Ethical dilemmas in a randomized trial of asthma treatment: can Bayesian statistical analysis explain the results?

OBJECTIVES: The original objective was to determine whether the use of bilevel positive airway pressure (BiPAP) ventilation would reduce the need for endotracheal intubation, the length of hospital stay, and hospital charges in patients with status asthmaticus. The development of physician treatment bias made patient enrollment difficult. The article subsequently describes the use of Bayesian statistics to explain study results when this bias occurs. METHODS: This study was a prospective, randomized controlled clinical trial conducted over a 34.5-month period at an urban university hospital with an emergency department census of 94,000 annual visits. Patients remaining in status asthmaticus after initial standard treatment with inhaled beta-agonists and steroids were randomized to receive BiPAP ventilation plus standard treatment versus standard treatment alone (non-BiPAP), with intubation for either group as needed. Patients with concurrent cardiac or other pulmonary diseases were excluded. The primary outcome measures were endotracheal intubation rate and length of hospital stay. Secondary outcome measures included vital signs (respiratory rate, pulse rate, blood pressure), changes in expiratory peak flow, changes in pulse oximetry values, and hospital charges. Data were analyzed using Fisher's exact test, Mann-Whitney tests, and Bayesian statistics. For patients enrolled in the study more than once, data analysis was performed on the first enrollment only. RESULTS: Nineteen patients were enrolled in the BiPAP group and 16 patients in the non-BiPAP group. Patients were frequently enrolled more than once and the data from the subsequent enrollments were excluded from the analysis. A marked decrease in enrollment, due to physician treatment bias, led to a premature termination of the study. Demographics showed that the groups were similar in age, sex, initial peak flow rate, and arterial blood gas measurements. There was a 7.3% increase (95% CI = -22 to +45) in the intubation rate in the non-BiPAP group (n = 2) compared with that for the BiPAP group (n = 1). No significant difference was seen in length of hospital stay or hospital charges, although there was a favorable trend toward the BiPAP group. Complications encountered in the BiPAP group included one patient with discomfort associated with the nasal BiPAP mask. Bayesian analysis demonstrated that in order for the collected data to be convincing at the 95% confidence level, the prior conviction among treating physicians that BiPAP was a successful treatment modality would have had to be 98.9%. CONCLUSIONS: In this study, BiPAP appeared to have no deleterious effects in patients with status asthmaticus, with a trend toward decreased endotracheal intubation rate, decreased length of hospital stay, and decreased hospital charges. Although further study with more patients is needed to determine the clinical and statistical significance of this intervention, ethical concerns regarding withholding BiPAP treatment from the patients in the control group forced a premature termination of the study in the authors' institution.

Use of a cardiocerebral-protective drug cocktail prior to countershock in a porcine model of prolonged ventricular fibrillation.

OBJECTIVES: This was the third study in a series exploring whether the use of combination pharmacotherapy with delayed countershock would produce higher rates of return of spontaneous circulation (ROSC) and one-hour survival when compared with standard advanced cardiac life support (ACLS) therapy in prolonged ventricular fibrillation (VF). METHODS: Twenty-four female, mixed-breed, domestic swine (ranging in mass from 22 to 25 kg) were used in this prospective, blinded, randomized, experimental trial. Animals were sedated (ketamine/xylazine), anesthetized (alpha-chloralose), paralyzed (pancuronium), mechanically ventilated with room air, and monitored with electrocardiography, arterial pressure, and Swan-Ganz catheters. VF was induced with a 3 s, 60 Hz, 100 mA transthoracic shock, and remained untreated for 8 min. One minute of basic life support followed (standardized by use of a mechanical device). At 9 min, animals were treated with one of three regimes: Group 1, cardiocerebral-protective cocktail (antioxidant U-74389G (3.0 mg/kg), epinephrine (0.2 mg/kg), lidocaine (1.0 mg/kg), bretylium (5.0 mg/kg), magnesium (2.0 g), and propranolol (1.0 mg)); Group 2, magnesium (2.0 g); and Group 3, standard ACLS. Groups 1 and 2 received drugs at minute nine (first countershock at minute 11), while Group 3 received first countershock at minute nine. Data were analyzed with two-tailed Fisher's exact tests. RESULTS: ROSC was achieved in Group 1, 7/7 (100%); Group 2, 3/9 (33%, P versus Group 1=0.01); and Group 3, 3/8 (38%; P versus Group 1=0.02). One-hour survival was attained in Group 1, 7/7 (100%); Group 2, 3/9 (33%; P versus Group 1=0.01), and Group 3, 1/8 (13%; P versus Group 1=0.001). CONCLUSIONS: Combination pharmacotherapy with a cardiocerebral-protective drug cocktail prior to countershock produced superior rates of ROSC and one-hour survival when compared with singular drug therapy (Group 2) and standard ACLS (Group 3) in this porcine model of prolonged VF.

The utility of routine electrolytes and blood cell counts in patients with chest pain.

The objective of the study was to determine whether the routine screening laboratory tests, electrolytes (E) and hemoglobin and hematocrits (H/H), could have been safely avoided in ED patients presenting with chest pain. Retrospective case series over a 3-month period were reviewed. Acceptable indications for E and H/H test ordering were prospectively developed using American College of Emergency Physicians (ACEP) guidelines. Charts were reviewed to determine whether any indications were present, whether the E and H/H laboratory tests were abnormal, and whether the abnormal test affected treatment or disposition. Patients were excluded if they had an abnormal electrocardiogram (ECG) indicating ischemia or electrolyte abnormality. This resulted in 191 patients meeting inclusion criteria. Sixty-eight patients did not have any criteria for ordering E or H/H. Of these, only one (1.5%) had an abnormal laboratory test. The sensitivity of the guidelines for predicting an abnormal laboratory test was 96% (95 CI, 82 to 100%) and the specificity was 46% (95CI, 38 to 54%). There were no interventions, consultations, or any change in disposition based on E or H/H testing in these patients. Of the 123 patients that met at least one criteria for ordering laboratory tests, 27 (18%) had an abnormal result. It was concluded that practice guidelines may reduce the number of E and H/H tests ordered without missing significant abnormalities. A prospective trial is needed to validate these findings.

Teaching patient empathy: the ED visit program.

OBJECTIVE: Residency programs only are not challenged with developing competent emergency clinicians, but should strive to develop caring, empathetic, and community-minded physicians. An exercise was designed to help residents experience emergency department (ED) visits from the patient's perspective. METHODS: This study occurred in emergency medicine residency program at an urban teaching institution with an annual ED census of 94,000. On the first day of residency orientation, each resident was given a clinical scenario and registered through triage into the ED. Nurses were blinded to the study. The study concluded when the examining physician entered the exam room. Residents were then presented with a simulated bill based on their scenario. Residents completed a survey initially and at six months. Survey ratings were measured using a 100-mm visual analog scale (VAS) (0 = not at all; 100 = a great deal). RESULTS: Twenty-five residents participated over two years. Sixty-four percent had never been an ED patient before. Median length of stay was 139 minutes. This exercise was found to improve resident empathy for patients on initial survey, 66 mm (range 16-71), and at follow-up, 66 mm (range 23-91). Residents found the exercise useful both initially, 50 mm (range 4-86), and at follow-up, 49 mm (range 15-81). Ninety-two percent of the residents thought the goals of the exercise had been met. Residents also stated the study changed their approach to patient care (45 mm, range 4-76) and made them a better physician (49 mm, range 5-80). CONCLUSIONS: The ED visit study enhanced patient empathy within residents and was useful in improving patient care attitude.

Effect of exogenous melatonin on mood and sleep efficiency in emergency medicine residents working night shifts.

OBJECTIVE: To determine whether melatonin taken prior to attempted daytime sleep sessions will improve daytime sleep quality, nighttime sleepiness, and mood state in emergency medicine (EM) residents, changing from daytime to nighttime work schedules. METHODS: A prospective, randomized, double-blind crossover design was used in an urban emergency department. Emergency medicine residents who worked two strings of nights, of at least three nights' duration each, and separated by at least one week of days were eligible. Subjects were randomized to receive either melatonin 1 mg or placebo, 30 to 60 minutes prior to their daytime sleep session, for three consecutive days after each night shift. Crossover to the other agent occurred during their subsequent night shifts. Objective measures of quality of daytime sleep were obtained using the Actigraph 1000. This device measures sleep motion and correlates with sleep efficiency, total sleep time, time in bed, and sleep latency. The Profile of Mood States (POMS) and the Stanford Sleepiness Scale (SSS) were also used to quantify nighttime mood and sleepiness. RESULTS: Among the 19 volunteers studied, there was no difference in sleep efficiency (91.16% vs 90.98%, NS), sleep duration (379.6 min vs 342.7 min, NS), or sleep latency (7.59 min vs 6.80 min, NS), between melatonin and placebo, respectively. In addition, neither the POMS total mood disturbance (5.769 baseline vs 12.212 melatonin vs 5.585 placebo, NS) nor the SSS (1.8846 baseline vs 2.2571 melatonin vs 2.1282 placebo, NS) demonstrated a statistical difference in nighttime mood and sleepiness between melatonin and placebo. CONCLUSIONS: There are no beneficial effects of a 1-mg melatonin dose on sleep quality, alertness, or mood state during night shift work among EM residents.

Effect of state legislation prohibiting denial of emergency department patient claims.

STUDY OBJECTIVE: On July 1, 1996, two Florida state laws were implemented to prevent denial of legitimate patient claims. Our objective was to determine whether the laws have been effective in reducing inappropriate denials as measured by the proportion of claims and charges denied. METHODS: A comprehensive set of claims for in-state emergency physician services from a physician billing company were analyzed for the period January 1996 through June 1997, covering 6 months before and 12 months after the effective date of state legislation. The number of facilities included in the data varied from 55 to 67 (mean 63). Denials were classified into 6 categories by payer type. Gross denials were those claims that were completely not paid by the payer, net denials represented the amount denied after patient payments. Downcoding was not examined in this study. Main outcome measures were the proportion of claims and charges denied before and after July 1, 1996. RESULTS: The classification of relative proportions of primary payers did not change appreciably over the study period. The proportion of denied claims decreased significantly (Kruskal-Wallis P <.001), starting 2 months after implementation. CONCLUSION: After initiation of state legislation, payers continue to inappropriately deny claims, although the number of claims and total charges denied has decreased. In response to this legislation, payers are denying larger claims and patient copayments have increased.

Combination pharmacotherapy with delayed countershock vs standard advanced cardiac life support after prolonged ventricular fibrillation.

OBJECTIVE: To test the hypothesis that combination pharmacotherapy with delayed countershock would produce higher rates of return of spontaneous circulation (ROSC) and one-hour survival when compared with standard Advanced Cardiac Life Support (ACLS) therapy. METHODS: A prospective, block-randomized, blinded, laboratory experiment was conducted in an established swine model of prolonged ventricular fibrillation (VF). Fifty-six female domestic swine were anesthetized, instrumented, and shocked into VF with a bipolar pacing catheter. The VF was untreated for 8 minutes, then basic CPR was done mechanically for 1 minute. At 9 minutes of VF, the animals were randomized to treatment with one of seven therapies: group 1, combination pharmacotherapy with epinephrine (0.20 mg/kg), lidocaine (1.0 mg/kg), bretylium (5.0 mg/kg), propranolol (1.0 mg), and U-74389G (3.0 mg/kg); group 2, epinephrine (0.20 mg/kg); group 3, lidocaine (1.0 mg/kg) and bretylium (5.0 mg/kg); group 4, propranolol (1.0 mg); group 5, U-74389G (3.0 mg/kg); group 6, normal saline solution (volume equal to that for group 1); and group 7, standard ACLS (first countershock at 9 minutes of VF). Initial countershocks for groups 1-6 were given after 11 minutes of VF. Data were analyzed with two-tailed Fisher's exact test, with alpha set at 0.05. RESULTS: Return of spontaneous circulation occurred in group 1 = 8/8 (100%); group 2 = 7/8 (88%); group 3 = 3/8 (38%); group 4 = 3/8 (38%); group 5 = 5/8 (63%); group 6 = 4/8 (50%); and group 7 = 3/8 (38%). One-hour survival occurred in group 1 = 8/8 (100%); group 2 = 5/8 (63%); group 3 = 2/8 (25%); group 4 = 2/8 (25%); group 5 = 3/8 (38%); group 6 = 2/8 (25%); and group 7 = 1/8 (13%). CONCLUSIONS: Combination pharmacotherapy with delayed countershock (group 1) produced significantly higher rates of ROSC (p = 0.03) and one-hour survival (p = 0.001) when compared with standard ACLS in this porcine model of prolonged VF.

Multicenter comparison of two clinical decision rules for the use of radiography in acute, high-risk knee injuries.

STUDY OBJECTIVE: Two separate clinical decision rules, one developed in Ottawa and the other in Pittsburgh, for the use of radiography in acute knee injuries have been previously validated and published. In this study, the rules were prospectively validated and compared in a new set of patients. METHODS: A prospective, blinded, multicenter trial was conducted in the emergency departments of three urban teaching hospitals. A convenience sample of 934 patients with knee pain requiring radiographs was enrolled. A standardized data form was completed for each patient, comprising the 10 clinical variables included in the two rules. Standard knee radiographs were then taken in each patient. The rules were interpreted by the primary investigator on the basis of the data sheet and the final radiologist radiograph reading. RESULTS: In the 745 patients in whom the Pittsburgh rules could be applied there were 91 fractures (12.2%). The use of the Pittsburgh rule missed one fracture, yielding a sensitivity of 99% (95% confidence interval [CI], 94% to 100%); the specificity was 60% (95% CI, 56% to 64%). The Ottawa inclusion criteria were met by 750 patients, with 87 fractures (11.6%). The Ottawa rule missed three fractures, for a sensitivity of 97% (95% CI, 90% to 99%); specificity was 27% (95% CI, 23% to 30%). CONCLUSION: Prospective validation and comparison found the Pittsburgh rule for knee radiographs to be more specific without loss of sensitivity compared with the Ottawa rule.

Correlation between triage nurse and physician ordering of ED tests.

Triage nurse ordering of tests can reduce patient time in the emergency department. This benefit must be balanced with the extra expense and potential morbidity of unnecessary tests. This study examined whether triage nurse ordering of lab tests and X-rays correlated with physician ordering. In Phase I, a prospective study, triage nurses completed an order form for each patient triaged during randomly selected 2-hour blocks over a 3-week period. In Phase II, Phase I was repeated with test ordering guidelines derived from Phase I results. Correlations of nurse and physician orders were analyzed by kappa statistics and paired t tests. In Phase I, 176 patients were enrolled of a possible 204 (82%). In Phase II, 350 patients were enrolled of a possible 418 (84%). The percentage of nurse overordering of tests in Phase I was 35%; In Phase II, it was 34% (P NS). Kappa values were determined for blood testing (Phase I, 0.48; Phase II, 0.54 [P = .048]), for X-ray (Phase I, 0.68; Phase II, 0.65 [P = .523]), for urinalysis (Phase I, 0.76; Phase II, 0.67 [P = .338]), and for electrocardiogram (Phase I, 0.78; Phase II, 0.77 [P = .438]). The improved lab correlation in Phase II resulted from the nurses ordering fewer tests (P = .046). Nurses deviated from the test ordering guidelines in 37% of patients. Triage nurse ordering of X-rays, electrocardiogram, and urinalysis has moderate to substantial correlation with physician ordering. Correlation of triage nurse and physician ordering of tests can be improved with the use of test ordering guidelines.

Radiographic interpretation in the emergency department.

This study examined the concordance of radiographic readings between emergency department (ED) attending physicians and radiologists in a community teaching hospital. In addition, the incidents of misinterpretations leading to an alteration in patient care were also reviewed. All radiographs obtained from January through October 1993 were initially interpreted by ED attending physicians with subsequent final review by attending radiology staff. Misread radiographs were placed into one of three categories. The groupings included overread radiographs with no change in treatment, underread radiographs with no change in treatment, and radiograph misinterpretations with a change in treatment. Of 15,585 radiographs obtained during the study period, there were 120 misreads; 12,099 (77.6%) of the 15,585 radiographs had an initial emergency physician interpretation. Radiographic misinterpretations included 7 (5.78%) overreads, 57 (47.1%) underreads, and 57 (47.51%) misreads requiring follow-up (MR-FU). The five most frequently misread radiographs were: abdominal, 12/247 (4.4%); rib, 3/99 (3.0%); foot, 13/621 (2.1%); hip, 3/152 (1.9%); and ankle 11/758 (1.4%). The most frequently obtained radiographs included: chest, 7,012 (0.33% MR-FU); cervical spine, 1,112 (0.18% MR-FU); ankle, 758 (0.66% MR-FU); knee, 633 (0.32% MR-FU); and foot, 621 (0.97% MR-FU). In this study, 99.0% of all emergency department radiographs were read correctly on initial review by ED attending physicians. Of all misread radiographs, less than half (46%) were deemed clinically significant and required a follow-up intervention.

The use of propofol for sedation in the emergency department.

OBJECTIVE: To evaluate the effectiveness of propofol and fentanyl when used by emergency physicians (EPs) for systemic sedation and analgesia in the ED. METHODS: Over an eight-month period, a convenience sample of consenting patients at an urban teaching hospital ED who required sedation for painful procedures were enrolled in a descriptive study of therapeutic propofol sedation. After receiving 2 micrograms/kg of fentanyl i.v. the patients received a continuous infusion of propofol at 0.21 mg/kg/min i.v. to the desired level of sedation. A maintenance infusion of 3-6 mg/kg/hr was administered during the remainder of the procedure. Sedation time, total procedure time, and recovery time were recorded and the presence of side effects was noted. The patients and the physicians rated their satisfaction with the technique on a 100-mm visual analog scale. The patients were contacted at 24 hours for follow-up information. RESULTS: Twenty patients (aged 19-62 years) received propofol infusion for procedures that included ten abscess incision and drainages, eight orthopedic reductions, one chest tube placement, and one wound exploration. The mean (+/- SD) total dose of propofol was 200 +/- 160 mg. The mean time to onset of sedation was 6.6 +/- 3.2 min, the mean procedure interval was 9.6 +/- 6.3 min, and the mean interval until recovery was 6.1 +/- 4.1 min. The median patient satisfaction score was 97 mm (range 40-100 mm); the median rating by physicians was 88 mm (range 4-100 mm). Side effects were noted in six patients: three had pain on injection, one had pain on injection and brief (< 1 min) hypotension (systolic blood pressure = 80 mm Hg), and two had apnea (< 30 sec). Seven patients had some recall of the procedure immediately afterwards, but only five had any recall at 24 hours. All 20 patients stated they would be willing to receive the same sedation again for a future procedure. CONCLUSION: Propofol appears to be an effective agent to provide systemic sedation and analgesia when used in conjunction with fentanyl for procedures in the ED. Given its rapidity of onset and elimination, it may become a useful agent to EPs for i.v. sedation in the ED.

Demand valve ventilation in a swine pneumothorax model.

Bag-valve-mask (BVM) and oxygen-powered demand valve (OPDV) are two available adjuncts for artificial ventilation. Use of OPVDs has been limited by concern for causing or worsening pneumothorax. This study examined the effect of OPDV and BVM ventilation in swine with pneumothorax. This was a randomized, crossover design. Five female swine (16 to 23 kg) were sedated (ketamine/xylazine), anesthetized (alpha chloralose), and intubated. Right chest injury was induced, through an incision, using a 14-gauge needle to produce disruption of underlying lung parenchyma. An 18F Foley catheter was inserted and sealed in the right pleural space at the cut-down site. For each trial, a pneumothorax was induced by instilling 300 mL of air. Each subject underwent 10-minute ventilation trials with OPDV and with BVM by the same investigator, blinded to instrumentation. Post-trial pneumothorax volume, heart rate, blood pressure, end-tidal CO2 (ETCO2), and arterial blood gases were measured. There was no significant difference in pneumothorax volume after OPDV ventilation (310.8 mL) versus BVM ventilation (315.4 mL) (P = .821). There were no significant differences between heart rates, blood pressures, or arterial blood gases, although ETCO2 showed statistically significant decreases from baseline in both OPDV (33.0 to 28.6) and BVM (33.8 to 30.2) trials. However, multiple comparisons showed no significant differences between OPDV and BVM at any time points. In this model, OPDV and BVM ventilation did not differ in their effects on pneumothorax volume or hemodynamic variables. No animal showed signs of tension pneumothorax.

The esophageal detector device: accuracy and reliability in difficult airway settings.

INTRODUCTION: The esophageal detector device (EDD) recently has been found to assess endotracheal (ET) tube placement accurately. This study describes the reliability of the EDD in determining the position of the ET tube in clinical airway situations that are difficult. METHODS: This was a prospective, randomized, single-blinded, controlled laboratory investigation. Two airway managers (an emergency-medicine attending physician and a resident) determined ET-tube placement using the EDD in five swine in respiratory arrest. The ET tube was place in the following clinical airway situations: 1) esophagus; 2) esophagus with 1 liter of air instilled; 3) trachea; 4) trachea with 5 ml/kg water instilled; and 5) right mainstem bronchus. Anatomic location of the tube was verified by thoracotomy of the left side of the chest. RESULTS: There was 100% correlation between the resident and attending physician's use of the EDD. The EDD was 100% accurate in the determining tube placement in the esophagus, in the esophagus with 1 liter of air instilled, in the trachea, and in the right mainstem bronchus. The airway managers were only 80% accurate in detecting tracheal intubations when fluid was present. CONCLUSIONS: The EDD is an accurate and reliable device for detecting ET-tube placement in most clinical situations. Tube placement in fluid-filled trachea, lungs, or both, which occurs in pulmonary edema and drowning, may not be detected using this device.

Right lung ventilation in a porcine open-chest shock model.

OBJECTIVE: To determine the effect of selective right lung ventilation on gas exchange and hemodynamics when compared with bilateral lung ventilation in a porcine open-chest shock model. METHODS: A randomized, controlled laboratory investigation was performed using a static hemorrhagic shock model in 12 adult swine undergoing thoracotomy. The animals were subjected to a fixed 40% circulating blood volume hemorrhage over 20 minutes. Each animal was then assigned to either a tracheal (control) or a right mainstem (experimental) intubation group. Minute ventilation was held constant in both groups and tidal volumes were decreased by 33% in the right mainstem intubation group. Following intubation and left lateral thoracotomy, another 20% fixed-volume hemorrhage was instituted simultaneously with IV crystalloid and whole blood resuscitation for both groups over 30 minutes. Heart rate, blood pressure, and arterial blood gases were measured at 5-minute intervals. RESULTS: There was no significant difference between the control and experimental groups for any of the measured variables, including mean arterial pressure, pH, partial arterial pressure of CO2 (PaCO2), and PaO2, over time. All animals survived the study protocol. CONCLUSION: Selective right lung ventilation has no detrimental effect on gas exchange or hemodynamics when compared with standard bilateral lung ventilation in a porcine open-chest shock model.

Intramuscular ketorolac versus oral ibuprofen in acute musculoskeletal pain.

STUDY OBJECTIVE: To evaluate the efficacy of IM ketorolac versus that of oral ibuprofen in acute musculoskeletal pain. DESIGN: Randomized, prospective, double-blind clinical trial. SETTING: Urban teaching emergency department with an annual census of 43,000. PARTICIPANTS: Convenience sample of 82 patients aged 18 to 70 years with acute musculoskeletal pain due to trauma. INTERVENTIONS: Forty-two subjects each received 60 mg ketorolac by IM injection and ingested a placebo capsule. Forty subjects each ingested 800 mg ibuprofen and received a placebo (saline) IM injection. Pain was evaluated with a 100-mm visual analog scale at baseline and 15, 30, 45, 60, 75, 90, and 120 minutes after dosing. The prevalence of side effects was elicited in each patient. RESULTS: Mean pain scores improved in each group during the course of the study but did not significantly differ between groups at baseline or at any subsequent interval. The numbers of dropouts due to inadequate analgesia and prevalence of side effects in the two groups did not differ significantly. CONCLUSION: IM ketorolac and oral ibuprofen provide comparable analgesia in ED patients with acute musculoskeletal pain.