| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Waikato Hospital, PO Box 3200, Hamilton, New
Zealand.
2 Department of Paediatrics, Waikato Hospital, Hamilton, New Zealand.
3 Department of Psychology, University of Waikato, Hamilton, New Zealand.
Received July 14, 2005;
accepted after revision October 25, 2005.
The Waikato Medical Research Foundation provided funding for this study.
The Health Research Council of New Zealand provided funding for N. A.
Keene.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Our study was a randomized double-blind controlled trial performed at a university teaching hospital; our study group consisted of children over the age of 1 year who been referred for their first VCU examination from July 2001 to July 2003. Participants were randomized to receive a placebo or midazolam syrup (0.5 mg/kg) before the examination. The primary outcome measures were the Groningen Distress Rating Scale (GDRS) and grading of VUR, as defined by the international grading system established by the International Reflux Study Group.
RESULTS. There were no serious adverse events. One hundred thirty-nine children were randomized in the study, and 117 underwent complete assessment. Eight who underwent VCU after the study day were included in a "complete case" intention-to-treat analysis. In the placebo group, 34 children (61%) experienced serious distress or severe distress (GDRS score, 3 or 4). In the midazolam group, 16 children (26%) experienced the same degree of distress. There was a significant difference between the GDRS scores (nonlinear mixed-model analysis, p < 0.001) of the two study groups. The number needed to treat to reduce serious or severe distress in one child was 2.9 (95% CI, 1.9-5.5). VUR was identified in 16% of all children. There was no difference in VUR grading between the groups (nonlinear mixed-model analysis, p =0.31).
CONCLUSION. Routine use of oral midazolam (0.5 mg/kg) for conscious sedation of children undergoing VCU reduces distress and does not alter the ability of VCU to show VUR well enough to allow diagnosis.
Keywords: conscious sedation • Groningen Distress Rating Scale • midazolam • pediatric radiology • urinary tract • voiding cystourethrography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The objectives of our study were to assess whether conscious sedation using oral midazolam is effective when given routinely and to ensure that conscious sedation using oral midazolam does not alter the ability of VCU to show vesicoureteric reflux (VUR).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Intervention
The intervention was to give oral midazolam syrup (0.5 mg/kg) to patients
while they were in a comfortable setting 30 minutes before catheter insertion.
The patients in the midazolam group were monitored continuously and fasted
before the procedure (no solid food for 6 hours and no liquid for 4 hours).
The procedure for the patients in the placebo group did not differ from a
normal single-void VCU examination performed in our department apart from
increased monitoring. No restraints were used, and a caregiver or caregivers
were present throughout the procedure.
All caregivers were offered the opportunity for their child to participate in play therapy before the procedure, which included a visit to the department during which the child laid down on the table and a doll was catheterized. Caregivers received a four-page pamphlet describing the procedure, and its benefits and risks, and they were given an opportunity to talk to the principal investigators before signing informed consent. The study had the approval of the Waikato Hospital Ethics Committee.
Eligibility
Eligible children were those undergoing their first VCU examination over
the age of 1 year. Children who had undergone VCU when younger than 1 year and
were now having their first VCU examination when 1 year old or older were also
eligible. We excluded children with a previous reaction to midazolam and those
with a systemic illness (American Society of Anesthesiology [ASA] grade 3 or
4).
Assignment
Individuals were randomized centrally by the pharmacy using blocks of 20
randomly sorted sealed envelopes that were opened sequentially. Group
allocation occurred at the time the appointment for VCU was made. The
pharmacist preparing the midazolam syringes and the placebo syringes and a
research assistant were aware of the allocation to ensure that the children
receiving the midazolam appropriately had fasted. An emergency disclosure
mechanism also existed.
If VCU could not be completed on the study day and the child came back for VCU on another day within 3 months of the study day, the results of that examination were included in the "complete case" intention-to-treat analysis.
Masking (Blinding)
A midazolam solution of 15 mg/3 mL was reconstituted in 3 mL of a 50 g/20
mL solution of a commercial mango-and-orange-flavored drink sachet to create a
final solution of 2.5 mg/mL. The dose given was 0.5 mg/kg with a maximum dose
of 15 mg (6 mL). The solutions were identical in packaging and appearance. The
midazolam solution still tasted slightly bitter but was well tolerated by
patients.
The principal investigators, the person performing the procedure, the person assisting with the procedure, the caregivers, the behavioral observers, and radiologists reviewing the films were unaware of group allocation. Blinding was assessed on the last 20% of caregivers (n = 24). They were asked which group they thought their child was in before the procedure and after the procedure.
Outcome Measures
The first primary outcome measure was distress as measured by the Groningen
Distress Rating Scale (GDRS): 1 = calm; 2 = mild distress; 3 = serious
distress, in control; 4 = severe distress, out of control; and 5 = panic
[2]. The second primary outcome
measure was VUR grade, as defined by the International Reflux Study Group
[11].
A questionnaire was completed by the caregiver present to identify potential confounders and the coping style of the child. We also asked the caregivers to complete a pre- and postprocedure assessment of their perceptions about the VCU examination [12]. The heart rate average for each phase of the examination and the baseline heart rate were measured. A score on a newly devised scale to measure caregiver and child interactions was recorded, and those data have been reported separately [13]. The volume infused into the bladder, the time for the examination overall, and the time between specific phases and events were recorded.
Data Collection
The VCU procedure was divided into five phases for data recording and
analysis. These phases were as follows: entering room, catheterization of the
bladder, filling bladder, voiding, and leaving room. A trained observer was
seated unobtrusively in the room where they could see the child and caregiver
or caregivers. Observations of the child, the caregiver, and their
interactions were recorded for each phase. Each phase was composed of at least
three steps; therefore, a minimum of 15 observations was taken for each
participant. In 20% of the cases, a second observer completed the same
observations. Interobserver reliability was calculated by dividing the number
of agreements with the total number of assessments.
Heart rate and oxygen saturation were recorded every 10 seconds using a Nellcor NPB-290 monitor (Nellcor Puritan Bennett). Blood pressure was measured before, once during, and after the procedure.
VUR was graded on the screening images by one of the principal investigators. A pediatric radiologist later reviewed all films. Interrater reliability testing was calculated by dividing the number of agreements with the total number of assessments.
Statistical Analysis
The GDRS scores and VUR grades are recorded on an ordinal scale. For
calculation of the sample size, the situation was simplified by using a
Mann-Whitney U test to assess the difference in outcome measures in
the two arms of the trial.
A clinically important difference in GDRS was considered a true shift of one grade down by two thirds of the subjects with the use of midazolam. The distribution of GDRS scores observed in patients undergoing VCU by Phillips et al. [1] was used as the expected distribution of GDRS scores. A sample size of 70 per group was required to have 80% power to detect a difference at the 5% level of significance.
A clinically important difference in VUR was considered a true shift of one grade down by half of the subjects with the use of rnidazolam. To determine distribution of VUR grades, we performed an audit of our local VUR. Our grading was similar to published distributions [14]. Because it is important to detect a difference if, in fact, one exists, a 90% power was used. There would be 90% power to detect a difference at the 5% level of significance with 60 per group if there was a real shift of 50% of the patients down one grade.
For the ordinal outcomes that were recorded at several different time periods and for VUR grading, which was recorded for each side, a nonlinear mixed-model analysis was performed. The ordinal outcome measures were used as repeated measures, and the treatment group was used as an explanatory variable. The possible confounders of age, sex, participation in play therapy, whether the caregiver with the child had been catheterized previously, whether the child had been catheterized previously, whether the child was toilet-trained, and the difficulty of the procedure were also included in the analysis. The interaction of age and time period with the group effect was also investigated.
For the continuous measures, which were recorded at the different time periods, a linear mixed model was used with the same explanatory variables as mentioned earlier. For the once-only outcomes, recorded multiple regression or ordinal logistic regression was used depending on the type of measure, including the same explanatory variables as described earlier.
Statistical analyses were performed using SAS software (version 8.0, SAS Institute).
Safety
Safety guidelines for sedation of a pediatric patient established by the
American Academy of Pediatricians
[15], the American Society of
Anesthesiologists [16], and
the Australian and New Zealand College of Anesthetists
[17] were followed. Children
were discharged 60 minutes after the procedure or if they were fully alert and
orientated 30 minutes after the procedure (60-90 minutes after ingestion).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Seventy-two children were randomized to the placebo group and 67, to the midazolam group. In the placebo group, 56 completed the VCU examination on the study day and complete behavior data were collected. Seven other patients went on to undergo micturating cystourethrography (MCU) within 3 months of VCU, and the VUR grades were collected on those individuals. In the midazolam group, 61 completed the VCU examination on the study day and one other patient completed VCU within 3 months of the study day. A total of 117 patients (Table 1) completed VCU on the study day and behavior data were recorded; VUR grades were recorded in a total of 125 cases.
|
|
|
Estimate of Treatment Effect Size
For descriptive purposes, peak distress at any stage of VCU was charted for
each group (Fig. 3). To
estimate treatment effect size, the number of children who experienced serious
or severe distress (i.e., GDRS score > 2) at any phase of the VCU study was
calculated. Overall 34 (61%) of 56 children in the placebo group experienced
distress scored on the GDRS as > 2 at any stage during the procedure. In
the midazolam group, 16 (26%) of 61 children distress scored on the GDRS as
> 2 at any stage during the procedure. Therefore, the number needed to
treat (NNT) to reduce either serious distress under control or severe distress
not under control (i.e., GDRS > 2) at any stage during the procedure in one
child was 2.9 (95% CI, 1.9-5.5).
Four children failed to complete the VCU examination on the study day due to panic. All were in the placebo group and were not included in the NNT calculation; inclusion of those cases would have increased the estimate of treatment effect size.
VUR
VUR was identified in 16% of all children. There appeared to be more cases
of reflux in the placebo group than in the treatment group (18.4% vs 13.6%,
respectively), but there was no effect of treatment group on reflux grading
(p = 0.31). There was no evidence of a difference in volume infused
between the two groups (p = 0.8). Interobserver reliability for VUR
grading was 97.2%.
Secondary Measures of Distress
For the single GDRS score produced by the nurse, there was strong evidence
of an effect of treatment (p < 0.0001), with the placebo group
patients being more distressed than the treatment group patients. There was
strong evidence of a difference in heart rate in the placebo and treatment
groups (p < 0.0001), with heart rate being higher in the placebo
group. The baseline heart rate (i.e., premedication) was controlled for in
this analysis. There was also evidence of a difference in the effect of
treatment at different phases (p = 0.02), with the difference at
recovery being less than at the other phases, but this difference was only in
magnitude rather than in direction.
Blinding Analysis
Before the procedure, 24 caregivers were asked to guess which group their
child was randomized to be in, 17 (71%) answered "Don't know." Of
the five who thought their child was in the midazolam group, one (20%) was
correct. Of the two who thought that their child was in the placebo group, one
was correct (50%).
After the procedure, of the 24 caregivers asked which group their child was in, six answered "Don't know." Of the 11 who thought their child had been given midazolam, eight were correct (73%). Of the five who thought their child was in the placebo group, four (80%) were correct. In total, 10 caregivers could not correctly identify which group their child had been in after the procedure (42%).
Timing of Events
The median time between the administration of medication and insertion of a
catheter was 25 minutes in the placebo group and 30 minutes in the midazolam
group; this difference was not significant. The median time between insertion
of the catheter and voiding was 14 minutes in the placebo group and 11 minutes
in the midazolam group. Because distribution was skewed, the log of the time
was used. There was evidence of a difference in time to voiding in the two
groups (p = 0.04), with the time to voiding for the treatment group
patients being shorter. There was no other significant time period difference
between the two groups.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Midazolam has been shown to be effective at inducing conscious sedation and reducing the distress patients feel about many procedures [5]. Midazolam can also cause amnesia, which may reduce the distress of repeatedly undergoing procedures such as VCU [21]. Elder and Longenecker [8] found that midazolam was effective in a highly selected group of children who had been previously distressed or were likely to be distressed by VCU. That study was not randomized or blinded. The authors concluded that midazolam had no appreciable effect on voiding dynamics.
Stokland et al. [22] performed a randomized controlled trial using intranasal midazolam, which was shown to reduce patient distress without affecting voiding dynamics. We routinely offered oral midazolam to all children presenting to our department for VCU. We chose oral administration for our study because we thought that oral midazolam would be tolerated better by children. Midazolam is a bitter medication. When given nasally, it is deposited to the back of the mouth and creates a bitter sensation. We were able to disguise the bitter taste of midazolam using a mango-and-orange flavor, and we do not think midazolam's bitter taste could have been disguised with a nasal preparation.
There was the potential for selection bias, but we asked all caregivers who declined to participate their reason for declining the study. Eight nonparticipants declined the study because they did not want sedation, and eight declined because they wanted sedation (Fig. 1). The largest group declined to undergo the VCU examination, but this number of refusals is thought to be similar to incidence before the study. We enrolled 76% of all the eligible children who went on to undergo VCU. This would reduce any potential selection bias. Four children in the placebo group could not complete the procedure due to panic, and this is consistent with previous experience. No child in the midazolam group failed to complete the examination because of distress.
The decision to fast only the treatment group made blinding difficult but made the results more clinically relevant. We compared our current VCU method with the proposed VCU method, which included a period of fasting recommended for safety [23]. From our blinding analysis, we noted that caregivers were unaware of the group to which their child was assigned. After VCU was performed, many caregivers had recognized the sedative effects of midazolam or lack thereof. Although our behavioral observers were unaware of allocation, they could have recognized the sedative effects and therefore biased reporting. The GDRS interobserver reliability was low, but more than 90% of our observations were from a single observer. The secondary outcome measures all suggest that there was a real reduction in distress, especially the objective heart rate measures.
There was a difference between the time of insertion of the catheter to voiding of urine between the two groups of 3 minutes. There was no difference in the volume infused, and this time difference was thought to be not clinically significant.
The filling and voiding phases of the VCU were the most distressing phases for children, which was contrary to our preconception that catheterization would have been the most distressing phase.
Our study results cannot be applied to children younger than 1 year old who made up 141 (33%) of the 429 requests for VCU. We may have been able to sedate these children in the presence of staff skilled in pediatric airway management.
We were fortunate to be able to offer play therapy, to have an experienced staff, and to work in a child-friendly environment. We were disappointed to find that 61% of the placebo group experienced serious or severe distress at some stage during the procedure. We achieved only a 30% use of play therapy, and this area is one in which we will look to improve. This study showed a number needed to treat to reduce serious or severe distress in one child was 2.9, which we consider to be clinically significant.
The practicalities of introducing any sedative procedure in the radiology department are considerable but are not insurmountable. Many departments already offer sedation for other procedures. The major requirements are skilled staff, assessment and recovery areas, and appropriate resuscitation and monitoring equipment as defined by sedation guidelines [15, 16, 23]. We created a dedicated recovery area in our department specifically for this study, and a sedation recovery area is included in future plans for our department. A VCU examination performed with patient sedation adds approximately 60 minutes to the entire procedure, mostly as recovery time.
VUR was identified in 16% of all children examined, and we found that the treatment group had no effect on reflux grading (p = 0.31). In addition, there was no evidence of a difference in volume infused between the two groups (p = 0.8). This finding is consistent with other previously reported indirect measures of bladder function during sedation [8, 22]. We conclude that oral midazolam does not affect the ability of VCU to allow diagnosis of VUR. The routine use of conscious sedation using oral midazolam (0.5 mg/kg) in pediatric patients older than 1 year reduces distress and does not alter the ability of VCU to diagnose VUR.
Acknowledgments
We thank Joanna Stewart for her valuable statistical advice. We especially
thank the children and their caregivers.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
