Critical Pediatric Equipment Availability in Canadian Hospital Emergency Departments

PEDIATRICS/SURVEY ARTICLE Critical Pediatric Equipment Availability in Canadian Hospital Emergency Departments From the Departments of Pediatrics, Division of Emergency Medicine, * and Epidemiology and Biostatistics, Research Institute, Montreal Children s Hospital, McGill University, Montreal, Quebec, Canada; and the Department of Pediatrics, Division of Emergency Medicine, Alberta Children s Hospital, University of Calgary, Calgary, Alberta, Canada. Received for publication July 9, 1999. Revisions received February 9, 2000, August 18, 2000, and September 18, 2000. Accepted for publication October 4, 2000. Oral presentation at the Ambulatory Pediatric Association/Emergency Medicine meetings, New Orleans, LA, May 1998, and poster presentation at the Canadian Pediatric Association/ Emergency Medicine annual meeting, Hamilton, Ontario, Canada, June 1998. Supported by a Canadian Association of Emergency Medicine, Hoffmann-La Roche grant. Address for reprints: David McGillivray, MD, Montreal Children s Hospital, 2300 Tupper Street, Room A-103, Montreal, Quebec, H3H 1P3; E-mail mdmg@musica.mcgill.ca. Copyright 2001 by the American College of Emergency Physicians. 0196-0644/2001/$35.00 + 0 47/1/112253 doi:10.1067/mem.2001.112253 David McGillivray, MD * Cheri Nijssen-Jordan, MD Michael S. Kramer, MD * Hong Yang, MSc Robert Platt, PhD * See editorial, p. 388. Study objective: Of all child visits to emergency departments, 1% to 5% involve critically ill children who require cardiopulmonary resuscitation. Numerous versions of pediatric equipment lists for EDs have been published. Despite these efforts, many EDs remain unprepared for pediatric emergencies. The objectives of this study were to assess the availability of pediatric resuscitation equipment items in Canadian hospital EDs and to identify risk factors for the unavailability of these items. Methods: Using the updated database of the Canadian Association of Emergency Physicians (CAEP), a questionnaire survey was sent to 737 Canadian hospital EDs with a maximum of 3 mailings to nonresponders. On-site visits to a selected subset of hospital EDs were completed to validate the results obtained by the mailed questionnaire. Results: The response rate was 88.3% (650/737). Results showed the following overall equipment unavailability: intraosseous needle, 15.9%; pediatric drug dose guidelines, 6.6%; infant blood pressure cuff, 14.8%; pediatric defibrillator paddles, 10.5%; infant warming device, 59.4%; infant bag-valvemask device, 3.5%; infant laryngoscope blade, 3.5%; 3-mm endotracheal tube, 2.5%; and pediatric pulse oximeter, 18.0%. Low percentage of pediatric visits, lack of an on-call pediatrician for the ED, and lack of a pediatric advanced life support trained physician on staff were independently associated with equipment unavailability. Conclusion: This study demonstrated that essential pediatric resuscitation equipment is unavailable in a disturbingly high number of EDs across Canada and has identified several determinants of this unavailability. [McGillivray D, Nijssen-Jordan C, Kramer MS, Yang H, Platt R. Critical pediatric equipment availability in Canadian hospital emergency departments. Ann Emerg Med. April 2001;37:371-376.] APRIL 2001 37:4 ANNALS OF EMERGENCY MEDICINE 371

INTRODUCTION Approximately 30% of visits to general hospital emergency departments are by children, and 10% of all paramedic calls are for children. However, only 1% to 5% of all child visits to EDs involve critically ill children who require cardiopulmonary resuscitation. Because ambulances and other emergency transport vehicles often go to the nearest available ED for assistance, every ED is at risk for receiving critically ill children. Preparation for the arrival of critically ill children is essential. 1 Numerous versions of pediatric equipment lists for EDs have been published. 2-9 Despite these efforts, previous studies have documented regional problems with pediatric equipment unavailability, and the 1995 Institute of Medicine Report on emergency medical services for children found that many EDs were unprepared for pediatric emergencies. 10,11 In Canada, recommended guidelines for EDs are of a general nature and do not list significant specific pediatric resuscitation equipment items that should be available. 12 Our study had 2 objectives: first, to assess the availability of essential pediatric resuscitation equipment in Canadian hospital EDs, and second, to identify the determinants of equipment unavailability. To our knowledge, ours is the first large national study to examine these issues. MATERIALS AND METHODS The updated database of the Canadian Association of Emergency Physicians (CAEP) was used to identify all hospital EDs in the 12 provinces and territories of Canada. We developed a multiple-choice questionnaire to assess the availability of essential resuscitation equipment. The questionnaire defined available equipment as equipment that was present and readily available in the ED for use of a physician or nurse. If an equipment item was only available elsewhere in the hospital, it was considered unavailable for immediate resuscitation. Items were selected on the basis of previously published equipment lists, 2-4 with priority given to those affecting the possibility of a successful resuscitation outcome: intraosseous needles, infant-sized endotracheal tubes, pediatric-sized defibrillator paddles, pediatric drug dosing guidelines, pediatric pulse oximeter device, infant bag-valve-mask system, and infant warming device. The absence of these equipment items is most likely to seriously impair the ability of the physician to perform a successful resuscitation. The Committee on Pediatric Equipment and Supplies for Emergency Departments of the National Emergency Medical Services for Children Resource Alliance has listed all the equipment in this study as part of guidelines for minimum equipment and supplies for the care of pediatric patients in EDs. 7 To achieve the second objective of the study, we inquired about potential determinants of equipment unavailability. The potential determinants included annual ED patient census, percentage of pediatric visits, presence or absence of consultant pediatricians and/or physicians trained in pediatric advanced life support (PALS), distance from a university center, number of pediatric resuscitations (defined as cardiac arrest, status epilepticus, or respiratory failure) per year, and level of care given in the ED (as defined by the American Medical Association Commission on Emergency Medical Services and endorsed by the American Academy of Pediatrics). 13 A level I center has comprehensive pediatric resources (people, equipment, facilities) available for definitive pediatric care. Level II centers have extensive pediatric resources available but lack comprehensive subspecialty expertise, whereas level III centers are first-line care hospitals with minimal pediatric resources but have the ability to stabilize seriously ill and injured children before transport. The levels of care were defined on the questionnaire. These potential determinants were chosen a priori based on our hypothesis that insufficient patient load, complexity of patients seen, or training were most likely to adversely affect equipment unavailability. A maximum of 2 additional mailings were sent after 1 to 2 months to nonresponders. If the director of the ED did not respond to the questionnaire, one of the mailings was sent to the head nurse of the ED. All EDs were assured of the confidentiality of their individual responses. To assess the validity of the responses received, a sample of 38 hospitals were selected for an on-site visit to verify the equipment availability reported on the mailed questionnaire. The validation study occurred within 1 year of the completion of mailings. Seven emergency physicians from university centers in 5 of the 12 provinces in Canada participated in the validation study. The hospitals were selected according to the feasibility of the participating physician or his or her nurse representative or research assistant knowledgeable in emergency medicine to perform an on-site visit. All 3 categories of individuals were used in the site validation study. The site visitors used the same equipment list as the mailed questionnaire. They were asked to assess at least 5 hospitals within their geographic area. The selected site visit hospitals were contacted before the visit to obtain 372 ANNALS OF EMERGENCY MEDICINE 37:4 APRIL 2001

permission for the visit and arrange an appropriate time, but the specific reason for the visit was not disclosed to the hospital s ED personnel. An SAS personal computer statistical software program was used for all statistical analyses. 14 Hospital demographic information is presented in a descriptive fashion, as are crude rates of unavailability (calculated as a percentage of responding hospitals). Polychotomous determinants were dichotomized by combining categories with homogeneous effects on equipment unavailability, while ensuring adequate numbers in each category. Multiple logistic regression was used to assess the independent effects of the potential determinants. All potential determinants were included in the regression model to minimize confounding. Bias-corrected (ad- Table 1. Description of independent variables for responding hospitals. Potential DeterminantsNo. * % Hospital level 579 Level I 6.4 Level II 19.0 Level III 74.6 Community population 616 <20,000 54.2 20,000 <50,000 17.9 50,000 <100,000 9.6 100,000 18.3 Annual ED census 589 <10,000 43.5 10,000 30,000 28.5 30,000 50,000 21.6 50,000 6.5 Percentage of pediatric patients 541 <10 35.3 10 <20 31.1 20 <30 22.0 30 11.6 Pediatrician on call for ED 605 Yes 70.4 No 29.6 PALS-trained physician in ED 563 Yes 29.7 No 70.3 Distance from university center 619 <200 km 59.5 200 km 40.5 Pediatric resuscitations/y 615 <3 65.5 3 5 13.8 6 10 9.3 >10 11.4 * No.=total number of respondents to this question. justed) bootstrap confidence intervals (200 bootstrap repetitions) were computed for the logistic models for 3 of the equipment items (intraosseous needle, pulse oximeter, and pediatric defibrillation paddles). 15,16 For the other 3 items (drug dosing guidelines, endotracheal tube, and infant bag-valve-mask system), unavailability was too rare to compute a bootstrap confidence interval. Therefore, confidence intervals for these items were calculated by Wald s method. 17 The goodness of fit of these 6 models was examined by the Hosmer-Lemeshow test. 18 Because this study did not involve human subjects, it was given an exemption by the institutional review board. RESULTS After excluding the hospitals or EDs that had closed since the development of the CAEP registry (N=65), 650 (88.3%) of 737 remaining hospital EDs returned the questionnaire. The lowest response rate from any of the 12 provinces or territories was 71.4% and the highest was 100%. The 2 lowest responding provinces (one 71.4% and the other 73.7%) have 7 and 95 EDs per province, respectively. The remaining 10 provinces all had response rates greater than 82.5%. Descriptive data on the responding hospitals as a group are shown in Table 1. The results of the validation study are shown in Table 2. Percentage agreement was generally high. Equipment unavailability at the time of the site visit was higher for all equipment items except for pediatric defibrillation paddles. Crude rates of equipment unavailability are shown in Table 3 for the overall study sample and according to the potential determinants of equipment unavailability. The Table 2. Results of validation study. Mailed Site Survey % Visit % Agreement Variable Unavailability Unavailability (%) Intraosseous needle 24.3 37.8 86 Drug dosing guidelines 7.9 13.2 89 Infant blood pressure cuff 10.8 16.2 73 Pediatric defibrillation paddles 18.9 5.4 81 Warming device 64.9 81.1 68 Infant bag-valve-mask system 2.7 5.4 92 Infant laryngoscope blade 5.4 8.1 86 3-mm endotracheal tube 5.3 13.2 82 Pediatric pulse oximeter 24.2 24.2 82 APRIL 2001 37:4 ANNALS OF EMERGENCY MEDICINE 373

intraosseous needle was the equipment item most frequently unavailable. However, deficiencies were noted for all types of equipment, even for drug dosing guidelines. Results of the logistic models and bootstrap 95% confidence intervals for selected essential equipment unavailability and potential determinants are shown in Table 4. The Hosmer-Lemeshow goodness-of-fit test of the 6 models showed that none was significant (P>.1), indicating an excellent fit for all models. DISCUSSION An ED s ability to handle a pediatric emergency requires 3 levels of preparation: equipment availability, knowledge of when and how to use that equipment, and finally, the psychomotor skills necessary to resuscitate a critically ill child. This study addresses only the first level of preparation. Our data indicate that a substantial number of EDs across Canada have major deficiencies in the availability of essential pediatric resuscitation equipment. Unavailability of equipment could significantly affect the outcome of resuscitation efforts. The conclusion of a recent inquest in Ontario noted that the staff members in charge were unaware of the availability of intraosseous needles, even though they were on the resuscitation cart, that the defibrillator used to reverse the cardiac arrest of the child had a minimum energy capacity that was 6 times the recommended dose for the child, that inappropriate adult-sized defibrillator paddles were used in the resuscitation, and that an inappropriate dose of lidocaine was administered. 19 Our study indicates that deficiencies in knowledge, equipment availability, and pediatric drug dose charts remain despite the availability of numerous publications outlining the required pediatric resuscitation equipment for EDs. 2-9 Despite the excellent response rate to our questionnaire, concern remains that nonresponding hospitals may have influenced the results. To assess the potential impact of the nonresponding group, we examined their effect on Table 3. Crude rates (percentages) of pediatric emergency equipment unavailability. Equipment Unavailability Drug Infant Blood Pediatric Infant Infant 3-mm Pediatric Intraosseous Dosing Pressure Defibrillator Warming Bag-Valve- Laryngoscope Endotracheal Pulse Needle GuidelinesCuff PaddlesDevice Mask Blade Tube Oximeter Potential Determinants (No.) * (627) (637) (636) (637) (628) (636) (635) (642) (628) Overall group 15.9 6.6 14.8 10.5 59.4 3.5 3.5 2.5 18.0 Hospital level Level III 18.8 7.5 17.4 12.3 60.3 4.0 4.0 2.6 19.5 Level I/II 10.1 4.7 9.5 7.0 57.5 2.4 2.4 2.3 15.1 Annual ED census <10,000 visits 30.5 12.5 22.9 17.9 54.9 6.1 6.9 3.6 25.4 10,000 visits 6.8 2.8 9.6 5.7 62.3 1.8 1.3 1.8 13.3 Percentage of pediatric visits <10% of total visits 32.4 12.2 16.9 18.5 63.0 7.1 6.0 1.8 27.0 10% of total visits 9.4 4.4 13.9 7.3 57.9 2.0 2.4 4.3 14.4 Pediatrician on call for ED No 21.3 7.7 17.5 12.8 58.9 4.8 4.8 3.1 19.2 Yes 2.9 2.3 6.9 3.5 59.2 0.6 0 0.6 11.6 PALS-trained physician on staff No 32.3 14.8 24.1 19.9 57.4 6.1 6.8 5.4 27.2 Yes 10.5 3.8 11.6 7.2 60.0 2.6 2.3 1.5 14.8 Distance from university center >200 km 21.8 8.6 21.1 12.2 64.2 3.7 4.9 3.2 21.4 200 km 12.4 5.0 10.5 9.4 56.0 3.6 2.5 1.9 15.5 Pediatric resuscitations/y <3 30.0 12.8 25.5 17.5 63.6 6.4 6.9 3.9 29.0 3 8.8 2.5 8.2 6.7 56.6 1.8 1.2 1.2 11.8 * No.=the number of hospitals responding to this part of questionnaire. 374 ANNALS OF EMERGENCY MEDICINE 37:4 APRIL 2001

the availability of an intraosseous needle and drug dose guidelines. If all nonresponders had had both of these items available, the unavailability of the intraosseous needle and the drug dose guidelines would have decreased only slightly: from 15.9% to 14.0% and 6.6% to 6.0%, respectively. If none of the nonresponders had had either of these items available, the unavailability would have increased substantially, from 15.9% to 26.2% for the intraosseous needle and 6.6% to 17.8% for drug dose guidelines. It is likely that equipment availability among nonresponders would be similar to or worse than that among responders. This conclusion is supported by the results of our validation study. Discordance between the results of the mailed questionnaire and the site visit are consistent with potential problems of finding the equipment needed at the time of an acute resuscitation. Such disagreement may indicate that certain individuals do not know where or what equipment is available, or that equipment is not being checked or replaced on a consistent basis. Thus, true unavailability is likely to be more prevalent than suggested by our results. Given the poor response to publicized lists of necessary resuscitation equipment, other solutions to the problem should be sought to ensure that pediatric equipment is available. The most efficacious solution would be a requirement by hospital-accrediting organizations that pediatric resuscitation equipment be available in all hospital EDs. Given that most hospitals already have a defibrillator and pulse oximeter but merely require the pediatric adapters, the estimated cost of equipping an ED for most of the items listed in this study is very low. Most of the equipment can be purchased for less than $1,000 (Canadian funds). 20 Our results suggest that on-call pediatricians, with or without at least one PALS-trained physician on staff in the ED, have beneficial impacts on equipment availability. In addition to the effect on the availability of equipment, PALS training should also have an impact on knowledge and psychomotor skills necessary for use of the equipment. The specialty of emergency medicine in Canada consists of either a 5-year Royal College Emergency Medicine program or a 3-year combined program of family medicine (2 years) and emergency medicine (1 year). These specialty programs are relatively new in Canada. Our questionnaire inquired about the type of postgraduate training in emergency medicine by the ED staff. The numbers were too small to draw any conclusions about the effect of these programs on equipment availability because of the concentration of these physicians in the larger university centers. Most emergency medicine is practiced by physicians who obtained their experience while in practice or during their family medicine training. Although it may be preferable to have emergency medicine physicians who have undergone extensive subspecialty training in emergency medicine to ensure that all standards are met, this is not a realistic short-term objective. The number of training positions is currently insufficient to guarantee the presence of emergency medicine trained physicians in all EDs across Canada. At present, the PALS course, like the adult advanced cardiac life support course, is the most realistic way to influence equipment unavailability in the near future. Table 4. Adjusted odds ratios (and 95% CIs) for selected pediatric resuscitation equipment unavailability. Equipment Items Pediatric 3-mm Infant Intraosseous Drug Dosing Pulse Defibrillation Endotracheal Bag-Valve-Mask Potential DeterminantsNeedle * Guide Oximeter * Paddles * Tube System Hospital level III 1.38 (0.57, 6.40) 3.27 (0.38, 28.22) 0.90 (0.46, 1.81) 1.72 (0.49, 11.50) NA 1.50 (0.16, 13.71) Annual census <10,000 2.00 (0.84, 4.47) 1.90 (0.57, 6.30) 1.22 (0.62, 2.64) 1.87 (0.54, 5.70) 0.34 (0.06, 1.93) 2.62 (0.58, 11.74) <10% of pediatric visits 2.47 (1.36, 5.35) 1.08 (0.43, 2.70) 1.61 (0.95, 3.04) 1.75 (0.79, 4.25) 3.98 (0.62, 25.56) 1.44 (0.45, 4.57) No pediatrician on call for ED3.12 (1.12, 11.34) 1.20 (0.31, 4.63) 1.05 (0.49, 1.97) 1.62 (0.57, 6.31) 2.58 (0.28, 23.80) 3.47 (0.40, 30.14) No PALS-trained physician on staff 2.18 (1.33, 3.50) 3.76 (1.45, 9.78) 1.85 (1.03, 3.27) 1.84 (0.88, 3.91) 4.66 (1.03, 21.00) 2.41 (0.75, 7.77) Distance >200 km from university center 1.66 (1.05, 3.24) 1.37 (0.62, 3.03) 1.26 (0.75, 1.91) 0.85 (0.46, 1.74) 1.53 (0.43, 5.46) 0.75 (0.28, 1.20) <3 pediatric resuscitations/y 1.61 (0.87, 3.08) 2.44 (0.98, 6.10) 1.85 (1.07, 3.27) 1.42 (0.69, 3.15) 1.44 (0.34, 6.16) 1.41 (0.48, 4.15) NA, Not available. * CIs calculated by the bootstrap method; other CIs calculated by Wald s method. The logistic regression model for the 3-mm endotracheal tube excludes hospital level, as the model including this potential determinant did not converge. APRIL 2001 37:4 ANNALS OF EMERGENCY MEDICINE 375

In summary, this study has demonstrated that essential pediatric resuscitation equipment is unavailable in a disturbingly high number of EDs across Canada. Anecdotal reports are available to suggest that these deficiencies have played a role in difficult pediatric resuscitations. 19 A systematic review of past coroners inquests across Canada would help to confirm the adverse consequences of unavailable emergency equipment. In the meantime, physicians and hospital administrators should ensure that basic pediatric resuscitation equipment is available in all EDs. 19. Wilson T. Inquest into the death of Jad El-Khatib deceased March 19, 1995. File number 14850. Province of Ontario: Office of the Coroner; April 1996. 20. Emergency care providers lack right equipment for kids. Health Technology Trends. 1993;5(9):4-5. REFERENCES 1. Seidel J, Henderson D. Approach to the pediatric patient in the emergency department. In: Barkin R, ed. Pediatric Emergency Medicine: Concepts and Clinical Practice. 2nd ed. Philadelphia, PA: Year Book; 1997;1:13-14. 2. Strange G, ed. APLS, The Pediatric Emergency Medicine Course. 3rd ed. Elk Grove Village, IL: American Academy of Pediatrics and the American College of Emergency Physicians; 1998:231-232. 3. Committee on Pediatric Emergency Medicine, American Academy of Pediatrics. Guidelines for pediatric emergency care facilities. Pediatrics. 1995;96:526-537. 4. Seidel J, Tittle S, Hodge D, et al. Guidelines for pediatric equipment and supplies for emergency departments. Committee on Pediatric Equipment and Supplies for Emergency Departments. National Emergency Medical Services for Children Resource Alliances. J Emerg Nurs. 1998;24:45-48. 5. Durch J, Lohr K, eds. Emergency Medical Services for Children. Committee on Pediatric Emergency Medical Services. Washington, DC: Institute of Medicine Report. National Academy Press; 1993. 6. Seidel J. Emergency medical services and the pediatric patient: are the needs being met? II. Training and equipping emergency medical services providers for pediatric emergencies. Pediatrics. 1986;78:808-812. 7. Committee on Pediatric Equipment and Supplies for Emergency Departments, National Emergency Medical Services for Children Resource Alliance. Guidelines for pediatric equipment and supplies for emergency departments. Ann Emerg Med. 1998;31:54-57. 8. Seidel J. Preparing for pediatric emergencies. Pediatr Rev. 1995;16:466-472. 9. American College of Emergency Physicians. Pediatric emergency guidelines. Ann Emerg Med. 1995;25:307-309. 10. Zaritsky A, French J, Schafermeyer R, et al. A statewide evaluation of pediatric prehospital and hospital emergency services. Arch Pediatr Adolesc Med. 1994;148:76-81. 11. Knapp J, ed. A call to action: The Institute of Medicine report on emergency medical services for children. Pediatrics. 1995;96(1 Pt 2):173-210. 12. Emergency Units in Hospital: Guidelines. Ottawa, Ontario, Canada: Health and Welfare Canada; 1998. 13. American Medical Association Commission on Emergency Medical Services. Pediatric emergencies. An excerpt from Guidelines for the Categorization of Hospital Emergency Capabilities. Pediatrics. 1990;85:879-887. 14. Language Guide for Personal Computers. 6th ed. Cary, NC: SAS Institute Inc; 1985. 15. Efron B, Gong G. A leisurely look at the bootstrap, the jackknife and cross-validation. Am Stat. 1983;37:36-48. 16. Efron B, Tibshirani R. Statistical data analysis in the computer age. Science. 1991;253:390-395. 17. Kleinbaum D. Logistic Regression: A Self-Learning Text. Statistics in Health Sciences. in: Dietz K, Gail M, Krickeberg K, et al, eds. New York, NY: Springer; 1994. 18. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 1989:135-145. 376 ANNALS OF EMERGENCY MEDICINE 37:4 APRIL 2001