Table 2 Summary of findings

Comparison 1a: Intranasal analgesia (IN) vs intravenous administration (IV):

Single dose IN fentanyl vs IV morphine in children aged 7–15 with acute fractures

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty of the evidence (GRADE)

What happens

Risk with IV morphine

Risk with IN fentanyl

Pain 10 min

Mean pain score: 41 mm

Mean 5 mm higher

(7 lower to 16 higher)

65

(1 RCT) [29]

Low^a

Children receiving intranasal fentanyl may achieve similar reductions in pain scores at 10 and 30 min as those receiving intravenous morphine

Pain 30 min

Mean pain score: 33 mm

Mean 4 mm higher

(8 lower to 16 higher)

65

(1 RCT) [29]

Low^a

Rescue medication

59 per 1000 children

61 per 1000 children

(9 to 405)

RR 1.03

(0.15 – 6.89)

67

(1 RCT)³²

Low^a

Mild adverse events

29 per 1,000 children

91 per 1,000 children

(10 to 830)

RR 3.091

(0.338 to 28.281)

67

(1 RCT) [29]

Low^a,b

Comparison 1b: Intranasal analgesia (IN) vs intravenous administration (IV):

Multiple doses IN fentanyl vs IV fentanyl in children aged 4–8 with postoperative pain

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty of the evidence (GRADE)

What happens

Risk with IV fentanyl

Risk with IN fentanyl

Pain 10 min (Hannallah score)

MD -0.47

(-1.51 – 0.37)

32

(1 RCT)³³

Low^a

Patients receiving intranasal fentanyl may achieve similar reductions in pain scores at 10 and 15 min as those receiving intravenous fentanyl

Pain 15 min

(Hannallah score)

MD 0

(-0.35 – 0.35)

32

(1 RCT)³³

Low^a

Mild adverse events

125 per 1000 children

187.5 per 1000 children

RR 1.50

(0.29 – 7.81)

32

(1 RCT)³³

Low^a

Comparison 1c: Intranasal analgesia (IN) vs intravenous administration (IV):

Single dose IN ketorolac vs IV ketorolac in children aged migraine headache with migraine

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty of the evidence (GRADE)

What happens

Risk with IV ketorolac

Risk with IN ketorolac

Pain 10 min

MD 0.9

(-0.4 – 2.2)

56

(1 RCT)³¹

Low^a

Patients receiving intranasal ketorolac may achieve similar reductions in pain scores at 10 and 30 min as those receiving intravenous ketorolac

Pain 30 min

MD 0.8

(-0.4 – 1.9)

56

(1 RCT)³¹

Low^a

Rescue medication

170 of 1,000 children

220 per 1,000 children

RR 1.29

(0.44 – 3.73)

56

(1 RCT)³¹

Low^a

Mild adverse events

207 per 1,000 children

148 per 1,000 children

RR 0.716

(0.23 – 2.26)

56

(1 RCT)³¹

Low^a

Acceptability (child)

955 per 1,000 children

792 per 1,000 children

RR 0.83

(0.66 – 1.04)

46

(1 RCT)³¹

Low^a

a – single study, no published protocol

b – underpowered to find differences in uncommon, but serious adverse events

Comparison 2: Intranasal analgesia (IN) vs intramuscular administration (IM)

IN fentanyl or IN diamorphine vs IN morphine in children aged 3–17 with acute moderate to severe pain

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty

What happens

Risk with IM morphine

Risk with IN fentanyl/diamorphine

Pain 10 min

Kendall 2001 found lower mean pain scores in the IN diamorphine group than in the IM morphine group (MD 0.35, 95%CI -0.01, 0.71);

Wilson 1997 found no difference in median pain scores between the IN diamorphine group and the IM morphine group (MD 0, 95%CI -0.24, 0.24)

Younge 1999 found a 1-point lower median pain score in the IN fentanyl group than in the IM morphine group

490 (3 RCTs) [32,33,34]

Moderate^a

Patients receiving intranasal fentanyl or diamorphine probably experience similar or more pain reduction at 10 and 30 min than those receiving intramuscular morphine. Any difference is unlikely to be clinically relevant

Pain 30 min

Kendall 2001 found lower mean pain scores in the IN diamorphine group than in the IM morphine group (MD 0.43, 95%CI 0.08, 0.78)

Wilson 1997 found no difference in median pain scores between the IN diamorphine group and the IM morphine group (MD 0, 95% CI -0.55, 0.55)

Younge 1999 found no difference in median pain score between the IN fentanyl group and the IM morphine group

483

(3 RCTs) [32,33,34]

Moderate^a

Rescue medication

45 per 1,000 children

50 per 1,000 children

(23 – 108 children per 1,000)

RR 1.11

(0.51 – 2.40)

503

(3 RCTs)

^28–30

Low^a,b

Patients receiving intranasal fentanyl or diamorphine may experience similar risk of requiring rescue analgesia, as those receiving intramuscular morphine

Adverse events

Kendall 2001 reported total adverse events 49/203 (24%) in the IN diamorphine group, and 37/200 (18.5%) in the IM morphine group. All events were mild, except one case of vomiting in the IN diamorphine group; and over halt the events were local irritation at the site of administration. Neither Kendall 2001, Wilson 1997 or Younge 1999 found a difference in pulse, respiratory rate or GCS, or clinically significant in O2-saturations, but a high dropout rate in the IM morphine group

502

(3 RCTs)

[32,33,34]

Low^{a, b}

Patients receiving intranasal fentanyl or diamorphine may experience similar risk of adverse events, as those receiving intramuscular morphine

Acceptability (child): uncooperative / negative reaction

505 per 1,000 children

31 per 1,000 children

(15 to 65 per 1,000 children)

RR 0.0611 (0.0291 to 0.1282)

449

(2 RCTs) [32, 34]

High^d

Children are less likely to be uncooperative or have a negative reaction to intranasal analgesia than intramuscular morphine

Acceptability (child)

588 per 1,000 children

941 per 1,000 children

(835 to 1,000 children)

RR 1.60 (1.42 to 1.81)

402

(1 RCT) [32]

Moderate^c

Children probably find intranasal diamorphine more acceptable than intramuscular morphine

Acceptability (parents)

721 per 1,000 chilrden

966 per 1,000 children (887 to 1,000 children)

RR 1.34

(1.23 to 1.47)

389

(1 RCT) [32]

Moderate^c

Parents probably find intranasal diamorphine more acceptable than intramuscular morphine

Acceptability (providers)

322 per 1,000 children

981 per 1,000 children

(801 to 1,000 children)

RR 3.05

(2.49 to 3.73)

402

(1 RCT) [32]

High^d

Providers find intranasal diamorphine more acceptable than intramuscular morphine

a – Wilson 1997 had high risk of bias, and some concerns in Kendall 2001 and Younge 1999

b – Imprecision: underpowered to find differences in uncommon events including rescue medication and adverse events

c – Imprecision—single study

d – Single study/only two studies, but upgraded because of large effect size

Comparison 3a: Comparison of different intranasal (IN) agents

IN ketamine vs IN fentanyl in children aged 3–17 with acute moderate to severe pain

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty

What happens

Risk with IN fentanyl

Risk with IN ketamine

Pain 10–15 min

SMD 0.05

(-0.19 to 0.28)

263

(4 RCTs) [36, 37, 39, 40]

High

Patients receiving intranasal ketamine achieve similar pain reduction at 10–15 min, as those receiving intranasal fentanyl

Pain 30 min

SMD 0.05

(-0.20 to 0.29)

254

(4 RCTs) [36, 37, 39, 40]

High

Patients receiving intranasal ketamine achieve similar pain reduction at 30 min as those receiving intranasal fentanyl

Rescue medication

396 of 1,000 children

336 per 1,000 children

(245 – 461 of 1,000 children)

RR

0.85

(0.62 – 1.17)

268

(4 RCTs)

^36–39

Moderate^a

Patients receiving intranasal ketamine probably experience a similar risk of requiring rescue analgesia as those receiving intranasal fentanyl

Mild adverse events

389 per 1,000 children

841 per 1,000 children

(670 to 1,000 children)

RR 2.16

(1.72 to 2.71)

263

(4 RCTs)

^{[[36, 37, 39, 40]]}

High

Patients receiving intranasal ketamine probably experience a higher risk of adverse events than those receiving intranasal fentanyl

Adverse events – Sedation

315 per 1,000 children

550 per 1,000 children

(408 to 740 per 1000 children)

RR 1.74*

(1.30 to 2.35)

261

(4 RCTs)

^{[[36, 37, 39, 40]]}

High

Patients receiving intranasal ketamine experience a higher risk of sedation than those receiving intranasal fentanyl

Acceptability (child)

722 per 1,000 children

829 per 1,000 children (644 to 1,000 children)

RR 1.15 (0.89 to 1.48)

71

(1 RCT) [36]

Low^b

Children may find intranasal ketamine equally acceptable to intranasal fentanyl

Comparison 3b: Comparison of different intranasal (IN) agents

IN fentanyl vs IN placebo in children aged 3–20 with vaso-occlusive crisis of sickle cell disease

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty

What happens

Risk with IN placebo

Risk with IN fentanyl

Pain 10 min

Mean pain reduction

1.0

(SD 2.0)

Mean pain reduction

2.2

(SD 2.6)

49

(1 RCT)³⁴

Low^b

Children receiving IN fentanyl may experience a 2 point greater reduction in pain than those receiving placebo at 10 min. This difference represents an appreciable change

Pain 30 min

Mean pain reduction

1.8

(SD 2.5)

Mean pain reduction

2.3

(SD 2.8)

49

(1 RCT)³⁴

Low^b

Children receiving IN fentanyl may experience no difference in reduction of pain compared to those receiving placebo at 30 min

Adverse events (in VOC)

There were no differences in serious adverse events. Two subjects in either group had hypotension that resolved spontaneously, and three in the intranasal fentanyl group had transient hypoxia

49

(1 RCT)³⁴

Low^b

There may be little or nor difference in adverse events between intranasal fentanyl and placebo in vaso-occlusive crises

Comparison 3c: Comparison of different intranasal (IN) agents

Standard concentration IN fentanyl vs high concentration IN fentanyl in children aged 7–15 with fractures

Outcomes

Absolute effect (95% CI)

Relative effect

No of participants (studies)

Certainty

What happens

Risk with high concentration IN fentanyl

Risk with standard concentration IN fentanyl

Pain 10 min

Median pain reduction

20 mm

Median pain reduction

20 mm

Median difference

0

(-5.2 – 5.2 mm)

189

(1 RCT)³⁵

Low^b

Patients receiving standard concentration (50mcg/ml) intranasal fentanyl may achieve similar pain relief at 10 min, as those receiving high concentration (300mcg/ml) intranasal fentanyl

Rescue medication

275 per 1,000 children

429 of 1,000 children

(286 – 642 per 1,000 children)

RR

1.56

(1.04 – 2.34)

189

(1 RCT)³⁵

Low^a

Patients receiving standard concentration (50mcg/ml) intranasal fentanyl may experience a higher risk of requiring rescue analgesia than those receiving high concentration (300mcg/ml) intranasal fentanyl

Adverse events

320 per 1,000 children

225 per 1,000 children

RR 0.76 (0.47 – 1.23)

189

(1 RCT)³⁵

Low^a

Patients receiving standard concentration (50mcg/ml) intranasal fentanyl may experience similar risk of adverse events, as those receiving high concentration (300mcg/ml) intranasal fentanyl

a – Imprecision: underpowered to find differences in uncommon events, including rescue medication and serious adverse events

b – Imprecision: single study