Abstract
Objectives
The objective of this study was to evaluate the association of emergency department (ED) crowding factors with the quality of pain care.
Methods
This was a retrospective observational study of all adult patients (≥18 years) with conditions warranting pain care seen at an academic, urban tertiary care ED from July 1 to July 31, 2005, and December 1 to December 31, 2005. Patients were included if they presented with a chief complaint of pain and a final ED diagnosis of a painful condition. Predictor ED crowding variables studied were: 1) census, 2) number of admitted patients waiting for inpatient beds (boarders), and 3) number of boarders divided by ED census (boarding burden). The outcomes of interest were process of pain care measures: documentation of clinician pain assessment, medications ordered, and times of activities (e.g., arrival, assessment, ordering of medications).
Results
A total of 1,068 patient visits were reviewed. Fewer patients received analgesic medication during periods of high census (>50th percentile) (Parameter estimate = −0.47 [95% CI = −0.80 to −0.07]). There was a direct correlation with total ED census and increased: time to pain assessment (Spearman r = 0.33, p < 0.0001), time to analgesic medication ordering (r = 0.22, p < 0.0001), and time to analgesic medication administration (r = 0.25, p < 0.0001). There were significant delays (>1-hour) for pain assessment and the ordering and administration of analgesic medication during periods of high ED census and number of boarders, but not with boarding burden.
Conclusions
ED crowding as measured by patient volume negatively impacts patient care. Greater numbers of patients in the ED, whether as total census or number of boarders, were associated with worse pain care.
Keywords: emergency department, quality of care, pain
INTRODUCTION
The 2006 Institute of Medicine (IOM) report, Hospital Based Emergency Care – at the Breaking Point,1 discusses how the U.S. health care system and emergency medicine (EM) are under increasing strain from emergency department (ED) crowding, threatening the ability of our medical centers to deliver timely patient care. Studies have demonstrated, both directly and indirectly, that ED crowding delays care. Examples include increased time to thrombolytics with network ambulance diversion,2 fewer patients receiving timely antibiotics during lengthened ED stays,3 and adverse 30-day outcomes for patients with acute coronary syndromes who arrive concurrently with a trauma patient.4
Pain serves as a model condition to understand how ED care is rendered and where improvements may potentially be directed. Now considered the “fifth vital sign,”5 pain is one of the most common chief complaints of patients presenting to the ED.6,7 Its management has been targeted as an area for quality improvement since the mid 1990s, with several organizations, including the Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations), the IOM, the American Pain Society (APS), and the Agency for Healthcare Research and Quality (AHRQ), issuing standards regarding timely, tailored, and adequate pain assessment and treatment.8–15 These guidelines acknowledge that patients have a right to appropriate assessment and effective pain treatment; they underscore the importance of effective pain management as a component of quality patient care. The process of pain care standards require that pain be assessed in patients; that the assessments be recorded such that there is regular re-assessment and follow-up; that policies be established to support appropriate ordering and administration of analgesic medication; and that data be collected to monitor the appropriateness and effectiveness of pain management.16 For these reasons, pain care can function as a surrogate for studying a facet of how ED care is delivered. Two previous studies indicate that ED crowding negatively impacts pain care.17,18 These studies, however, were limited in their scope of evaluating the effects of crowding on general ED pain care, as they only evaluated patients with hip fractures or with severe pain scores.
The objective of this study was to evaluate the association of three ED crowding factors with the quality of general ED pain care. We evaluated the effect of ED census (all patients in the ED), number of boarders (number of admitted patients waiting for inpatient beds), and boarding burden (number of boarders divided by ED census) on process of pain care measures for adult patients seen in an ED.
METHODS
Study Design
This was a retrospective observational study of adult patients (≥18 years) with conditions warranting pain care seen at an urban, academic, tertiary care ED. To capture seasonal variation with ED crowding, all visits during the months of July and December 2005 (i.e., 7/1/05–7/31/05 and 12/1/05–12/31/05) were reviewed. This study was approved by the Institutional Review Board at the study site.
Study Setting and Population
Conditions involving the abdomen, extremities, back, falls, headache, flank, and neck were included. Patients with complaints of chest pain were excluded as these patients typically follow a cardiac algorithm for patient care instead of pain care at ED presentation. Adult patients (≥18 years age) presenting with a chief complaint of pain and a final primary ED diagnosis of a painful condition were reviewed for their ED pain care using a text filter algorithm. The filter included word such as “ache”, “appendicitis”, “appendix”, “arthritis”, “biliary”, “burn”, “cancer”, “cholangitis”, “cholecystitis”, “colic”, “contusion”, “Crohn”, “diverticulitis”, “fall”, “fell”, “fracture”, “fx”, “gout”, “hernia”, “injury”, “kidney”, “laceration”, “ligament”, “-lithiasis”, “meniscus”, “obstruction”, “pain”, “pancreatitis”, “perforation”, “problem”, “pyelonephritis”, “sprain”, “stone”, “strain”, “tear”, and “tendon.” An example of a patient included in the study would be: chief complaint of toe pain, and final ED diagnosis of gout. An example of a patient excluded from the study would be: chief complaint of toe pain, final ED diagnosis of congestive heart failure.
Study Protocol
The ED utilizes a comprehensive electronic medical record (ED Pulsecheck, PICIS Inc., Wakefield, MA) for patient tracking, physician and nurse documentation, and order entry. All data entered into the system are time stamped. Patient-related, clinician, and pain care data were collected from the medical record; ED crowding data were collected from hospital administrative registration data.
All data were extracted by the investigators or by trained research assistants (RAs) following 12 recommended criteria for medical record review studies.22 All RAs had at minimum a 4-hour training session of the ED medical record abstraction process, with shadowing of a minimum of 10 record reviews. A minimum of 20 test abstractions were independently completed by each RA and then compared with that of the investigator. The RA was qualified to abstract independently when test abstractions were completed with 95% agreement. Performance on abstractions was monitored, with 10% of the charts randomly reviewed by the investigator. There was 100% inter-observer reliability of the variables of interest for this cohort.
Measurements
The predictor variables of interest were ED census (total number of patients in the adult ED), number of admitted patients waiting for beds (boarders), and number of boarders divided by ED census (boarding burden). These crowding data were collected of the first hour of the index patient’s ED visit using hospital administrative registration data (Cerner, Inc., Kansas City, MO). Continuous crowding variables were also dichotomized at median levels to represent HIGH (>50th percentile) and LOW (≤50th percentile) values. Covariates for multivariable analyses included patient-related demographic and clinical information (age (older age (≥65 years)), gender, race/ethnicity, triage score using the Emergency Severity Index (ESI),19 Charlson comorbidity score (this is a widely used weighted-prognostic score that classifies comorbid conditions associated with risk of mortality; the higher the score, the greater the risk),20 number of prior medications, if the patient reported considerable pain (>5 out of 10), time of day (8:00-15:59, 16:00-11:59, 12:00-7:59), and clinician discipline (emergency medicine physicians [EPs], emergency medicine [EM] physician assistants, and non-EM physicians [e.g., Internal Medicine]).
The pain care outcomes studied were process of pain care measures: documentation of nurse and physician pain assessment, medications ordered, and times of activities (e.g., arrival, assessment, ordering, and administration of medication). For this ED, patients are assessed for pain using an 11-point verbal numeric scale for pain (0 = none, 10 = most severe pain). Because disparities have recently been documented in use of opioid analgesia,21 we also investigated types of analgesic medication used. We defined “analgesic medication” as any oral, intravenous or intramuscular opioid (e.g., morphine, fentanyl, hydromorphone, oxycodone, propoxyphene, codeine) or non-opioid, (e.g., acetaminophen, ibuprofen, ketorolac, tramadol); for patients with abdominal pain we also included antacids (e.g., magnesium hydroxide, famotidine, lansoprazole) as a non-opioid analgesic medication. Combination opioid and non-opioid medications (e.g. acetaminophen/oxycodone) were classified as opioids. Aspirin, ibuprofen, indomethacin, naproxen, ketorolac, and any COX-2 inhibitors were subcategorized as Non-Steroidal Anti-Inflammatory Drugs (NSAIDs). Times to pain care were defined as the first recorded time of documented pain assessment, or time the first analgesic medication was ordered, or the time the first analgesic medication was administered minus the time of arrival (triage or registration time, which ever came first).
Because process of pain care measures were studied, any missing data for outcomes were treated as not having that outcome done or given. (e.g., if a pain score was absent from the chart or missing, this was regarded as not having any documented pain assessment.).
Data Analysis
Univariate descriptive statistics of patient and clinician characteristics and ED crowding factors were completed. Bivariate analyses using chi-square, t-test, Spearman’s correlation coefficient, and linear and logistic regression evaluated ED crowding factors against pain care received. Analyses are presented for predictor and outcome variables dichotomized at median levels (50th percentile). Multivariable logistic and generalized linear regression models were created for adjusted pain outcomes using covariates that were significant (p < 0.10) or of construct validity. Generalized Estimating Equations (GEE),23 were used to account for the possible correlation of observations within EP. GEE can be used with models that, except for the correlation, can be modeled using generalized linear models.
The “time to care” dependent variables were log-transformed to achieve homoscedasticity and normality of residuals. The procedure described by Afifi, et al.24 was followed to estimate the effect of the independent variables in the original scale (i.e. minutes). For each observation, the predicted value was adjusted using a smear factor that corrects for the bias caused by the non-normality of the original distribution of the outcome prior to exponentiation. Bootstrapping was used to estimate the empirical 95% confidence limits for the effects.
All analyses were completed in SAS (version 9.1, Cary, NC) Results presented, whether using variables as continuous, categorical, or dichotomous, are those producing best-fit models (c-statistic and convergence model fit statistics) and demonstrating evidence of data fit via Hosmer-Lemeshow goodness of fit.
RESULTS
During the study period, a total of 9,149 adult patients were seen in this ED; 1,068 (12%) of these patients met study inclusion criteria and their ED medical records were reviewed for pain care. Study population characteristics are presented in Table 1, and ED and clinician characteristics are presented in Table 2.
Table 1.
Patient characteristics
| Characteristics | N=1,068 |
|---|---|
| Age, mean (SD), years | 47 (19) |
| Female | 65% |
| Race/Ethnicity | |
| White | 24% |
| African American | 31 |
| Hispanic | 36 |
| Other | 9 |
| Emergency severity index [1=acute, 5=non-acute], mean (SD) | 3.2 (0.6) |
| ED medical conditions of pain involving | |
| Abdomen | 44.9% |
| Extremity | 18.5 |
| Back | 15.5 |
| Fall | 7.4 |
| Headache | 6.7 |
| Neck | 3.7 |
| Flank | 3.3 |
| Charlson comorbidity score, | |
| mean (SD) | 0.86 (1.72) |
| median (quartiles) | 0 (0, 1) |
| Number of prior medications, | |
| mean (SD) | 2.38 (3.10) |
| median (quartiles) | 1 (0, 3) |
| Levels of reported pain at triage [0-none, 10-severe] | |
| 9–10 [Severe] | 20% |
| 6–9 [Moderate-severe] | 22 |
| 3–5 [Moderate] | 23 |
| 1–2 [Mild] | 26 |
| None | 9 |
SD = standard deviation
Table 2.
Emergency department (ED) characteristics
| Square footage area | 4500 sq. ft. |
| Bed capacity | 41 beds |
| Census, median (quartiles) | 54 (40, 68) |
| Number of admitted patients waiting for beds (boarders), median (quartiles) | 18 (11, 25) |
| Number of clinicians evaluating study patients | 122 |
| mean number of patients seen by individual clinician (SD) | 8.78 (9.73) |
| median number of patients (quartiles) | 6 (2, 12) |
| Clinician discipline | |
| EM physicians | 53% |
| EM physician assistant | 13% |
| Non-EM physicians | 34% |
SD = standard deviation; EM = emergency medicine
There were no significant differences in physician documentation of pain assessment or pain follow-up during periods of HIGH (>50th percentile) census when compared to LOW (≤50th percentile) census. There were also no differences in pain assessment during periods of HIGH and LOW number of boarders or non-boarders. A significant difference in pain treatment was found when comparing the effects of HIGH and LOW ED crowding factors. Fewer patients received any form of analgesic medication, and in particular NSAIDs, during periods of high census and high number of boarders. Boarding burden, however, had no significant impact on any pain care outcomes. These differences are quantified in Table 3, and remained statistically significant with multivariable analyses that accounted for clinician clustering effects and adjusted for patients' older age, gender, race/ethnicity, emergency severity index (ESI) score, comorbidity, number of prior medications, if the patient reported considerable pain at triage (>5 out of 10), time of day, and clinician type. Additional analyses adjusting for physician type (emergency vs. non-emergency) also found no differences in pain care outcomes, and thus were not included in analyses because multivariable methods accounting for clustering were used instead.
Table 3.
Pain care received (N=1,068)
| Percent of patients receiving: | Total ED Census | # Boarders | Boarding burden | ||||||
|---|---|---|---|---|---|---|---|---|---|
| LOW ≤50 percentile | HIGH >50 percentile | * Parameter estimate when HIGH (95% CI) | LOW ≤50 percentile | HIGH >50 percentile | * Parameter estimate when HIGH (95% CI) | LOW ≤50 percentile | HIGH >50 percentile | * Parameter estimate when HIGH (95% CI) | |
| Nursing documented pain assessment (n=1,068) | 86% | 87% | 0.11 (−0.24, 0.47) |
87% | 86% | 0.07 (−0.32, 0.46) |
86% | 87% | 0.11 (−0.22, 0.44) |
| Physician documented pain assessment (n=1,068) | 86 | 85 | −0.17 (−0.54, 0.19) |
86 | 85 | −0.15 (−0.53, 0.22) |
86 | 85 | −0.13 (−0.57, 0.30) |
| Any documented pain assessment (n=1,068) | 90 | 90 | 0.05 (−0.35, 0.45) |
90 | 89 | −0.05 (−0.51, 0.42) |
90 | 90 | 0.07 (−0.30, 0.45) |
| Follow-up pain assessment by MD, RN, or PA (n=961) | 50 | 47 | −0.27 (−0.59, 0.05) |
49 | 48 | −0.08 (−0.40, 0.23) |
47 | 50 | 0.08 (−0.20, 0.35) |
| Received analgesic medication (n=642) | 65 | 55 | −0.47 (−0.80, −0.07) |
65 | 54 | −0.36 (−0.70, −0.01) |
61 | 59 | −0.03 (−0.31, 0.26) |
| No ED analgesic medication, received prescription at discharge (n=117) | 25 | 29 | 0.00 (−0.55, 0.55) |
28 | 27 | 0.07 (−0.47, 0.61) |
32 | 28 | 0.30 (−0.22, 0.82) |
| Received opioid (n=358) | 35 | 32 | −0.36 (−0.73, 0.02) |
35 | 32 | −0.27 (−0.62, 0.09) |
32 | 35 | 0.11 (−0.17, 0.39) |
| Received NSAID (n=148) | 18 | 10 | −0.40 (−0.75,−0.05) |
17 | 10 | −0.43 (−0.86, −0.01) |
16 | 12 | 0.40 (−0.81, 0.01) |
[*adjusted for older age (≥65 years), gender, race/ethnicity, ESI, comorbidity, number of prior medications, considerable pain (>5 out of 10), time of day (8:00-15:59, 16:00-11:59, 12:00-7:59), and clinician type]
NSAID = non-steroidal anti-inflammatory drug; ESI = emergency severity index; ED = emergency department; CI = confidence interval; MD = physician; RN = registered nurse; PA = physician assistant
Emergency department census was directly associated with delays to patient care. There was a direct correlation with total ED census and increased time to pain assessment (Spearman r = 0.33, p < 0.0001), time to analgesic medication ordering (r = 0.22, p < 0.0001), and time to analgesic medication administration (r = 0.25, p < 0.0001). There were significant delays of over 1-hour for pain assessment, the ordering of analgesic medication, and the administration of analgesic medication during periods of HIGH ED census, HIGH number of boarders, but not for HIGH boarding burden. When adjusted for patient characteristics, time of day, and clinician type in multivariate analyses accounting for clinician clustering effects, patients seen during periods of HIGH census took up to 55 minutes longer to have documented pain assessment, and up to 43 minutes longer to have their administration of analgesic medication (Table 4).
Table 4.
Times of pain care activities
| Total ED Census | # Boarders | Boarding burden | |||||||
|---|---|---|---|---|---|---|---|---|---|
| LOW ≤50 percentile mean time in minutes | HIGH >50 percentile mean time in minutes | * Parameter estimate in minutes (95% CI) | LOW ≤50 percentile mean time in minutes | HIGH >50 percentile mean time in minutes | * Parameter estimate in minutes (95% CI | LOW ≤50 percentile mean time in minutes | HIGH >50 percentile mean time in minutes | * Parameter estimate in minutes (95% CI) | |
| Time to 1st clinician pain assessment (number of patients) | 106 (455) |
174 (444) |
55 (40, 74) |
112 (500) |
174 (399) |
43 (27, 64) |
125 (435) |
153 (464) |
16 (−3, 31) |
| Time to 1st analgesic medication ordering (number of patients) | 104 (326) |
136 (278) |
41 (23, 64) |
107 (360) |
136 (244) |
33 (15, 57) |
112 (295) |
125 (309) |
12 (−7, 30) |
| Time to 1st analgesic medication administration (number of patients) | 125 (316) |
167 (274) |
43 (25, 67) |
130 (351) |
167 (239) |
35 (16, 59) |
137 (125) |
152 (299) |
16 (−3, 37) |
[*adjusted for older age (≥65 years), gender, race/ethnicity, ESI, comorbidity, number of prior medications, considerable pain (>5 out of 10), time of day (8:00-15:59, 16:00-11:59, 12:00-7:59), and clinician type]
ESI = emergency severity index; ED = emergency department; CI = confidence interval
DISCUSSION
Although pain is one of the most common complaints in the ED,7 and surveys of ED patients indicate they expect complete relief of their pain,25 this study demonstrates barriers such as ED crowding may preclude the ability to achieve this expectation.26 In our study, we found high ED census negatively impacts the quality of pain care delivered, and results in significant delays to pain care. Periods of greater ED patient volume (high levels of crowding) result in untreated pain and delays to documented pain care. The very nature of ED patient evaluation, with multi-tasking and frequent interruptions27,28 coupled with ED crowding, in fact contributes to poorer pain management.
It is known that patient-related factors, such as age, gender, and race/ethnicity, are associated with disparities in pain care.21,29–34 In multivariable analyses, when adjusting for these covariates and controlling for ESI (triage score), comorbid illness, number of concurrent medications, complaints of considerable pain, time of day, and type of clinician (by training), ED census and number of boarders continued to remain independently associated with patients not receiving any analgesic medication for their pain, and significant delays to pain assessment, analgesic medication ordering, and analgesic medication administration (Table 3 and Table 4).
There are two primary implications of this study. The first is that ED crowding (whether as census or number of boarders) is a condition that negatively impacts the quality of patient care delivery and results in negative outcomes. ED crowding prolongs times to documented pain care and increases the risk of not being treated at all. Increased pain and suffering, while not directly measured in this study, is a logical consequence of not receiving analgesia or delays in pain care. Using the findings of this study, and a conceptual model of ED crowding that partitions the phenomenon into “three interdependent components: input, throughput, and output,”35 researchers, administrators, and policymakers may be guided to develop solutions focused on decreasing the number of patients in the ED to improve the quality of pain care. Although targeting patients who come to the ED (input) may not be reasonable or even feasible, efforts to facilitate the processing of patients while in the ED (throughput), and increase the number of patients being discharged from the ED (output), whether for outpatient follow-up or to inpatient beds, may be better options that should be considered and emphasized.
A second implication of this study is the impact of ED census and number of boarders as ED crowding factors. Patient volume, whether measured as total number of patients in the ED or only boarders, has a significant effect on quality of pain care delivered. The association of these with not receiving any analgesic medication and with delays to care demonstrates that volume is an important factor and contributor to what ED crowding is as an entity and phenomenon if conceptualized for its outcomes.
Previous studies have been hindered by difficulties in studying the effects of ED crowding because of the lack of a gold standard measure. All of the previously mentioned studies evaluating “ED crowding” or factors associated with crowding used different definitions or even outcomes of crowding itself (i.e., diversion) to study its impact.36 Although many ED crowding definitions have been proposed, most consisting of multi-component measures incorporating similar factors in their characterization of crowding,37–41 they have been derived using ED clinician perception of crowding rather than patient-centered outcomes. Very few studies have evaluated the effects of summative crowding measures or their individual component factors, such as ED census or number of boarders, on the quality of patient care delivered.
It is interesting to note that this study also demonstrates the similar effects of ED census and boarding patients, but not boarding burden (number of boarders divided by ED census) on the patient-centered outcome of pain care. As future attempts to approach a working definition of ED crowding develop, this study provides some insight as to how patient volume and its components are associated with patient centered outcomes and delays to care.
LIMITATIONS
Limitations of this study include the single-institution setting, and our findings may not generalize to other settings. Future multi-center studies evaluating ED crowding factors and their impact on pain care are needed to validate these findings. Other potential limitations of this study are the retrospective study design and awareness of the study hypotheses by a minority of the abstractors. The investigators attempted to diminish expected biases by following 12 recommended criteria for medical record review studies.22 It is possible that patients may have had their pain treated, yet not documented. For this ED, however, this is highly unlikely because all medications used are recorded by an automated medication dispensing device. Actual patient pain levels and the degree of success in relieving pain may not be accurate or precise. The data for this study were also based on time stamps of when documentation or ordering of medications were entered into the computer system, not the actual time of the events. These outcomes, however, were not the focus of this investigation. Instead, process of pain care measures were evaluated for this study. Documentation of patient care is necessary to demonstrate and guide clinical care of a condition, in this case pain. Just as documentation of glucose levels or blood pressure readings is needed to guide the management of uncontrolled hyperglycemia or hypertension, pain assessment scores and response to pain treatment also require documentation for proper management. Documentation of care by clinical staff via medical record data abstraction is the most commonly used method of measuring quality.22,42,43 The National Committee on Quality Assurance, the Health Care Financing Administration, and local and national providers routinely use abstraction of data from medical records to assess the quality of inpatient care.42 Finally, while it would have been ideal to evaluate data from the entire year, only the months of July and December were reviewed for this study. There may be limitations of the data secondary to holidays or residency patterns. It was our opinion, however, that differences in ED patient volume traditionally accompany summer (troughs) and winter (peaks) months were more critical to evaluate.
CONCLUSIONS
Emergency department crowding as measured by patient volume was associated with poorer pain care. Higher numbers of patients in the ED, whether as total census or number of admitted patients waiting for inpatient beds (boarders), were associated with not receiving analgesic medication and delays to documented pain care.
ACKNOWLEDGMENTS
The authors would like to thank Susan Holland and Amy Kossoy for their assistance with data abstraction.
Funding and support: This study was supported by a Jahnigen Career Development Award administered by the American Geriatrics Society (Ula Hwang) and a Mid-Career Investigator Award in Patient-Oriented Research (K24 AG022345) from the National Institute on Aging (R. Sean Morrison).
Footnotes
Presentations: None
REFERENCES
- 1.Institute of Medicine. [Accessed Jun 29, 2007];Hospital-based emergency care: at the breaking point. 2006 Available at: http://www.nap.edu/catalog/11621.html#toc.
- 2.Schull M, Vermeulen M, Slaughter G, Morrison L, Daly P. Emergency department crowding and thrombolysis delays in acute myocardial infarction. Ann Emerg Med. 2004;44:577–585. doi: 10.1016/j.annemergmed.2004.05.004. [DOI] [PubMed] [Google Scholar]
- 3.Pines J, Hollander J, Localio R, Metlay JP. The association between emergency department crowding and hospital performance on antibiotic timing for pneumonia and percutaneous intervention for myocardial infarction. Acad Emerg Med. 2006;13:873–878. doi: 10.1197/j.aem.2006.03.568. [DOI] [PubMed] [Google Scholar]
- 4.Fishman P, Shofer F, Robey J, et al. The impact of trauma activations on the care of emergency department patients with potential acute coronary syndromes. Ann Emerg Med. 2006;48:347–354. doi: 10.1016/j.annemergmed.2006.02.021. [DOI] [PubMed] [Google Scholar]
- 5.Lanser P, Gesell S. Pain management: the fifth vital sign. Healthc Benchmarks. 2001;8:68–70. [PubMed] [Google Scholar]
- 6.McCaig L, Nawar E. National Hospital Ambulatory Medical Care Survey: 2004 Emergency Department Summary. Adv Data. 2006;372:1–29. [PubMed] [Google Scholar]
- 7.Burt C, McCaig LF. Trends in hospital emergency department utilization: United States, 1992–1999. Adv Data. 2001;13:1–34. [PubMed] [Google Scholar]
- 8.Gordon D, Dahl J, Miaskowski C, et al. American Pain Society recommendations for improving the quality of acute and cancer pain management. Arch Int Med. 2005;165:1574–1580. doi: 10.1001/archinte.165.14.1574. [DOI] [PubMed] [Google Scholar]
- 9.Joint Commission. [Accessed Jan 11, 2007];Setting the Standard for Quality in Health Care. Available at: http://www.jointcommission.org/NR/rdonlyres/6C33FEDB-BB50-4CEE-950B-A6246DA4911E/ 0/setting_the_standard.pdf.
- 10.Institute of Medicine. Crossing the Quality Chasm. Washington, DC: National Academy Press; 2001. p. 50. [Google Scholar]
- 11.American Pain Society Quality of Care Committee. Quality improvement guidelines for the treatment of acute pain and cancer pain. JAMA. 1995;274:1874–1880. doi: 10.1001/jama.1995.03530230060032. [DOI] [PubMed] [Google Scholar]
- 12.Brownfield E. Making Health Care Safer: A Critical Analysis of Patient Safety Practices. Rockville, MD: Agency for Healthcare Research and Quality; 2001. [Accessed Jul 30, 2008]. Pain Management. Contract No.290-97-0013, Available at http://www.ahrq.gov/clinic/ptsafety/pdf/chap37.pdf. [Google Scholar]
- 13.Acute pain management: operative or medical procedures and trauma. Part 1. Agency for Health Care Policy and Research. Clin Pharm. 1992;11:309–331. [PubMed] [Google Scholar]
- 14.Acute pain management: operative or medical procedures and trauma, Part 2. Agency for Health Care Policy and Research. Clin Pharm. 1992;11:391–414. [PubMed] [Google Scholar]
- 15.Jaycox A, Carr D, Payne R, et al. Management of cancer pain. Clinical practice guideline. Agency for Health Care Policy and Research; 1994. No. 9. Vol AHCPR publication 94-0592. [PubMed] [Google Scholar]
- 16.Dahl J. Medical Student-JAMA: Improving the practice of pain management. JAMA. 2000;284:2785. [PubMed] [Google Scholar]
- 17.Hwang U, Richardson LD, Sonuyi TO, Morrison RS. The Effect of Emergency Department Crowding on the Management of Pain in Older Adults with Hip Fracture. J Amer Geriatr Soc. 2006;54:270–275. doi: 10.1111/j.1532-5415.2005.00587.x. [DOI] [PubMed] [Google Scholar]
- 18.Pines J, Hollander J. Emergency department crowding is associated with poor pain care for patients with severe pain. Ann Emerg Med. 2008;51:1–5. doi: 10.1016/j.annemergmed.2007.07.008. [DOI] [PubMed] [Google Scholar]
- 19.Wuerz R, Milne L, Eitel D, Travers D, Gilboy N. Reliability and validity of a new five-level triage instrument. Acad Emerg Med. 2000;7:236–242. doi: 10.1111/j.1553-2712.2000.tb01066.x. [DOI] [PubMed] [Google Scholar]
- 20.Charlson M, Pompei P, Ales K, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 21.Pletcher M, Kertesz S, Kohn M, Gonzales R. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA. 2008;299:70–78. doi: 10.1001/jama.2007.64. [DOI] [PubMed] [Google Scholar]
- 22.Worster A, Bledsoe R, Cleve P, Fernandes CM, Upadhye S, Eva K. Reassessing the methods of medical record review studies in emergency medicine research. Ann Emerg Med. 2005;45:448–451. doi: 10.1016/j.annemergmed.2004.11.021. [DOI] [PubMed] [Google Scholar]
- 23.Liang K, Zeger S. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22. [Google Scholar]
- 24.Afifi A, Lotlerman J, Ettner S, Cowan M. Methods fro improving regression analysis for skewed continuous or counted responses. Ann Rev Pub Health. 2007;28:95–111. doi: 10.1146/annurev.publhealth.28.082206.094100. [DOI] [PubMed] [Google Scholar]
- 25.Fosnocht D, Heaps N, Swanson E. Patient expectations for pain relief in the ED. Am J Emerg Med. 2004;22:286–288. doi: 10.1016/j.ajem.2004.04.011. [DOI] [PubMed] [Google Scholar]
- 26.Fosnocht D, Swanson E, Barton E. Changing attitudes about pain and pain control in emergency medicine. Emerg Med Clin N Amer. 2005;23:297–306. doi: 10.1016/j.emc.2004.12.003. [DOI] [PubMed] [Google Scholar]
- 27.Chisholm C, Collison E, Nelson D, Cordell WH. Emergency department workplace interruptions: are emergency physicians "interrupt-driven" and "multitasking"? Acad Emerg Med. 2000;7:1239–1243. doi: 10.1111/j.1553-2712.2000.tb00469.x. [DOI] [PubMed] [Google Scholar]
- 28.Manias E, Botti M, Bucknall T. Observation of pain assessment and management -- the complexities of clinical practice. J Clin Nurs. 2002;11:724–733. doi: 10.1046/j.1365-2702.2002.00691.x. [DOI] [PubMed] [Google Scholar]
- 29.Bernabei R, Gambassi G, Lapane K, et al. Management of pain in elderly patients with cancer. SAGE study group, Systematic Assessment of Geriatric Drug use via Epidemiology. JAMA. 1998;279:1877–1882. doi: 10.1001/jama.279.23.1877. [DOI] [PubMed] [Google Scholar]
- 30.Brown J, Klein E, Lewis C, Johnston BD, Cummings P. Emergency department analgesia for fracture pain. Ann Emerg Med. 2003;42:197–205. doi: 10.1067/mem.2003.275. [DOI] [PubMed] [Google Scholar]
- 31.Cleeland C, Gonin R, Hatfield A, et al. Pain and its treatment in outpatients with metastatic cancer. N Eng J Med. 1994;330:592–596. doi: 10.1056/NEJM199403033300902. [DOI] [PubMed] [Google Scholar]
- 32.Cleeland C, Gonin R, Hatfield A, et al. Pain and treatment of pain in minority patients with cancer. The Eastern Cooperative Oncology Group Minority Outpatient Pain Study. Ann Intern Med. 1997;127:813–816. doi: 10.7326/0003-4819-127-9-199711010-00006. [DOI] [PubMed] [Google Scholar]
- 33.Todd K, Lee R, Hoffman J. The effect of ethnicity on physician estimates of pain severity in patients with isolated extremity trauma. JAMA. 1994;271:925–928. [PubMed] [Google Scholar]
- 34.Todd K, Samaroo N, Hoffman J. Ethnicity as a risk factor for inadequate emergency department analgesia. JAMA. 1993;269:1537–1539. [PubMed] [Google Scholar]
- 35.Asplin B, Magid D, Rhodes K, Solberg LI, Lurie N, Camargo CA., Jr A conceptual model of emergency department crowding. Ann Emerg Med. 2003;42:173–180. doi: 10.1067/mem.2003.302. [DOI] [PubMed] [Google Scholar]
- 36.Hwang U, Concato J. Care in the emergency department: how crowded is overcrowded? Acad Emerg Med. 2004;11:1097–1101. doi: 10.1197/j.aem.2004.07.004. [DOI] [PubMed] [Google Scholar]
- 37.Hoot N, Zhou C, Jones I, Aronsky D. Measuring and forecasting emergency department crowding in real time. Ann Emerg Med. 2007;49:747–755. doi: 10.1016/j.annemergmed.2007.01.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Epstein S, Tian L. Development of an emergency department work score to predict ambulance diversion. Acad Emerg Med. 2006;13:421–426. doi: 10.1197/j.aem.2005.11.081. [DOI] [PubMed] [Google Scholar]
- 39.Bernstein S, Verghese V, Leung W, Lunney AT, Perez I. Development and validation of a new index to measure emergency department crowding. Acad Emerg Med. 2003;10:938–942. doi: 10.1111/j.1553-2712.2003.tb00647.x. [DOI] [PubMed] [Google Scholar]
- 40.Weiss S, Derlet R, Arhndahl J, et al. Estimating the degree of emergency department overcrowding in academic medical centers: results of the National ED Overcrowding Study (NEDOCS) Acad Emerg Med. 2004;11:38–50. doi: 10.1197/j.aem.2003.07.017. [DOI] [PubMed] [Google Scholar]
- 41.Reeder T, Garrison H. When the safety net is unsafe: real time assessment of the overcrowded emergency department. Acad Emerg Med. 2001;8:1070–1074. doi: 10.1111/j.1553-2712.2001.tb01117.x. [DOI] [PubMed] [Google Scholar]
- 42.Luck J, Peabody J, Dresselhaus T, Lee M, Glassman P. How well does chart abstraction measure quality? A prospective comparison of standardized patients with the medical record. Am J Med. 2000;108:642–649. doi: 10.1016/s0002-9343(00)00363-6. [DOI] [PubMed] [Google Scholar]
- 43.Gilbert E, Lowenstein S, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med. 1996;27:305–308. doi: 10.1016/s0196-0644(96)70264-0. [DOI] [PubMed] [Google Scholar]