Department of Paediatrics, Starship Children's Hospital, University of Auckland, Private Bag 92024, Auckland, New Zealand
Background. Voiding cystourethrograms are distressing for children and parents. Nonpharmacological methods reduce distress. Pharmacological interventions for VCUG focus on sedation as well as analgesia, anxiolysis, and amnesia. Sedation has cost, time, and safety issues. Which agents and route should we use? Are we sure that sedation does not influence the ability to diagnose vesicoureteric reflux? Methods. Literature search of Medline, EMBASE, and the Cochrane Database. Review of
comparative studies found. Results. Seven comparative studies including two randomised controlled trials were reviewed. Midazolam given orally (0.5-0.6 mg/kg) or intranasally (0.2 mg/kg) is effective with no apparent effect on voiding dynamics. Insufficient evidence to recommend other sedating agents was found. Deeper sedating agents may interfere with voiding dynamics. Conclusion. Midazolam reduces the VCUG distress, causes amnesia, and does not appear to interfere with voiding dynamics. Midazolam combined with simple analgesia is an effective method to reduce distress to children undergoing VCUG.
1. Introduction
The voiding cystourethrogram
(VCUG) is the gold standard for diagnosing vesicoureteric reflux (VUR) and a
number of other bladder conditions [1, 2]. The VCUG is a procedure performed
mainly on infants and young children in the Radiology Department [3]. There is increasing belief that
interventions for VUR are less effective than anticipated, but much debate
remains [4–11]. The child is required to be conscious,
a urinary catheter is inserted, and the bladder is
filled with radio-opaque material, then the child is asked to spontaneously
void [12]. This procedure creates distress in the
child, the parents, and occasionally staff [13–15]. Nonpharmacological methods to reduce
this distress include education prior to the procedure, distraction during, and
rewards after [14, 16–28].
Pharmacological interventions primarily
focus on sedation but also could include beneficial analgesic, anxiolytic, and
amnesic effects [29, 30]. Sedation brings with it cost, time,
and safety concerns [1, 29–33]. It is unknown whether we can predict
which children will go on to have distress or whether we should sedate
routinely [34]. Which agents should we use, and what
is the best route of administration? The
majority of children having VCUG would not have had one previously. Coping
styles and parent-child interaction are important determinants of distress
during a medical procedure [34]. Safety of sedating agents is excellent
in the context of a sedation service with the necessary staff and equipment to
manage sedation in young children [35–37]. Those who sedate children should be
prepared for inadvertent deeper sedation, basic life support, and airway
management [29–31, 38, 39, 40]. Advanced help should be available.
Time and cost factors limit the introduction of this distress-reducing
intervention. Sedation recovery area and staff time are being the primary cost
factors. The medications themselves are relatively inexpensive. Finally, are we
sure that sedation does not influence the VCUGs ability to diagnose
vesicoureteric reflux?
2. Methods
These methods include a sensitive search of PubMed (1950–2007), EMBASE
(1980–2007), Cochrane Database of Systematic Reviews, and Cochrane Randomised Controlled Trials Register. Articles on VCUG were identified through the terms
urography (MESH heading exploded), micturating, or voiding cystourethrogram
using wildcard search for variations of spelling. Acronyms VCUG and MCUG were
also used. To identify sedation articles, the following exploded MESH terms
were used: “hyponotics and sedatives,” conscious sedation, midazolam, propofol,
chloral hydrate, and nitrous oxide. “Sedation” was searched for as a title
word. Results of the VCUG search and sedation search were combined. There were
no limits on language for search, but only English language articles were
reviewed. Further studies were identified from bibliographies. Unpublished
studies were not actively sought.
3. Results
Medline search (2008) found 234 papers
of which 17 were considered to be of interest [6, 8, 13, 14, 16, 41–52]. EMBASE search found 416 papers of which
additional 8 papers were of interest [27, 35, 53–59]. Cochrane Randomised Controlled Trials Register
found no further articles of interest. Cochrane Database of Systematic Reviews
found one review on interventions for primary vesicoureteric reflux, but none
on sedation for this procedure [7]. A review of the bibliographies
identified further 39 papers of interest [2–5, 9, 11, 15, 17, 18, 19, 20, 21, 22, 26, 28, 34, 37, 40, 60–81]. Four papers on anesthesiology for VCUG were found
and included for discussion [82–85].
Any study comparing a
sedative against another sedative, placebo, or standard treatment for VCUG
was reviewed. One French language article [77] and one Polish language article [56] were not included. Study designs are summarised in
Table 1. Outcome measures and results are in Table 2 [41, 43, 45, 47, 52, 59]. Quality assessment is shown in Table 3. Only two
of the studies [45, 52] were of high quality with Jadad scores [86] of 4 or more. One unpublished comparative study
was found, but not included [68].
Table 1: Studies comparing a sedative against another sedative, placebo, or standard treatment
for VCUG. Design and interventions. (NSD: no significant difference; VCUG: voiding cystourethrogram; RNC: radionuclide cystography; VUR: vesicoureteric reflux; SD: standard deviation).
Table 2: Studies comparing a sedative against another sedative, placebo, or standard treatment for VCUG. Outcomes,
results, and follow-up. (NSD: no significant difference; VCUG: voiding cystourethrogram; RNC: radionuclide cystography; VUR: vesicoureteric reflux; SD: standard deviation).
Table 3: Quality assessment of studies of sedation for VCUG, including Jadad score [
86].
4. Discussion
The best way to avoid the distress of
the VCUG is not to do the procedure. A better way to image vesicoureteric
reflux has recently been discussed in an editorial by Elder [55]. As much evidence becomes available to show that we
are not influencing the outcome of VUR, less VCUGs may be ordered [7, 93]. Possible alternatives include Doppler ultrasound [94] or ultrasound with contrast [61]. A suprapubic approach to avoid catheterisation
seems promising but still requires filling and voiding [49]. Methods to detect reflux without voiding are
impaired as some reflux may be present only on voiding [95], although the fact that whether this is important
or not is debated [96]. Nuclear medicine scanning may be an alternative
or may be able to select those who are more likely to benefit from VCUG [67]. Nuclear medicine cystoscopy replaces radio-opaque
contrast with pharmacolabelled material with lower radiation, but otherwise it is
very similar to the VCUG. Currently, VCUG remains the gold standard until less
invasive tests are developed [1]. At the very least, we should be perfecting our
current technique [12].
5. Distress, Pain, and Anticipatory Anxiety
Distress is an all encompassing term
that may or may not include a painful stimulus. This can be evidenced by fear
of a procedure, pain responses prior to nociceptive stimuli, or anxiety
behaviours before a planned event. Distress during the voiding cystourethrogram
has been reported in observational and controlled studies. Phillips et al. [13] showed that 52 out of 73 children (71%)
experienced serious distress, severe distress, or panic on the Groningen distress
rating scale [90]. Herd et al. found that serious or severe distress
was detected in 61% of all unsedated children at some stages during the VCUG.
This level of distress may have been brief but is generally considered
unacceptable. This distress is caused not only by urethral catheterisation, but
also by the distension of the bladder and the subsequent voiding of urine in a
socially abnormal situation (Figure 1). Nociceptors related to urethral mucosa
and stretch receptors in the bladder provide the peripheral pain signals, but
the majority of the distress is cortical.
Figure 1: Bar graph shows the number of children (
) who experienced serious or severe distress
(Groningen distress rating scale (GDRS) score
2 [
90]) at each phase of voiding cystourethrography. 56 received placebo (black bars) and 61
received midazolam (white bars), from [
45].
Distress can also be manifest in the
parent. Parents' perceptions of fear, distress, and pain in their children are
anticipated to be greater than the reality [51].
6. Psychological Therapy (Nonpharmacological Treatments)
Psychological treatments should be
considered first as they often have little or no risk. There are many ways to
reduce the distress of procedures in children [19] and even more challenges researching and
implementing psychological interventions for controlling children’s pain [97]. Interventions may range between
simple educational [28] and more
structured play therapy sessions[14, 76]
and hypnoses [42]. Preparation, distraction, and reassurance before,
during, and after the procedure are likely to reduce the distress of the
procedure [22, 69, 75]. Parental presence is comforting for children
during painful procedure and should be encouraged despite the lack of specific
VCUG evidence [23].
Those who have been previously
distressed by VCUG would seem to be ideal candidates for sedation, but the
majority of children would not have had a previous VCUG. Factors which may
reduce distress in children during VCUG include “effortful control” by the
child and coping and distress-promoting behaviours by the parent [34]. No validated prediction tool exists for VCUG
distress.
7. Pharmacotherapy
Pharmacotherapy includes sedation, anxiolysis, analgesia, amnesia, and anesthesia.
7.1. Sedating Agents
Sedation continues to be difficult to
define[63]. In the case of the VCUG, a degree of
consciousness is required. This may be defined as light sedation, and the use
of oxymoron “conscious sedation” is being discouraged [98].
Midazolam
Of the selected studies, 5 had midazolam
as a treatment arm and 4 of which were oral and one intranasal. Oral midazolam
dose was 0.5 mg/kg in two studies [45, 47] and 0.6 mg/kg in two other studies [41, 43]. Maximum dose was 15 mg in all. Time
between ingestion and procedure ranged from 15 to 30 minutes. Intranasal dose
in one study was 0.2 mg/kg with a maximum of 5 mg, and it was administered 5
minutes before the procedure [52]. A number of behavioural measures were
employed (Table 1). All the studies demonstrate significantly less distress
with midazolam in a variety of measurement tools. Few adverse effects were
encountered. Midazolam may cause adverse paradoxical agitated reactions in less
than 5% of children [99]. These reactions have been shown in
case reports to be ameliorated using the antidote for midazolam (flumazenil)
both in adults [100] and children [101]. Ketamine, a dissociative anesthetic,
has been shown to be more effective than increased doses of midazolam or
placebo in a randomised controlled trial [99].
The study by Stockland et al. [52] on 95 children compared intranasal
midazolam (0.2 mg/kg with a
maximum of 5 mg) to placebo. Nurses reported a trend to easier procedure in the
midazolam group (), with girls reported easier than boys (). No serious adverse events were reported. Parents felt that the
administration of midazolam was more uncomfortable than that of placebo ().
Parents felt that midazolam made catheterisation, voiding, and the overall
procedure more comfortable (, , and ,
resp.). The authors report -values and no absolute scores, which
makes it impossible to estimate treatment effect size or clinical relevance.
A study by the current author and
colleagues [45] compared oral midazolam (0.5 mg/kg with a maximum of 15 mg)
to placebo in 125 children who had VCUG. Behavioural observations were
completed in 117. This was the only study that calculated a priori power requirement or attempted
to quantify the treatment effect. We rated our paper highly using the Jadad
score [86]. We found no serious adverse events.
The number of children experiencing serious or severe distress (Groningen
distress rating scale
(GRDS) 2) at any stage of the procedure was 34 (61%) in the placebo group
and 16 (26%) in the midazolam group. Number needed to treat to reduce serious or severe severe distress in one child was
2.9 (95%CI 1.9–5.5).
VUR was identified in 16% of all children. This study was limited to children
above the age of one year.
Nitrous oxide (O)
Two studies evaluated nitrous oxide
given with continuous flow devices at 50% and 70%. Keidan et al. compared 50%
nitrous oxide in 23 children to 0.5 mg/kg oral midazolam in 24 children without
a placebo group [47]. They found no difference between
midazolam and 50% nitrous oxide although they did not design this as an
equivalence study, and no power calculation was done. There was a trend for the
time to micturition to be longer in the nitrous group (15.3 minutes versus 7.2
minutes), but it did not reach statistical significance (). Nitrous
oxide was significantly faster with regard to recovery time, with recovery in 29 minutes
versus 63 minutes (). Zier et al. used 70% continuous nitrous oxide only for urethral
catheterisation phase of VCUG in an older group of 107 children, and compared
this to standard treatment in other 107 children [59]. The authors chose not to randomise the
study based on difficulties with recruitment and parental expectations. Brief
behavioural distress scores (BBDSs) were demonstrated by the observational tool
selected[102]. Wong-Baker FACES scale was the self-report tool used [103]. For the VCUG group (), BBDS was
44 (range of 11–100) in the
nonsedated group versus 11 (range of 0–67) for the
sedated group (). Immediately after catheterisation, the Wong-Baker
FACES scale median was 6 for the nonsedated group and 0 for the sedated group ().
Both studies reported time of
completion, but neither study reported VUR grading or residual volume.
Other agents
Choral hydrate was compared to oral
midazolam and placebo in one study [41]. A dose of 25 mg/kg was not found to be
statistically different from placebo in reducing distress. This may have been
due to inadequate dose or lack of power in the study. The sedation scale was
also not significant for chloral versus placebo and, therefore, it suggests too low a dose was selected. There is no enough data to
make any assessment of effect on voiding dynamics.
One retrospective study of propofol
using historical controls was selected for review [48]. While this was an attempt to create a
sedative state using low-dose propofol, the study required the presence of an
anesthetist. During this study, low-dose propofol infusion followed sevoflurane
gas induction and intravenous cannula insertion. This study found that propofol
reduced the ability of children to completely void, which may interfere with
the diagnosis of VUR.
7.2. Anxiolysis
Midazolam in the doses used in the
reviewed studies is anxiolytic. Many children may appear fully conscious yet
more cooperative, while another child given the same dose may appear sleepy.
Where anxiolysis ends and sedation begins is unclear, but there would be a
large overlap.
7.3. Analgesia
There is a wide range of analgesics
available for children [104]. Midazolam does not provide any
analgesia and, therefore, should be supplemented with a simple analgesic.
Acetaminophen
Acetaminophen is the most commonly
provided childhood analgesic with low side effects and cost. It is routinely
offered prior to other potentially painful procedures in children such as
vaccination. Acetaminophen is usually provided in a sweet syrup base, and could
be used to disguise the bitter taste of midazolam. There are many formulations
of acetaminophen syrup, and palatability may vary [105].
Oral sucrose
Oral sucrose is an effective analgesic
in new-born babies, and has been subject to several controlled trials and a
Cochrane review [106]. While no studies have examined its
effect for VCUG distress, it seems a simple likely effective intervention with
low risk for children under 3 months of age.
Nitrous oxide
Nitrous oxide is a strong analgesic
antagonising central NMDA receptors, and this is a potential advantage over
midazolam.
Study of Keidan et al. comparing
midazolam to continuous flow nitrous oxide found no difference in FLACC scores [91], a measure of pain used more recently
for procedural distress [107]. Study of Keidan et al. was not
designed as an equivalence study, and no power calculations were done; so a
true difference may not have been detected by the study.
Opiates
No studies have looked at opiate use for
VCUG distress. Intranasal midazolam has proven effective, and opiates may also
be administered by this route. Intranasal fentanyl shows promise as a rapid,
easy-to-administer analgesic for severe pain in the children’s Emergency Department [108]. Opiates may interfere with bladder
function [109].
Local anesthetics
Lignocaine gel has been shown to reduce
the pain of catheterisation for VCUG, but a 10-minute process of repeated
application of lignocaine gel to the urethral meatus is required. The authors
did not measure the effect of this procedure but only the reduced pain of
catheterisation that followed. It would seem reasonable to use it with low risk
of harm but at added cost [66]. Further study on children is required.
7.4. Anesthesia
There is increasing use of deeper
sedation outside the operating room by nonanesthesiologists [110]. There is debate about which agents
should be used outside the operating room and who should provide this service [63]. For VCUG, anesthetics have been given
to avoid the trauma associated with urethral catheterisation, and then the
child is allowed to wake and complete the VCUG. This does not avoid the
distress caused by bladder distension or micturition. It also requires an
anesthetist and the full costs associated with anesthesia and recovery.
8. Who Should Receive Sedation?
Many children do not experience distress
during the VCUG. This may be related to previous experience, coping style,
parental influence, staff skill, and empathy. Developmental considerations and
education level of the child and parent are important. Nevertheless, many
children, who would not have been predicted, may go on to experience distress.
Parental perceptions of the procedure are such that most parents would request
some medication if it were effective, safe, and available [43, 47].
9. Does Sedation Affect the Ability of the VCU to Diagnose Reflux?
Effect of sedation on ability to void can be measured with
indirect or direct measures. Indirect measures include filling volume, residual
volume, and time of
micturition. Bozkurt et al. carefully examined urodynamic variables under the
influence of midazolam [62]. They used a high-intranasal dose of
0.5 mg/kg. Stockland et al. used intranasal midazolam at a dose of 0.2 mg/kg, and
found no difference in reflux grading between the groups [52]. They did not perform a power
calculation, so there is still the possibility of missing a true effect. Herd
et al. considered a clinically important difference in VUR to be a true shift
of one grade down by half of the subjects with the use of midazolam [45]. It was important to detect a
difference, so a 90% power was used. There was no difference in VUR grading
between the groups (nonlinear mixed model analysis, ). There was no
evidence of a difference in volume infused between the two groups ().
10. Conclusions
Sedation reduces distress of the micturating cystourethrogram in children previously
distressed or likely to be distressed. Midazolam is the agent most studied, and
has an excellent safety profile. An oral dose of 0.5-0.6 mg/kg or
intranasal dose of 0.2 mg/kg seems effective. Most children have not had a VCUG
previously, and it may be difficult to predict which of them will go on to have
distress. When giving oral midazolam of 0.5 mg/kg to children routinely, the
number needed to treat them is 2.9
(95%CI 1.9–5.5) to eliminate serious or severe distress.
Continuous flow nitrous oxide appears promising, particularly with a fast onset
and recovery time, but it has greater potential for deeper sedation. This may
interfere with voiding, and further studies are required. Midazolam appears not
to interfere with the VCUG's ability to diagnose vesicoureteric reflux using
indirect (residual volume) and direct (VUR grading) measures. There are many
children who would avoid distress if they were given sedation. Local sedation
services should be engaged, and safety guidelines should be followed to ensure
that this effective treatment might be implemented safely.
Acknowledgment
A grant from the Waikato Medical
Research Foundation for the original VCU sedation study made this review
possible.