- Email: [email protected]
Parenteral analgesic and sedative use among ED patients in the United States: Combined results from the National Hospital Ambulatory Medical Care Survey (NHAMCS) 1992-1997
Parenteral Analgesic and Sedative Use Among ED Patients in the United States: Combined Results From the National Hospital Ambulatory Medical Care Survey (NHAMCS) 1992-1997 MARK A. HOSTETLER, MD, MPH,* PEGGY AUINGER, MS‡ AND PETER G. SZILAGYI, MD, MPH† The objective of the study was to describe parenteral analgesic and sedative (PAS) use among patients treated in US emergency departments (EDs). Data representing 6 consecutive years (1992-1997) from the National Hospital Ambulatory Medical Care Survey (NHAMCS) were combined and analyzed. Patients were identified as having received PAS if they received fentanyl, ketamine, meperidine, methohexital, midazolam, morphine, nitrous oxide, or propofol. Patients were stratified according to age (pediatric <18 years), race, gender, insurance, type of hospital, urgency of visit, and ICD-9 diagnostic codes. Logistic regression was performed to determine independent associations and calculate odds ratios (OR) for receiving analgesia or sedation. A total of 43,725 pediatric and 114,207 adult ED encounters were analyzed and represented a weighted sample of 555.3 million ED visits. For patients with orthopedic fractures, African American children covered by Medicaid insurance were the least likely to receive PAS (OR 0.2, 95% confidence interval 0.1-0.6). These results suggest that variations may be occurring among ED patients receiving PAS. (Am J Emerg Med 2002;20:83-87. Copyright 2002, Elsevier Science (USA). All rights reserved.)
The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) requires that all patients receive an assessment for, and appropriate management of, pain.1 Sedation provided by nonanesthesiologists in the emergency department (ED) is both safe and effective.2-4 Detailed guidelines and recommendations have been developed and instituted by a wide variety of governmental and professional organizations.5-9 The provision of safe and adequate analgesia and sedation in the ED has now become the standard of care for both children and adults.10-13
Several studies have nonetheless suggested that ED patients on the whole receive inadequate amounts of analgesia or sedation.14-16 In addition, there may also be evidence that certain groups of patients receive substantially less analgesia or sedation than others.17-23 These studies have at times been contradictory, but have continued to suggest differential proportions of patients in the ED receiving analgesia or sedation based on age, race, gender, or ethnicity.18-26 The pediatric age group may be particularly at risk for inadequate analgesia or sedation.14,16,27-31 Although some of these studies have documented broad global variation based on physician training or preference,27-29,32 others have shown that when comparing similar illness and injury, children appear to receive substantially less analgesia or sedation than adults.17,33,34 We sought to explore whether significant variations occurred among the proportions of patients receiving parenteral analgesia and sedation (PAS). Rather than using data from a single institution, however, we chose to combine multiple years worth of data from a large, nationally representative database, the National Hospital Ambulatory Medical Care Survey (NHAMCS). By combining multiple years’ worth of data we were able to create a sample large enough to allow for detailed subgroup analyses using multivariate logistic regression techniques. METHODS Data Source
From the *Department of Pediatrics, Division of Emergency Medicine, St. Louis Children’s Hospital, Washington University School of Medicine, St. Louis, MO; the †Department of Pediatrics, Rochester General Hospital, Rochester, NY and the ‡Departments of Emergency Medicine and Pediatrics*, University of Rochester School of Medicine and Dentistry, Rochester, NY. Presented at the Society for Academic Emergency Medicine Annual Conference, San Francisco, CA, May 2000 and the New York State Public Health Association’s 50th Annual Meeting, Albany, NY, May 2000. Manuscript received August 29, 2001, accepted September 20, 2001. Address reprint requests to Mark A. Hostetler, MD, MPH, St. Louis Children’s Hospital, Washington University School of Medicine, One Children’s Place, St Louis, MO 63110. E-mail: mark_ [email protected] Key Words: analgesia, sedation, emergency, fracture, race, insurance Copyright 2002, Elsevier Science (USA). All rights reserved. 0735-6757/02/2002-0004$35.00/0 doi:10.1053/ajem.2002.31578
This study consisted of a secondary analysis of data obtained from the NHAMCS for EDs. Six years of data were compiled for 1992 through 1997. Data collection for NHAMCS is coordinated by the Ambulatory Care Statistics Branch of the National Center for Health Statistics, Centers for Disease Control and Prevention. NHAMCS data are available for public use and does not contain patient identifiers. NHAMCS is a probability sample of US emergency and outpatient departments designed to produce nationally representative statistics. Hospitals from all 50 states and the District of Columbia were eligible for sampling with the exception of Federal, military, and Veterans Administration hospitals. The survey is conducted using a multistage sampling design that begins with primary sampling units (PSUs) that were developed from the 1985-1994 National Health Interview Survey (NHIS). PSUs are geographically defined areas that are then stratified by socioeconomic and demo83
84
AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 20, Number 2 ■ March 2002
graphic variables. Individual patient encounters selected for data acquisition are the result of a step-wise series of probability samples obtained by selecting a given PSU, a hospital within the PSU, an ED within the hospital, and then an individual patient encounter within that ED. More detailed descriptions of this multi-stage sampling design may be found elsewhere.35 Primary data collection occurred during any of 13 randomly assigned 4-week periods during any individual year. Data were collected in a systematic block design calculated to provide an average of 50 patient encounters per ED based on that hospital’s average census. All data were compiled at the time of visit, or shortly thereafter, in a standardized format by hospital staff specially trained by field representatives from the NCHS and the Bureau of the Census. The surveys involved sampling approximately 25,000 patient encounters and 400 hospitals per year. The overall response rate for the surveys averaged 95%. The NCHS maintains an internal quality control system for monitoring accuracy. Field representatives from the Bureau of the Census and the NCHS made weekly followups for missing and inconsistent data. Nonresponse rates for individually missing items averaged 5% or less. Missing items were imputed using a “hot deck” procedure by assigning the missing value from a randomly selected patient record with similar characteristics. For the urgency variable, imputation was based on ED size, geographic region, and ICD-9-CM code for principal diagnosis. For visit time, birth year, sex, and race, imputations were based on ED size, geographic region, urgency, and ICD-9-CM code for principle diagnosis. The average error rate for coding of nonmedical items was 0.1 percent, and for medical items averaged less than 3%. NHAMCS collects information on approximately 200 variables for each patient encounter. Routine demographics contained within the survey included age, race, gender, insurance, metropolitan statistical area (MSA), type of hospital, and urgency of visit.
or nonteaching based on the presence of resident physicians in training. Visits were classified as either urgent or nonurgent based on predefined standardized definitions. An urgent visit was defined as “any acute medical illness or injury that required immediate attention such that a delay in care would be harmful.” Nonurgent medical conditions were defined as “medical conditions that would not necessitate immediate attention and could be safely handled within a few hours.”36 Patients were stratified into two diagnosis groups based upon their diagnosis codes from the International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM):orthopedic and wound-related injuries.37 Orthopedic diagnoses included any fracture, but did not include sprains, strains, or contusions (ICD-9-CM ⫽ 800.0829.9). Wound diagnoses included lacerations involving any part of the body (ICD-9-CM ⫽ 870.0-897.9). Data Analysis Data were analyzed using SAS software (SAS Institute Inc., Cary, NC). Unweighted data were not used for analysis. Analyses were weighted to obtain national estimates by taking into account the disproportionate sampling design used by NHAMCS (oversampling techniques of low frequency minority groups) based on the previously cited 4-stage sampling process. Standard errors were calculated using the method proposed by Potthoff and endorsed by NHAMCS officials.38 Multivariate logistic regression was performed to determine independent associations with PAS. Odds ratios were calculated and then used to compare the proportions of patients receiving PAS among subgroups. We hypothesized that the three highest risk factors for inadequate PAS might include the age, race, and insurance variables. We therefore also combined these 3 variables into a single variable called an interaction, and performed the logistic regression again with all of the other variables in the model remaining the same. The institutional review board at the University of Rochester approved this research.
Data Extraction Patients were identified as having received PAS if they received any of the following drugs: fentanyl, ketamine, meperidine, methohexital, midazolam, morphine, nitrous oxide, or propofol. Patients 18 years of age and younger were defined as pediatric. Race was originally coded as white, African American, Asian/Pacific Islander, or American Indian/Eskimo/Aleutian. For our analysis, Asian/Pacific Islander and American Indian/Eskimo/Aleutian were collapsed into the white variable. Information on insurance status was coded in a wide variety of formats. Groups were recoded into 4 categories: Private (Blue Cross/Blue Shield, Worker’s Compensation, Health Maintenance Organization [HMO]/other prepaid, fee-for-service, other/unspecified insurance), Medicaid, Medicare, and other (self-pay, no charge, other/unspecified/unknown). All types of insurance were placed into the private category unless specifically linked to Medicaid, for example Medicaid HMO, which was placed into the Medicaid category. Locations of hospitals were dichotomized into urban or nonurban based on their metropolitan statistical area (MSA) as coded by the National Census Bureau. Hospitals were classified as teaching
RESULTS A total of 43,725 pediatric and 114,207 adult ED encounters were analyzed and represented a weighted sample of 555.3 million ED visits. Three different groups of patients were examined: the entire population, orthopedic injuries (fractures), and wound-related injuries (lacerations). The proportions of patients receiving PAS (percentage rates) and the calculated independent odds ratios (OR) are presented in Table 1. Interactions were calculated with the three highest risk factors combined into a single variable (age, race, and insurance), and are presented in Table 2. For patients with orthopedic injuries, 5.6% of pediatric patients versus 7.8% of adults received PAS (OR 0.8, 95% CI 0.6-1.6, P ⫽ .05), 4.5% of African American patients vs 7.5% of white patients received PAS (OR 1.0, 95% CI 0.8-1.2, p⫽0.84), and 4.6% of patients on Medicaid versus 6.6% of patients with private insurance received PAS (OR 0.8, 95% CI 0.6-1.1, P ⫽ .19). Using the combined interactions of age, race, and insurance, African American children covered by Medicaid insurance were the least likely to receive PAS (OR 0.2, 95% CI 0.1-0.6).
HOSTETLER, AUINGER, AND SZILAGYI ■ PARENTERAL ANALGESIA AND SEDATION IN US
TABLE 1.
85
Proportions of Patients and Associated Odds Ratios for Receiving Parenteral Analgesia or Sedation Entire Population
Age ⬍ 18 years ⱖ 18 years (referent) Race African American White (referent) Gender Female Male (referent) Insurance Medicare Medicaid Other Private (referent) Hospital Teaching Nonteaching (referent) Urbanization status Urban Nonurban (referent) Urgency of Visit Urgent/emergent Nonurgent (referent)
Orthopedic Injuries
Wound-Related Injuries
%
OR
95% CIs
P
%
OR
95% CIs
P
%
OR
95% CIs
P
1.3 5.0
0.3 1.0
0.2-0.3 1.0
⬍.001
5.6 7.8
0.8 1.0
0.6-1.4 1.0
.05
2.3 2.8
0.8 1.0
0.6-1.0 1.0
.05
2.3 4.4
0.9 1.0
0.8-1.0 1.0
.002
4.5 7.5
1.0 1.0
0.8-1.2 1.0
.84
2.7 2.7
1.0 1.0
0.8-1.2 1.0
.97
4.2 3.7
1.7 1.0
1.4-2.0 1.0
⬍.001
7.4 7.0
1.7 1.0
1.2-2.5 1.0
.001
2.7 2.7
1.2 1.0
0.9-1.7 1.0
.19
5.6 2.4 3.2 4.6
0.9 0.7 0.7 1.0
0.8-1.0 0.7-0.8 0.6-0.8 1.0
.06 ⬍.001 ⬍.001
14.1 4.6 6.0 6.6
2.0 0.8 0.8 1.0
1.7-2.5 0.6-1.1 0.7-1.1 1.0
⬍.001 .19 .13
2.9 2.7 2.7 2.6
1.0 1.2 1.0 1.0
0.6-1.7 0.8-1.7 0.7-1.3 1.0
.86 .25 .78
3.9 4.0
1.1 1.0
1.0-1.2 1.0
.24
10.7 6.7
1.7 1.0
1.4-2.5 1.0
⬍.001
4.3 2.5
1.7 1.0
1.3-2.5 1.0
⬍.001
3.8 4.5
0.9 1.0
0.8-1.0 1.0
.01
7.1 7.2
0.9 1.0
0.8-1.2 1.0
.5
2.7 2.5
1.1 1.0
0.8-1.7 1.0
.97
5.4 2.5
2.0 1.0
1.9-2.5 1.0
⬍.001
10.2 3.7
2.5 1.0
2.0-3.3 1.0
⬍.001
3.8 1.3
2.5 1.0
2.0-3.3 1.0
⬍.001
Abbreviations: OR, odds ratio; 95% CI, 95% confidence intervals for odds ratio.
The proportion of female patients receiving PAS was 7.4% compared with 7.0% of males (OR 1.7, 95% CI 1.2-2.5, P ⫽ .001), 10.7% for patients from teaching hospitals compared with 6.7% of patients from nonteaching hospitals (OR 1.7, 95% CI 1.4-2.5, P ⬍ .001). The proportion of patients categorized as urgent receiving PAS was 10.2% compared with 3.7% of nonurgent visits (OR 2.5, 95% CI 2.5-3.3, P ⬍ .001), and 7.1% of urban patients received PAS compared with 7.2% of nonurban patients (OR 0.9, 95% CI 0.8-1.2, P ⫽ .5). DISCUSSION These results suggest that variations may be occurring in the proportion of patients receiving PAS in the ED. Children, African American patients, and those covered by TABLE 2.
Medicaid insurance tended to have the lowest proportions receiving PAS. Although multiple logistic regression showed only mixed results when evaluating for independent associations, when combined into interactions, it became clear that being a child, African American, or covered by Medicaid insurance were all additive risk factors for receiving less PAS. In contrast, their counterparts, being an adult, white, or covered by private insurance were all protective for receiving higher proportions of PAS. African American children covered by Medicaid insurance are at greatest risk for undertreatment. In contrast, several factors appeared to have a protective effect on patients with relatively higher proportions receiving PAS. These included females, patients categorized as urgent/emergent, patients treated at teaching hospitals, and those covered by Medicare insurance. We hypothesize that
NHAMCS 1992-1997: Interactions Between Age, Race, and Insurance Entire Population
Age, Race, Insurance Peds, A-A, Medicaid Peds, A-A, private Peds, white, Medicaid Peds, white, private Adult, A-A, Medicaid Adult, A-A, private Adult, A-A, Medicare Adult, white, Medicaid Adult, white, private
Orthopedic Injuries
Wound-Related Injuries
OR
95% CIs
P
OR
95% CIs
P
OR
95% CIs
P
0.1 0.2 0.1 0.3 0.5 0.6 0.7 0.8 1.0
0.1-0.2 0.1-0.3 0.1-0.2 0.2-0.3 0.4-0.6 0.4-0.7 0.5-0.9 0.7-0.9 1.0
⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 .003 ⬍.001
0.2 0.6 0.5 0.8 0.5 0.6 0.6 0.9 1.0
0.1-0.6 0.3-1.1 0.3-0.9 0.6-1.1 0.1-2.0 0.3-1.0 0.2-2.0 0.6-1.4 1.0
.002 .10 .02 .22 .32 .05 .38 .78
0.6 0.9 0.9 0.7 0.6 0.9 0.4 1.8 1.0
0.3-1.1 0.4-20 0.6-1.7 0.5-1.0 0.2-1.1 0.4-2.0 0.1-2.0 1.0-2.5 1.0
.11 .78 .77 .06 .12 .78 .27 .03
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; A-A, African American.
86
AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 20, Number 2 ■ March 2002
patients in the urgent/emergent category and those seen at teaching hospitals were most likely associated with an even higher acuity or complexity of injury and therefore required higher levels of PAS. It would seem unlikely however the females or those covered by Medicare insurance suffered more severe or complex injuries than males or those covered by private insurance. An alternative hypothesis may be that females are better at communicating their pain, and the elderly (generally covered by Medicare) may be assumed to have pain even when unable to communicate that need. Overall, NHAMCS appears well suited for epidemiological studies into emergency medicine practice variations, particularly from a national public health perspective. In addition to the availability of large numbers of patients, NHAMCS combines highly sophisticated sampling techniques with detailed clinical information to create nationally representative data that could otherwise never be obtained. There are a number of difficulties, however, inherent in analyzing and interpreting this data. First, reasonable comparisons are difficult when analyzing the entire population because of the potential for distinctly different diagnostic categories to occur across groups, particularly with respect to age. In addition there may be access issues affecting the relative proportions of patients seen in the ED among different groups. For this reason we performed subgroup analyses of those patients most likely to require PAS independent of other factors. We looked specifically at 2 groups of traumatic injuries, orthopedic and wound-related injuries, and chose orthopedic fractures as our anchor diagnosis for the best comparison group for study. Although the relative frequencies of fractures might vary within a subgroup (for example males having more fractures than females), the association with a need for PAS should remain relatively constant. In addition, this is a secondary data analysis and as such there are several methodological limitations that can be associated with its use and interpretations. First, this study is admittedly using a database not originally intended for studying national variations in the use of analgesia or sedation. There was no single variable for determining whether or not a patient received PAS. In our case we used whether a patient received any of the drugs we listed as being most commonly associated with PAS. We chose to include a rather limited list of analgesic and sedative agents. Rather than a compendium of all possible agents, we chose a group of parenteral agents most likely to be used relatively equally across all groups of patients. We did not include agents such as toradol or etomidate because they have in the past been used almost exclusively in the adult population. Although this has recently been changing, during the time period under study their use in children was limited. We also did not include oral analgesics such as acetaminophen, ibuprofen, or codeine. Although effective, they are more frequently used for lower intensities of pain. In addition, the NHAMCS database could not differentiate reliably between a medication being given therapeutically in the ED, versus a prescription being given for home use. Although the data provided much of the key information necessary for determining the use of analgesic and sedative agents in conjunction with most of the demographic variables, NHAMCS lacks more precise indicators of severity.
We attempted to compensate for this first by controlling for diagnosis (orthopedic fractures), and to a certain extent this should have placed patients into a higher level of acuity. Second, we were also able to approximate severity by using the triage acuity variable contained within NHAMCS. We know from looking at our data that the acuity variable was internally valid in that significantly higher proportions of patients in the urgent/emergent category received PAS and therefore did more than likely represent the more severe end of the spectrum. This categorization, however, is admittedly crude. We cannot make any assumptions about the variability of acuity occurring within that subgroup of emergent/ urgent patients. Newer revisions of NHAMCS are now including a more detailed triage classification system as well as subjective patient ratings of pain. Both of these changes should allow for more detailed analyses in the future. In the interim, further measures may be necessary to ensure that adequate PAS is provided to all ED patients. Our analysis of the data regarding age, race, and insurance is somewhat controversial. We have taken great time however to analyze all of the race and socioeconomic data separately, in combination, and in context with the other data. We believe that a complete analysis of the issues surrounding race and socioeconomic status requires consideration of all of these combinations. In a theme that is illustrated and repeated across all combinations of analysis, risk factors for receiving less PAS include being a child, African American, and covered by Medicaid insurance. We note that these differences are not a new phenomenon and have been reported in multiple other studies examining variations in the delivery of healthcare in the United States.39-46 These results are not meant to be an indictment of emergency medicine, rather a preliminary exploration into some of the epidemiological variations that may be occurring in the delivery of emergency care in the United States. REFERENCES 1. Comprehensive Accreditation Manual for Hospitals, Joint Commission on Accreditation of Healthcare Organizations (JCAHO), Update 3, 1999. 2. Pena BM, Krauss B: Adverse events of procedural sedation and analgesia in a pediatric emergency department. Ann Emerg Med 1999;34:483-91 3. Bauman LA, Kish I, Baumann RC, et al: Pediatric sedation with analgesia. Am J Emerg Med 1999;17:1-3 4. Malviya S, Voepel-Lewis T, Tait AR: Adverse events and risk factors associated with the sedation of children by nonanesthesiologists. Anesth Analg 1997;85:1207-13 5. American Academy of Pediatrics Committee on Drugs: Guidelines for the elective use of conscious sedation, deep sedation, and general anesthesia in pediatric patients. Pediatrics 1992;89:11101115 6. American College of Emergency Physicians: Clinical policy for procedural sedation and analgesia in the emergency department. Ann Emerg Med 1998;31:663-77 7. American Society of Anesthesiologists: Practice guidelines for sedation and analgesia by non-anesthesiologists. Anesthesiology 1996;84:459-71 8. US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, Acute Pain Management Guideline Panel: Clinical practice guideline - Acute pain management: Operative or medical procedures and trauma. Washington, DC, Author, 1992
HOSTETLER, AUINGER, AND SZILAGYI ■ PARENTERAL ANALGESIA AND SEDATION IN US
9. Innes G, Murphy M, Hijssen-Jordan C, et al: Procedural sedation and analgesia in the emergency department. Canadian Consensus Guidelines. J Emerg Med 1999;17:145-56 10. Ilkhanipour K, Juels CR, Langdorf MI: Pediatric pain control and conscious sedation: A survey of emergency medicine residencies. Acad Emerg Med 1994;1:368-72 11. Sacchetti A, Schafermeyer R, Gerardi M, et al: Pediatric analgesia or sedation. Ann Emerg Med 1994;23:237-50 12. Litman RS: Sedation and anesthesia outside the operating room: Answers to common questions. Semin Pediatr Surg 1999;8: 34-9 13. Holzman RS, Cullen DJ, Eichhorn JH, et al: Guidelines for sedation by nonanesthesiologists during diagnostic and therapeutic procedures. The Risk Management Committee of the Department of Anaesthesia of Harvard Medical School. J Clin Anesth 1994;6:26576. 14. Friedland LR, Pancioli AM, Duncan KM: Pediatric emergency department analgesic practice. Pediatr Emerg Care 1997;13:103-6 15. Lewis LM, Lasater LC, Brooks CB: Are emergency department physicians too stingy with analgesics? South Med J 1994;87: 7-9 16. Wilson JE, Pendleton JM: Oligoanalgesia in the emergency department. Am J Emerg Med 1989;7:620-3 17. Petrack EM, Christopher NC, Kriwinsky J: Pain management in the emergency department: Patterns of analgesic utilization. Pediatrics 1997;99:711-4 18. Todd KH, Deaton C, D’Adamo AP, et al: Ethnicity and analgesic practice. Ann Emerg Med 2000;35:11-16 19. Martin ML: Ethnicity and analgesic practice: An editorial. Ann Emerg Med 2000;35:77-79 20. Goldfrank LR, Knopp RK: Racially and ethnically selective oligoanalgesia: Is this racism? Ann Emerg Med 2000;35:79-82 21. Todd KH, Lee T, Hoffman JR: The effect of ethnicity on physician estimates of pain severity in patients with isolated extremity trauma. JAMA 1994;271:925-8 22. Todd KH, Samaroo N, Hoffman JR: Ethnicity as a risk factor for inadequate emergency department analgesia. JAMA 1993;269: 1537-9 23. Karpman RR, Del Mar N, Bay C: Analgesia for emergency centers’ orthopaedic patients: Does an ethnic bias exist? Clin Orthop 1997;334:270-5 24. Raftery KA, Smith-Coggins R, Chen AH: Gender-associated differences in emergency department pain management. Ann Emerg Med 1995;26:414-21 25. Jones JS, Johnson K, McNinch M: Age as a risk factor for inadequate emergency department analgesia. Am J Emerg Med 1996;14:157-60 26. Portenoy RK, Kanner RM: Patterns of analgesic prescription and consumption in a university-affiliated community hospital. Arch Intern Med 1985;145:439-41 27. Schweich PJ, Smith KM, Dowd MD, et al: Pediatric emergency medicine practice patterns: A comparison of pediatric and general emergency physicians. Pediatr Emerg Care 1998;14:89-94 28. Sacchetti A, Carraccio C, Warden T, et al: Community hospital management of pediatric emergencies: Implications for pediatric emergency medical services. Am J Emerg Med 1986;4:10-13
87
29. Krauss B, Zurakowski D: Sedation patterns in pediatric and general community hospital emergency departments. Pediatr Emerg Care 1998;14:99-103 30. Hawk W, Crockett RK, Ochsenschlager DW, et al: Conscious sedation of the pediatric patient for suturing: a survey. Pediatr Emerg Care 1990;6:84-8 31. Conners GP, Sacks WK, Leahey NF: Variations in sedating uncooperative, stable children for post-traumatic head CT. Pediatr Emerg Care 1999;15:241-4 32. Hauswald M, Anison C: Prescribing analgesics: The effect of patient age and physician specialty. Pediatr Emerg Care 1997;13: 262-3 33. Selbst SM, Clark M: Analgesic use in the emergency department. Ann Emerg Med 1990;19:1010-3 34. Schechter NL, Allen DA, Hanson K: Status of pediatric pain control: A comparison of hospital analgesic usage in children and adults. Pediatrics 1986;77:11-5 35. Stussman BJ: National hospital ambulatory medical care survey: 1995 emergency department summary. Adv Data 1997;285: 1-20 36. McCaig LF: National hospital ambulatory medical care survey: 1996 emergency department summary. Adv Data 1997;294: 1-17 37. Public Health Service and Health Care Financing Administration. International Classification of Diseases, ninth revision, Clinical Modification. Washington, DC, Public Health Service, 1991 38. Potthoff RF, Woodbury MA, Manton KG: Equivalent sample size and equivalent degrees of freedom refinements for inference using survey weights under superpopulation models. J Amer Statist Assn 1992;87:383-96 39. Chen J, Rathore SS, Radford MJ, et al: Racial differences in the use of cardiac catheterization after acute myocardial infarction. N Engl J Med 2001;344:1471-3 40. Rathore SS, Berger AK, Weinfurt KP, et al: Race, sex, poverty, and the medical treatment of acute myocardial infarction in the elderly. Circulation 2000;102:943-4 41. Cooper R, Cutler J, Desvigne-Nickens P, et al: Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: Findings of the national conference on cardiovascular disease prevention. Circulation 2000;102: 3137-47 42. Harris MI: Racial and ethnic differences in health care access and health outcomes for adults with type 2 diabetes. Diabetes Care 2001;24:454-9 43. Wolfe RA, Ashby VB, Milford EL, et al: Differences in access to cadaveric renal translplantation in the United States. Am J Kidney Dis 2000;36:1025-33 44. Erickson LC, Wise PH, Cook EF, et al: The impact of managed care insurance on use of lower-mortality hospitals by children undergoing cardiac surgery in California. Pediatrics 2000; 105:1271-8 45. Gadomski A, Jenkins P: Ruptured appendicitis among children as an indicator of access to care. Health Serv Res 2001;36: 129-42 46. Volmer T: The socio-economics of asthma. Pulm Pharmacol Ther 2001;14:55-60
The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) requires that all patients receive an assessment for, and appropriate management of, pain.1 Sedation provided by nonanesthesiologists in the emergency department (ED) is both safe and effective.2-4 Detailed guidelines and recommendations have been developed and instituted by a wide variety of governmental and professional organizations.5-9 The provision of safe and adequate analgesia and sedation in the ED has now become the standard of care for both children and adults.10-13
Several studies have nonetheless suggested that ED patients on the whole receive inadequate amounts of analgesia or sedation.14-16 In addition, there may also be evidence that certain groups of patients receive substantially less analgesia or sedation than others.17-23 These studies have at times been contradictory, but have continued to suggest differential proportions of patients in the ED receiving analgesia or sedation based on age, race, gender, or ethnicity.18-26 The pediatric age group may be particularly at risk for inadequate analgesia or sedation.14,16,27-31 Although some of these studies have documented broad global variation based on physician training or preference,27-29,32 others have shown that when comparing similar illness and injury, children appear to receive substantially less analgesia or sedation than adults.17,33,34 We sought to explore whether significant variations occurred among the proportions of patients receiving parenteral analgesia and sedation (PAS). Rather than using data from a single institution, however, we chose to combine multiple years worth of data from a large, nationally representative database, the National Hospital Ambulatory Medical Care Survey (NHAMCS). By combining multiple years’ worth of data we were able to create a sample large enough to allow for detailed subgroup analyses using multivariate logistic regression techniques. METHODS Data Source
From the *Department of Pediatrics, Division of Emergency Medicine, St. Louis Children’s Hospital, Washington University School of Medicine, St. Louis, MO; the †Department of Pediatrics, Rochester General Hospital, Rochester, NY and the ‡Departments of Emergency Medicine and Pediatrics*, University of Rochester School of Medicine and Dentistry, Rochester, NY. Presented at the Society for Academic Emergency Medicine Annual Conference, San Francisco, CA, May 2000 and the New York State Public Health Association’s 50th Annual Meeting, Albany, NY, May 2000. Manuscript received August 29, 2001, accepted September 20, 2001. Address reprint requests to Mark A. Hostetler, MD, MPH, St. Louis Children’s Hospital, Washington University School of Medicine, One Children’s Place, St Louis, MO 63110. E-mail: mark_ [email protected] Key Words: analgesia, sedation, emergency, fracture, race, insurance Copyright 2002, Elsevier Science (USA). All rights reserved. 0735-6757/02/2002-0004$35.00/0 doi:10.1053/ajem.2002.31578
This study consisted of a secondary analysis of data obtained from the NHAMCS for EDs. Six years of data were compiled for 1992 through 1997. Data collection for NHAMCS is coordinated by the Ambulatory Care Statistics Branch of the National Center for Health Statistics, Centers for Disease Control and Prevention. NHAMCS data are available for public use and does not contain patient identifiers. NHAMCS is a probability sample of US emergency and outpatient departments designed to produce nationally representative statistics. Hospitals from all 50 states and the District of Columbia were eligible for sampling with the exception of Federal, military, and Veterans Administration hospitals. The survey is conducted using a multistage sampling design that begins with primary sampling units (PSUs) that were developed from the 1985-1994 National Health Interview Survey (NHIS). PSUs are geographically defined areas that are then stratified by socioeconomic and demo83
84
AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 20, Number 2 ■ March 2002
graphic variables. Individual patient encounters selected for data acquisition are the result of a step-wise series of probability samples obtained by selecting a given PSU, a hospital within the PSU, an ED within the hospital, and then an individual patient encounter within that ED. More detailed descriptions of this multi-stage sampling design may be found elsewhere.35 Primary data collection occurred during any of 13 randomly assigned 4-week periods during any individual year. Data were collected in a systematic block design calculated to provide an average of 50 patient encounters per ED based on that hospital’s average census. All data were compiled at the time of visit, or shortly thereafter, in a standardized format by hospital staff specially trained by field representatives from the NCHS and the Bureau of the Census. The surveys involved sampling approximately 25,000 patient encounters and 400 hospitals per year. The overall response rate for the surveys averaged 95%. The NCHS maintains an internal quality control system for monitoring accuracy. Field representatives from the Bureau of the Census and the NCHS made weekly followups for missing and inconsistent data. Nonresponse rates for individually missing items averaged 5% or less. Missing items were imputed using a “hot deck” procedure by assigning the missing value from a randomly selected patient record with similar characteristics. For the urgency variable, imputation was based on ED size, geographic region, and ICD-9-CM code for principal diagnosis. For visit time, birth year, sex, and race, imputations were based on ED size, geographic region, urgency, and ICD-9-CM code for principle diagnosis. The average error rate for coding of nonmedical items was 0.1 percent, and for medical items averaged less than 3%. NHAMCS collects information on approximately 200 variables for each patient encounter. Routine demographics contained within the survey included age, race, gender, insurance, metropolitan statistical area (MSA), type of hospital, and urgency of visit.
or nonteaching based on the presence of resident physicians in training. Visits were classified as either urgent or nonurgent based on predefined standardized definitions. An urgent visit was defined as “any acute medical illness or injury that required immediate attention such that a delay in care would be harmful.” Nonurgent medical conditions were defined as “medical conditions that would not necessitate immediate attention and could be safely handled within a few hours.”36 Patients were stratified into two diagnosis groups based upon their diagnosis codes from the International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM):orthopedic and wound-related injuries.37 Orthopedic diagnoses included any fracture, but did not include sprains, strains, or contusions (ICD-9-CM ⫽ 800.0829.9). Wound diagnoses included lacerations involving any part of the body (ICD-9-CM ⫽ 870.0-897.9). Data Analysis Data were analyzed using SAS software (SAS Institute Inc., Cary, NC). Unweighted data were not used for analysis. Analyses were weighted to obtain national estimates by taking into account the disproportionate sampling design used by NHAMCS (oversampling techniques of low frequency minority groups) based on the previously cited 4-stage sampling process. Standard errors were calculated using the method proposed by Potthoff and endorsed by NHAMCS officials.38 Multivariate logistic regression was performed to determine independent associations with PAS. Odds ratios were calculated and then used to compare the proportions of patients receiving PAS among subgroups. We hypothesized that the three highest risk factors for inadequate PAS might include the age, race, and insurance variables. We therefore also combined these 3 variables into a single variable called an interaction, and performed the logistic regression again with all of the other variables in the model remaining the same. The institutional review board at the University of Rochester approved this research.
Data Extraction Patients were identified as having received PAS if they received any of the following drugs: fentanyl, ketamine, meperidine, methohexital, midazolam, morphine, nitrous oxide, or propofol. Patients 18 years of age and younger were defined as pediatric. Race was originally coded as white, African American, Asian/Pacific Islander, or American Indian/Eskimo/Aleutian. For our analysis, Asian/Pacific Islander and American Indian/Eskimo/Aleutian were collapsed into the white variable. Information on insurance status was coded in a wide variety of formats. Groups were recoded into 4 categories: Private (Blue Cross/Blue Shield, Worker’s Compensation, Health Maintenance Organization [HMO]/other prepaid, fee-for-service, other/unspecified insurance), Medicaid, Medicare, and other (self-pay, no charge, other/unspecified/unknown). All types of insurance were placed into the private category unless specifically linked to Medicaid, for example Medicaid HMO, which was placed into the Medicaid category. Locations of hospitals were dichotomized into urban or nonurban based on their metropolitan statistical area (MSA) as coded by the National Census Bureau. Hospitals were classified as teaching
RESULTS A total of 43,725 pediatric and 114,207 adult ED encounters were analyzed and represented a weighted sample of 555.3 million ED visits. Three different groups of patients were examined: the entire population, orthopedic injuries (fractures), and wound-related injuries (lacerations). The proportions of patients receiving PAS (percentage rates) and the calculated independent odds ratios (OR) are presented in Table 1. Interactions were calculated with the three highest risk factors combined into a single variable (age, race, and insurance), and are presented in Table 2. For patients with orthopedic injuries, 5.6% of pediatric patients versus 7.8% of adults received PAS (OR 0.8, 95% CI 0.6-1.6, P ⫽ .05), 4.5% of African American patients vs 7.5% of white patients received PAS (OR 1.0, 95% CI 0.8-1.2, p⫽0.84), and 4.6% of patients on Medicaid versus 6.6% of patients with private insurance received PAS (OR 0.8, 95% CI 0.6-1.1, P ⫽ .19). Using the combined interactions of age, race, and insurance, African American children covered by Medicaid insurance were the least likely to receive PAS (OR 0.2, 95% CI 0.1-0.6).
HOSTETLER, AUINGER, AND SZILAGYI ■ PARENTERAL ANALGESIA AND SEDATION IN US
TABLE 1.
85
Proportions of Patients and Associated Odds Ratios for Receiving Parenteral Analgesia or Sedation Entire Population
Age ⬍ 18 years ⱖ 18 years (referent) Race African American White (referent) Gender Female Male (referent) Insurance Medicare Medicaid Other Private (referent) Hospital Teaching Nonteaching (referent) Urbanization status Urban Nonurban (referent) Urgency of Visit Urgent/emergent Nonurgent (referent)
Orthopedic Injuries
Wound-Related Injuries
%
OR
95% CIs
P
%
OR
95% CIs
P
%
OR
95% CIs
P
1.3 5.0
0.3 1.0
0.2-0.3 1.0
⬍.001
5.6 7.8
0.8 1.0
0.6-1.4 1.0
.05
2.3 2.8
0.8 1.0
0.6-1.0 1.0
.05
2.3 4.4
0.9 1.0
0.8-1.0 1.0
.002
4.5 7.5
1.0 1.0
0.8-1.2 1.0
.84
2.7 2.7
1.0 1.0
0.8-1.2 1.0
.97
4.2 3.7
1.7 1.0
1.4-2.0 1.0
⬍.001
7.4 7.0
1.7 1.0
1.2-2.5 1.0
.001
2.7 2.7
1.2 1.0
0.9-1.7 1.0
.19
5.6 2.4 3.2 4.6
0.9 0.7 0.7 1.0
0.8-1.0 0.7-0.8 0.6-0.8 1.0
.06 ⬍.001 ⬍.001
14.1 4.6 6.0 6.6
2.0 0.8 0.8 1.0
1.7-2.5 0.6-1.1 0.7-1.1 1.0
⬍.001 .19 .13
2.9 2.7 2.7 2.6
1.0 1.2 1.0 1.0
0.6-1.7 0.8-1.7 0.7-1.3 1.0
.86 .25 .78
3.9 4.0
1.1 1.0
1.0-1.2 1.0
.24
10.7 6.7
1.7 1.0
1.4-2.5 1.0
⬍.001
4.3 2.5
1.7 1.0
1.3-2.5 1.0
⬍.001
3.8 4.5
0.9 1.0
0.8-1.0 1.0
.01
7.1 7.2
0.9 1.0
0.8-1.2 1.0
.5
2.7 2.5
1.1 1.0
0.8-1.7 1.0
.97
5.4 2.5
2.0 1.0
1.9-2.5 1.0
⬍.001
10.2 3.7
2.5 1.0
2.0-3.3 1.0
⬍.001
3.8 1.3
2.5 1.0
2.0-3.3 1.0
⬍.001
Abbreviations: OR, odds ratio; 95% CI, 95% confidence intervals for odds ratio.
The proportion of female patients receiving PAS was 7.4% compared with 7.0% of males (OR 1.7, 95% CI 1.2-2.5, P ⫽ .001), 10.7% for patients from teaching hospitals compared with 6.7% of patients from nonteaching hospitals (OR 1.7, 95% CI 1.4-2.5, P ⬍ .001). The proportion of patients categorized as urgent receiving PAS was 10.2% compared with 3.7% of nonurgent visits (OR 2.5, 95% CI 2.5-3.3, P ⬍ .001), and 7.1% of urban patients received PAS compared with 7.2% of nonurban patients (OR 0.9, 95% CI 0.8-1.2, P ⫽ .5). DISCUSSION These results suggest that variations may be occurring in the proportion of patients receiving PAS in the ED. Children, African American patients, and those covered by TABLE 2.
Medicaid insurance tended to have the lowest proportions receiving PAS. Although multiple logistic regression showed only mixed results when evaluating for independent associations, when combined into interactions, it became clear that being a child, African American, or covered by Medicaid insurance were all additive risk factors for receiving less PAS. In contrast, their counterparts, being an adult, white, or covered by private insurance were all protective for receiving higher proportions of PAS. African American children covered by Medicaid insurance are at greatest risk for undertreatment. In contrast, several factors appeared to have a protective effect on patients with relatively higher proportions receiving PAS. These included females, patients categorized as urgent/emergent, patients treated at teaching hospitals, and those covered by Medicare insurance. We hypothesize that
NHAMCS 1992-1997: Interactions Between Age, Race, and Insurance Entire Population
Age, Race, Insurance Peds, A-A, Medicaid Peds, A-A, private Peds, white, Medicaid Peds, white, private Adult, A-A, Medicaid Adult, A-A, private Adult, A-A, Medicare Adult, white, Medicaid Adult, white, private
Orthopedic Injuries
Wound-Related Injuries
OR
95% CIs
P
OR
95% CIs
P
OR
95% CIs
P
0.1 0.2 0.1 0.3 0.5 0.6 0.7 0.8 1.0
0.1-0.2 0.1-0.3 0.1-0.2 0.2-0.3 0.4-0.6 0.4-0.7 0.5-0.9 0.7-0.9 1.0
⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 .003 ⬍.001
0.2 0.6 0.5 0.8 0.5 0.6 0.6 0.9 1.0
0.1-0.6 0.3-1.1 0.3-0.9 0.6-1.1 0.1-2.0 0.3-1.0 0.2-2.0 0.6-1.4 1.0
.002 .10 .02 .22 .32 .05 .38 .78
0.6 0.9 0.9 0.7 0.6 0.9 0.4 1.8 1.0
0.3-1.1 0.4-20 0.6-1.7 0.5-1.0 0.2-1.1 0.4-2.0 0.1-2.0 1.0-2.5 1.0
.11 .78 .77 .06 .12 .78 .27 .03
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; A-A, African American.
86
AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 20, Number 2 ■ March 2002
patients in the urgent/emergent category and those seen at teaching hospitals were most likely associated with an even higher acuity or complexity of injury and therefore required higher levels of PAS. It would seem unlikely however the females or those covered by Medicare insurance suffered more severe or complex injuries than males or those covered by private insurance. An alternative hypothesis may be that females are better at communicating their pain, and the elderly (generally covered by Medicare) may be assumed to have pain even when unable to communicate that need. Overall, NHAMCS appears well suited for epidemiological studies into emergency medicine practice variations, particularly from a national public health perspective. In addition to the availability of large numbers of patients, NHAMCS combines highly sophisticated sampling techniques with detailed clinical information to create nationally representative data that could otherwise never be obtained. There are a number of difficulties, however, inherent in analyzing and interpreting this data. First, reasonable comparisons are difficult when analyzing the entire population because of the potential for distinctly different diagnostic categories to occur across groups, particularly with respect to age. In addition there may be access issues affecting the relative proportions of patients seen in the ED among different groups. For this reason we performed subgroup analyses of those patients most likely to require PAS independent of other factors. We looked specifically at 2 groups of traumatic injuries, orthopedic and wound-related injuries, and chose orthopedic fractures as our anchor diagnosis for the best comparison group for study. Although the relative frequencies of fractures might vary within a subgroup (for example males having more fractures than females), the association with a need for PAS should remain relatively constant. In addition, this is a secondary data analysis and as such there are several methodological limitations that can be associated with its use and interpretations. First, this study is admittedly using a database not originally intended for studying national variations in the use of analgesia or sedation. There was no single variable for determining whether or not a patient received PAS. In our case we used whether a patient received any of the drugs we listed as being most commonly associated with PAS. We chose to include a rather limited list of analgesic and sedative agents. Rather than a compendium of all possible agents, we chose a group of parenteral agents most likely to be used relatively equally across all groups of patients. We did not include agents such as toradol or etomidate because they have in the past been used almost exclusively in the adult population. Although this has recently been changing, during the time period under study their use in children was limited. We also did not include oral analgesics such as acetaminophen, ibuprofen, or codeine. Although effective, they are more frequently used for lower intensities of pain. In addition, the NHAMCS database could not differentiate reliably between a medication being given therapeutically in the ED, versus a prescription being given for home use. Although the data provided much of the key information necessary for determining the use of analgesic and sedative agents in conjunction with most of the demographic variables, NHAMCS lacks more precise indicators of severity.
We attempted to compensate for this first by controlling for diagnosis (orthopedic fractures), and to a certain extent this should have placed patients into a higher level of acuity. Second, we were also able to approximate severity by using the triage acuity variable contained within NHAMCS. We know from looking at our data that the acuity variable was internally valid in that significantly higher proportions of patients in the urgent/emergent category received PAS and therefore did more than likely represent the more severe end of the spectrum. This categorization, however, is admittedly crude. We cannot make any assumptions about the variability of acuity occurring within that subgroup of emergent/ urgent patients. Newer revisions of NHAMCS are now including a more detailed triage classification system as well as subjective patient ratings of pain. Both of these changes should allow for more detailed analyses in the future. In the interim, further measures may be necessary to ensure that adequate PAS is provided to all ED patients. Our analysis of the data regarding age, race, and insurance is somewhat controversial. We have taken great time however to analyze all of the race and socioeconomic data separately, in combination, and in context with the other data. We believe that a complete analysis of the issues surrounding race and socioeconomic status requires consideration of all of these combinations. In a theme that is illustrated and repeated across all combinations of analysis, risk factors for receiving less PAS include being a child, African American, and covered by Medicaid insurance. We note that these differences are not a new phenomenon and have been reported in multiple other studies examining variations in the delivery of healthcare in the United States.39-46 These results are not meant to be an indictment of emergency medicine, rather a preliminary exploration into some of the epidemiological variations that may be occurring in the delivery of emergency care in the United States. REFERENCES 1. Comprehensive Accreditation Manual for Hospitals, Joint Commission on Accreditation of Healthcare Organizations (JCAHO), Update 3, 1999. 2. Pena BM, Krauss B: Adverse events of procedural sedation and analgesia in a pediatric emergency department. Ann Emerg Med 1999;34:483-91 3. Bauman LA, Kish I, Baumann RC, et al: Pediatric sedation with analgesia. Am J Emerg Med 1999;17:1-3 4. Malviya S, Voepel-Lewis T, Tait AR: Adverse events and risk factors associated with the sedation of children by nonanesthesiologists. Anesth Analg 1997;85:1207-13 5. American Academy of Pediatrics Committee on Drugs: Guidelines for the elective use of conscious sedation, deep sedation, and general anesthesia in pediatric patients. Pediatrics 1992;89:11101115 6. American College of Emergency Physicians: Clinical policy for procedural sedation and analgesia in the emergency department. Ann Emerg Med 1998;31:663-77 7. American Society of Anesthesiologists: Practice guidelines for sedation and analgesia by non-anesthesiologists. Anesthesiology 1996;84:459-71 8. US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, Acute Pain Management Guideline Panel: Clinical practice guideline - Acute pain management: Operative or medical procedures and trauma. Washington, DC, Author, 1992
HOSTETLER, AUINGER, AND SZILAGYI ■ PARENTERAL ANALGESIA AND SEDATION IN US
9. Innes G, Murphy M, Hijssen-Jordan C, et al: Procedural sedation and analgesia in the emergency department. Canadian Consensus Guidelines. J Emerg Med 1999;17:145-56 10. Ilkhanipour K, Juels CR, Langdorf MI: Pediatric pain control and conscious sedation: A survey of emergency medicine residencies. Acad Emerg Med 1994;1:368-72 11. Sacchetti A, Schafermeyer R, Gerardi M, et al: Pediatric analgesia or sedation. Ann Emerg Med 1994;23:237-50 12. Litman RS: Sedation and anesthesia outside the operating room: Answers to common questions. Semin Pediatr Surg 1999;8: 34-9 13. Holzman RS, Cullen DJ, Eichhorn JH, et al: Guidelines for sedation by nonanesthesiologists during diagnostic and therapeutic procedures. The Risk Management Committee of the Department of Anaesthesia of Harvard Medical School. J Clin Anesth 1994;6:26576. 14. Friedland LR, Pancioli AM, Duncan KM: Pediatric emergency department analgesic practice. Pediatr Emerg Care 1997;13:103-6 15. Lewis LM, Lasater LC, Brooks CB: Are emergency department physicians too stingy with analgesics? South Med J 1994;87: 7-9 16. Wilson JE, Pendleton JM: Oligoanalgesia in the emergency department. Am J Emerg Med 1989;7:620-3 17. Petrack EM, Christopher NC, Kriwinsky J: Pain management in the emergency department: Patterns of analgesic utilization. Pediatrics 1997;99:711-4 18. Todd KH, Deaton C, D’Adamo AP, et al: Ethnicity and analgesic practice. Ann Emerg Med 2000;35:11-16 19. Martin ML: Ethnicity and analgesic practice: An editorial. Ann Emerg Med 2000;35:77-79 20. Goldfrank LR, Knopp RK: Racially and ethnically selective oligoanalgesia: Is this racism? Ann Emerg Med 2000;35:79-82 21. Todd KH, Lee T, Hoffman JR: The effect of ethnicity on physician estimates of pain severity in patients with isolated extremity trauma. JAMA 1994;271:925-8 22. Todd KH, Samaroo N, Hoffman JR: Ethnicity as a risk factor for inadequate emergency department analgesia. JAMA 1993;269: 1537-9 23. Karpman RR, Del Mar N, Bay C: Analgesia for emergency centers’ orthopaedic patients: Does an ethnic bias exist? Clin Orthop 1997;334:270-5 24. Raftery KA, Smith-Coggins R, Chen AH: Gender-associated differences in emergency department pain management. Ann Emerg Med 1995;26:414-21 25. Jones JS, Johnson K, McNinch M: Age as a risk factor for inadequate emergency department analgesia. Am J Emerg Med 1996;14:157-60 26. Portenoy RK, Kanner RM: Patterns of analgesic prescription and consumption in a university-affiliated community hospital. Arch Intern Med 1985;145:439-41 27. Schweich PJ, Smith KM, Dowd MD, et al: Pediatric emergency medicine practice patterns: A comparison of pediatric and general emergency physicians. Pediatr Emerg Care 1998;14:89-94 28. Sacchetti A, Carraccio C, Warden T, et al: Community hospital management of pediatric emergencies: Implications for pediatric emergency medical services. Am J Emerg Med 1986;4:10-13
87
29. Krauss B, Zurakowski D: Sedation patterns in pediatric and general community hospital emergency departments. Pediatr Emerg Care 1998;14:99-103 30. Hawk W, Crockett RK, Ochsenschlager DW, et al: Conscious sedation of the pediatric patient for suturing: a survey. Pediatr Emerg Care 1990;6:84-8 31. Conners GP, Sacks WK, Leahey NF: Variations in sedating uncooperative, stable children for post-traumatic head CT. Pediatr Emerg Care 1999;15:241-4 32. Hauswald M, Anison C: Prescribing analgesics: The effect of patient age and physician specialty. Pediatr Emerg Care 1997;13: 262-3 33. Selbst SM, Clark M: Analgesic use in the emergency department. Ann Emerg Med 1990;19:1010-3 34. Schechter NL, Allen DA, Hanson K: Status of pediatric pain control: A comparison of hospital analgesic usage in children and adults. Pediatrics 1986;77:11-5 35. Stussman BJ: National hospital ambulatory medical care survey: 1995 emergency department summary. Adv Data 1997;285: 1-20 36. McCaig LF: National hospital ambulatory medical care survey: 1996 emergency department summary. Adv Data 1997;294: 1-17 37. Public Health Service and Health Care Financing Administration. International Classification of Diseases, ninth revision, Clinical Modification. Washington, DC, Public Health Service, 1991 38. Potthoff RF, Woodbury MA, Manton KG: Equivalent sample size and equivalent degrees of freedom refinements for inference using survey weights under superpopulation models. J Amer Statist Assn 1992;87:383-96 39. Chen J, Rathore SS, Radford MJ, et al: Racial differences in the use of cardiac catheterization after acute myocardial infarction. N Engl J Med 2001;344:1471-3 40. Rathore SS, Berger AK, Weinfurt KP, et al: Race, sex, poverty, and the medical treatment of acute myocardial infarction in the elderly. Circulation 2000;102:943-4 41. Cooper R, Cutler J, Desvigne-Nickens P, et al: Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: Findings of the national conference on cardiovascular disease prevention. Circulation 2000;102: 3137-47 42. Harris MI: Racial and ethnic differences in health care access and health outcomes for adults with type 2 diabetes. Diabetes Care 2001;24:454-9 43. Wolfe RA, Ashby VB, Milford EL, et al: Differences in access to cadaveric renal translplantation in the United States. Am J Kidney Dis 2000;36:1025-33 44. Erickson LC, Wise PH, Cook EF, et al: The impact of managed care insurance on use of lower-mortality hospitals by children undergoing cardiac surgery in California. Pediatrics 2000; 105:1271-8 45. Gadomski A, Jenkins P: Ruptured appendicitis among children as an indicator of access to care. Health Serv Res 2001;36: 129-42 46. Volmer T: The socio-economics of asthma. Pulm Pharmacol Ther 2001;14:55-60