SGEM#287: Difficult to Breathe – It Could Be Pneumonia

Date: March 18th, 2020 Reference: Ebell et al. Accuracy of Biomarkers for the Diagnosis of Adult Community-Acquired Pneumonia: A Meta-analysis. AEM March 2020 Guest Skeptic: Dr. Chris Bond is an emergency medicine physician and assistant Professor at the University of Calgary. He is also an avid FOAM supporter/producer through various online outlets including TheSGEM. Disclaimer: This is Not an Episode on COVID19 Things are changing quickly with the COVID19 pandemic.  Here are five basic things you can do to help flatten the curve as of this blogpost: Wash your hands well and often (at least 20 seconds with soap and water) Try not to touch your face Physically isolate yourself from large gatherings but stay socially connected electronically Cough into your elbow or use a tissue, throw the tissue out and go to #1 Disinfect objects or surfaces with a regular household cleaning wipe or spray If you are unsure of what to do or for more information, here are five websites to get up-to-date information about COVID19: COVID19 Centre for Disease Control and Prevention Health Canada Public Health Ontario World Health Organization Food and Drug Administration   Case: A 47-year-old healthy, non-smoker, presents to the emergency department (ED) with a productive cough, fever and says it has been difficult to breathe for the past four days. He appears well, with a temperature of 38.7 Celsius, heart rate of 90 beats per minute, respiratory rate of 20 breaths per minute and room air oxygen saturation of 91%. On auscultation you hear some fine crackles at the bases. You wonder if there is value in ordering any bloodwork, particularly a biomarker such as C-reactive protein (CRP), procalcitonin (PCT) or a complete blood count for white blood cell count (WBC) in addition to doing a chest x-ray (CXR). Background: Community-acquired pneumonia (CAP) is a significant source of morbidity and mortality in adults (1,2).  We have covered this issue a couple of times on the SGEM. One episode looked at β-Lactam monotherapy vs. β-Lactam plus macrolide combination therapy in adult patients admitted to hospital with moderately severe CAP (SGEM#120). This study supported the combination therapy in these patients. More recently, we looked at the question of whether steroids improve morbidity and mortality in patients admitted to hospital with CAP (SGEM#216). The bottom line was that corticosteroids appear to improve mortality and/or morbidity in patients admitted to hospital with CAP. There is evidence that an accurate diagnosis of CAP may lead to earlier treatment while avoiding unnecessary antibiotics for patients who do not have CAP. Pervious research has demonstrated that individual signs and symptoms have limited accuracy in the diagnosis of CAP. The diagnosis of CAP is usually based on an abnormal chest x-ray in a patient with signs and symptoms of a lower respiratory tract infection (3,4). White blood cell count (WBC), C-reactive protein (CRP), and procalcitonin are biomarkers associated with an increased likelihood of CAP. There are also clinical prediction rules that include CRP for the diagnosis of CAP (5,6). Procalcitonin is another potential biomarker that may help in the diagnosis of bacterial pneumonia (7).  Guidelines such as the National Institute for Health and Care Excellence (NICE) recommend the use of CRP at the point of care to reduce inappropriate antibiotic when diagnosing CAP (8) These various biomarkers are readily available in the ED setting in the US, as well as in the primary care setting in other countries in Europe. The study we are reviewing on this SGEM episode performs an updated systematic review and meta-analysis (SRMA) of the diagnostic accuracy of biomarkers for CAP. Clinical Question: What is the accuracy of biomarkers for the diagnosis of community acquired pneumonia? Reference: Ebell et al. Accuracy of Biomarkers for the Diagnosis of Adult Community-Acquired Pneumonia: A Meta-analysis. AEM March 2020 Population: Adult patients presenting with symptoms of acute respiratory infection and patients with clinically suspected pneumonia based on physician order of a chest radiograph, reporting sufficient information to calculate sensitivity and specificity for the diagnosis of CAP for at least one biomarker. Exclusions: Studies of dyspnea or sepsis rather than suspected CAP. Studies limited to patients with chronic lung disease, patients in skilled nursing facilities, or immunosuppressed/HIV patients. Ventilator or hospital acquired pneumonia. Studies of the diagnosis of a specific pathogen (i.e. mycoplasma or legionella). Studies that did not use a cohort design (i.e. recruited patients with known CAP and healthy controls). Intervention: C-reactive protein (CRP), procalcitonin or white blood cell (WBC) count Comparison: Chest imaging with CXR or CT scan Outcome: Diagnosis accuracy of biomarkers for pneumonia This is an SGEMHOP episode which means we have the lead author on the show.  Dr. Mark Ebell is a Family Physician and Professor at the University of Georgia in Athens. He is a co-founder of POEMs, editor-in-chief of Essential Evidence, deputy editor of American Family Physician, and co-host of the podcast Primary Care Update.   Dr. Mark Ebell Authors’ Conclusions: Biomarkers can be useful for the diagnosis of community-acquired pneumonia. The cutoff chosen will determine whether the test is most useful for ruling out pneumonia (CRP < 10 or 20 mg/L) or for ruling in pneumonia (e.g., CRP > 50 or 100 mg/L). CRP is the most accurate of the three studied biomarkers that are currently being used to assist in the diagnosis of community acquired pneumonia. We note that CRP is inexpensive and readily available in many settings and may be easily integrated into the clinical workflow for diagnosis of community acquired pneumonia in appropriate patients. Quality Checklist for Therapeutic Systematic Reviews: The clinical question is sensible and answerable. Yes The search for studies was detailed and exhaustive. Yes The primary studies were of high methodological quality. Yes The methodological quality of primary studies were assessed for bias. Yes The assessment of studies were reproducible. Yes The outcomes were clinically relevant. Yes There was low heterogeneity for estimates of sensitivity or specificity. No There was low statistical heterogeneity for the primary outcomes. No  The treatment effect was large enough and precise enough to be clinically significant. Yes Key Results: They screened 829 studies and found 14 that met inclusion and exclusion criteria with a total number of 6,599 patients. The study time periods ranged from 1986 to 2016, with 12 studies being performed in Europe, and one in each of the United States and Chile. Half of the studies were performed in ED patients and the other half in primary care settings. CRP was studied in 13 of the 14 studies, PCT in seven and leukocytosis in five. One study used the combination of CRP and PCT. Eight studies were felt to be at low risk of bias using the QUADAS-2 tool while six studies were felt to be at moderate risk of bias. None of the studies appeared to have been industry funded. Key Result: Diagnostic accuracy for community acquired pneumonia  was greatest with CRP. All of these biomarkers have a threshold effect, meaning that sensitivity increases as specificity decreases. As a result, summary estimates of sensitivity, specificity and likelihood ratio are shown for different cutoffs for each test. Primary Outcome: Diagnostic accuracy of community acquired pneumonia C-Reactive Protein: A CRP cutoff of 10 mg/L had the highest sensitivity at 90% and lowest negative likelihood ratio of 0.27. CRP > 20 mg/L CRP > 50 mg/L and CRP > 100 mg/L had positive likelihood ratios of 2.08, 3.68 and 5.79 respectively, with poor negative likelihood ratios. Procalcitonin: PCT > 0.25 mcg/L and PCT > 0.50 mcg/L had good positive likelihood ratios (5.43 and 8.25 respectively), negative likelihood ratios were worse than for CRP Leukocytosis: This was defined as a white blood cell count (WBC) > 9.5 to 10.5 x 10^9 cells/L had modest accuracy (LR+ 3.15, LR- 0.54) with good homogeneity around this estimate. We asked Mark five questions to get a greater understand of his publication. Listen to the SGEMHOP podcast to hear all of his answers. 1) External Validity: Less than 1/3 of patients came from the ED setting. This limits the application of these results to this clinical setting. The NICE guideline recommends the use of CRP in the primary care setting, presumably as a point of care test to help decide whether or not to order a CXR. Is this a rational use of resources in an ED setting where a CXR could be done as the initial test? 2) Point Estimates and 95% Confidence Intervals: There have been some conventional cut offs for likelihood ratios. None of the positive likelihood ratios were >10 to confidently rule in pneumonia and none of the negative likelihood rations were <0.1 to confidently rule out pneumonia. There were generally wide confidence intervals around the point estimates. 3) Post-Hoc Cut Offs: It is not clear in some of the studies used a post-hoc cutoff. We have discussed this before on the SGEM of potentially overfitting the data. How do you think this could affect your results and the interpretation? 4) Imperfect Gold Standard Bias (Copper Standard Bias):The biomarkers were compared to CXR in 13 of the 14 studies. We know that CXRs is less accurate in diagnosing CAP than a CT scan. How do you think that could have impacted the results? 5) Clinically Significant: A positive CXR does not mean a patient has a bacterial pneumonia.