OBJECTIVE: To determine if pediatric procedural sedation-provider medical specialty affects major complication rates when sedation-providers are part of an organized sedation service.
METHODS: The 38 self-selected members of the Pediatric Sedation Research Consortium prospectively collected data under institutional review board approval. Demographic data, primary and coexisting illness, procedure, medications used, outcomes, airway interventions, provider specialty, and adverse events were reported on a self-audited, Web-based data collection tool. Major complications were defined as aspiration, death, cardiac arrest, unplanned hospital admission or level-of-care increase, or emergency anesthesia consultation. Event rates per 10 000 sedations, 95% confidence intervals, and odds ratios were calculated using anesthesiologists as the reference group and were then adjusted for age, emergency status, American Society of Anesthesiologists physical status > 2, nil per os for solids, propofol use, and clustering by site.
RESULTS: Between July 1, 2004, and December 31, 2008, 131 751 pediatric procedural sedation cases were recorded; there were 122 major complications and no deaths. Major complication rates and 95% confidence intervals per 10 000 sedations were as follows: anesthesiologists, 7.6 (4.6–12.8); emergency medicine, 7.8 (5.5–11.2); intensivist, 9.6 (7.3–12.6); pediatrician, 12.4 (6.9–20.4); and other, 10.2 (5.1–18.3). There was no statistical difference (P > .05) among provider's complication rates before or after adjustment for potential confounding variables.
CONCLUSIONS: In our sedation services consortium, pediatric procedural sedation performed outside the operating room is unlikely to yield serious adverse outcomes. Within this framework, no differences were evident in either the adjusted or unadjusted rates of major complications among different pediatric specialists.
WHAT'S KNOWN ON THIS SUBJECT:
Pediatric procedural sedation is provided by a variety of pediatric specialists. Low patient numbers, single-institution experience, and evaluation of single-specialty groups have limited previous reports on complications.
WHAT THIS STUDY ADDS:
We used a large multiinstitution procedural sedation database to compare the major complication rates among providers. We found that provider specialty did not affect major complication rates in our consortium of organized sedation systems.
Thousands of children receive sedation for diagnostic and therapeutic interventions annually, and this number is expected to increase.1 Children are at higher risk for sedation-related complications than adults.2 Several factors contribute to this increased risk, including the common need for a deeper level of sedation, the wider range of procedures for which a child will require sedation, and the physiologic and anatomic differences present in children that make them more vulnerable to respiratory depression and hypoxemia. Because of these risks, it is essential that personnel skilled in the administration of sedative agents and the management of sedation related complications attend to these patients. Multiple specialty groups, including the American Academy of Pediatrics,3 the American Society of Anesthesiologists (ASA),4 the American College of Emergency Physicians,5 and the American Academy of Pediatric Dentistry,6 have published guidelines on the provision of care during pediatric procedural sedation (PPS). The Joint Commission on Accreditation of Healthcare Organizations has incorporated these guidelines into their recommendations, including the need for a sedation practitioner to be able to rescue a patient from a level of sedation deeper than intended and the need for a consistent standard of care in all areas of the hospital where sedation is administered.7
Anesthesiologists have been important providers of PPS; however, the growth in demand has outpaced the availability of pediatric anesthesiologists in many areas of the country. As a result, during the past 2 decades, other pediatric specialists, including intensivists, emergency medicine physicians, radiologists, and hospitalists have increasingly provided PPS. An important question that arises from this shift in practice is whether the use of different pediatric specialists in the provision of PPS affects the associated complication rates. The experiences of those specialists have been reported in several studies, but low patient numbers, single-institution experience, and evaluation of single-specialty groups have limited them. Comparing these studies has also been difficult because of varying definitions of what constitutes a sedation related complication. In 2003, the Pediatric Sedation Research Consortium (PSRC) organized a collaborative group of institutions interested in prospectively collecting data on pediatric procedural sedation/anesthesia outside of the operating room to better understand this practice and its safety. The PSRC has recorded an extraordinary number of cases that reflect the experience of multiple pediatric specialists at multiple institutions. We sought to determine if the frequency of major complications during pediatric procedural sedation/anesthesia outside of the operating room differs when provided by different pediatric specialists as reflected by the PSRC database and the sedation systems used by participating PSRC members.
PATIENTS AND METHODS
The data collection methodology used by the PSRC has been described in 2 previous reports: the first on the initial 30 000 sedation cases8 and the second on the use of propofol in nearly 50 000 sedation cases by a variety of providers.9 Thirty-eight sites participated in the consortium database; the sites included large children's hospitals, children's hospitals within hospitals, and general/community hospitals. The sites were self-selected for involvement in the PSRC data-sharing group. Each site agreed to identify a primary investigator and to use a standardized methodology for data collection and quality oversight. The site could elect to include all PPS performed within the institution or to include PPS from a specific area only, such as magnetic resonance imaging. However, submission of >90% of PPS cases performed in the selected area was required to minimize selection bias. Each institution agreed to submit an independent count of PPS cases in their selected area and to perform audits of at least 5% of their PPS cases to ensure data integrity. Each site used its own institutional criteria to determine whether a patient was an appropriate candidate for PPS by the sedation service or whether elective referral to the anesthesia (or another) service was indicated. The institutional review board or equivalent at each participating institution approved this study. All participating sites are listed in the Appendix.
The consortium database uses a Web-based tool to collect information at the time of each sedation event. The Web-based tool consists of 25 “screens” that ask 1 question per screen with subsequent screens on the basis of the answer entered. Standard-answer sets facilitate clear coding and interpretation of responses as well as faster data entry. The tool includes computer code designed to validate data at the time of entry and branching logic to ensure that only relevant questions are asked in each survey. A Sybase database management system (Sybase, Dublin, CA) is used to manage the collected data. The study Web site and data-entry portal are secured using the Secure Sockets Layer, whereby each study participant is authenticated each time they use the system and is then only allowed to access the portion of the Web site relevant to his or her institution. All collected data met the Health Insurance Portability and Accountability Act requirements for deidentification.
The data tool includes 3 separate screens, or question sets, related to complications during sedation/anesthesia. Included in the list of possible complications are aspiration (defined as suspected or witnessed tracheal aspiration of gastric contents associated with a worsened respiratory status compared with presedation status), death, cardiac arrest, unplanned admission to the hospital (outpatient study requiring admission) or an increase in a patient's level-of-care (prolonged recovery/monitoring period, transfer to step-down unit or ICU), and the requirement for emergency anesthesia consultation (unplanned difficulty with airway management). Unplanned admission and level-of-care increase were joined into a combined variable for statistical analysis. We focused the study on these major complications because of their serious nature, the unambiguous nature of the events, and the likelihood of consistent identification of these events by multiple observers. Although multiple complications might occur during 1 patient encounter, each event was counted separately to maximize their capture. And, whereas a requirement for an increased level-of-care and emergency anesthesia consultation are not patient complications themselves, we believe that they are surrogates for the occurrence of significant physiologic alterations in patients.
All participating institutions and primary investigators were blinded to the data submitted from any individual institution other than their own. Study authors were also blinded to referring institutions. All primary investigators were charged both with selecting locations within their institutions where collection of sedation/anesthesia data were feasible and with assuring nonselective data submission through regular inventories of their data submissions. The data for this study were collected by the PSRC between July 1, 2004, and December 31, 2008.
Our current evaluation focused on a comparison of major complication frequency during procedural sedation/anesthesia performed by pediatric specialists outside of the operating room. Our null hypothesis was that the frequency of major complications among pediatric specialists, when operating within the confines of an organized sedation system, would be equivalent. We calculated event rates per 10 000 sedations for each type of provider. Provider groups were anesthesiologist (both pediatric and general), pediatric intensivist, pediatric emergency medicine, pediatrician, and other (radiologist, surgeon, dentist, pediatric resident or fellow, advanced practice nurse, certified registered nurse anesthetist, or registered nurse). We then calculated odds ratios (ORs) and 95% confidence intervals (CIs) using the anesthesiologist event rate as the reference group. Odds ratios and CI were further adjusted for age, emergency status, ASA physical status > 2, nil per os (NPO) for solids, propofol use, and clustering by site via multiple regression analysis because these factors have been suggested to have an effect on complication frequency during PPS. We recognize that observations within an institution may not be independent, so clustering adjustments over site were incorporated in our models using a sandwich estimator. Stata 11 (Stata Corp, College Station, TX) was used for the statistical analysis. P ≤ .05 was considered statistically significant.
Pediatric specialists submitted 133 941 procedural sedation records during the study period. Of those, 2190 cases were excluded because of lapsed institutional review board status or inadequate data audits. Thus 131 751 records were evaluated. No deaths were recorded, and other major complications occurred rarely. The demographic data of the analyzed records are listed in Table 1.
A variety of procedures were performed while procedural sedation was provided. The most common procedures were radiology (62.4%), followed by hematology/oncology (10.8%), minor surgical procedures (8.1%), endoscopy/colonoscopy (6.9%), and bronchoscopy/laryngoscopy (1.2%). Medications used individually or in combination for procedural sedation were varied; the most common medications employed were propofol (59.7% of cases), midazolam (25.5% of cases), pentobarbital (10.2% of cases), ketamine (9.8% of cases), and chloral hydrate (9.6% of cases).
The total number of cases analyzed, the number of major complications by provider type, the rate of major complications per 10 000 cases, 95% CI, and OR with 95% CI (unadjusted and adjusted) are listed in Table 2. There was no statistical difference between providers' major complication rates either before or after adjustment for possible confounding variables (age, emergency status, ASA > 2, NPO for solids, propofol use, and clustering by site). A graphical representation of the adjusted OR and resultant 95% CI are illustrated in Fig 1.
A breakdown of the type of major complication and the associated provider category is shown in Table 3. No statistical difference among providers was present.
Pediatric procedural sedation/anesthesia performed outside of the operating room is necessary to minimize pain and anxiety in children during procedures performed in hospitals and clinics. Requests for this service seem to be increasing, indicated by surveys of the members of the PSRC, currently at ∼10% per year. The supply of pediatric anesthesiologists in both developed and developing nations is inadequate to meet the demand of all pediatric sedation needs. As a result, pediatric specialists other than anesthesiologists increasingly staff sedation services and sedation units.10,–,12 It is essential that PPS be provided consistently and in the safest manner possible within and across institutions that care for children, even when a variety of pediatric specialists are involved in the provision of sedation.
To date, systematic evaluation of the risks associated with PPS have mostly been limited by single-institution experiences of a relatively small number of cases performed by single-provider types. Agrawal et al13 reported a complication rate prevalence of 6.7% in an urban emergency department procedural sedation case series of 1014 patients. Pitetti et al14 reported a 7.6% hospital-wide complication rate over a 3-year period in 14 386 pediatric procedural sedations using a protocol, with hypoxia representing 40% of the complications. In a meta-analysis of pediatric ketamine sedations performed by Green et al,15 an airway and respiratory complication rate of 3.9% was reported after pooling patient data from 32 studies. In none of these studies were any serious complications such as aspiration, unanticipated intubation, or death reported. They were, however, underpowered to detect these rare events. It is also uncertain whether certain events reported as adverse, such as minor oxygen desaturation, are of clinical importance.
The PSRC has collected data on >131 000 procedural sedations. They reported on the first 30 000 cases in 2006.8 In a second report by the PSRC, the safety of propofol used for PPS outside of the operating room was studied.9 In both of these reports the overall incidence and nature of adverse events during PPS were evaluated, but the effect of provider type was not examined. Shavit et al16 published a dual-site study comparing complication rates in 984 cases between pediatric residents and pediatric emergency physicians, and they did not find any difference in complication rates. No other publications have reported on complication rates on the basis of sedation provider specialty.
We used the PSRC database to better understand whether a difference in the frequency of major complication rates during PPS exists when different pediatric specialists provide sedation. As cited above, previous studies have defined “complications” in terms of various physiologic findings, such as minor oxygen desaturation events. These end points are not ideal because it is uncertain whether a brief period of oxygen saturation in the 89% range represents a clinically important event that would be a marker for unsafe clinical practice and increased likelihood of adverse outcomes. We sought to avoid these ambiguities and defined major complications as aspiration, death, cardiac arrest, unplanned admission to the hospital or an increase in a patient's level of care, or emergency anesthesia consultation. Using these definitions, no statistically significant difference in the rate of major complications among pediatric specialists was found. This is the first study, to our knowledge, that has evaluated the rates of major complications among pediatric specialists who provide sedation/anesthesia outside of the operating room using multi-institutional data with enough cases to power a meaningful evaluation of these rare complications.
A total of 122 major complications were reported. These events are expected to be rare; a critical incident analysis that evaluated cases over a 27-year period identified only 60 cases of death or severe neurologic injury attributable to sedation.2 In another study a pediatric perioperative cardiac arrest rate of 1.4 per 10 000 cases was reported.17 The 2 factors contributing to increased risk of arrest were an ASA physical status > 2 and age younger than 1 year. As a result, we adjusted our statistical analysis to consider the effect of age and ASA physical status. We also adjusted for NPO status for solids and whether the procedure was considered emergent because these factors may also influence the occurrence of major adverse events. Propofol use was also adjusted because emergency medicine physicians, intensivists, and anesthesiologists are more likely to use this agent than are pediatric hospitalists, advanced practice nurses, or radiologists. We clustered by site because we found that there was more variability between providers within a site than across specialties in general; however, we did not find that any of these factors influenced the provider specialty complication rate. We also examined the type of complications experienced by each provider type and did not find any statistical difference.
The strengths of our database are that information on more than 131 000 cases has been prospectively collected and audits have been performed to ensure the integrity of the data. The audits helped make certain that all sites have reported >90% of the cases from their selected area and have not excluded cases with unfavorable outcomes. An additional strength is that a variety of pediatric specialists in hospitals throughout the United States and Canada contributed to this database. This variety of participation enhances the ability to apply our results more globally to institutions with similar sedation systems as those used by PSRC members.
A limitation of an observational database used by a wide variety of providers (and applied to a variety of sedation systems) is the inability to ensure that definitions of adverse events have equal meaning between all providers. The major complications that we chose to evaluate, however, are unlikely to be defined differently by different providers. These complications are also clinically relevant, and the occurrence of each complication would have a significant impact on the care of the patient.
A second potential limitation to this study is that an analysis of the incidence of major complications in relationship to the level of sedation achieved by the patient was not performed. It is likely that the risk of complications increases as the level of sedation is deepened. Moderate sedation was not separated from deep sedation as potential risk factors in our major complications analysis. We did not record the level of sedation achieved by the patient in the database because such a determination requires patient stimulation. It was not practical to interrupt and potentially delay procedures that otherwise do not require stimulation. In addition, during the course of procedural sedation in the majority of children, deep sedation is likely achieved at least during part of the encounter. We also did not associate any particular drug with deep sedation versus moderate sedation because any sedative medication used in sufficient doses can induce deep sedation. We chose to evaluate the occurrence of major complications in global PPS practice as reflected in our database. This approach is congruent with ASA statements that emphasize the continuum of sedation and the need for practitioners to be able to recognize and potentially rescue patients from the effects of a deeper-than-intended level of sedation.18
A third limitation to our study is that a selection or exclusion bias could be introduced if institutions selected areas of their practice that were less likely (or more likely) to include patients with existing comorbidities that could contribute to the development of a major complication, if cases with unfavorable outcomes were not selected for data entry, or if the referral practice of the site did not refer all potential cases to the sedation service. The use of blinded data submission ensures anonymity of the submitting institution and allows more candid disclosure, which should help limit selection bias. In addition, the requirement for regular audits of submitted cases by the site principle investigator should further contribute to a reduction in selection bias. However, it is possible that in certain institutions patients are referred for care to pediatric specialists who are not part of the sedation service. The present study cannot completely resolve this issue, although the lack of difference in provider's major complication rates and type of complications after adjustment for ASA scores, age, NPO status, and emergency status argues against this being a major source of potential bias.
A final limitation to our data collection is the bias in site selection. Participating sites were highly motivated to set up systems with adequately trained staff and safeguards in place. Pronovost et al,19,20 in their work on catheter-associated bloodstream infections, have demonstrated that a culture of high patient-safety has a proportionate effect on outcomes. Hence, the favorable comparison of major complication rates between provider specialties may not apply in institutions with dissimilar sedation systems. However, our results do provide reassurance that when motivated providers are providing PPS in an established system with rigorous systemic safeguards, there will be an extremely low rate of major complications.
The rapid growth in the use of sedation services by nonanesthesiologists has been the subject of some concern.21 Our data reveal that, within our consortium, there was no increased danger associated with PPS provided by nonanesthesiologists. The application of our data to sites outside of our consortium will require rigorous evaluation of the skill level of the providers and the institution's systemic safeguards for the care of a sedated pediatric patient.
Avera Mckennan Hospital
Backus Children's Hospital
Cape Fear Valley Medical Center
Children's Healthcare of Atlanta (Egleston and Scottish Rite campuses)
Children's Hospital of Philadelphia
Children's Hospital and Medical Center, Omaha
Children's Hospitals and Clinics of Minnesota
Children's Memorial Hospital (Chicago, IL)
Children's Mercy Hospitals and Clinics, Kansas City
Children's Mercy Hospital, Emergency Department
Chris Evert Children's Hospital
Columbus Children's Hospital
Dartmouth-Hitchcock Medical Center
The Children's Hospital (Denver, CO)
Dr Alan R. Milnes
East Tennessee Children's Hospital
Eastern Maine Medical Center
Helen DeVos Children's Hospital
Jackson Memorial Hospital, University of Miami School of Medicine
Joe DiMaggio Children's Hospital
Kentucky Children's Hospital
Kosair Children's Hospital, University of Louisville
Le Bonheur Children's Medical Center
Medical University of South Carolina
Nemours/Alfred I. duPont Hospital for Children
New York University School of Medicine
Palmetto Health Children's Hospital
Rainbow Babies and Children's Hospital
The Children's Hospital at Providence
Tod Children's Hospital
UMass Memorial Medical Center
University of Florida
University of Virginia
Yale-New Haven Children's Hospital
This study was supported by the National Patient Safety Foundation.
We thank the Pediatric Sedation Research Consortium for contributions to the data collection for this project, and Dustin Samples for editorial assistance.
- Accepted January 24, 2011.
- Address correspondence to James H. Hertzog, MD, Department of Anesthesiology and Critical Care Medicine, Alfred I. duPont Hospital for Children, 1600 Rockland Road, Wilmington, DE 19803. E-mail:
Dr Couloures, Dr Hertzog, Dr Cravero, Dr Beach, and Dr Monroe participated extensively in the conception, design, analysis, interpretation of data, and preparation of the final manuscript, and approved the final version of the manuscript to be considered for publication.
This work was presented in part at the annual meeting of the Society of Critical Care Medicine's 39th Critical Care Congress; January 10, 2010; Miami, FL.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- ASA =
- American Society of Anesthesiologists •
- PPS =
- pediatric procedural sedation •
- PSRC =
- Pediatric Sedation Research Consortium •
- OR =
- odds ratio •
- CI =
- confidence interval •
- NPO =
- nil per os
- Coté CJ,
- Notterman DA,
- Karl HW,
- Weinberg JA,
- McCloskey C
- Coté CJ,
- Wilson S
- Mace SE,
- Brown LA,
- Francis L,
- et al
- 6.↵American Academy of Pediatric Dentistry. Guideline for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures. Available at: www.aapd.org/media/policies_guidelines/g_sedation.pdf. Accessed June 29, 2010
- 7.↵Joint Commission on Accreditation of Healthcare Organizations. Comprehensive Accreditation Manual for Hospitals (CAMH): The Official Handbook. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2006. PC41–PC43
- Cravero JP,
- Blike GT,
- Beach M,
- et al
- Cravero JP,
- Beach ML,
- Bilke GT,
- Gallagher SM,
- Hertzog JH
- Lowrie L,
- Weiss AH,
- Lacombe C
- Yamamoto LG
- Green SM,
- Roback MG,
- Krauss B,
- et al
- 18.↵American Society of Anesthesiologists. Statement on granting privileges for deep sedation to nonanesthesiologist sedation practitioners. Available at: www.asahq.org/for-healthcare-professionals/standards-guidelines-and-statements.aspx. Accessed December 13, 2010
- Copyright © 2011 by the American Academy of Pediatrics