Use of Common Clinical Laboratory Tests to Assess Infectious and Inflammatory Diseases

Physiology and Measurement

The erythrocyte sedimentation rate (ESR) measures in vitro the fall of the red blood cell (RBC) column in a vertical suspension of anticoagulated plasma over 60 minutes. A surrogate of in vivo inflammation, the ESR generally rises over 24–48 hours after stimulus and returns to baseline more slowly than other acute phase reactants, thus lagging resolution of illness. Sedimentation depends on RBC mass, volume, shape, RBC−RBC forces, and the protein constitution of plasma. Electrostatic forces normally cause RBCs to repel each other and inhibit their aggregation. Increased amounts of plasma fibrinogen or globulins coat the RBCs, foster aggregation, and hasten settling, consequently elevating the ESR. Large-molecular-weight, needle-shaped fibrinogen has the greatest effect, followed by β-globulins and distantly by α ₂ -globulins, γ-globulins, and albumin.

Normal ESR values are age and sex dependent (androgens lowering the ESR). Top normal are 2 mm/hr in the first few days of life, 15 mm/hr for children 1 month to puberty, 15 mm/hr for boys >12 years of age and men, and 20 mm/hr for girls >12 years of age and women. After puberty, the normal top value rises 0.85 mm/hr for every 5-year increase in age. Obese children have higher mean ESRs than lean children (e.g., 20 vs. 10 mm/hr at 9–10 years of age). ^,

Elevated Sedimentation Rate

The ESR is elevated in most bacterial, mycobacterial, and fungal infections and is normal or mildly elevated in uncomplicated viral, rickettsial, and ehrlichial infections. An elevated ESR (>20 or >30 mm/hr, depending on the study) has poor discriminating power as a single test (i.e., sensitivity and specificity ≤50%) to predict bacterial infection in children with nonspecific febrile illnesses of short duration. The mean ESR during most viral upper respiratory tract infections is approximately 20 mm/hr (with 90% of values <30 mm/hr), with noteworthy exceptions such as during adenovirus infections, in which mean ESR can be 40 mm/hr. ^, For patients with fever of unknown origin, a normal or mildly elevated ESR is predictive of a nonserious or viral disease in >90% of cases; ^, an ESR >50 mm/hr in such patients can be an impetus for more extensive evaluation. ^,

Usefulness of the ESR in children with osteoarticular conditions has been studied extensively. Several studies of children with bacterial arthritis have documented ESRs >30 mm/hr in 80% to 90% of cases (mean, ∼65 mm/hr). Prospective validation of Kocher criteria (fever ≥38.5°C, inability to bear weight, ESR ≥40 mm/hr, and systemic white blood cell [WBC] count >12,000 cells/μL) for distinguishing pyogenic arthritis of the hip from toxic synovitis showed an area under the receiver operating characteristic (ROC) curve of 0.86. In a retrospective case series in a primary referral hospital, positivity of all 4 Kocher criteria in addition to C-reactive protein (CRP) ≥2 mg/dL had a predictive value of only 60% for bacterial arthritis, and in series from the US and Europe, Kocher scores were low in children with Lyme and Kingella arthritis of the hip. ^, In children with bacterial arthritis, the ESR frequently rises on days 2–4 of effective therapy, and an ESR >30 mm/hr persists in at least 50% of cases at the conclusion of successful therapy. ^, The peak ESR in acute osteomyelitis generally is lower than in bacterial arthritis (mean, ∼45 mm/hr), with expectations of a gradual fall and normal values at an appropriate conclusion of therapy. The ESR can be normal in acute osteomyelitis or bacterial arthritis of the small bones/joints of the hand and foot. In subacute or chronic osteomyelitis at any site, the ESR usually is only mildly elevated (25–40 mm/hr) but can be normal.

An elevated ESR characteristically occurs in multiple noninfectious inflammatory conditions, as well as in noninflammatory states that affect RBC stearic properties or plasma volume, fibrinogen, globulins, or protein concentrations ( Box 288.2 ). Conversely, during infectious diseases, certain underlying conditions or complications can result in a low ESR. The ESR usually exceeds 30 mm/hr during febrile episodes in children with periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis (PFAPA) syndrome and other autoinflammatory periodic fever symptoms. An ESR <50 mm/hr does not exclude Kawasaki disease but makes the diagnosis unlikely. Immune globulin intravenous therapy causes acute elevation of the ESR. ^,

Discordant results between the ESR and other markers of inflammation, particularly CRP, can be clues to infectious and noninfectious diseases. Relatively high CRP with mildly elevated ESR can be a clue to Kawasaki disease, severe inflammatory bowel disease, multisystem inflammatory syndrome associated with SARS-CoV-2 in children (MIS-C) and e-cigarette (vaping) associated lung injury. High ESR with normal or mildly elevated CRP can be a clue to underlying autoimmune disease (especially systemic lupus erythematosus), ^, scurvy associated with severely restricted diet, chronic indolent infections such as endocarditis and tuberculosis, or treatment with an agent such as tocilizumab that inhibits activity of IL-6.

Extreme Elevation of the Sedimentation Rate

Box 288.3 shows infectious and noninfectious conditions that should be considered especially when the ESR exceeds 100 mm/hr. Extreme elevation of the ESR is characteristic of certain noninfectious conditions. In children (unlike adults), however, infection still is a likely cause, accounting for 56% of cases in 1 study of 156 children; miliary tuberculosis is notable among infectious diseases. Collagen vascular, renal, and macrophage disorders and neoplastic disease rise among differential diagnoses. Extreme elevation of the ESR at the diagnosis of bacterial infection does not in itself indicate complicated disease or a guarded prognosis.

Low Sedimentation Rate

Causes of a low ESR are shown in Box 288.2 . In children with infectious diseases, an abnormally low ESR (e.g., <4 mm/hr) is most frequently a sign of disseminated intravascular coagulopathy and reflects low plasma fibrinogen concentration; elevated fibrinolytic markers (e.g., D-dimer) are expected. Administration of glucocorticoids and possibly high-dose (but not low-dose) salicylates and l -asparaginase lower the ESR, as does performance of peritoneal dialysis or hemodialysis. Sickling of RBCs prevents aggregation, lowering the ESR. The ESR is not valueless in children with sickle cell disease, however. ^, In one study, 72% of hospitalized children with an ESR >20 mm/hr had clinical infection (and 38% had vaso-occlusive crisis) compared with 23% with an ESR of ≤20 mm/hr, 74% of whom had vaso-occlusive crisis.

Physiology and Measurement

The production of IL-6, and less so IL-1β, stimulates hepatocytes to produce CRP, a highly conserved protein so named because of precipitation with C-polysaccharide of the pneumococcal cell wall. Although the liver is the main source of CRP, small amounts are produced at extrahepatic sites, including in macrophages, monocytes, lymphocytes, neurons, respiratory and renal epithelial cells, and smooth muscle cells. CRP is a marker of infection, cell damage and inflammation. The CRP level increases within 4–6 hours of an inflammatory stimulus, doubles every 8 hours, peaks at levels 100–1000 times normal within 1–3 days, and then falls relatively rapidly (with a half-life of 19 hours) once the triggering stimulus ceases. CRP also participates functionally in inflammatory processes and host responses to infection through activation of the complement pathway, apoptosis, phagocytosis, nitric oxide (NO) release, the production of cytokines (particularly interleukin-6 and TNF). A main role of CRP in inflammation is its activation of the C1q molecule in the complement pathway leading to the opsonization of pathogens. Nephelometric assays remain the most widely used techniques for measuring CRP; a value <1 mg/dL (<10 mg/L) is considered “normal” for clinical predictive purposes. Normal range is variable by the test used. Modest long-term elevations (>0.3 mg/dL) using high-sensitivity CRP (hsCRP) tests have been used as a marker of increased risk for atherosclerosis, ^{, ,} but value of CRP compared with clinical risk factors, or IL-6, has been questioned.

Compared with the ESR, CRP rises more rapidly, peaks earlier, and returns to normal levels more quickly when the stimulus abates. Factors that affect the ESR, such as anemia, polycythemia, protein levels, immunoglobulin infusion and RBC shape, do not affect CRP values. Any tissue injury, however, such as trauma, burn injury, ischemia, or infarction, can elicit production of CRP. Because CRP is degraded rapidly, serial determinations can provide additional information on progression of an insult or the effectiveness of therapy. ^, Liver failure and certain rare inherited immunodeficiencies (e.g., nuclear factor-κB [NF-κB] pathway defects) can cause “falsely” low CRP levels in the presence of invasive bacterial or fungal infection.

Clinical Use

In general, CRP levels are elevated substantially (5–35 mg/dL) in bacterial infections with a tissue site of infection and only modestly (<2–4 mg/dL) in most acute viral infections; a value >10 mg/dL is more likely to be associated with bacterial infection, although infections with certain viruses (e.g., adenovirus, cytomegalovirus [CMV], influenza, measles, and mumps), fungi and protozoa can be associated with CRP levels >10 mg/dL. ^, Limitations in validating the accuracy of elevated CRP to predict serious bacterial infection (SBI) in neonates and infants with fever but without a clinical focus of infection include variable cutoff values used and the low incidence of SBI in studies. Pooled estimates in studies with cutoff values of 1–7 mg/dL showed sensitivity for SBI of 63%–95% and specificity 40%–91%, and 6 studies, 3 of which used a cutoff value of 4 mg/dL, showed sensitivity of 77% and specificity of 79%. In neonates, sensitivity for culture-proven early onset sepsis (EOS) was only 39%. Although combination of CRP with other biomarkers such as IL-6, IL-8, and PCT improve diagnostic accuracy for EOS, ^, currently no biomarker has sufficiently exquisite predictive value to drive clinical management. “Elevated” CRP can occur spontaneously in neonates in the first 48 hours of life (≤1 mg/dL), in the 48 hours after injections of multiple vaccines (>1 mg/dL), and 1–2 days following major surgical procedures (median peak, 2 mg/dL). CRP appears to be less sensitive than IL-8 or PCT for detecting SBI in oncologic patients with fever and neutropenia.

CRP level has better predictive value for distinguishing bacterial infection when a tissue focus of infection is identified. A meta-analysis of 8 studies including 1230 children with pneumonia showed a positive predictive value (PPV) of CRP >4 mg/dL for bacterial pneumonia of 64%. However, the Clinical Practice Guidelines of the Pediatric Infectious Diseases Society conclude only that CRP may help to assess response to therapy (with low-quality evidence). In one study, CRP level >2 mg/dL was an independent risk factor for pyogenic arthritis (odds ratio, 14.5; 95% confidence interval [CI], 3.2–64.9) compared with toxic synovitis. In adults, CRP-guided decision making for use of antibiotics in exacerbations of COPD led to reduction in unnecessary antibiotic use.

Discordantly high CRP levels compared with other acute phase reactants (especially WBCs and ESR) have been noted in Kawasaki disease, focal bacterial infections (in 1 study 26% of cases had a WBC count <15,000/μL and CRP ≥8 mg/dL), Crohn disease, neoplasms of the liver and reticuloendothelial system, Castleman disease, allograft rejection and graft-versus-host disease, and severe burns. ^, Tocilizumab and other inhibitors of IL-6 lower the CRP, even in the face of bacterial infection, but do not acutely affect the ESR. In 66 children with MIS-C in two New York studies, median CRP levels were 21 and 25 mg/dL ESR 53 mm/hr and total WBC 9,000 and 11,000 cells/μL, respectively. ^,

Sequential measurement of CRP is valuable as an assessment of therapeutic control of focal infections, including wound infections following spinal operations and infective endocarditis, and as a clue to ineffective therapy or complicated infection such as undrained abscess and tissue necrosis. ^{, ,} Serial CRP levels that increase or fail to decrease substantially after 48 hours of therapy suggest treatment failure, uncontrolled primary site of infection, a complication, or a noninfectious diagnosis. A meta-analysis of CRP values following abdominal operations (using variable threshold values by surgical procedure) showed an 84% NPV for infectious complication when measured on postoperative day 4. Serially measured normal CRP values have a better NPV for bacterial infection in certain clinical settings, including early onset neonatal sepsis ^{, ,} and fever without a source in infants and young children. A CRP value of <2 to 3 mg/dL in an afebrile, clinically improving patient has been used as an adjunct marker for safe transition from parenteral to oral antibiotic therapy for patients with uncomplicated pediatric osteoarticular infections.

Serial concomitant testing of both ESR and CRP during acute infection likely is not beneficial compared with CRP measurements alone. In patients with chronic autoimmune diseases, such as systemic lupus erythematosus (SLE) with serositis, ESR usually is elevated, and CRP can be >4 mg/dL; ratio of ESR:CRP rises with SLE activity and falls with infection (because CRP rises disproportionately with superimposed infection). Following both values in SLE may assist in differentiating disease progression from acute infection during acute febrile episodes.