Pediatric Anesthesia ISSN 1155-5645
ORIGINAL ARTICLE
Procedural sedation for MRI in children with ADHD Eimear Kitt1,2, Jennifer Friderici2,3, Reva Kleppel2,3 & Michael Canarie2,4,5 1 2 3 4 5
Department of Medicine-Pediatrics, Baystate Medical Center, Springfield, MA, USA Department of Medicine and Department of Pediatrics, Tufts University School of Medicine, Boston, MA, USA Department of Medicine, Baystate Medical Center, Springfield, MA, USA Department of Pediatric Critical Care, Baystate Medical Center, Springfield, MA, USA Department of Pediatric Critical Care, Yale University School of Medicine, New Haven, CT, USA
What is already known
• Despite ADHD being the most common neurobehavioral disorder in children, little is known about the
necessary sedative doses for procedural sedation, or about the potential for interactions with prescribed psychostimulants.
What this article adds
• We examined the association between ADHD and effective weight-based propofol dose in children aged
5–12 years and found that children with ADHD do not require higher doses to achieve a successful MRI.
Keywords Attention-deficit hyperactivity disorder; humans; propofol; midazolam; hypnotics and sedatives; child Correspondence Dr. E. Kitt, Department of MedicinePediatrics, Baystate Medical Center, 759 Chestnut Street, S2575 Springfield, MA 01199, USA Email:
[email protected] Section Editor: Joseph Cravero Accepted 30 May 2015 doi:10.1111/pan.12721
1026
Summary Background: Attention-deficit hyperactivity disorder (ADHD) is the most common neurobehavioral disorder of childhood, affecting 5–8% of children. It has been observed that these children have poor sedation experiences; however, to date there is minimal research on procedural sedation in this population. Aim: To examine whether children with ADHD required larger doses of propofol for magnetic resonance imaging (MRI) sedation. Methods: The hospital’s administrative billing database was used to identify all billing codes for MRI brain scans (with and without contrast) in children aged between 5 and 12 years over the preceding 5.5 years. The hospital’s electronic medical record database provided baseline demographics. The sedation record was reviewed for propofol dose, psychostimulant use, and prescribed dose. All children received a standard weight-based dose of midazolam prior to receiving the necessary amount of propofol. Primary outcome was the dose of propofol administered (mgkg 1) to achieve adequate sedation. Results: A total of 258 procedures met the inclusion criteria. The sample was 52% male, 74% White, 7.8% Black, 7.8% Hispanic, 4.3% Asian, and 6.2% other. ADHD was documented for 49 procedures with a prevalence of 18.5%. Patients with ADHD were older, more likely to be male, Hispanic, or to report race as ‘Refused/Unknown’. Indications for MRI for patients with ADHD varied significantly, with ‘Behavioral’ and ‘Neurocutaneous’ being significantly overrepresented in the ADHD group. The average sedative dose for all patients was 2.8 mgkg 1 (95% CI 2.62–2.94). Sedative dose was similar among children with and without ADHD diagnosis. Conclusions: Our study illustrates that children with ADHD do not have higher sedative requirements to achieve a successful brain MRI.
© 2015 John Wiley & Sons Ltd Pediatric Anesthesia 25 (2015) 1026–1032
E. Kitt et al.
Introduction The need for pediatric sedation outside of the operating room setting has grown substantially over the past decade (1). The expanding use of magnetic resonance imaging (MRI), the evolution of interventional radiology procedures, and an increasing sensitivity to pain and anxiety in children have contributed to this trend. A variety of agents are typically employed to achieve the goals of pediatric sedation, namely to reduce anxiety, minimize discomfort, and maximize amnesia. Among those are midazolam and propofol, which are known to work synergistically to achieve the deep sedation that is often required for imaging and procedures (1–3). Propofol is an ideal agent with a rapid onset and short duration of action, although it has been described as having a wide therapeutic range with some groups of children requiring significantly different dosages to others. In addition, it has been associated with a higher rate of side effects including hypotension, respiratory depression, and apnea. The majority of adverse events do not require intervention, but do appear to be dose related. Therefore, providers should aim to administer the minimum dose of propofol necessary to achieve adequate sedation (3–5). Attention-deficit hyperactivity disorder (ADHD) is currently the most common pediatric neurobehavioral disorder diagnosed, affecting 5–8% of school-aged children (6,7). As part of a spectrum of neurobehavioral disorders, children with ADHD are, by definition, hyperactive and/or inattentive. Their behavior has been associated with a dysfunctional dopamine system where altered reinforcement and extinction processes lead to continued symptoms of hyperactivity, inattention, and impulsivity (8). As a result, this population has anecdotally been associated with increased sedative requirements. Also contributing may be that this population frequently have a coexisting behavioral disorder, with as many as 25% diagnosed with an anxiety disorder, and 50% reported to be diagnosed with oppositional defiant disorder or conduct disorder (7,9). One prior prospective study investigating the perioperative behaviors of children with ADHD illustrated that they were significantly less cooperative at induction of anesthesia compared with controls (10). Children with ADHD may also react differently to procedural sedating agents leading to potential increased requirements in dosing, be that either related to an inherent alteration in brain function, concurrent behavioral disorder, or potential drug interaction with a prescribed psychostimulant. Psychostimulant therapy is the mainstay of the management of ADHD, with benefits reported in numerous © 2015 John Wiley & Sons Ltd Pediatric Anesthesia 25 (2015) 1026–1032
Procedural sedation in ADHD
randomized controlled trials (5). Despite the large volume of diagnosed school-aged children currently prescribed such medications, little is known about potential drug interaction with procedural sedating medications such as propofol. In fact, minimal research has been carried out on procedural sedation in children with ADHD overall. Available data have come both from Emergency Department literature focusing primarily on sedation with combinations of midazolam with fentanyl or ketamine (6). Sedation for CT scanning in children with ADHD has also been investigated, with one study comparing midazolam and trazodone for efficacy in this population. In this study, trazodone appeared to be promising as a potential sedative with one-third of the patients failing to achieve the adequate sedation with midazolam (11). In our procedure unit at Baystate Children’s Hospital, the procedural sedation is performed by Pediatric Intensivists who utilize propofol as the primary sedating medication in children undergoing MRI, after a standardized dose of midazolam (0.05 mgkg 1) is administered. The primary aim of this study was to determine whether children with ADHD require larger doses of propofol for MRI sedation. The secondary aim was to determine the effect of psychostimulant medication prescribed for ADHD on this dose. We hypothesized that patients with ADHD would require larger propofol doses. As the secondary aim was descriptive in nature, no hypotheses were tested. Methods The hospital’s administrative billing database was used to identify all billing codes for MRI brain scans (with and without contrast) performed in children aged between 5 and 12 years. From this, 388 subjects were identified. Of this group, we included only those sedated using the standard protocol; propofol and midazolam. Patients who received other agents were excluded from the study. The hospital’s electronic medical record database was then queried to extract demographic, medical, and procedural data. The sedation record was examined and data concerning drug doses and successful sedation was extracted securely using our devised online abstraction form. This form was used to document ADHD diagnosis, weight, propofol dose in mg per kg and indication for MRI. If the patient was on a psychostimulant, it was noted with the dose prescribed at time of MRI. To confirm the diagnosis of ADHD, several methods were employed. Firstly, only children who had a primary care provider and electronic medical record within Baystate Medical Practice (BMP) were included. This ensured all provider notes and listed diagnoses were 1027
E. Kitt et al.
Procedural sedation in ADHD
available for review. The diagnosis was also confirmed using either an ICD-9 diagnosis or billing code, psychostimulant prescription at time of MRI, or provider notes. Only those patients with a confirmed diagnosis of ADHD in the medical records of this academic pediatric medical center were included. Unconfirmed diagnoses, even in the setting of parental report or psychostimulant use, were not included. The sedation records were reviewed for total propofol dose administered in mgkg 1 and complications. Our MRI pediatric sedation protocol remained the same over the time period reviewed: After placement of IV access, midazolam (0.05 mgkg 1) was administered approximately 15 min prior to the study, then propofol was administered as an intravenous bolus just prior to MRI. The initial propofol bolus was calculated at 1–2 mgkg 1, with bolusing repeated as necessary to achieve and maintain deep sedation (defined as a druginduced depression of consciousness during which patients cannot be easily aroused but respond purposefully after repeated verbal or pain) (1). At the time of this study, our institution used propofol and midazolam rather than a continuous propofol infusion as we have found it to be safe and effective for the shorter (20– 30 min) brain MRI sedations. Children requiring longer sedations were referred to anesthesiology. Subjects underwent cardiorespiratory and endtidal CO2 monitoring throughout the study and during recovery. Subjects were also continuously monitored by the pediatric intensive care unit (PICU) physician, registered nurse, and MRI technologist for the duration of the procedure. The propofol was rebolused if the child displayed movements or changes in vital signs (increased HR, BP, or respiratory rate) consistent with inadequate sedation. Each MRI study was reviewed by a technologist and/or radiologist at the completion of the study and adequate deep sedation was assumed if the MRI was successfully performed. If a study was terminated or unsuccessful due to inadequate sedation or other clinical concerns, the subject was excluded. This included only two subjects, one who had a seizure prior to the sedation and one who had multiple procedures performed on the same day. Our primary exposure was documentation consistent with the ADHD diagnosis. Our primary outcome was the amount of propofol administered in mgkg 1 necessary to achieve deep sedation. Statistical analysis Descriptive statistics (means/SD; N/proportions) were used to characterize the sample. Boxplots and histograms were used to examine the distribution of the principal outcome (sedative dose in mg per kg). After 1028
confirming the assumptions of linearity, the outcome and possible covariates were compared between exposure groups (ADHD yes vs no) using Fisher’s exact test (categorical) or one-way ANOVA (Gaussian), with Sidak’s adjustment for multiple comparisons. Multivariable linear regression was used to predict sedative dose (mgkg 1) as a function of ADHD diagnosis, adjusting for covariates. Covariates were defined as baseline characteristics associated in univariable analyses with both the exposure (ADHD) and outcome (sedative dose) at P ≤ 0.2. Covariates were tested and maintained in the model if their removal changed ADHD’s coefficient by ≥10%. The study’s full sample size provided 83% power to detect, as statistically significant (two-sided P-value ≤0.05), a standardized difference in means of 0.5; considered medium in magnitude. To avoid Type II error, subgroup analyses are descriptive only, with effect sizes estimated as g2 which describes the proportion of variability in sedative dose explained by a given predictor. STATA 13.1 for WINDOWS (©2014, StataCorp LP, College Station, TX, USA) was used for all analyses (12,13). Results A total of 257 procedures met our inclusion criteria. The mean/SD age in years was 7.3/1.8 (range of 5.0–12.7). The sample was 52% male (n = 134), 74% White (n = 190), 7.8% Black (n = 20), 7.8% Hispanic (n = 20), 4.3% Asian (n = 11), and 6.2% Other/Refused (n = 16) (See Table 1). ADHD was documented in a total of 48 patients (18.7% 95% CI 13.9%, 23.5%). Patients with ADHD were older, more likely to be male, Hispanic, or of ‘Refused/Unknown’ ethnicity. Indications for MRI varied significantly in children with ADHD, with ‘Behavioral’ and ‘Neurocutaneous’ being significantly more common in the ADHD group. The average sedative dose for all patients was 2.8 mgkg 1 (95% CI 2.6, 2.9) (see Table 2). The difference in sedative dose between the subjects with and without ADHD diagnosis was negligible and was not statistically significant in univariable and/or multivariable analyses (see Table 3). In the multivariable model, the covariates ‘race’ and ‘indication’ maintained the highest explanatory value (g2 = 0.08 for each, vs 0.00 ADHD) (Data not shown). Methylphenidate was the most commonly prescribed psychostimulant in the ADHD group with 12 (25%) children taking this medication at the time of MRI. Of the remainder, 10 (20.8%) were prescribed dextroamphetamine, 3 (6.3%) were prescribed dexmethylphenidate, 3 (6.3%) were prescribed clonidine, 2 (4.2%) were prescribed atomoxetine, 6 (12.5%) were prescribed multiple psychostimulants, and 12 (25.0%) were taking no © 2015 John Wiley & Sons Ltd Pediatric Anesthesia 25 (2015) 1026–1032
E. Kitt et al.
Procedural sedation in ADHD
Table 2 Mean sedative dose (mgkg 1) by patient characteristic
Table 1 Patient characteristics, by ADHD diagnosis ADHD diagnosis All subjects n = 257 n (%) Age in years 5 to
