Status Asthmaticus, Questions

1. When do you use non-invasive positive pressure ventilation (NPPV) in status asthmaticus?

EML Asthma Questions2. Do you start inhaled corticosteroids on asthma patients who are going to be discharged from the ED?

3. When, if ever, do you use ketamine for induction, or for treatment without intubation, in status asthmaticus?

4. When do you use epinephrine in status asthmaticus and when do you avoid it?


Syncope, Answers

  1. In which patients with syncope do you get a NCHCT?

Syncope is defined as a transient loss of consciousness and postural tone. It has a rapid onset, short duration, spontaneous recovery and is due to transient global cerebral hypoperfusion. It may have a prodromal phase. In the Emergency Room, about 3% of patients present with a chief complaint of syncope. As part of the work-up, a non-contrast Head CT-scan (NCHCT) is often ordered. The question is whether such a test is necessary and who should get it?

A 2007 prospective observational study looked at 293 adult ED patients with syncope, of which 113 (39%) underwent head CT and of those, 5 patients (5%) had an abnormal head CT (Grossman 2007). These abnormal findings included 2 subarachnoid hemorrhages, 2 intracranial hemorrhages, and 1 stroke. Each of these patients either had a focal neurologic finding, headache, or signs of trauma. Of the patients who did not have a head CT, none were found to have a new neurologic disease during hospitalization or 30-day follow-up. The results from this study suggested that limiting head CT to patients with neurologic signs or symptoms, trauma above the clavicle, or use of warfarin would potentially reduce scans by over 50%.

Cartoon Faint

Several other small retrospective studies have shown a low yield for obtaining a head CT in syncope patients. In a 2005 study in Emergency Radiology, 128 patients presented to a community hospital with syncope of which 44 received a head CT scan (Giglio 2005). Only 1 CT (2%) showed acute evidence of a posterior circulation infarction. In another retrospective review of patients who got a head CT for syncope alone, none of the 117 patients had CT findings that were clinically related to the syncopal event (Goyal 2006). The authors concluded that a head CT in the absence of focal neurologic findings may not be necessary. It is important to note that the reason for obtaining head CTs in these retrospective studies is unclear. Additionally, syncope patients who did not get a head CT were not analyzed.

A more recent prospective study looked at 254 head CT patients (out of 292 with syncope) and classified them into four groups: 1) normal CT / normal neuro exam; 2) abnormal CT / abnormal neuro exam; 3) abnormal CT not related to their syncope presentation; 4) abnormal CT / normal neuro exam. The last group (abnormal head CT with normal neuro exam) which we are interested in capturing in the ED only included 2 patients (0.7%) (Al-Nsoor 2010).

Bottom Line: For patients presenting to the Emergency Department with a chief complaint of syncope, a NCHCT is of low yield and should only be considered in patients with focal neurologic deficits, complaints of headache, or signs of head trauma. This is consistent with the ACEP clinical policy for syncope, which states that no test should be routinely used in the absence of specific findings on physical exam or history (ACEP 2007).

  1. In which patients with syncope do you get a troponin?

Cardiovascular etiologies are at the top of the differential for dangerous causes of syncope. Identifying those at risk for adverse cardiac outcome after syncope is challenging.  Cardiac markers such as troponins are often sent on patients with syncope as part of a standard work up, presumably as a general screen for cardiac etiology such as acute myocardial infarction (AMI).  As with any test, its results are only relevant if interpreted correctly. To that end, the diagnostic and prognostic utility of troponin in syncope patients is examined here.

AMI is a relatively rare cause of syncope, accounting for approximately 1.4-3.5% of cases (Link 2001, Grossman 2003, Hing 2005).  The diagnosis of AMI in patients without EKG changes on presentation is even less common, likely due to the significant infarct required to induce either a non-perfusing dysrrhythmia or severely impaired cardiac output, which manifests in syncope. The utility of ED troponins for diagnosis of AMI is quite limited, especially in the patient without EKG changes.  One prospective cohort study evaluating troponin-I at 12 hours post-syncope for diagnosis of AMI diagnosed four AMIs, 1.5% of their 289 patients (Reed 2010).   Of these four patients, all had ischemic changes on their presenting EKG.

A larger prospective cohort study of 1474 patients found a 3.1% incidence of AMI within 30 days of a syncopal event (McDermott 2009).  Of these patients, 80% had abnormal EKGs on presentation (any change from baseline or any abnormality-overtly ischemic or not- if no comparison). A normal EKG showed a negative predictive value of 99%.  Of these AMI patients, only 50% had positive troponins obtained in the ED.  Along with an abnormal EKG, being male gender, having history of coronary artery disease (CAD) were both significantly more sensitive than a positive troponin in detecting AMI.

Although the utility of troponins in patients with syncope and normal EKGs looking for AMI is limited, a troponin may be useful in risk stratification of patients with alternate causes of syncope. Pulmonary embolism, type A dissection, and intracranial hemorrhage can all cause syncope and can cause type II ischemia (e.g. supply/demand ischemia).  Additionally, an elevated troponin in the setting of syncope has been associated with worse outcomes. In one study, 50% of patients with a positive troponin (at 12 hours post-syncope) had a serious outcome (not including AMI) or all-cause death at 1 month, as opposed to only 6% of patients without the positive biomarker (Reed 2010).

A similar study design performed by the same research group used a lower cutoff for a diagnostically positive troponin-I and examined the outcomes in patients with any detectable troponin level (Reed 2012).  They found the majority of syncope patients (77%) to have detectable values and 20% to have troponin levels above the new diagnostic threshold (although only 2.9% diagnosed with AMI).  At both one month and one year, patients with any detectable troponin level were at higher risk of adverse outcomes and mortality.  This risk increased with higher troponin values.

Another prospective study similarly showed a positive troponin-T (taken at least 4 hours post-syncope) to strongly predict adverse cardiac outcome (Hing 2005). However, a positive troponin proved to have no added value in predicting adverse cardiac outcomes over the OESIL* score (Colivicci 2003).  The study also reports troponin to have a sensitivity of only 13% and a low negative predictive value.

*The OESIL score is a risk stratification score to predict recurrent syncope or adverse outcome and includes the following risk factors:

  1. Age >65years old
  2. History of cardiovascular disease (CAD, CHF, cerebrovascular or peripheral vascular disease)
  3. Syncope without prodrome
  4. Abnormal EKG

The benefit of a troponin’s prognostic ability in syncope patients has not been clearly determined.  A recent study reports on the significantly improved clinical outcomes associated with using troponin-I at a lower threshold to detect positive values in patients with suspected acute coronary syndrome (Mills 2011).  While this is a different subset of patients with a clearly defined disease, the results raise the question of whether the test’s prognostic value may translate into improved outcomes in syncope patients.

Bottom Line

As a diagnostic screening test for AMI in syncope patients without chest pain or EKG changes, a single troponin is inadequate and does not appear to be helpful in risk stratification.  Admitting syncope patients for serial troponins, or ‘rule-out AMI,’ is also low-yield and should be considered only in conjunction with patients’ symptoms and significant risk factors such as known CAD or CHF, older age, syncope preceded by palpitations or without prodrome.  However, the value of a positive troponin is not limited to diagnosis of AMI.  The value of a troponin as a predictor of adverse outcome may have utility for an inpatient team and potentially in the ED as high sensitivity troponins become more ubiquitous.  Whether obtaining this prognostic data significantly improves outcomes is not clear.

  1. Do you get orthostatic measurements in patients with syncope and how do you use them?

Volume depletion (i.e. dehydration) and blood loss are two of the myriad reasons for patients to present with syncope in the Emergency Department. While it isn’t difficult for us to determine whether the patient with an active upper gastrointestinal bleed has significant volume loss, this determination can be more challenging in other patients (i.e. the elderly patient with a urinary tract infection). Thus, clinicians would benefit from having an easy bedside test that assesses volume status, particularly, one that improves our ability to pick up patients with moderate volume depletion or blood loss.

Orthostatic blood pressure measurements have historically been taught to be useful in determination of volume status. It is defined as either:

  1. A drop in systolic blood pressure (SBP) > 20 mm Hg OR
  2. Increase in heart rate (HR) by > 30 beats per minute (bpm)

when a patient stands from a supine position (McGee 1999). It is unclear from the available literature how these numbers were originally derived, but they are likely based on consensus rather than empirical data. Even available consensus statements differentiate the entities of symptomatic and asymptomatic orthostatic hypotension, bringing the overall utility of the test in to question. (Kaufmann 1996).

Despite the traditional teaching, orthostatic measurements have little if any proven utility. There are two major criticisms:

  1. Many patients without signs or symptoms of intravascular volume depletion will demonstrate orthostatic vital signs when measured

AND

  1. Many patients with clear evidence of intravascular volume depletion will not exhibit orthostatic vital signs.

How prevalent are orthostatic vital sign measurements among asymptomatic patients? A number of studies investigated elderly patients living in nursing facilities. Results of these studies are inconsistent. Mader et al and Aronow et al found relatively low prevalence (6.4% and 8% respectively) of orthostatic vital signs in absence of symptoms (Mader 1987, Aronow 1988). These studies, however, were small and excluded elderly patients on medications that may cause hypotension reducing generalizability to the general population of elderly patients. More recent larger studies on unselected elderly patients showed higher rates ranging from 28 – 50% (Raiha 1995, Ooi 1997). Studies in adolescents show similarly poor numbers with approximately 44% of patients exhibiting orthostatic changes (Stewart 2002).

Witting et al attempted to define vital sign thresholds that would decrease false positives. They performed tilt-table testing in healthy volunteers after blood donation (moderate blood loss). In patients < 65 years of age, a change in pulse > 20 bpm or a change in SBP > 20 mm Hg had a sensitivity/specificity of 47%/84% (Witting 1994). This yields a (+) LR = 2.94 and a (-) LR = 0.63. Sensitivity and specificity were similar in patients > 65 (41%/86%) with similarly poor (+) LR = 2.93 and (-) LR = 0.69. McGee et al performed a systematic review in 1999 showing similarly dismal sensitivity for moderate blood or fluid loss (McGee 1999)

Sensitivity Specificity
Blood Loss – Pulse Change (> 30 bpm) 22% NA
Blood Loss – SBP Change (> 20 mm Hg) 7-27% NA
Fluid Loss – Pulse Change (> 30 bpm) 43% 75%
Fluid Loss – SBP Change (> 20 mm Hg) 29% 81%

Mendu et al performed a retrospective study that stands as one of the few studies supporting the use of orthostatic blood pressure management in patients with syncope (Mendu 2009). The researchers found that in 18% of patients, orthostatics affected the final diagnosis and affected management in 25% of patients. However, the study is deeply flawed. The utility of the measurements was determined by the clinician with no gold standard for diagnosis with which to compare. Additionally, 55% of patients in whom orthostatics were measured were found to have abnormal results but far less of these findings were thought to be relevant. Finally, the average age in this study was near 80 years of age, the exact population in which prior studies (previously discussed) have shown poor sensitivity and specificity of these measurements (Raiha 1995, Ooi 1997).

Based on the available literature, orthostatic vital signs do not appear to be either sensitive for screening patients for moderate blood or fluid loss or specific.

Bottom Line: Many asymptomatic patients will have positive orthostatic vital signs and many patients with moderate volume loss won’t have orthostatic vital signs. This makes checking orthostatic vital signs of questionable utility. More important is to see what the patient’s symptoms are. If the patient feels lightheaded or dizzy when they go from lying supine to sitting or from sitting to standing, they are orthostatic and this should be addressed.

  1. Do you manage patients with near-syncope differently than those with syncope?

Pre-syncope is a chief complaint commonly evaluated in the Emergency Department (ED). Defined as the sense of impending loss of consciousness, its symptoms can include lightheadedness, weakness, visual disturbances, “feeling faint”, and other nonspecific complaints. While there have been several attempts to develop and derive clinical prediction tools for syncope, most have been unsuccessful due to poor sensitivity and specificity and large performance variability (Constantino 2014, Birnbaum 2008, Serrano 2010). As one can imagine, if validation of prediction rules for the objective finding of syncope is fraught with difficulty, prediction rules for the more subjective and vague symptoms of pre-syncope would be an arduous task. As such, there is a lack of guidance when it comes to management and disposition decisions for patients who present to the ED with pre-syncope.

Though classically pre-syncope was thought to be benign with many of these patients being discharged from the ED, this may not be the case. In a study of approximately 200 patients presenting to the ED with nonspecific complaints (including weakness, dizziness, and feeling unwell) and Emergency Severity Index (ESI) scores of 2 or 3 with normal vital signs, 59% had a serious condition diagnosed within 30 days, and 30-day mortality was 6% (Nemec 2010). The median age of the study’s cohort was 82 years, and most had co-morbidities. While this study did not specifically assess patients with pre-syncope, given the overlap of the included symptoms with the symptoms seen in pre-syncope, it suggests that pre-syncope may similarly be a harbinger of serious disease.

Early syncope studies often excluded pre-syncope since its definition is poorly defined. However, recent literature corroborates the potential severity of pre-syncope. Some experts purport that the pathophysiologic mechanism for pre-syncope is the same as that for syncope except that the global cerebral hypoperfusion is not significant enough to cause complete loss of consciousness (Quinn 2014). A prospective observational pilot study of 244 patients with pre-syncope and 293 patients with syncope found similar ED hospitalization and 30-day adverse outcome rates in the two groups- 23% and 20% respectively (Grossman 2012). One of the reasons that the rates were so high may have been the broad and inclusive definition of adverse outcome (which included, amongst other conditions, cortical stroke, carotid stenosis and endarterectomy, and alterations in antidysrhythmics medications).

A larger prospective cohort study in 2014 found a significant number of adverse outcomes in pre-syncope patients. Of 881 adult patients with pre-syncope (which constituted 0.5% of total ED visits), 5.1% had serious outcomes at 30-day follow-up (Thiruganasambandamoorthy 2014). Furthermore, physicians were not accurate in predicting which patients were high risk for serious outcomes after their ED visit, with an area under the receiver operating characteristic (ROC) curve of 0.58-slightly better than a coin flip.

Should we be managing patients with pre-syncope similarly to those with syncope? Despite the paucity of literature on outcomes for ED patients presenting with pre-syncope, it appears as though the potential severity of pre-syncope has been under-appreciated. Once thought to be low-risk, recent literature challenges this dogma and suggests that a significant proportion of patients with pre-syncope suffer adverse outcomes similar to those who present with syncope. Intuitively it makes sense that true pre-syncope, syncope, and cardiac arrest exist on the same spectrum, differentiated by severity and duration of hypoperfusion, and thus should be risk stratified and managed similarly. However, to date no evidence exists on whether managing pre-syncope patients the same as syncope patients improves outcomes. As such, future studies are needed to further explore which patients with pre-syncope are at higher risk for adverse outcomes, with the ultimate goal to derive and validate a clinical decision rule for this patient population.

Bottom Line:

Previously thought to be a benign diagnosis, recent literature suggests that, like syncope, a non-insignificant proportion of patients with pre-syncope suffer serious adverse outcomes. Further studies are needed to determine which patients with pre-syncope are at higher risk for adverse outcomes, as we currently do not have clinical decision rules to guide our management for this patient population.

References

Grossman SA, Fisher C, Bar JL, Lipsitz LA, Mottley L, Sands K, Thompson S, Zimetbaum P, Shapiro NI. The yield of head CT in syncope: a pilot study. Intern Emerg Med. Mar 31 2007. PMID: 17551685

Giglio P, Bednarczyk EM, Weiss K, Bakshi R. Syncope and head CT scans in the emergency department. Emergency Radiology. Dec 12 2005. PMID: 16292675

Goyal N, Donnino MW, Vachhani R, Bajwa R, Ahmed T, Otero R. The utility of head computed tomography in the emergency department evaluation of syncope. Intern Emerg Med. 2006. PMID: 17111790

Al-Nsoor NM, Marat AS. Brain computed tomography in patients with syncope. Neurosciences (Riyadh). 2010 April 15. PMID: 20672498

ACEP Clinical Policy Subcommittee on Syncope. Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department with Syncope. Annals of Emerg Med. April 2007. PMID: 18035161

Link MS, Lauer EP, Homoud MK et al. Low yield of rule-out myocardial infarction protocol in patients presenting with syncope. Am J Cardiol 2001;88:7067. PMID: 11566406

Grossman SA, Van Epp S, Arnold R et al. The value of cardiac enzymes in elderly patients presenting to the emergency department with syncope. J Gerontol A Biol Sci Med Sci 2003;58:1055–8. PMID: 14630890

Hing R, Harris R. Relative utility of serum troponin and the OESIL score in sycnope. Emerg Med Australas 2005;17:31-8. PMID: 15675902

Reed MJ, Newby ED, Coull AJ, et al. Diagnostic and prognostic utility of troponin estimation in patients presenting with syncope: a prospective cohort study. Emerg Med J. 2010;27:272-276. PMID: 20385677

Colivicchi F, Ammirati F, Melina D, et al. Development and prospective validation of a risk stratification system for patients with syncope in the emergency department: the OESIL risk score.  Eur Heart J. 2003 May;24(9):811-9. PMID: 12727148

McDermott D, Quinn JV, Murphy CE. Acute myocardial infarction in patients with syncope. CJEM. 2009 Mar;11(2):156-60. PMID: 19272217

Reed MJ, Mills NL, Weir CJ. Sensitive troponin assay predicts outcome in syncope. Emerg Med J. 2012;29:1001-1003. PMID: 22962048

Mills NL, Churchouse AM, Lee KK, et al. Implementation of a sensitive troponin I assay and risk of recurrent myocardial infarction and death in patients with suspected acute coronary syndrome. JAMA. 2011 Mar 23;305(12):1210-6. PMID: 21427373

McGee S, Abernethy WB, Simel DL. The rational clinical examination. Is this patient hypovolemic. JAMA 1999; 281(11): 1022-9. PMID: 10086438

Kaufmann H. Consensus statement on the definition of orthostatic hypotension, pure autonomic failure and multiple system atrophy. Clin Auto res 1996; 6: 125-6. PMID: 8726100

Witting MD, Wears RL, Li S. Defining the positive tilt test: a study of healthy adults with moderate acute blood loss. Ann Emerg Med 1994; 23(6): 1320-3. PMID: 8198307

Mader SL, Josephson KR, Rubenstein LZ. Low prevalence of postural hypotension among community-dwelling elderly. JAMA 1987; 258: 1511-14. PMID: 3625952

Aronow WS, Lee NH, Sales FF, Etienne F. Prevalence of postural hypotension in elderly patients in a long-term health care facility. Am J Cardiology 1988; 62(4): 336. PMID: 3135742

Raiha I, Luutonen S, Piha J, Seppanen A et al. Prevalence, predisposing factors and prognostic importance of postural hypotension. Arch Intern Med 1995; 155: 930-935. PMID: 7726701

Ooi WL, Barrett S, Hossain M, Kelley-Gagnon M, Lipsitz LA. Patterns of orthostatic blood pressure change and the clinical correlates in a frail, elderly population. JAMA 1997; 277: 1299-1304. PMID: 9109468

Stewart JM. Transient orthostatic hypotension is common in adolescents. J Pediatr 2002; 140: 418-24. PMID: 12006955

Mendu ML et al. Yield of diagnostic tests in evaluating syncopal episodes in Older patients. Arch Intern Med 2009; 169: 15: 1299-1305. PMID: 1963031

Birnbaum A, et al. Failure to Validate the San Francisco Syncope Rule in an Independent Emergency Department Population. Ann Emerg Med. 2008 Aug;52(2):151-9. PMID: 18282636

Constantino G, et al. Syncope Risk Stratification Tools vs Clinical Judgement: An Individual Patient Data Meta-analysis. Am J Med. 2014 Nov;127(11):1126.e13-25. PMID: 24862309

Grossman SA, et al. Do Outcomes of Near Syncope Parallel Syncope? Am J Emerg Med. 2012 Jan;30(1):203-6. PMID: 21185670

Nemec M et al. Patients Presenting to the Emergency Department with Non-specific Complaints: The Basel Non-specific Complaints (BANC) Study. Acad Emerg Med. 2010 Mar;17(3):284-92. PMID: 20370761

Quinn, JV. Syncope and Presyncope: Same Mechanism, Causes, and Concern. Ann Emerg Med. 2014 Oct 31. pii: S0196-0644(14)01257-8. PMID: 25441246

Serrano LA, et al. Accuracy and Quality of Clinical Decision Rules for Syncope in the Emergency Department: A Systematic Review and Meta-analysis. Ann Emerg Med. 2010 Oct;56(4):362-373. PMID: PMC2946941/

Thiruganasambandamoorthy V, et al. Outcomes in Presyncope Patients: A Prospective Cohort Study. Ann Emerg Med. 2014 Aug 30. pii: S0196-0644(14) 1115-9. PMID: 25182542

Answers Created By:


Syncope, Questions

Syncope1. In which patients presenting with syncope do you get a Non-Contrast Head CT (NCHCT)?

2. In which patients presenting with syncope do you get a troponin?

3. Do you get orthostatic vital sign measurements in patients presenting with syncope? How do you use them?

4. Do you manage patients presenting with near-syncope differently than those presenting with syncope?


Spinal Cord Injury, “Answers”

1. What imaging do you use for patients with possible acute, traumatic spinal cord injury?

Patients who can be cleared using the Nexus or Canadian C-spine criteria should be cleared clinically. However, those with moderate to high risk of a cervical spine (C-spine) injury should have cross sectional imaging, based on substantial amounts of data. The Eastern Association for the Surgery of Trauma (EAST) referenced 52 articles to construct guidelines recommending against plain radiographs in the assessment of potential C-spine injuries (Como, 2009). Although the C-spine is actually only injured in approximately 3% of all major trauma patients, (Crim, 2001) these tend to be some of the most disabling injuries. There is less data regarding the type of imaging preferred for patients with possible thoracic and lumbar spine injuries.

EML Spinal Cord AnswersIn patients in whom you cannot clear their C-spine clinically and you suspect an acute, traumatic spinal cord injury, computerized tomography (CT) scanning is the most appropriate initial imaging. There is a growing body of literature stating that plain films miss many clinically significant injuries, and have little to no role in evaluating spinal fractures, particularly those of the C-spine.

The argument against this is that some patients may actually be appropriate for plain film imaging of the C-spine. These are patients who are deemed low risk mechanism, younger, and in whom good views can be obtained. There will still be some fractures that are missed, but these have a low chance of being clinically significant, and in a patient with low pre-test probability, this may be appropriate utilization of plain film imaging. The most compelling argument for this comes from the original NEXUS study. Of 34,069 patients with blunt trauma, 1,496 had C-spine injuries. Plain films missed 564 injuries in 320 patients. However, for 436 of these injuries, 0.80% of all patients, the plain films were interpreted as abnormal (although non diagnostic) or abnormal. Only 23 (0.07%) of all patients in the studies had injuries that were not seen on plain films that were also read as negative. Three of these patients had unstable c-spine injuries (Mower, 2001). In a retrospective review of a trauma database of 3,018 patients, 116 (9.5%) had a C-spine fracture. The injury was only seen in 75 of these patients on plain film. In the remaining 41 patients (3.2%), the injury was detected on CT scan, and in all cases these injuries required treatment. It is important to recognize that the mean Glasgow coma score (GCS) of this patient population was 13, thus, they were likely a high acuity patient population overall. These authors concluded that there was no role for plain imaging in patients in whom a C-spine injury is suspected in the Emergency Department population (Griffen, 2003).  Most recently, Mathen, et al. performed a prospective cohort study of 667 patients who required C-spine imaging. They found that plain radiography missed 15 of 27 (55.5%) clinically significant c-spine injuries (Mathen, 2007).

In a cost effectiveness analysis, CT was found to be the preferred modality for imaging in moderate to high risk patients, given that missed C spine fractures and resultant paralysis may be devastating to both the patient and society (Blackmore, 1999).

Unlike C-spine fractures, we have no clinical decision rules to help guide our care of patients with potential thoracolumbar fractures. There is considerably less data on thoracolumbar injuries. According to a literature review published in the Journal of the American College of Radiology, which looked at studies on TL spine imaging comprising several thousand patients, those who should be evaluated for thoracic or lumbar spine injuries are those with high force mechanism and any of the following findings: back pain or midline tenderness, local signs of thoracolumbar injury, abnormal neurologic signs, C-spine fracture, GCS < 15, major distracting injury, or drug or alcohol intoxication (Daffner, 2007).

There is evidence to suggest that many injuries may be missed by plain films of the thoracolumbar regions. One study of seventy intubated trauma patients found that thin slice CT discovered 100% of unstable fractures in comparison with 56-80% (depending on spinal level) seen by conventional radiographs (Herzog, 2004). A prospective evaluation of 1.915 trauma patients presenting to a Level I trauma center compared the sensitivity of CT versus plain film in the detection of the 78 thoracic or lumbar spine fractures sustained by the group. CT sensitivity was 97% and 95% for thoracic and lumbar injuries, respectively. For plain films those numbers were 62% and 82% (Sheridan, 2003). Yet another study in high acuity trauma patients found the sensitivity of CT to be 97% and that for plain films to be an abysmal 33.3% (Wintermark, 2003). In a retrospective review of 3,537 patients, the only fractures missed by CT scan were a cervical compression fracture identified on MRI, and a thoracic compression fracture identified by plain films. This study recommended that plain films of the spine are unnecessary in the evaluation of blunt trauma patients. Furthermore, after a panel reviewed all literature on the topic, the American College of Radiology Appropriateness Criteria recommended that patients with potential thoracic or lumbar spine injury undergo CT scan, as opposed to plain films (Daffner, 2007). The ACR grades on a scale of 1-9, with 1-3 being usually not appropriate, 4-6 may be appropriate, and 7-9 usually appropriate. Their level of recommendation in this indication is a 9.

Due to the data regarding imaging, most patients should undergo CT imaging for possible spinal trauma. Only the lowest risk patients who have adequate plain films should be cleared without CT imaging.

2. How do you treat neurogenic shock?

Neurogenic shock is a form of distributive shock unique to patients with spinal cord injuries. Fewer than 20% of patients with a cervical cord injury have the classic diagnosis of neurogenic shock upon arrival to the emergency department, and it is a relatively uncommon form of shock overall (Guly, 2008). Patients with injuries at T4 or higher are most likely to be affected by neurogenic shock (Wing, 2008). It is caused by the loss of sympathetic tone to the nervous system, ultimately leading to an unopposed vagal tone (Stein, 2012). Many times the terms “spinal shock” and “neurogenic shock” are used interchangeably, although they are two separate entities. Spinal shock consists of the loss of sensation and motor function immediately following a spinal cord injury (Nacimiento, 1999). During this period of spinal shock, reflexes are depressed or absent distal to the site of the injury. Spinal shock may last for several hours to several weeks post injury (Nacimiento, 1999).

Symptoms of neurogenic shock consist of bradycardia and hypotension (Grigorean, 2009). Bradycardia is typically not present in other forms of shock, and may provide a clue to clinicians that a patient has sustained a spinal cord injury. However, emergency physicians should recognize that hemorrhagic shock needs to be first ruled out, even in patients with bradycardia, many patients with hemorrhagic shock are not tachycardic (Stein, 2012). Cardiac dysfunction is another feature of neurogenic shock, and patients may present with dysrhythmias following injury to the spinal cord (Grigorean, 2009).

The American Spinal Injury Association (ASIA) has classified injuries based on motor and sensory findings at the time of injury. ASIA A and B injuries are the worst; with A being a complete motor and sensory loss with no preserved function in the sacral segments S4-S5. ASIA B includes patients who have sacral sparing, meaning that they have function of S4 and S5 (Marino, 2003). Neurogenic shock is rarely encountered in the emergency department, however, it is important to recognize that almost 100% of patients who sustain complete motor cervical ASIA A or ASIA B injuries develop bradycardia. Thirty five percent of these patients ultimately require vasopressors, so management of neurogenic shock is imperative for emergency physicians (McKinley, 2006). There is no conclusive data regarding the optimal time to start vasopressors, however, it is important to maintain appropriate hemodynamic goals in patients with spinal cord injuries.

Hemodynamic goals in patients with spinal cord injuries are unique. A systolic blood pressure <90 mmHg must be corrected immediately (Muzevich, 2009). The American Association of Neurological Surgeons and the Congress of Neurological Surgeons Guidelines for the Acute Management of Spinal Cord Injuries both recommend a MAP at 85 to 90 mm Hg for the first seven days following a spinal cord injury based on observational descriptions of the hemodynamics in spinal cord injured patients (Levi, 1993; Licina, 2005).

Patients who are suspected of being in neurogenic shock should receive adequate fluid resuscitation prior to initiating vasopressors (Wing, 2008). However, there are no current recommendations regarding the first line vasopressor for neurogenic shock (Stein, 2012). Depending on a patient’s hemodynamics, this vasopressor will likely be norepinephrine, phenylephrine, or dopamine.

Norepinephrine is an excellent first line vasopressor in neurogenic shock due to its alpha and some beta activity, thus leading to its ability to improve blood pressure and heart rate (Stein, 2012). Phenylephrine is another common choice because it is easy to titrate and can be given through a peripheral line. A disadvantage of phenylephrine is the fact that it can lead to reflex bradycardia due to its lack of beta agonism. This drug may be most appropriate in patients who are not bradycardic (Wing, 2008). Dopamine is another option, however, it may lead to diuresis and ultimately worsened hypovolemia (Stein, 2012). It does have beta agonism, and in bradycardic patients may be favored over phenylephrine (Muzevich, 2009). Dopamine is unlikely to be tolerated in patients who are experiencing dysrhythmias.

3. What is your management and disposition for elderly patients with vertebral compression fractures?

Vertebral compression fractures of the thoracic and lumbar vertebrae are extremely common in the elderly population, with an annual incidence of 1.5 million vertebral compression fractures per year (Barr, 2000). They are most commonly seen in patients with osteoporosis, although may be seen in younger patients, particularly those with malignancy. The majority of patients are treated non-surgically, usually with bed rest and hyperextension bracing (Gardner, 2006). Pain is typically the presenting symptom, and neurologic deficits are rare unless there is retropulsion of bone into the vertebral canal. This presentation is rare in compression fractures, but it does constitute a surgical emergency (Kavanagh, 2013).

Although some patients may experience mild or minor symptoms related to a vertebral compression fractures, many patients will have a significant degree of pain and decreased quality of life associated with their fracture (Adachi, 2002). At a minimum, patients with very mild symptoms and a normal neurologic exam may be discharged home with adequate pain control and spine surgery follow up established.

Therapy should be tailored toward avoiding a prolonged period of bed rest as well as adequate pain control (Wong, 2013). Prolonged immobilization may lead to poor pulmonary toilet, venous thromboembolism, and deconditioning, especially in elderly patients. Non-steroidal anti-inflammatory drugs (NSAIDS) are first line therapy since they are non-sedating, but may be poorly tolerated in certain groups of patients such as the elderly or those with underlying peptic ulcer disease (Wong, 2013). Opiates and muscle relaxers may be necessary for pain control, but should be used with caution in the geriatric population, especially those at risk of falls.

Although admission of elderly patients with vertebral compression fractures may not result in surgical management, it may provide other avenues of therapy that are unavailable or difficult to arrange in the ED setting. One of these therapies includes physical therapy, which may help patients regain early mobility if introduced appropriately. Physical therapy is also helpful in training patients to strengthen other extra-axial muscles, particularly the spine extensors (Wong, 2013). Several trials have demonstrated effectiveness of physical therapy in patients with vertebral compression fractures. Malmros et al evaluated a 10-week physical therapy program in a placebo-controlled, randomized, single-blinded study that demonstrated improved quality of life and reduction in pain and analgesic use (Malmros, 1998). Papioannou, et al. conducted a randomized controlled trial consisting of a 6 month home exercise program and found that patients in the physical therapy arm had significant improvement of quality of life scores and improved balance at one year (Papaioannou, 2003).

The decision of when to use thoracolumbosacral orthosis (TLSO) brace in a patient with a vertebral compression fracture is somewhat controversial. Pfeifer, et al. demonstrated in a randomized trial that the use of a brace increased trunk muscle strength and was associated with an improved quality of life, decreased pain, and improved daily functioning in patients with compression fractures (Pfeifer, 2004). However, an electromyelography study demonstrated increased muscle spasming in patients with brace placement (Lantz, 1986). Furthermore, braces may contribute to skin breakdown, especially in geriatric patients. If a TLSO brace is given to patients, it should be done in consultation with a spine surgeon.

Surgical management options for vertebral compression fractures include kyphoplasty and vertebroplasty. Both procedures are minimally invasive, but are traditionally only performed if patients are in pain several weeks following diagnosis of a compression fracture (Wong, 2013).

Patients who are discharged from the ED with compression fractures need to be able to ambulate and perform activities of daily living prior to discharge. If pain control limits these activities, they will likely require admission for pain control, physical therapy, and potentially rehabilitation.

4. How do you clear a C-spine after a negative CT in a trauma patient who is awake, neuro intact, wearing a collar?

According to EAST guidelines, there are multiple appropriate options in patients who are awake, neurologically intact, and still have midline tenderness after a negative CT (Como, 2009). Although CT scans will pick up the majority of injuries, it is well documented that they specifically may miss ligamentous injuries, subluxations, and dislocations (Woodring, 1992).

The first option is to obtain an MRI within 72 hours post injury. Very little data exists in the literature regarding this option. Schuster, et al. evaluated prospectively collected registry data for 2854 blunt trauma patients, 93 of whom had a normal neurologic exam at admission, a negative CT result, and persistent C-spine pain. These patients all had an MRI. In all 93 of these patients, the MRI was negative for clinically significant injury. However, the argument could also be made that since no clinically significant injury was detected by MRI in this case, that there was no need for any further imaging (Schuster, 2005).

The second option is to continue the C-spine immobilization until there is no midline tenderness and the patient has been followed up as an outpatient. This is not ideal in centers where there is no trauma team to assist in outpatient management of these patients. Furthermore, the collar itself poses a risk of skin breakdown and decubitus ulcers when worn for a prolonged period of time. This option may work best for patients who can rapidly be seen in a trauma clinic.

The third option is to obtain flexion-extension films in patients with a negative CT of the C-spine. Although studies have evaluated the utility of flexion-extension films in patients with negative plain films of the C spine, no study has completely evaluated flexion-extension films following a negative CT of the C-spine. Insko et al reviewed 106 patients with negative plain films or negative CT imaging in areas that were not visualized by plain films. This study demonstrated a false negative rate of zero in diagnosing C spine fractures when flexion extension films were performed in patients who were persistently tender (Insko, 2002).

Ultimately, more information is needed to determine the best course of action to take in a patient with persistent pain following negative CT imaging of the C-spine (Como, 2009). However, at this time, there are three potential options in ruling out a C-spine injury in these patients. The decision may largely depend on local practice patterns, clinical suspicion for injury, as well as a patient’s ability to follow up.


Spinal Cord Injury, Questions

1. What imaging do you use for patients with possible acute, traumatic spinal cord injury?

2. How do you treat neurogenic shock?

3. What is your management and disposition for elderly patients with vertebral compression fractures?

4. How do you clear a C-spine after a negative CT in a trauma patient who is awake, neuro intact, wearing a collar?


Infectious Diseases, “Answers”

1. Which patients with neutropenic fever do you consider for outpatient management?

Neutropenic fever is a common presentation to the Emergency Department, especially in tertiary hospitals where many oncology patients are undergoing chemotherapy. According to the Infectious Disease Society of America (IDSA), fever in neutropenic patients is defined as a single oral temperature of >38.3°C (101°F) or a temperature of >38.0°C (100.4°F) sustained for >1 hour. Rectal temperature measurements (and rectal exams) are not recommended by the IDSA to prevent colonizing gut organisms from entering the surrounding mucosa and soft tissues (Freifeld, 2011). The definition of neutropenia varies from institution to institution; the IDSA defines it as an (ANC) <500 cells/microL or an ANC that is expected to decrease to <500 cells/microL over the next 48 hours. Profound or severe neutropenia occurs when the ANC is < 100 cells/micromol. The National Cancer Institute defines neutropenia as an ANC < 1000 cells/micromol (HHS 2010).

Patients with neutropenic fever are usually started on broad spectrum IV antibiotics and admitted to the hospital, however there are a subgroup of patients who can be safely managed as outpatients. The official wording from the IDSA guidelines is that “Carefully selected low-risk patients may be candidates for oral and/or outpatient empirical antibiotic therapy (B-I)”. Grade B is defined as moderate evidence to support a recommendation for use, and Level I is evidence from ≥1 properly randomized, controlled trial. The data which they derived these recommendations include one large series of patients, where oral outpatient treatment for low-risk fever and neutropenia was deemed successful in 80% of patients, with 20% patient requiring readmission. Factors predicting readmission include age > 70, grade of mucositis >2, poor performance status, and ANC < 100 cells/microL at onset of fever (Escalante 2006). Klastersky et al studied 178 low-risk patients who were treated with oral antibiotics. Only 3 patients were readmitted resulting in a 96% success rate (Klastersky 2006) .

The IDSA formally risks stratifies using the Multinational Association for Supportive Care in Cancer (MASCC) scoring system. The adult guidelines from Australia, European Society for Medical Oncology (ESMO), the American Society of Clinical Oncology (ASCO) also recommend the use of the MASCC index (Gea-Banacloche 2013). Low-risk patients have a MASCC score ≥ 21.

MASCC

The index has been validated in multiple settings and performs well, although it may function better in solid tumors than in hematologic malignancies (Klastersky 2013). An issue with the major criteria of “burden of febrile neutropenia” is that there is no standardized definition for this criteria, making uniform application of the MASCC confusing (Kern 2006).

Based on the best available data prior to the 2010 guidelines release, the IDSA developed an algorithm to risk stratify neutropenic fever patient and their appropriate management:

Untitled

 

Ciprofloxacin plus amoxicillin-clavulanate is recommended for oral empirical treatment (Friefeld 2010). Other oral regimens, including levofloxacin or ciprofloxacin monotherapy or ciprofloxacin plus clindamycin, are less well studied but are commonly used. In low-risk patients, the risk of invasive fungal infection is low, and therefore routine use of empirical antifungal therapy is not recommended. Respiratory virus testing and chest radiography are indicated for patients with upper respiratory symptoms and/or cough .

A systematic review and meta-analysis of 14 RCTs was published in 2011 and not, therefore, included in the 2010 guidelines (Teuffel 2011). The meta-analysis concluded that inpatient versus outpatient management was not significantly associated with treatment failure; death occurred with no difference between the two groups; and outpatient oral versus outpatient parenteral antibiotics were similarly efficacious with no association between route of administration and treatment failure.

It must be said without questions that the patients whom you are treating as outpatients need to have easily accessible and very close follow up with their oncologists. They should be vigilantly examined for a source of infection, including a thorough skin, mucosal, and neurologic exam, as any obvious focal infection may necessitate inpatient treatment.

Bottom Line: Based on the IDSA guidelines and other meta-analyses, it is reasonable to treat a certain subset of low risk febrile neutropenic patients with oral antibiotics as an outpatient with good follow up.

2. Which patients with community-acquired pneumonia do you admit?

Community-acquired pneumonia (CAP) is defined as an acute infection of the pulmonary parenchyma in a patient who has acquired the infection in the community, as distinguished from hospital-acquired (nosocomial) pneumonia, which occurs 48 hours or more after hospital admission and is not present at time of admission. A third category of pneumonia, designated “healthcare-associated pneumonia,” is acquired in other healthcare facilities such as nursing homes, dialysis centers, and outpatient clinics or within 90 days of discharge from an acute or chronic care facility. The most common validated prediction rules for prognosis in community-acquired pneumonia include the Pneumonia Severity Index, CURB, and CURB-65 severity scores.

Question 2

 

According to the Pneumonia Severity Index (PSI), patients in risk classes I-III are defineQuestion 2 2d as low risk for short-term mortality and are considered for outpatient treatment. In the original derivation study of the PSI, Mortality ranged from 0.1 to 0.4 percent for class I patients, from 0.6 to 0.7 percent for class II, and from 0.9 to 2.8 percent for class III (Fine 1997). This older but well validated rule can be difficult to remember and apply; as a result, a relatively simpler CURB score with a total point score ranging from 0-4 was described and externally validated (Ewig 2004). A modified version, the CURB-65, which added age ≥ 65 as another positive risk factor was internally validated to stratify short term mortality for patients with CAP (Lim 2003). Patients scoring CURB < 1 and CURB-65 < 2 are considered low-risk and candidates for outpatient treatment. The CURB and CURB-65 scores are easier to remember and apply, however it does require laboratory data (BUN), whereas the lowest PSI risk class I can be attained without getting blood draws, a benefit appreciated in outpatient settings.

The three prediction rules were pitted against each other in a prospective study of over 3000 patients with community-acquired pneumonia from 32 hospital EDs to see which one was better at predicting 30-day mortality (Aujesky, 2005). Inclusion criteria included age ≥ 18 with a clinical diagnosis of pneumonia and new radiographic pulmonary infiltrate. Exclusion criteria included hospital acquired pneumonia, immunosuppression, psychosocial problems incompatible with outpatient treatment, or pregnancy.

The PSI classified a significantly greater proportion of patients as low risk (68%) than the CURB (51%) and the CURB-65 (61%). Of the low-risk by PSI (class I-III), the aggregate 30-day mortality was 1.4%, which is less than the CURB (score < 1) low-risk mortality rate of 1.7% and also 1.7% mortality for CURB-65 low-risk group (score <2). High-risk patients based on the PSI (class IV-V) had a higher mortality of 11.1% compared with high-risk CURB (≥1) and high risk CURB-65 (≥2), with respective mortality rates of 7.6% and 9.1%.

The PSI had a slightly higher sensitivity and negative predictive value across each risk cut-off point compared to the CURB and CURB-65. In addition, by comparing the areas under the receiver operating characteristic (ROC) curves, the PSI had a statistically significantly greater discriminatory power to predict 30-day mortality. CURB-65 showed a higher overall discriminatory power than the original CURB score.

Based on the estimation of 4 million annual cases of community acquired pneumonia in the USA (DeFrances, 2007), and the average cost of inpatient versus outpatient care of $7500 vs $264; using the PSI would identify an additional 650,000 low-risk patients vs. CURB and 250,000 vs. CURB-65, saving a significant amount of healthcare dollars, while still maintaining a low 30-day mortality rate (Aujesky, 2005) .

The 2007 (most current) Infectious Diseases Society of America (IDSA) recommendations on managing CAP revolve around the initial assessment of severity. “Severity-of-illness scores such as the CURB-65 criteria, or prognostic models, such as the Pneumonia Severity Index (PSI), can be used to identify patients with CAP who may be candidates for outpatient treatment. (Strong recommendation; level I evidence).” In addition to objective data, the IDSA also recommends supplementing with subjective factors, including the ability to safely and reliably take oral medications and the availability of outpatient support resources (Strong recommendation; level II evidence). (Mandell, 2007).

Bottom Line: Based on the validated prognostic scoring systems and recommendations from the IDSA, a subgroup of low-risk patients with community acquired pneumonia can safely be managed as outpatients. Their prognosis can be reliably predicted by 3 scoring systems, where the PSI performs slightly better but is more complex to apply than the CURB and CURB-65.

  1. Which patients with influenza do you treat with oseltamivir?

Neuraminidase inhibitors, specifically oseltamivir, have been increasingly used in the last 5 years for the treatment of patients with symptoms of influenza. In fact, between May and December of 2009, around 18.3 million prescriptions were written for the drug in seven countries (Australia, Canada, France, Germany, Japan, UK and USA) (Muthuri 2014). Additionally, hundreds of millions of doses have been stockpiled in various countries as a safe guard against pandemic influenza and the World Health Organization (WHO) lists oseltamivir as an essential drug. Physicians have been encouraged to prescribe osletamivir to patients by drug companies, professional society recommendations, hospitals and by patient pressure. Despite widespread use, the evidence of benefit for oseltamivir has never sat on firm evidence-based grounds. Much of this stems from the reluctance of the pharmaceutical giant Roche to release all relevant study data on the drug. This changed in 2013 when all of the data was made available for analysis.

Before we delve into the recently released data surrounding oseltamivir, let’s look at the prior recommendations and the basis for these recommendations. In 2009, the BMJ published a review of a number of observational studies looking at the effect of oseltamivir in the treatment of influenza (Freemantle 2009). This publication looked at randomized controlled studies provided by Roche pharmaceutical at that time. The limited available evidence supported the role of oseltamivir in reducing the rate of post-influenza pneumonia in otherwise healthy adults. There was no evidence of a mortality benefit and limited safety data (Freemantle 2014). Additionally, the available data supported the idea of using oseltamivir for chemoprophylaxis in patients who were at risk of exposure. A Cochrane group review in 2010 echoed these results but also stated that there was extensive bias present in the available studies and without a full disclosure of all the research, no strong recommendation could be made (Jefferson 2014). In spite of the limited evidence, broad recommendations were made including treatment of patients with multiple comorbidities, pregnant patients, those with immunocompromise and chemoprophylaxis for close contacts (Harper 2009).

In 2013, Roche pharmaceuticals released all of the study data. The Cochrane Respiratory group subsequently published an updated systematic review of all of the randomized controlled trials (Jefferson 2014) as well as a summary statement in the BMJ (Jefferson 2014). A number of statements from the prior review stand: there are minimal studies on efficacy and safety in pregnant patients, and no mortality benefit was seen. They did not find a reduction in post-influenza pneumonia and posited that this prior finding was likely due to publication bias. The below table summarizes the major outcomes of the 2014 Cochrane systematic review:

Outcome Measure Finding
Alleviation of Symptoms Shortened by 16.8 hrs with oseltamivir
Admission to Hospital No Difference
Reduction in Confirmed Pneumonia No Difference
Other Complications No Difference
Transmission in Prophylaxis Group No Reduction

Additionally, the group reported on a number of common side effects:

Side Effect Results
Nausea Increased (NNH 28)
Vomiting Increased (NNH 22)
Psychiatric Events Increased (NNH 94)
Headache Increased (NNH 32)

Overall, we see a mild shortening in the duration of symptoms with no reduction in admission, confirmed, post-influenza pneumonia or other complications. The same findings were seen in pediatric patients as well. There is minimal evidence in regards to efficacy or safety in pregnancy as pregnancy was an exclusion criteria in most of the studies. Side effects were common. Also, chemoprophylaxis did not reduce transmission of the disease. These results call into question the utility of oseltamivir for the treatment of influenza in any patient.

In June 2014, the PRIDE Consortium Investigators published a study challenging the Cochrane group findings (Muthuri 2014). In this large observational cohort (n=29,234 patients) Murthi et al found an association with decreased mortality (adjusted OR = 0.81) and an additional benefit to early (< 2 days) treatment versus later treatment (adjusted OR = 0.48). This study, however, has major flaws and biases that question the validity of their conclusions. Only 19% of centers that were contacted agreed to contribute data to the Consortium. Thus there is a high potential for bias. Additionally, the researchers do not assess the quality of the studies included in their meta-analysis (Antes 2014). Regardless, observational data should not be used to trump the RCT data included in the Cochrane review.

Bottom Line: The best available evidence demonstrates that oseltamivir leads to a mild reduction in the duration of symptoms of influenza. There is no proven benefit for mortality, hospital admissions or confirmed influenza related complications including pneumonia. The frequency of side effects may outweigh the mild symptom reduction benefit of the drug. The results of the 2014 Cochrane meta-analysis should be used to update the current CDC recommendations.

4. Which adult patients getting worked up for a urinary tract infection do you send a urine culture on?

A Urinary tract infection (UTI) is a condition in which bacteriuria is present with evidence of host invasion (presence of dysuria, frequency, flank pain, or fever). The “gold standard” in defining significant bacteriuria is the detection of any microorganisms by suprapubic aspiration. Since this method is not typically employed, many sources utilize a definition of more than 105 cfu/ml on a midstream urine culture to indicate true infection.

The American College of Emergency Physicians has recently kicked off the “Choosing Wisely” campaign (ACEP 2013) in an attempt to limit unnecessary testing in the Emergency Department. Although urine culture was not one of the five tests that was addressed, it is one of the most commonly sent laboratory tests in the ED making it a potential place to curb costs. In 1997, UTIs accounted for one million ED visits in the United States (Foxman 2002). Numerous publications and practice guidelines have recommended against the use of routine urine cultures in uncomplicated UTIs. Despite this, a 1999 survey of 269 EM physicians showed that 24% of them would order a urine culture on a 30-year-old non-pregnant woman with an uncomplicated UTI with dysuria of recent onset (Wigton 1999).

Werman et al in 1986 tackled the question of the “Utility of Urine Cultures in the Emergency Department” (Werman 1986). They concluded that urine cultures should only be obtained in patients at high risk for pyelonephritis or bacteremia/urosepsis, as well as in those expected to have uncommon or resistant organisms. This paper cited studies showing that routine urine cultures in nonpregnant women with acute cystitis do not affect management. Morrow et al showed that treatment of women seen in the ED with suspected cystitis proceeded with little attention to urine culture results, despite the fact that cultures were obtained routinely (Morrow 1976). Winickoff et al demonstrated that patients who did not have follow-up urine cultures after a UTI had no greater risk for reinfection or complications than did patients in whom follow-up cultures were obtained (Winickoff 1981). Additionally, a positive culture does not necessarily indicate the absolute need for antibiotics. In a study of 53 women with culture-proven UTIs, no patient progressed to pyelonephritis or bacteremia in spite of the fact that all were treated with placebo (Mabeck 1972).

In 2011, Johnson et al. addressed the question “Do Urine Cultures for Urinary Tract Infections Decrease Follow-up Visits?” (Johnson 2011). This retrospective cohort study looked at 779 female patients age 18-65 diagnosed with a UTI or acute cystitis treated in a family medicine clinic (exclusion criteria: pregnancy, diabetes, UTI or antibiotic use in preceding 6 weeks, other medical condition making UTI complicated). The follow-up rate for patients without urine cultures was 8.4%, which showed no statistical difference between the follow-up rate of 8.7% for patients with urine cultures. Ordering a urine culture was not associated with a decreased rate of follow-up visits (adjusted OR 1.11 [CI 0.65-1.90]). Of all 447 urine cultures ordered, only 1 grew bacteria resistant to nitrofurantoin, a common antibiotic used in the ED for uncomplicated cystitis. A 2006 UK study found that 23 women required a urine culture to prevent one follow-up visit from resistance-based failure; thus, empiric treatment with no urine culture was recommended (McNulty 2006).

Bottom Line: Although there are no prospective randomized controlled trials looking specifically at ED patients with UTI symptoms, it is safe to say that a urine culture in healthy adult non-pregnant females with new onset urinary symptoms without concern for pyelonephritis or bacteremia is unlikely to change management or outcome.