An evidence-based emergency medicine podcast. The Washington University Emergency Medicine Journal Club podcast provides an excellent way to keep up with the medical literature, improve patient care, and increase understanding of evidence-based medicine. Each month we ask a clinical or operational question pertinent to the practice of emergency medicine. We then discuss the four articles most relevant to that question in an attempt to determine the most evidence-based approach to patient care.
Asymptomatic Hypertension in the Emergency Department
Jun 12, 2013
Show notes
Journal Club Podcast #2
My good friend Greg Polites and I sit down and discuss the evaluation and management of patients with asymptomatic
severely elevated hypertension in the emergency department...
Vignette:
You arrive in TCC for your 3p-11p shift and are taking sign out from Dr. Bavolek after one of her typical trauma shifts. After receiving report on the 4th level 1 trauma, the stroke patient receiving tPA and awaiting an NNICU bed, the “wound check” with turned out to be nec fasc, and “I’ve had abdominal pain for 6 months, I’m here for a third opinion,” you see that EM4 has a green zoom icon. “Oh, don’t worry about him,” she says, “that dental pain is going home.” On the way to see your first patient that was paged out as “unknown male found down, agonal respirations, CPR in progress TCC6. TY Cindi” you are stopped by one of our new nurses. “Hey, are you done with sign out? Did Dr. Bavolek leave already? I was just discharging EM4 but his blood pressure is 215/111, do you want to do anything about that?” In between shocks, CPR, and intubating the new patient, you manage to review EM4s chart. He’s a 51 year old African American gentleman who came in with left lower dental pain. He has no past medical history but has not sought medical care in years. After exam showed no abscess you see he received an inferior alveolar block, discharged on vicodin, penicillin, and given both a dental referral and IHN follow up. While placing an ICY line with your new electronic manometer you think to yourself, “huh, that’s kind of high, what DO I do? No, better yet, what would Chris Carpenter do? WWCCD… WWCCD…” As you’re waiting for radiology to pull up your chest X-ray confirming tube placement, you open PubMed at the computer between TCC4 and TCC5 and this is what you find… PICO Question:
Population: Patients presenting to the emergency department (ED) with severely elevated blood pressure and no signs or symptoms concerning for end-organ damage Intervention: Laboratory testing, electrocardiogram (ECG), chest x-ray, or rapid reduction in blood pressure Comparison: Outpatient referral for evaluation and initiation of antihypertensive therapy Outcome: Stroke, MI, renal failure, dialysis, death. Search Strategy:
In PubMed, the search terms “asymptomatic hypertension” and “emergency department” were entered (http://tinyurl.com/d44c5vq) resulting in 83 articles, from which 3 relevant articles are selected. A search of the ACEP clinical policies for hypertension yields the final article. Bottom Line:
Elevated BP remains a common finding in patients presenting to the emergency department (Karras 2005). While the detrimental effects of long-standing untreated hypertension have been well-documented with respect to the risk of stroke, myocardial infarction, and chronic kidney disease, the risks of untreated hypertension in the short-term have not been proven. For patients with signs or symptoms of end-organ damage, the course of action for Emegency Physicians is fairly straightforward, with work-up and treatment recommended. The dilemna lies with those patients presenting with asymptomatic elevated BP.
Currently, the ED management of patients with severe elevated blood is highly variable, with differences in both treatment (intravenous and oral anti-hypertensives to effect an immediate reduction in BP, inititiation or modification of outpatient antihypertensive regimen, discharge for follow-up with no antihypertensive medications) and evaluation (extensive ED testing. Options for treatment include intravenous and oral anti-hypertensives to effect an immediate reduction in BP, inititiation or modification of outpatient antihypertensive regimen, or discharge for follow-up with no antihypertensive medications.
Two articles were identified which addressed the utility of ED testing in asymptomatic hypertensive patients (Karras 2008, Nishijima 2010). Typically, testing involves evaluation of a urinalysis and basic metabolic profile to assess for renal injury, an electrocardiogram to assess for cardiac ischemia, and a complete blood count to evaluate for anemia. In these two studies, 6%-7.2% of patients had abnormal results that resulted in a change in management. Both of these were multicenter studies assessed at urban EDs, and therefore may not be externally valid to community EDs, managed healthcare institutions, or countries with national healthcare systems. As the majority of management changes involved hospital admission, the lack of primary care follow-up in urban EDs, with largely uninsured and Medicare/Medicaid patients (Asplin 2005, Owens 2009), may have artifically inflated the proportion of patients whose management was altered. Insurance status has been shown to influence admission rates in other disease processes, including TIA (Chaudhry 2013) and venous thromboembolism (Misky 2011). Additionally, these two studies on asymptomatic hypertension did not assess the long-term impact of ED testing on patient-important outcomes (stroke, MI, death, renal failure).
Only one article addressing the effect of rapid reduction of BP in patients with asymptomatic serevely elevation hypertension in the ED (Zeller 1989). This study demonstrated similar reductions in BP one week after ED presentation among patients discharged on an outpatient antihypertensive regimen in all 3 treatment groups: 1) patients given 0.1 mg of clonidine in the ED every hour for up to 5 hours until sufficient BP reduction was achieved; 2) patients given placebo every hour for up to 5 hours; 3) patients discharged from the ED immediately. While this study assessed a surrogate outcome (mean change in BP), it seems unlikely that short-term elevations in BP (less than one week) would cause a significant increased in the risk of more important outcomes previously described.
The American College of Emergency Physicians has a clinical policy addressing the evaluation and management of patients with asymptomatic hypertension. Unfortunately, this policy is based on limited available evidence, and no level A recommendations were made. With regards to the accuracy and reliability of BP readings in the ED in asymptomatic patients, 2 recommendations were made
1) A level B was made that patients with persistently elevated BP (sytolic BP > 140 mmHg, diastolic BP > 90 mmHg) should be referred for follow-up
2) A level C recommendation was made that patients with a single elevated BP may need further screening as outpatients.
With regards to the rapid lowering of BP in asymptomatic patients, 3 level B recommendations were made:
1) Initiation of treatment is not necessary in patients with follow-up.
2) Rapid lowering of BP is unnecessary and potentially harmful.
3) Management, when initiated, should attempt to gradually lower the BP, and should not be expected to normalize the BP during the ED stay.
This clinical policy was limited, not only by the lack of available evidence, but also in its failure to address patient values and preferences; the policy was based purely on physician and nursing input. Additionally, policies and guidelines are often difficult to interpret due to lack of a standardized system for grading the evidence and the strength of the recommendations. The GRADE working group (Grading of Recommendations Assessment, Development, and Evaluation) has been working to address these shortcomings and standardize grading across multiple organizations. This system recommends assigning a “strength of recommendation” based upon the uncertainty associated with risks and benefits of an intervention, rather than physician preference and gestalt.
The current literature supports a focused laboratory assessment of patients with asymptomatic hypertension in the ED, particular among patients with poor follow-up. It seems reasonable to initiate outpatient antihypertensive management in patients with persistently elevated BP in the ED, again particularly in those in whom early follow-up is not guaranteed. The rapid lowering of BP in the ED in patients without symptoms of end-organ damage does not seem to provide any benefit with respect to BP measurement at one week, and likely provides no decrease in the risk of patient-important outcomes. Additionally, this practice may actually lead to increased short-term risk, and should be avoided.
Synovial Lactate in the Diagnosis of Septic Arthritis
Jun 01, 2013
Show notes
Journal Club Podcast #1 This month, I sit down with Chris Carpenter, evidence-based medicine guru, and talk about the diagnostic accuracy of synovial lactate in septic arthritis... Download the Podcast Click Tab to Expand Articles: Article 1: Evidence Based Diagnostics: Adult Septic Arthritis, Acad Emerg Med 2011; 18(8):781-796. (http://pmid.us/21843213) Answer Key Article 2: D-lactic acid in synovial fluid. A rapid diagnostic test for bacterial synovitis, J Rheumatol 1995; 22: 1504-1508. (http://pmid.us/7473474) Answer Key Article 3: Synovial fluid lactic acid: A diagnostic aid in septic arthritis, Arthritis Rheum, 1978; 21(7):774-779. (http://pmid.us/697948) Answer Key Article 4: Synovial fluid lactic acid in septic arthritis, N Z Med J 1981; 93(678): 115-117. (http://pmid.us/6943453) Answer Key Vignette: While working another crowded EM2 shift, you evaluate a 40-year old construction worker with a painful and swollen right knee. He cannot recall any injury to the knee and has no significant past medical history, including no prior crystalloid arthropathy (gout), knee surgeries, or endovascular infections (endocarditis). The knee hurt yesterday, but is exquisitely painful today with a palpable effusion and no notable overlying erythema or warmth to the touch. You note no surgical scars or overlying abrasions. He does not take any daily medications and cannot recall his last primary care physician evaluation. His vitals are BP 140/80, P 60, RR 18, T 36.8°C, and pulse ox 100%. He is lean and muscular without any other remarkable findings on physical exam, including no genitourinary complaints. Passive or active range of motion of his right knee is extremely uncomfortable for him. You promptly order morphine analgesia for this patient and then contemplate the differential diagnosis of a unilateral swollen and painful knee. In the absence of traumatic injury, you doubt a hemearthrosis, although he is a construction worker so you cannot exclude an occult twisting injury or overuse syndrome. However, the two primary diagnoses you consider most pertinent to rule-in or rule-out today are crystalloid arthropathies or septic arthritis. Recent debates about the American College of Emergency Physician’s decision not to participate in the Choosing Wisely campaign linger in your mind as you ponder what resources exist to “choose wisely” (pros and cons) in ED diagnostic decision making (see http://tinyurl.com/bs2vjj7). You speculate on the diagnostic accuracy of serum tests (CBC, ESR, CRP) and ponder the risk/benefits of arthrocentesis. After reviewing Roberts and Hedges’ Clinical Procedures in Emergency Medicine 5th Edition chapter describing the methods to perform and interpret knee joint aspirations (page 980-983), you discuss the procedure with the patient. After obtaining informed consent, you prep the patient and perform a time-out before administering 7cc of 1% lidocaine. The subsequent joint aspirate, obtained without significant patient discomfort or other adverse sequelae, is a cloudy yellow hue. You send the fluid to the lab and request a synovial lactate because you read somewhere that is a valuable test to evaluate for septic arthritis. The lab calls you back 30-minutes later telling you that they have no protocol for synovial lactate and cannot run the specimen for lactate. You turn to the medical literature to explore this diagnosis further.PICO Question: PICO Question Population: Adult patients with suspected septic arthritis Intervention: History/physical exam, sWBC, sLactate Comparison: Unaided clinical gestalt Outcome: Diagnostic accuracy (sensitivity, specificity, LR’s)Search Strategy: EBM has taught you that well-done meta-analyses can be high-yield products for busy clinicians. Therefore, you turn to PUBMED to conduct a diagnostic study Clinical Query using the search term “septic arthritis” and select the systematic reviews (108 citations – see http://tinyurl.com/9822z3r). Among the first 10 citations are two high-quality reviews, but one of the reviews derives from the other so you choose the meta-analysis on this topic which yields the remainder of the manuscripts that you review.Bottom Line: The differential diagnosis for acute monoarticular arthritis presenting to the emergency department is broad and includes infections (bacterial, fungal, mycobacterial, viral), crystalloid arthropathy (gout, pseudogout), rheumatoid arthritis, and trauma. Based upon the sole emergency medicine-centric systematic review available on this topic, the best estimate pre-test probability for septic arthritis in the emergency department is 27%. This means that among all emergency department patients in whom the clinician believes that septic arthritis is sufficiently likely to merit an arthrocentesis, 27 in 100 will actually have non-gonococcal septic arthritis. Most experts agree that this is an overestimate, but there is currently no better estimate of pre-test probability to facilitate Bayesian reasoning. Clinicians are able to deduce the etiology of acute nontraumatic joint pain/swelling within 3-days in most cases, but in an era of “overdiagnosis” and “overtreatment” Emergency physicians lack the luxury of a 3-day admission for most monoarticular arthritis patients. Emergency physicians’ clinical decision making often skews to the “rule out worst case scenario” model, which in the case of an acutely painful/swollen joint includes septic arthritis. About 50% of septic arthritis cases involve the knee, but any synovial space can be infected. Septic arthritis management options include surgical drainage and systemic antibiotics, although needle aspiration has also been evaluated in select cases. Short-term mortality for treated septic arthritis ranges from 3% to 11% (Cooper 1986, Deesomchok 1990, Kaandorp 1995, Gupta 2001). As is true with the majority of emergency medicine diagnoses (and medicine/surgery diagnoses), there is a paucity of diagnostic research to illuminate best-evidence practices for history, physical exam, clinical gestalt, and lab testing for septic arthritis. In fact, diagnostic studies of history and physical exam to evaluate septic arthritis in any setting are virtually non-existent. We identified two studies assessing historical risk factors and found none that evaluated physical exam. None of the studies (including the lab test studies) adheres to the STARD criteria in design or intent. Therefore, the available evidence may provide skewed estimates of diagnostic accuracy because of various biases, including spectrum bias, double-gold standard bias, and incorporation bias. Nonetheless, the available evidence for historical risk factors identified just three factors with a likelihood ratio > 3: prosthetic joint with overlying skin infection (LR+ 15), joint surgery within the preceding 3-months (LR+ 6.9), and age > 80 (LR+ 3.5). The absence of risk factors does not significantly reduce the probability of septic arthritis (the range of LR- was 0.64-0.93). To make this diagnosis even more challenging, commonly available serum tests (WBC, ESR, CRP) for septic arthritis are inaccurate and probably worthless acutely (see below). Eliminating these serum tests from the lexicon of reflexive emergency department testing for septic arthritis is an attractive target to “Choose Wisely”(see Bukata essay 1 and essay 2) in reducing unhelpful, potentially expensive laboratory testing. During Journal Club, Orthopedic surgery opined that serial assessment of ESR and CRP is anecdotally helpful for the longitudinal diagnosis and prognosis in the evaluation of suspected septic arthritis. However, nobody could identify any evidence to support this practice. Although serum procalcitonin and tumor necrosis factor (as does PCR testing to identify the specific infecting organism within 3-hours) have promising positive likelihood ratios, these tests are not commonly available in the emergency department. LR+ LR- WBC* 1.4-1.7 0.28-0.84 ESR* 1.3-7.0 0.17-2.4 CRP* 1.1-4.5 0.3-0.7 Procalcitonin 5-∞ 0.3-0.7 TNF ∞ 0.7 IL-6 1.5 0.9 IL-β 3.2 0.8 The suboptimal diagnostic test characteristics for history, physical exam, and serum tests leaves synovial testing as the standard of care diagnostic strategy to evaluate for septic arthritis. The sole systematic review noted that the risk of arthrocentesis includes post-procedural iatrogenic infections with a range of 0.01% in healthy populations to 0.05% in immunocompromised patients. During this Journal Club, Orthopedic surgery noted that: 1. They would prefer to perform the first and only arthrocentesis for suspected post-op joint infections to avoid iatrogenic complications (i.e. arthrocentesis-related infections) 2. Prosthetic joints typically yield less pronounced (lower) synovial leukocytosis than do native joint infections. So how well does the synovial fluid distinguish non-gonococcal bacterial arthritis from other acute joint problems? Synovial gram stain has sensitivity 29% - 65% with an undefined specificity (Argen 1966, Goldenberg 1976, McCutchan 1990, Faraj 2002, McGillicuddy 2007). A synovial WBC > 100,000 cells/mm3 has an interval LR of infinity, whereas a synovial WBC 0-25,000 has interval LR of 0.33. Our review of the synovial lactate data in conjunction with Laboratory Medicine and Orthopedic Surgery for this Journal Club suggests that synovial lactate is a promising test for the future ED evaluation of suspected acute septic arthritis. Unfortunately, two of the three studies did not distinguish whether they assessed D- or L-lactate! Bacteria, not humans (with rare exceptions such as short gut syndrome), produce D-lactate. On the other hand, humans, not bacteria, produce L-lactate. Laboratory Medicine believes that D-lactate is biologically plausible as a diagnostic marker of bacterial arthritis, but no commercially available laboratory test currently exists for D-lactate. D-lactate is a 3-day mail out test to Mayo Clinic. On the other hand, Laboratory Medicine hypothesizes that L-lactate is correlated with the synovial white blood cell count with higher sWBC counts producing more L-lactate. This hypothesis has not been tested. In fact, none of the synovial lactate studies evaluated the L- or D-lactate real-time. Instead, specimens were frozen and tested in batches long after the patient care episode was past. Here is the study-by-study synopsis of the available data. Gratacós 1995 This is the only study to specify which form of lactate they assessed (D-lactate). Interval LR’s range from 0.16 (D-lactate 0-0.05 mmol/L) to 20 (D-lactate >0.15 mmol/L). This test is superior to synovial WBC > 50,000 cells/mm3 (LR+ 9.3, LR- 0.47) or sPMN > 90% (LR+ 2.7, LR- 0.37) and may be particularly useful for partially treated septic arthritis. Brook 1978 At a threshold of 5.5 mmol/L (note 10-fold higher than Gratacós’ study which may indicate that Brook was studying L-lactate) the synovial lactate LR+ is 5.9 and LR- 0.04. More importantly, the interval LR for 0-5.5 mmol/L is 0.06 and for >16.7 mmol/L it is infinity. Unfortunately, the authors do not describe whether they evaluated D-lactate or L-lactate and by failing to adhere to STARD criteria, the authors leave open the possibility for significant biases, most of which skew estimates of sensitivity and specificity upwards. Moss 1981 Based on this case-control, single-center, Rheumatology clinic study, synovial lactate assays (probably L-lactate based on the investigator’s discussion) using the Calbiochem-Behring Rapid Lactate Kit accurately discriminates non-GC septic arthritis from other etiologies of acute monoarticular joint pain/swelling. At a threshold of 10 mmoL/L (which is nearly two-fold the 5.5 mmoL/L threshold proposed by Brook 1978) the LR+ is 22 and the LR- is 0. Using the Brook 5 mmoL/L threshold, Moss demonstrates LR+ 2.7 and LR- 0 (compared with 5.9 and 0.04 for Brook, respectively) for lactate. The interval LR for 0-10 mmol/L is zero versus 17.2 for 10-20 mmoL/L and infinity for synovial lactate >20 mmoL/L. In summary, Lab Medicine, Ortho, and EM agreed that the available data on synovial lactate is insufficient to change the diagnostic management of these patients. However, all agreed that sufficient clinical equipoise exists to justify additional research. Future studies should assess diagnostic accuracy in a consecutive sampling of ED patients with monoarticular arthritis in whom there is sufficient suspicion of non-GC septic arthritis to perform an arthrocentesis. It will be essential for these future studies to follow the STARD criteria and to report interval LR’s. If these “level 2” diagnostic accuracy studies confirm synovial lactate as a useful adjunct to synovial WBC, the logical progression in research would be to assess the impact of awareness of synovial lactate on clinician decision-making.
Prior to 2005, standard practice in treatment of STEMI in many emergency departments included giving IV beta-blockers (typically metoprolol 5 mg IV q 5 minutes up to 3 doses, unless contraindicated by hypotension or bradycardia). Patients were then admitted and given oral beta-blockers.
In 2005, the COMMIT trial (aka 2nd Chinese Cardiac Study/CCS-2) was published in Lancet: 45,852 patients were randomized to receive either placebo or metoprolol (15 mg IV in the emergency department followed by 50 mg PO q 6 hours for 48 hours, followed by 200 mg extended release metoprolol daily for 4 weeks). There was no difference between the treatment and control groups for the primary outcome (composite of death, reinfarction, or cardiac arrest) (odds ratio [OR] 0.96, 95% CI 0.90-1.01) or for death alone (OR 0.99, 95% CI 0.92-1.05). However, there was a significant increase in the risk of cardiogenic shock in patients receiving metoprolol (OR 1.30, 95% CI 1.19-1.41).
The result for our emergency department (and many others) was that we stopped giving immediate metoprolol to STEMI patients (though they still received oral metoprolol once admitted). The AHA/ACC STEMI guidelines changed as well, though not so drastically as our practice. The wording in 2004 was that “oral beta-blocker therapy should be administered promptly to those patients without a contraindication,” compared to the current wording: “oral beta blockers should be initiated in the first 24 hours in patients with STEMI.” The guidelines remained unchanged with regards to IV beta-blockers, which they state “it is reasonable to administer…promptly to STEMI patients without contraindications.”
Methods:
This observational study used data collected prospectively in the Austrian Myocardial Infarction Network from June 1, 2006 to December 31, 2010. The use of bisoprolol (2.5 mg orally) was recommended by protocol to be given by the emergency physician either out-of-hospital or in the ED within 30 minutes of confirmation of an acute STEMI. If bisoprolol was not given within 30 minutes, the protocol recommended it be given 24 hours after the first ECG. The decision to give bisoprolol immediately was made by the physician on duty. Contraindications included:
1) Systolic blood pressure (SBP) < 100 mmHg
2) Bradycardia (HR < 60 bpm)
3) 2nd or 3rd degree AV block
4) Clinical signs of heart failure (bilateral rales, cyanosis).
All patient eligible for percutaneous intervention (PCI) also received aspirin, clopidogrel, bivalirudin, and abciximab. Patients receiving rtPA were given aspirin, clopidogrel, and unfractionated heparin.
The primary outcome was all-cause mortality greater than 48 hours after presentation: only patients who survived the first 48 hours were included in the survival analyses. Secondary outcomes included cardiovascular death at 30 days and 1 year. Mortality was also assessed for two predefined subgroups: elderly patients aged > 70 years and patients with ≥ 2 COMMIT shock index criteria (age > 70, symptoms > 12 hours, SBP < 120 mmHg, and HR > 110 bpm). Follow-up data on mortality was obtained from the Statistical Department of the Austrian government that collects data on the cause of death in all patients who die in a hospital in Austria.
Results:
664 patients with STEMI were analyzed: 343 (52%) received immediate β-blocker therapy while 321 (48%) received delayed β-blocker therapy. The average age was 64, with 70% male. The two groups were similar with respect to age, gender, previous medical history, time from onset of pain to first ECG, time from first ECG to PCI, and reperfusion strategy. The delayed β-blocker group had significantly lower initial mean SBP (129 vs. 146 mmHg, p < 0.001), lower initial mean diastolic blood pressure (77 vs. 85 mmHg, p < 0.001), and lower initial mean heart rate (75 vs. 82 bpm, p <0.001).
24 patients died within the first 48 hours (immediate β-blocker group: 5, delayed β-blocker group: 19), all related to the MI. For the remaining 640 patients, survival analysis revealed an overall mortality of 19.2% in the delayed β-blocker group vs. 10.7% in the immediate β-blocker group (p = 0.0022, RR = 0.56, NNT 11.7). Thirty-day and one-year cardiovascular mortality were lower in the immediate β-blocker group than the delayed β-blocker group, as were 30-day and one-year all-cause mortality (Table 1). The use of immediate β-blocker was not associated with an increase in all-cause mortality in either those with an increased risk for cardiogenic shock, or those over 70 years of age.
Table 1. 30-day and 1-year cardiovascular (CV) and all-cause mortality
Delayed βBL
Immediate βBL
P -value
30-day CV mortality
4.34%
0.89%
0.0058
1-year CV mortality
8.14%
1.88%
0.0003
Total CV mortality
13.41%
5.17%
0.0002
30-day all-cause mortality
4.66%
0.89%
0.0033
1-year all-cause mortality
9.79%
3.23%
0.0006
Total all-cause mortality
19.19%
10.73%
0.0022
Commentary:
If we believe this study, we need to give 12 STEMI patients 2.5 mf of oral bisoprolol (equivalent to 25 mg metoprolol) to save one life over the next 4.4 years, or 26 patients to save one life in the next month (these numbers would be even lower if we included deaths in the first 48 hours). Compare this with a NNT of 42 for aspirin (ISIS-2) and 50 for PCI (vs. thrombolysis) (Wang 2009) to prevent one death at one month.
It seems biologically implausible that giving a small dose of an oral blocker 24 hours earlier in STEMI could result in such a large reduction in mortality. This is especially true considering the large mortality difference seen in the first 48 hours (5 in the immediate group vs. 19 in the delayed group). The authors themselves note “The assumption is that patients dying within the first 48 hrs have died independently of immediate or delayed β-blocker treatment, e.g., death occurred prior to a sufficient β-blocker effect,” but go on to attribute the large mortality reduction seen to treatment effect. It would seem more plausible, given the observational non-randomized nature of the study, that a prognostic imbalance existed between the two groups at the beginning of the study that would explain the difference in outcomes (rather than an effect of the treatment), including:
· Reported: the only reported differences were SBP, diastolic blood pressure, and heart rate, which were all lower in the control group. While decreased SBP has been shown to increase mortality in STEMI, lower heart rates have been shown to be protective (Gale 2008).
· Not reported: location of MI (anterior STEMI has been shown to predict increased mortality compare to other locations [Stone 2008])
· Not reported - hx of CHF has been shown to increase mortality (Gale 2008).
· Hidden: the proportion of patients with ≥ 2 shock index criteria was significantly imbalanced: 79 (26.2%) patients in the delayed β-blocker group vs. 54 (16.0%) in the immediate β-blocker group (p = 0.0016). And that’s in patients who survived the first 48 hours!!!
· Hidden: there was in increase in both the incidence of hypotension (17 vs. 5, p = 0.008) and bradycardia (41 vs. 14, p < 0.001) in the delayed treatment group. It seems unlikely that withholding beta-blockers for 24 hours would result in an increase in these complications during that same time period. It is more likely that a prognostic imbalance put the control group at increased risk of both of these.
Bottom Line:
Despite a large reduction in mortality (19.19% vs. 10.73%) with immediate vs. delayed oral beta-blocker administration in STEMI, it seems unlikely that this difference was due to treatment effect, and more plausible that the difference was due to prognostic imbalance. As the authors state, a randomized-controlled trial to assess the efficacy of immediate beta-blockers would be helpful to resolve this issue. Randomization attempts to balance both known and unknown confounders between treatment groups, so that any difference in outcome can be attributed to treatment effect.
EBM Teaching Point:
The authors used a survival analysis, and presented Kaplan-Meier curves of the results for mortality over time. If we take the total all-cause mortality for the delayed beta-blocker group as an example, we see 41 deaths out of 302 participants. If we use these numbers to calculate a percent, we do not arrive at the 19.19% probability of death reported. This is because all of the patients could not be followed for the same duration. Imagine a patient enrolled on day 1 of the study (June 1, 2006), who is followed for approximately 4.5 years (the last point plotted on the Kaplan-Meier curve): we know whether or not that patient died over a 4.5-year period. Now imagine a patient enrolled on the last day of the study (December 31, 2010). If we followed this patient for 4.5 years, we would still be watching for the outcome (and would continue doing so until July of 2015!). Instead, the authors took the follow-up data they had until the point at which they stopped looking at outcomes. Entering this data into statistical software will yield a survival analysis and Kaplan-Meier curve, which is used to compute an estimate of survival probability.
Our TOPPODCAST Picks 
Stay Connected
