I completed my emergency medicine residency in 1975. Yes, it was the mesozoic age of emergency medicine. From then until now, I've never used IV fosphenytoin to load a seizure patient. I never understood what all the fuss was about when it came to giving IV phenytoin. I followed the directions, generally gave it about twice as slowly as advised, and never encountered any problems.
Then the new supposedly safer (but much more expensive) version of phenytoin came along, and the brand name IV product, Dilantin disappeared (made by the same company who makes fosphenytoin), apparently because its use was thought to be associated with all manner of adverse reactions (that I apparently had not noticed over the past 25 years). Fortunately, there's been a generic version of phenytoin around for a while, and it didn't matter that the Dilantin brand was no longer available for IV use.
The following absolutely fascinating paper serves to vindicate my practice. It is the second of a two-part study performed at Los Angeles County/USC Medical Center, my alma mater. The first part estimated how long it took to get patients with subtherapeutic levels into the therapeutic range, and compared oral phenytoin, IV phenytoin, and IV fosphenytoin. Oral loading took the longest, averaging 6.4 hours to get patients to a level permitting discharge from the ED (and was associated with one adverse event per patient on average, all minor). IV phenytoin took 1.7 hours on average to produce therapeutic levels (and was associated with 1.93 adverse events per patient), and IV fosphenytoin took 1.3 hours (and was associated with an average 2.13 adverse events per patient).
This paper examined the cost consequences associated with phenytoin loading with the three techniques. One analysis included nursing and physician time, and the other excluded these costs because no physicians or nurses are going to be laid off or go home because one technique is faster than the other.
The bottom line: The costs were $2.83 for oral phenytoin loading, $23.48 for IV phenytoin loading, and $176.79 for IV fosphenytoin loading. Because the time to achieve therapeutic blood levels and the time to discharge were comparable for IV phenytoin and IV fosphenytoin and given that the adverse events also were comparable, why would anyone ever use IV fosphenytoin?
The authors conclude: It is unlikely that intravenous fosphenytoin is justifiable in any setting. How's that for a non-wishy-washy, call-it-as-they-see-it, no-need-for-corroborative-studies, non-equivocating statement?
Yes, everybody knows that fosphenytoin can be given IM when an IV line cannot be established, and for that we are all grateful, but it is clear that routine use of IV fosphenytoin doesn't seem to be justified.
Cost-Effectiveness of Oral Phenytoin, Intravenous Phenytoin, and Intravenous Fosphenytoin in the Emergency Department Rudis MI, et al Ann Emerg Med 2004;43(3):386
METHODS: The authors from Los Angeles County/USC Medical Center analyzed the cost-effectiveness of three strategies for phenytoin loading in epileptic patients with subtherapeutic serum levels presenting to the ED following a recent seizure. The analysis was based on findings in a previous study of 45 patients who were randomized to loading with oral phenytoin, IV phenytoin, or IV fosphenytoin, which found a lower incidence of adverse drug events with oral phenytoin (mean adverse drug events per patient 1.06 vs. 1.93 and 2.13, respectively) but a longer time to safe ED discharge (therapeutic serum level with no adverse drug events precluding discharge; 6.4 hours vs. 1.7 and 1.3 hours, respectively).
RESULTS: In a base case analysis that did not include physician and nursing time, the cost per patient was $2.83 for oral phenytoin, $23.48 for IV phenytoin, and $176.79 for IV fosphenytoin. In a second analysis that included physician and nurse time, the per-patient costs were $40.06 for oral phenytoin, $84.31 for IV phenytoin, and $238.67 for IV fosphenytoin. In a final analysis that assumed that patients treated with oral phenytoin were managed in a triage or minor care area rather than a monitored bed, oral treatment was associated with a cost savings to the system of $297.17 per patient, while the per-patient costs for IV phenytoin and fosphenytoin remained at $23.48 and $176.79, respectively.
CONCLUSIONS: This study suggests that oral treatment is the most cost-effective approach to phenytoin loading in the ED. If one is willing to pay more in exchange for a more rapid interval to ED discharge, IV phenytoin appears to be the second most cost-effective approach.
Testing for Drugs of Abuse: Low Clinical Yield
I find that some emergency physicians have a very low threshold for obtaining drugs of abuse panels on patients while others rarely order them. I'm in the rarely order category. I view them as largely an unhelpful hassle: you have to get blood (no problem) and urine (often a problem), and it generally takes a long time to get the results. And what are you going to do if the results are positive? Generally nothing other than what you would have done anyway.
Fortunately, there is a wealth of papers confirming the nonvalue of most drug of abuse testing. Granted, there are specific antidotes for opioids (but an opioid OD diagnosis will be made clinically long before a drug test result is back), and there is specific treatment for acetaminophen (but this drug is not included in the traditional drugs of abuse panel), but for the vast majority of other drugs that are found, no specific treatment is needed if the patient is more or less awake and improving over time (the vast majority). To my way of thinking, a drug panel is only needed if the patient is already comatose or is becoming more obtunded rather than less. One of the real dangers I see is waiting for a drug screen to return in an obtunded, drunk trauma patient, and, in the meantime, missing the subdural (which can occur independent of the drug screen results).
The bottom line of this paper: Urine drug screens, at least in Canada, rarely, if ever, result in meaningful changes in therapy.
Screening Urine for Drugs of Abuse in the Emergency Department: Do Test Results Affect Physicians' Patient Care Decisions? Eisen JS, et al Can J Emerg Med 2004;6(2):104
BACKGROUND: Rapid qualitative urine drug screens are often ordered for ED patients in whom drug use is suspected.
METHODS: This study from Queen's University and McMaster University in Canada examined the impact of urine screening for eight drugs of abuse (the Triage Panel for Drugs of Abuse plus Methadone [Biosite Diagnostics]) in adult patients (not involved in vehicular trauma or sexual assault) presenting to one of two academic EDs during a nine-month period in 2001-2002. Drug screening was ordered at the discretion of the managing physicians. Results were immediately reported to a study investigator, who contacted the managing physician to determine his planned management before and after being advised of the drug screen results.
Children and adolescents on bed rest for 24 hours after an LP had a higher rate of complaints than those immediately mobilized
RESULTS: Of 271 patients undergoing urine drug screens during the study period, 110 were enrolled in the study. A review of hospital records suggests that these drug screens were ordered selectively rather than routinely. Upon learning the results of the drug screens, the managing physicians altered their plan of care for four patients (3.6%). Only one of these changes was judged to be substantive based on a priori criteria and on secondary review by an independent expert, none of these management changes was felt to be justified based on the results of the urine drug screen.
CONCLUSIONS: These findings confirm those of other studies that have reported that performing urine drug screens on ED patients seldom leads to significant changes in management.
Bed Rest after LP of No Benefit
ED rituals drive me a little nutty. Like the idea that patients have to be lying flat after their lumbar punctures or else all of their brain juice will drain out, and they'll have a whopping headache. In case you've missed every other study on this topic, here's just another that makes it clear that this practice is without merit, no matter for how many decades it's been done and by how many board certified neurologists.
Strict Bed Rest Following Lumbar Puncture in Children and Adolescents is of No Benefit Ebinger F, et al Neurology 2004;62:1003
BACKGROUND: Post-lumbar puncture (LP) headache has been reported to occur in one-third of adults undergoing this procedure but less frequently in children. Post-LP bed rest for 24 hours was first recommended to prevent headache more than 100 years ago, and continues to be standard practice in many facilities.
METHODS: In this German study, 111 children aged 2-17 undergoing diagnostic LP and without associated conditions that precluded immediate mobilization were randomized to 24 hours of bed rest or immediate mobilization post-LP.
Automated BP measurements at ED triage are often inaccurate
RESULTS: The two groups were similar in age, gender, body mass index, a history of headache before LP, canulla size and bevel orientation, volume of CSF removed and the presence of CSF pleocytosis. In the bed rest group, bed rest was maintained for at least 19 hours by 97 percent and for 24 hours by 73 percent. In the alternate group, 87 percent stood up immediately after LP, and all of the patients did so within 10 minutes. The immediate mobilization group had a statistically lower incidence of post-LP positional headache (2% vs. 15% in the bed rest group, p=0.018), all headaches (21% vs. 39%, p=0.042), and backaches (23% vs. 42%, p=0.031). There were no statistically significant differences between the groups in rates of nausea/vomiting (21% vs. 22%) or neck stiffness (15% vs. 27%).
CONCLUSIONS: Children and adolescents randomized to bed rest for 24 hours after diagnostic LP had a higher rate of post-LP complaints than those randomized to immediate mobilization. The authors cite the possible psychological influence of perceived debility that might be associated with instructions to remain in bed after LP.
Why Measure Blood Pressure?
Blood pressure is measured on every adult who presents to the ED independent of the patient's complaint. Strange. Why would we do a test when there may be no indication for its performance? All other tests have indications, but apparently not ED blood pressures; everybody gets one. There's a box for it on the chart, and the box needs to be filled in. It doesn't matter if the patient has a splinter he wants removed, a sprained ankle, a bleeding hemorrhoid - it doesn't matter. The blood pressure must be taken.
Given that blood pressure is being taken in situations when it clearly is not medically indicated, then we must be doing it for some other reason. Screening for hypertension? Yes, it must be screening for hypertension. The question then is just how accurate is the ED blood pressure if this is the purpose? The answer is not very. All sorts of papers have been published indicating that we use cuffs that are too small, that we don't support the arm, that we don't hold the arm extended at the level of the heart, etc.
The next paper nicely documents just how inaccurate ED blood pressures can be, and the important fact that just because there is a machine to do the test doesn't mean the test is being done better than if a human being took the time to do the test properly. The second important point this paper makes is the fact that blood pressures that are initially elevated often moderate after a period of rest.
I do think it is very important that blood pressure readings that are outside the range of normal (140/90 mm Hg) be rechecked, and if still abnormal, the patient be advised in writing on his aftercare instructions (along with the BP reading) to have his pressure rechecked by his family physician (and the patient should be given a list of physicians if he doesn't have one). It doesn't matter if the patient is a known hypertensive. If the pressure persists above 140/90, it is likely that the patient's blood pressure is inadequately controlled.
Studies have demonstrated that primary care physicians often do not adequately treat patients to reach recognized goal levels, whether it is hypertension or hypercholesterolemia, despite the availability of medication regimens to achieve the needed goals. The fact that patients get referred from the ED because of potential inadequate control may trigger the needed treatment supplementation. (Most hypertensives, for example, will need at least two drugs to achieve a blood pressure below 140/90 mm Hg.)
The Validity of Emergency Department Triage Blood Pressure Measurements Cienki JJ, et al Acad Emerg Med 2004;11(3):237
BACKGROUND: Blood pressure measurement is an integral component of ED triage, and is frequently performed using an automated device. The accuracy of these devices can be affected by both technical and environmental factors.
METHODS: In this prospective study from the University of Miami School of Medicine, adults presenting to an urban university ED underwent simultaneous manual and automated BP measurement at triage, and repeated manual BP measurement after five minutes of rest in a quiet environment. Automated BP measurements were made with the Critikon Dinamap Vital Signs Monitor 8100. Manual BP measurements were performed by carefully trained research assistants proficient in recommended techniques.
RESULTS: All comparisons were available for 104 of 170 patients undergoing automated BP measurement. In the comparison of automated and manual BP measurements at triage, automated measurements consistently overestimated systolic BP (mean difference compared with manual BP measurement, 3.8 mm HG, 95% confidence interval [CI] 2.1-5.4 mm Hg) and underestimated diastolic BP (mean difference [-]6.6 mm Hg, 95% CI [-]7.9 to [-]5.2 mm Hg). The authors observed that the automated device was not always correctly used by ED personnel, that ambient noise was a significant factor, and that triage nurses did not discourage patients from talking and moving during automated BP measurement. In the comparison of manual triage and repeat manual BP measurements, the mean difference between the two scenarios was 12 mm Hg for systolic BP and 9.9 mm Hg for diastolic BP.
New findings support concomitant truncal CT scanning at the time of head CT scanning in multiple trauma patients
CONCLUSIONS: Clinicians should be aware of the substantial potential for inaccuracy of automated BP measurement at the time of ED triage.
Total Body CT Scans in Trauma Fast and Cost-Effective
Finally, a paper extolling the value of the total body CT scan for significant head trauma associated with chest and/or abdominal trauma. This is just one of a growing number of such papers, and I, for one, think there is considerable merit to this practice when performed selectively. We've seen study after study clearly indicating that CT scans are substantially better than plain x-rays for head trauma, cervical trauma, chest trauma, and abdominal trauma. They're better and faster hands down. There is radiation to consider, but the magnitude of the improvement in diagnostic ability associated with CT scanning is such that it seems to clearly outweigh these concerns. And costs? The true costs are nominal - $150 in the current study - a bargain by any standard.
The Benefit of Routine Thoracic, Abdominal, and Pelvic Computed Tomography to Evaluate Trauma Patients with Closed Head Injuries Self ML, et al Am J Surg 2003;186:609
BACKGROUND: Clinical assessment for possible concomitant injuries is difficult in patients with closed head trauma, but there is no consensus regarding the optimal method to screen for truncal injuries in such patients.
METHODS: This study from Methodist Hospital of Dallas retrospectively assessed findings in 457 patients with blunt head trauma prompting cranial CT scanning who, at the discretion of their managing physicians, underwent IV contrast-enhanced thoracic, abdominal, and pelvic CT scanning at the same time.
RESULTS: The additional body scans were positive in 48.5 percent of the patients (222), identified unexpected findings in 38 percent, and prompted a change in management (admission to the hospital or ICU, additional imaging, tube thoracostomy or surgery) in 26 percent. Truncal CT scans were positive in 39 percent of the 170 patients with positive truncal physical examination and 37 percent of those with negative physical exams. In patients under the influence of drugs or alcohol, scans were positive in 23 percent of those with positive physical findings and 26 percent of patients with negative physical findings, and prompted a change in management in 15 percent. In intubated patients (17% of the total group), unexpected truncal CT findings were reported in 46 percent and CT findings prompted a change in management in 10 percent. The total recurring institutional costs of the additional truncal CT scans were estimated to be $150 per examination.
CONCLUSIONS: The authors suggest that their findings support concomitant truncal CT scanning at the time of head CT scanning in multiple trauma patients.
© 2004 Lippincott Williams & Wilkins, Inc.