lower limb temperature gradient, weak radial pulse volume or severe tachycardia (1.1% in the saline-bolus group and 1.3% in the no-bolus group). At 48 h, however, the children who received a saline bolus had a statistically significantly greater risk of dying than children randomized to receive no fluid bolus (RR, 1.44; 95% CI, 1.09–1.90; p = 0.01). The difference in mortality rates between the groups was maintained at 4 weeks (RR, 1.38; 95% CI, 1.07–1.78; p = 0.01). Few results were reported for children in stratum B. The mortality rates in the two groups were not statistically significantly different (RR, 1.23; 95% CI, 0.70–2.16; p = 0.45)
The GDG agreed that the FEAST trial demonstrated clear harm “if rapid infusions of IV fluids are given” to children who have febrile illnesses such as pneumonia, malaria and meningitis, or have severe anaemia and do not fulfil all the criteria for the WHO definition of “shock”. The adverse outcomes may have been due to fluid overload, including pulmonary oedema, heart failure and cerebral oedema, but other mechanisms might also have been involved, as the FEAST investigators rarely identified cerebral or pulmonary oedema.
The GDG noted that the trial excluded children with a history of diarrhoea and severe dehydration and children with clinical signs of severe acute malnutrition.
View in own window
| SUPPORTING EVIDENCE AND ADDITIONAL CONSIDERATIONS |
|---|
| BENEFITS AND HARM | Do the desirable effects outweigh the undesirable effects?
☑ Yes
☐ No
☐ Uncertain | The evidence is consistent with current understanding of the pathophysiology of impaired circulation and responses to IV fluids. In the absence of shock, excessive fluids can result in fluid overload. Children with febrile illness, severe pneumonia, severe malaria, meningitis, severe acute malnutrition or severe anaemia do not require fluids additional to those for maintaining normal hydration. Inappropriate administration of IV fluids to these children, especially if given rapidly, can precipitate heart failure and lung congestion, cause cerebral oedema and exacerbate anaemia by further haemodilution. Fluid boluses may also result in adverse outcomes by mechanisms other than direct fluid overload, including blunting protective cardiovascular responses, such as vasoconstriction and tachycardia due to high levels of circulating catecholamines. Instead, the initial care for these children must involve identification of the underlying cause of their condition and provision of appropriate treatment.
The GDG noted, however, that children presenting with some signs of impaired circulation are at risk for clinical deterioration. Early, careful assessment, especially to determine a history of diarrhoeal illness and dehydration, appropriate treatment and further monitoring of these children are essential for effective management and to prevent other complications. |
| VALUES AND PREFERENCES; ACCEPTABILITY | Is there important uncertainty about or variation in how people value the options?
☐ Major
☑ Minor
☐ Uncertain
Is the option acceptable to key stakeholders?
☑ Yes
☐ No
☐ Uncertain | The GDG considered that all health workers would wish to understand the basis for and provide the appropriate fluids to sick infants and children. Health workers would recognize the potential harm of inappropriate excess IV fluids while acknowledging that infants and children with a history of diarrhoea require adequate, timely replacement of fluid deficits, either with oral rehydration solution or with IV fluids in cases of severe dehydration. |
| FEASIBILITY AND RESOURCE USE | How large are the resource requirements?
☐ Major
☑ Minor
☐ Uncertain
Is the option feasible to implement?
☑ Yes
☐ No
☐ Uncertain | The GDG considered that targeted use of IV fluids in sick infants and children can be done easily and does not require special equipment or skills. More explicit, focused use of IV fluids will not only improve outcomes but also conserve valuable resources such as IV fluids and IV cannulae. Training will be required to define when fluids are indicated for children who are well and, especially, to ensure that children with severe dehydration are not inadvertently deprived of the required fluids. |
View in own window
RECOMMENDATIONS
- 2.1.
Children with only one or two signs of impaired circulation – either cold extremities or capillary refill time > 3 s or a weak and fast pulse – but who do not have the full clinical features of shock, i.e. all three signs present together, should not receive rapid infusions of fluids but should still receive maintenance fluids appropriate for their age and weight (WHO, 2013b). - 2.2.
In the absence of shock, rapid intravenous infusions of fluids may be particularly harmful to children with severe febrile illness, severe pneumonia, severe malaria, meningitis, severe acute malnutrition, severe anaemia, congestive heart failure with pulmonary oedema, congenital heart disease, renal failure or diabetic ketoacidosis. - 2.3.
Children with any sign of impaired circulation, i.e. cold extremities, or prolonged capillary refill or a weak and fast pulse, should be prioritized for full assessment and treatment and reassessed within 1 h.
|
|---|
| Strength of recommendations | Strong (for all recommendations) |
| Quality of evidence | High (for all recommendations) |
| Justification | The GDG agreed that the quality of the evidence for these recommendations was high because, although only one study was identified, there was a large effect in the population of direct interest in a high-quality RCT with a large sample.
The GDG agreed that the recommendations should be strong in view of the high quality of the evidence on a clinical outcome of critical importance in a specific population and that the recommendations could be generalized to all settings.
Populations not included in the study, i.e. children with diarrhoea and with severe acute malnutrition, are clearly described in other WHO recommendations; for this reason, they are not included in these recommendations. Health workers are directed to the other recommendations.
The GDG noted that the recommendation is “negative”; however, there was consensus that, given the serious consequences of giving unnecessary fluids to already sick children, it was important to stress the occasions on which additional IV fluids should not be given as well as providing guidance on when additional IV fluids are indicated. |
| Implementation considerations | The GDG noted that:
Full assessment should include re-evaluation of children's circulatory status to exclude progression to shock or signs of fluid overload related to fluid management as well as to identify signs of the underlying condition. The WHO Pocket book ( WHO, 2013b) specifies administration of maintenance fluids by age and weight and consideration of the underlying disease. Maintenance fluids might have to be given intravenously until the child is able to take and retain oral fluids.
|
| Research priorities | None |
3.2.2. Children in shock
Question 6b. In infants and children who are in shock, which intravenous fluids, at what rate and for how long are associated with the lowest risk for mortality?
Summary of evidence
The systematic review (Fluid resuscitation of critically ill children aged 2 to 59 months with impaired circulation) identified only one RCT (Maitland et al., 2011), the FEAST trial, that provided some evidence for the population of infants and children presenting with the WHO criteria for shock, who are directly addressed by this recommendation.
In the FEAST trial (see the summary of evidence for the above recommendation), 65 of 3141 (2%) infants and children with severe febrile illness enrolled into stratum A exhibited all three signs of severely impaired circulation, i.e. cold extremities with capillary refill time > 3 s and a weak and fast pulse, thereby fulfilling the WHO definition of shock. The outcomes of this small group of children and subgroup analyses, such as for children with moderate hypotension, were consistent with the findings in the main study population: boluses of additional fluids were associated with increased mortality rates. The study did not, however, have enough power to detect differences in the subgroup s, and any differences may have occurred by chance. Interpretation of this subgroup analysis was also difficult because, although the children in the study were randomized, the small number in shock were not equally distributed between the intervention and control arms, with 50 in the groups receiving boluses and 15 in the maintenance group. (Note that randomization was not stratified according to this criterion.) No generalizable conclusions could be drawn from these data about the management of children in shock. No other randomized trials of bolus fluids versus no bolus were identified.
Another systematic review (Signs of severe circulatory impairment in children) was conducted to determine whether clinical signs can predict whether a child with severely impaired circulation will respond to fluids. This review identified only observational and non-randomized studies.
A background paper on the physiology of shock was prepared for the guideline meeting (Physiological basis for the administration of intravenous fluids to children with shock). This and the second systematic review provided some additional evidence.
In high-resource settings, no clinical signs have been found to predict a response to fluid management. Some invasive measurements, such as stroke volume and cardiac output, may be helpful, and a haemodynamic response to passive leg raising is likely to be associated with increased blood pressure if fluids are given subsequently. Improvements in these signs or an initial response to IV fluids do not, however, necessarily predict recovery or survival.
Initial blood pressure is not predictive of outcome, but blood pressure measurements are helpful for monitoring responses to interventions.
Immediate responses to fluid bolus are not necessarily predictive of outcomes.
Children with severely impaired circulation are likely to have raised titres of antidiuretic hormone. Giving excess fluids to this group of children may precipitate fluid overload and congestive heart failure.
Observational data in both low- and high-resource settings, including the results of invasive monitoring, indicate that 20–40 mL/kg bw of IV fluid over 30–60 min are required to restore circulating volume in children with septic shock.
View in own window
| SUPPORTING EVIDENCE AND ADDITIONAL CONSIDERATIONS |
|---|
| BENEFITS AND HARM | Do the desirable effects outweigh the undesirable effects?
☐ Yes
☐ No
☑ Uncertain | The group agreed that the findings of the FEAST trial must be carefully considered in order to determine whether they are generalizable. They also agreed that fluid management of children in shock must be judicious, and the clinical condition of the children should be carefully monitored to identify positive or detrimental responses. Decisions on fluid management in children in shock should be guided by frequent reassessments after fluid provision. The group acknowledged that this may be difficult in settings where there are few staff and many children to manage.
The group noted that it might be difficult for health workers to determine rapidly whether a child in shock has a history of diarrhoea and also has severe dehydration, for which rapid fluid provision is an essential part of management. Similarly, health workers may not accurately assess children for severe acute malnutrition, which requires a conservative approach to fluid management, even when the children have had diarrhoea.
The GDG noted that international practice favours use of fluid boluses in the management of children in shock; however, this reported practice is common in settings in which invasive physiological monitoring is available to guide management and inotropic and ventilator support is available to manage fluid overload. |
| VALUES AND PREFERENCES; ACCEPTABILITY | Is there important uncertainty about or variation in how people value the options?
☑ Major
☐ Minor
☐ Uncertain
Is the option acceptable to key stakeholders?
☐ Yes
☐ No
☑ Uncertain | The GDG considered that the general population would be unable to judge the merits of one fluid strategy versus another. Communities would have difficulty in interpreting the specificity of the population studied in the FEAST trial, i.e. with a high prevalence of malaria and anaemia requiring transfusion, and the limits of the subgroup analyses and how they should be interpreted in relation to international guidelines.
Similarly, health workers are likely to have difficulty in interpreting the findings of the FEAST trial and their relation to international guidelines.
GDG members noted the risk associated with inappropriate use of very conservative fluid management for children with severe dehydration. Every effort should be made to send the consistent message that children with diarrhoea and dehydration require additional fluid in the form of oral rehydration solution or IV fluid for children with severe dehydration. |
| FEASIBILITY AND RESOURCE USE | How large are the resource requirements?
☐ Major
☑ Minor
☐ Uncertain
Is the option feasible to implement?
☐ Yes
☐ No
☑ Uncertain | Changes in fluid management recommendations are unlikely to require major additional resources. Health workers in low-resource settings require simple algorithms to guide emergency management of sick children. Health workers should be trained and supported in using new algorithms, which and will require funds.
While simplifying the clinical algorithms for use in low-resource settings is important for improving the outcomes of sick infants and children presenting to primary or second-level health facilities, it is equally or more important to improve the skills and competence of health workers for correct assessment, triage and initial management of such children. In low-resource settings, the outcomes of critically ill children are also influenced by access to oxygen and good monitoring, including with pulse oximetry, and the availability of other forms of respiratory support, such as simple methods of continuous positive airway pressure.
The GDG was uncertain about the feasibility of implementing the recommendation because of wide variation in staffing levels and infrequent training and supervision of front-line health workers |
View in own window
RECOMMENDATIONS
- 2.4.
Children in shock, i.e. who have all the following signs: cold extremities with capillary refill time > 3 s and a weak and fast pulse, should receive 10–20 mL/kg bw of isotonic crystalloid fluids over 30–60 min. They should be fully assessed, an underlying diagnosis made, receive other relevant treatment and their condition monitored. They should be reassessed at the completion of infusion and during subsequent hours to check for any deterioration. - —
If the child is still in shock, consider giving a further infusion of 10 mL/kg bw over 30 min. - —
If shock has resolved, provide fluids to maintain normal hydration status only (maintenance fluids). - —
If, at any time, there are signs of fluid overload, cardiac failure or neurological deterioration, the infusion of fluids should be stopped and no further intravenous infusions of fluids should be given until these signs resolve.
- 2.5.
Children in shock and with severe anaemia (erythrocyte volume fraction < 15 or haemoglobin < 5 g/dL, as defined by WHO, 2013b) should receive a blood transfusion as early as possible and receive other intravenous fluids only to maintain normal hydration. - 2.6.
Children with severe acute malnutrition who are in shock should receive 10–15 mL/kg bw of intravenous fluids over the first hour. Children who improve after the initial infusion should receive only oral or nasogastric maintenance fluids. Any child who does not improve after 1 h should be given a blood transfusion (10 mL/kg bw slowly over at least 3 h) (WHO, 2013a).
|
|---|
| Strength of recommendations | 2.4 Conditional
2.5 Strong
2.6 Strong |
| Quality of evidence | Low (for all recommendations |
| Justification | The GDG agreed that the quality of the evidence for these recommendations was low because there were minimal data for the population of direct interest – children who fulfil the WHO definition of shock. The evidence was downgraded to “low” because of indirectness; i.e. children with diarrhoea and severe acute malnutrition were excluded. The generalizability of the evidence from the one clinical trial was limited because children with severe dehydration or severe acute malnutrition were excluded, and there was a high prevalence of malaria and of severe anaemia.
The additional systematic review on clinical signs and the background paper provided largely observational data and did not significantly improve the quality of the overall evidence. The group noted that the observational data from both high- and low-resource settings and the outcomes reported in the multi-country RCT were inconsistent. The group could not reconcile these inconsistencies.
The GDG agreed that there was no evidence to support a change in the WHO clinical criteria for shock.
The GDG agreed by consensus that the first recommendation, on the range of volumes and time of fluid management, should be conditional in view of the low quality of evidence and the uncertainty about the generalizability of the evidence. The group agreed that the volume of fluid to be given to children who fulfil the WHO criteria for shock depends on the context; therefore, a more precise range of volumes could not be recommended in the absence of more evidence.
The GDG considered it inappropriate to generalize the finding in the one randomized trial (FEAST) of increased mortality rates among children who did not fulfil the WHO criteria for shock and who received IV boluses of fluids (noting that children with severe dehydration were excluded from the trial). The GDG considered that this finding was not necessarily applicable to children who do fulfil the WHO criteria for shock, even though subgroup analyses in the FEAST trial indicated increased mortality in response to fluid boluses in the study population with these characteristics.
The GDG noted that international practice, while based largely on observational data, is to give IV boluses of crystalloid fluid to children in shock, especially when inotropic and ventilation support are available. The data from the FEAST trial are inconsistent with this approach, but the populations and the context are very different. In many Australian, European and North American centres, the trend is to give a 10-mL/kg bw fluid challenge and reassess the child before any further boluses are given. The GDG concluded that a range of initial fluid volumes and rates should be recommended to permit national authorities and other expert and professional groups to determine those that are appropriate in their settings.
The two other recommendations were considered to be strong, even though the quality of the evidence was also low. These recommendations support either recognized best practice or current WHO recommendations and were considered to give health workers clear guidance for improving the safety and health outcomes of children presenting with emergency signs. |
| Implementation considerations | When adapting recommendations to the national context, country teams should consider the following points as they consider adopting either a more conservative approach, i.e. a smaller fluid volume over a longer time, or a more permissive approach, i.e. a larger fluid volume over a shorter time:
the disease profile of children presenting with emergency signs, e.g. the prevalence of malaria or severe anaemia; the competence of the health workers who will be trained in applying these recommendations, including their ability to differentiate the causes of shock and to detect congestive heart failure or other signs of fluid overload; the number of health workers at health facilities who will be available to treat and monitor children presenting with emergency signs; and the availability of monitoring and support equipment, such as blood pressure measuring devices and ventilation support.
Children in shock who respond partially or not at all to fluid boluses require a differentiated response, depending on the cause. A careful history should be taken, with a clinical examination, investigations (such as echocardiography if available) and treatment. Specific supportive treatment may include oxygen, adrenaline for anaphylaxis, inotropic drugs (such as adrenaline or dopamine) for poor cardiac function, vasoconstrictor drugs (such as noradrenaline) and antibiotics for septic vasodilatation (“warm shock” with low blood pressure), diuretics and positive airway pressure (including continuous) if congestive heart failure is present.
In all settings:
In considering the underlying diagnosis, health workers should check for a history of heart disease, ingestion of poisons or toxins, allergy, snake or spider bites and signs of heart failure. If any of these is present, fluid management should be reviewed and treated accordingly. Fluid therapy alone may not be sufficient for the management of many children presenting with shock. Early inotropic or ventilator support may be required. After resolution of shock, the child's condition should be continuously monitored and assessed to direct further management.
|
| Other remarks | The following points were discussed in plenary:
Decisions on fluid management in children presenting with severely impaired circulation must take into account the cause of shock and the ability of health workers to evaluate the children accurately. Health workers tend to interpret clinical signs in the clinical context in which they practise, e.g. the conditions that are especially prevalent and the clinical history of the child for whom they are caring. For this reason, health care workers must rapidly take a history from the caregiver when the child first presents, in addition to a full history and assessment after initial treatment. Even in high-resource settings, IV fluid management of children with severe circulatory impairment can result in fluid overload, which may be harmful. Fluid overload generally indicates the need for inotrope and ventilation support. Observational studies in high-resource settings demonstrate that children with septic shock or who are severely hypotensive require rapid expansion of the cardiac volume up to 30 mL/kg bw within 15 min in order to restore initial circulation; however, the correlation of this short-term outcome with long-term outcomes is not known. In high- and middle-income countries or settings, the availability of inotropes, ventilation support, technical monitoring equipment and adequate human resources greatly influence decisions about fluid management of children who present with shock. The absence of such facilities in many low-resource settings complicates decisions about what is generally advisable for children presenting with shock. Other aspects of care, such as early assessment and reassessment, substantially influence the outcomes of children, including their response to fluids within the first 30–60 min. Children presenting with severely impaired circulation require not only initial triage and treatment but also reassessment and continuous monitoring to guide further fluid management and other interventions. Children with severe anaemia should receive blood as soon as possible. The group noted, however, the practical difficulty of providing blood rapidly as part of emergency treatment in most low-resource settings. In some children with severe anaemia, the circulating volume is normal or even expanded, so, although a blood transfusion is urgently required, it should be given slowly. If it is given too rapidly, it can lead to circulatory overload. For children in haemorrhagic shock, blood should be infused rapidly, but for those in shock associated with severe anaemia, with no loss of blood volume, blood should be given over 2–4 h. In patients with normal or expanded circulating volume and shock associated with severe anaemia, furosemide should be given with the blood transfusion so that it can be infused more rapidly without causing circulatory overload. Additional research is required to provide further high-quality evidence for this specific population and greater insight into this complex question.
|
| Research priorities |
What is the optimal fluid management of children in shock (as defined by WHO) who also have severe acute malnutrition or severe anaemia? What is the role of blood transfusion in the management of children with shock and severe anaemia?
|
- a
Emergency signs described in WHO ETAT guideline include:
Obstructed or absent breathing
Severe respiratory distress
Central cyanosis
Signs of shock, defined as cold extremities with capillary refill time > 3 s and weak and fast pulse
Coma (or seriously reduced level of consciousness)
Seizures
Signs of severe dehydration in a child with diarrhoea with any two of the following signs: lethargy or unconscious, sunken eyes, very slow return after pinching the skin.
- b
In infants and children 6–59 months of age, severe acute malnutrition is defined as weight-for-height < −3 Z-score 1 of the median of the WHO growth standards or clinical signs of bilateral oedema of nutritional origin, even if other measures are above specified cut-off values (WHO, 2009a).
