My patient with COPD exacerbation on corticosteroids has an elevated white blood cell and neutrophil count. How can I tell if his elevated neutrophil count is caused by the corticosteroids or an acute infection?

The most helpful lab data favoring corticosteroid-induced granulocytosis (CIG) is the absence of a shift to the left in the peripheral WBC (ie, no more than 6% band forms) and toxic granulation.1 Although the total WBC itself is less helpful, experimental studies have reported a mean maximum neutrophil counts 2.4 times the base line after IV injection of hydrocortisone (200 mg) 2, and a mean increase of 4,000 neutrophils/mm3 after prednisone (20-80 mg). 3

Several possible mechanisms for CIG revolving around altered neutrophil characteristics and dynamics have been proposed4, including

  • Reduced egress from blood into tissues
  • Demargination from vascular endothelial surfaces
  • Delayed apoptosis
  • Enhanced release from the bone marrow.

An experimental animal study reported that only 10% of CIG is related to bone marrow release of neutrophils with the rest related to demargination (61%) and reduced egress from blood or delayed apoptosis (29%).4 This study may explain why high percentage of band forms would not be expected CIG.


  1. Shoenfeld Y, Gurewich Y, Gallant LA, et al. Prednisone-induced leukocytosis: influence of dosage, method, and duration of administration on the degree of leukocytosis. Am J Med 1981;71:773-78. Link
  2. Bishop CR, Athens JW, Boggs DR, et al. Leukokinetic studies: A non-steady-state kinetic evaluation of the mechanism of cortisone-induced granulocytosis. J Clin Invest 1986;47:249-60.
  3. Dale DC, Fauci AS, Guerry DuPont, et al. Comparison of agents producing a neutrophilic leukocytosis in man. J Clin Invest 1975;56:808-13. PDF
  4. Nakagawa M, Terashma T, D’yachkova YD, et al. Glucocorticoid-induced granulocytosis: Contribution of marrow release and demargination of intravascular granulocytes. Circulation 1998;98:2307-13. PDF

If you liked this pearl, sign up under menu and receive future pearls right into your mailbox!


My patient with COPD exacerbation on corticosteroids has an elevated white blood cell and neutrophil count. How can I tell if his elevated neutrophil count is caused by the corticosteroids or an acute infection?

Can native valve infective endocarditis be associated with hemolytic anemia?

Yes, but it’s rare!  Hemolytic anemia (HA) in the setting of infective endocarditis (IE) has only been described in a few case reports (1-3).  Although diseased valves may cause shearing stress that fragments RBCs, similar to that associated with mechanical heart valves, an autoimmune hemolytic process has also been implicated. 

A 2018 case report describes a patient with hypertrophic obstructive cardiomyopathy (HOCM) with left ventricular outflow tract (LVOT) obstruction who had HA secondary to subacute IE due to Actinomyces israelii (1).   The anemia completely resolved after treating the IE (1). The cause was most likely mechanical shearing (schistocytes or fragmented RBCs present on peripheral smear) by the diseased valves; autoimmune hemolysis was considered unlikely in this case due to consistently negative Coombs tests and failure to respond to corticosteroids (1). 

An autoimmune mechanism was invoked by a 1999 report reviewing 6 cases of HA associated with IE (3).  All patients had fragmented erythrocytes, but several also demonstrated an immune-mediated mechanism for their HA, supported by the presence of spherocytes, splenomegaly, and + Coombs test (2,3).  The production of anti-erythrocyte antibodies, modification of antigenicity of erythrocyte antigens, or unmasking of antigens in IE may play a role (1,3). Additional evidence in support of an immune-mediated mechanism of HA in IE has been provided by an experimental study demonstrating significantly shorter RBC half-life in rabbits with intact spleen compared to that of splenectomized animals (4).



1. Toom S, Xu Y. Hemolytic anemia due to native valve subacute endocarditis with Actinomyces israellii infection. Clin Case Rep 2018;6: 376-79. 

2. Hsu CM, Lee PI, Chen JM, et al. Fatal Fusarium endocarditis complicated by hemolytic anemia and thrombocytopenia in an infant. Pediatr Infect Dis 1994;13:1146-48. 

3. Huang HL, Lin FC, Hung KC, et al. Hemolytic anemia in native valve infective endocarditis. Jpn Circ J 1999;63:400-403. 

4. Joyce RA, Sand MA. Mechanism of anaemia in experimental bacterial endocarditis. Scand J Haematol 1975;15:306-11. 


Contributed by Scott Goodwin, Medical Student, Harvard Medical School, Boston, MA. 


If you like this post, sign up under menu and receive future pearls from P4P right into your mailbox! Thank you!

Can native valve infective endocarditis be associated with hemolytic anemia?

How can I tell if my febrile patient who uses IV drugs had cotton fever?

Although IV drug use (IVDU) is associated with febrile illness of numerous etiologies (eg, soft tissue infections, pneumonia, bacteremia, endocarditis), certain features of a febrile illness may be helpful in considering cotton fever (CF) as the cause.1-3

First, onset of fever—often associated with chills, shortness of breath, nausea, vomiting, headache, abdominal pain and myalgias—in CF is usually manifest within 10-30 minutes of drug injection. Second, infectious disease workup, including blood cultures and chest radiograph, are unrevealing despite clinical signs of systemic inflammatory response syndrome (SIRS), such as leukocytosis, tachypnea and tachycardia. Third, symptoms and clinical signs of inflammation usually resolve or improve within 6-12 h of onset (less commonly up to 24-48 h). Nevertheless, CF remains a diagnosis of exclusion.

As for the cause of CF, the most widely-held theory revolves around the endotoxin of Pentoea agglomerans (formerly Enterobacter agglomerans), a gram-negative rod that colonizes cotton plants. Since cotton is often used as a filter during injection of illicit substances, any endotoxin present in the cotton is also injected resulting in abrupt onset of a febrile illness. Of note, the toxin is water soluble and heating (often part of the preparation of the drug) enhances its toxic effect.3


  1. Zerr AM, Ku K, Kara A. Cotton Fever: a condition self-diagnosed by IV drug users. JABFM 2016;29: 276-279.PDF
  2. Xie Y, Pope BA, Hunter AJ. Cotton fever: does the patient know best? J Gen Intern Med 31:442-4. PDF
  3. Torka P, Gill S. Cotton fever: an evanescent process mimicking sepsis in an intravenous drug abuser. J Emerg Med 2013;44:e385-e387. PDF
How can I tell if my febrile patient who uses IV drugs had cotton fever?

Is it possible to have acute pancreatitis with normal serum lipase?

Yes! Although an elevated serum lipase has a negative predictive value of 94%-100% for acute pancreatitis (1), there are ample reports in the literature of patients with CT findings of pancreatitis in the presence of abdominal symptoms but with normal serum lipase and/or amylase (2,3).

A case series and review of literature of acute pancreatitis with normal lipase and amylase failed to reveal any specific risk factors for such observation (2). More specifically, the etiologies of acute pancreatitis in the reported cases have varied, including drug-induced, cholelithiasis, alcohol, hypertriglyceridemia, and postoperative causes.

But what accounts for this phenomenon? Many cases have been associated with the first bout of pancreatitis without evidence of pancreatic calcifications which makes the possibility of a “burned-out” pancreas without sufficient acinar cells to release lipase as a frequent cause unlikely. Other potential explanations for normal lipase in acute pancreatitis have included measurement of serum lipase at a very early phase of the disease before significant destruction of acinar cells has occurred (increases in 3-6 h, peaks at 24 h [4]) and more rapid renal clearance of serum lipase due to tubular dysfunction (2).

Of note, unlike amylase, lipase is totally reabsorbed by renal tubules under normal conditions (5). Thus, it’s conceivable that even a reversible tubular dysfunction may lead to increased clearance of serum lipase and potentially lower its levels.
1. Ko K, Tello LC, Salt J. Acute pancreatitis with normal amylase and lipase. The Medicine Forum. 2011;11 Article 4.
2. Singh A, Shrestha M. Acute pancreatitis with normal amylase and lipase-an ED dilemma. Am J Emerg Med 2016;940.e5-940.e7.
3. Limon O, Sahin E, Kantar FU, et al. A rare entity in ED: normal lipase level in acute pancreatitis. Turk J Emerg Med 2016;16:32-34.
4. Shah AM, Eddi R, Kothari ST, et al. Acute pancreatitis with normal serum lipase: a case series. J Pancreas (Online) 2010 July 5;11:369-72. PDF
5. Lott JA, Lu CJ. Lipase isoforms and amylase isoenzymes: assays and application in the diagnosis of acute pancreatitis. Clin Chem 1991;37:361-68.
If you liked this post, sign up under menu to receive future P4P pearls right into your mailbox!

Is it possible to have acute pancreatitis with normal serum lipase?

What is the significance of Terry’s nails in my hospitalized patient?

Terry’s nails were first described in 1954 in patients with hepatic cirrhosis (prevalence 82%, majority related to alcohol abuse) (1). Since then, they have been reported in a variety of other conditions, including adult-onset diabetes mellitus (AODM), chronic congestive heart failure, chronic renal failure, pulmonary tuberculosis, and Reiter’s syndrome (2).

A 1984 study found Terry’s nails in 25% of hospitalized patients (3).  In this study, cirrhosis, chronic congestive heart failure, and AODM were significantly associated with Terry’s nails, while pulmonary tuberculosis, rheumatoid arthritis and cancer were not. The presence of Terry’s nails may be particularly concerning in patients 50 y of age or younger as it increases the relative risk of cirrhosis, chronic congestive heart failure or AODM by 5-fold (18-fold for cirrhosis alone) in this age group (3).

Terry’s nails should be distinguished from Lindsay’s nails or “half and half” nails. Although both nail abnormalities are characterized by an opaque white proximal portion, Terry’s nails have a thinner distal pink to brown transverse band no more than 3 mm wide (3) (Fig 1), while the same anomaly is wider and occupies 20%-60% of the nail bed in Lindsay’s nails (Fig 2). Of interest, Lindsay’s nails have been reported in up to 40% of patients with chronic kidney disease (4,5).


1. Terry R. White nails in hepatic cirrhosis. Lancet 1954;266:757-59. 
2. Nia AM, Ederer S, Dahlem K, et al. Terry’s nails: a window to systemic diseases. Am J Med 2011;124:603-604. 
3. Holzberg M, Walker HK. Terry’s nails: revised definitions and new correlations. Lancet 1984;1(8382):896-99. 
4. Pitukweerakul S, Pilla S. Terry’s nails and Lindsay’s nails: Two nail abnormalities in chronic systemic diseases. J Gen Intern Med 31;970. 
5. Gagnon AL, Desai T. Dermatological diseases in patients with chronic kidney disease 2013;2:104-109.

Figure 1. Terry’s nails in a patient with end-stage liver disease

Figure 2. Lindsay’s nails in a patient with chronic kidney disease

If you liked this post, SELRES_9060f380-b0ce-41bb-b812-fe2595cb3460SELRES_4b9ffe76-4732-435c-a61e-cb3aba28fef9SELRES_055e8f9c-d15f-4b5c-8ddc-c9eb04539366sign upSELRES_055e8f9c-d15f-4b5c-8ddc-c9eb04539366SELRES_4b9ffe76-4732-435c-a61e-cb3aba28fef9SELRES_9060f380-b0ce-41bb-b812-fe2595cb3460 on the P4P home page and receive future pearls delivered directly into your mailbox!

What is the significance of Terry’s nails in my hospitalized patient?

My hypertensive patient needs hemodialysis. How dialyzable are common antihypertensives?

Among antihypertensives, most commonly used angiotensin converting enzyme inhibitors (ACE-Is) such as captopril, enalapril, lisinopril, and benazepril are at least partially removed by hemodialysis; ramipril and fosinopril are not appreciably removed.1,2

In contrast, none of the commonly used angiotensin receptor blockers such as losartan, valsartan, and irbesartan are removed by hemodialysis.

Among β-blockers and combined α- and β-blockers, atenolol and metoprolol are removed by hemodialysis while carvedilol, bisoprolol, propranolol and labetalol are not.

Many other antihypertensives such as calcium channel blockers, α-blockers, clonidine, and hydralazine are not appreciably removed by hemodialysis, while isosorbide dinitrate appears to be.

Of interest, a 2015 retrospective cohort study found that initiation of high- dialyzability β-blockers (atenolol, acebutolol, or metoprolol) was associated with a higher risk of death in the following 180 days compared to that of low-dialyzability  β-blockers (bisoprolol or propranolol), suggesting that perhaps we should be more selective in our choice of β-blockers in this patient population.2 In contrast, no significant difference in all-cause mortality was noted among older patients receiving ACE-Is with high vs low dialyzability potential.3



  1. Inrig JK, Antihypertensive agents in hemodialysis patients: A current perspective. Semin dial 2010;23:290-7.
  2. β-Blocker dialyzability and mortality in older patients receiving hemodialysis. J Am Soc Nephrol 2015;26:987-96.
  3. Weir MA, Fleet JL, Dixon SN, et al. Angiotensin converting enzyme inhibitor dialyzability and outcomes in older patients receiving hemodialysis. Blood Purif 2015;40:232-42.   

Contributed in part by Andrew Lundquist, MD, PhD, Mass General Hospital, Boston, MA.

If you like this pearl, sign up on P4P’s home page and get future pearls delivered right into your mailbox!

My hypertensive patient needs hemodialysis. How dialyzable are common antihypertensives?

Why has my patient with Clostridium difficile diarrhea developed Klebsiella bacteremia?

Although there are many potential sources for Klebsiella sp. bacteremia, C. difficile infection (CDI) itself may be associated with GI translocation of enteric organisms.

A retrospective study of over 1300 patients found an incidence of 1.8% for CDI-associated bacteremia. E. coli, Klebsiella sp. , or Enterococcus sp. accounted for 72% of cases. History of malignancy, neutropenia (at the time of CDAD), and younger age (mean 59 y) were among the risk factors.1 Another study reported over 20 cases of bacteremia caused by C. difficile plus other bacteria often of enteric origin such the aforementioned organisms, Bacteroides sp, Candida sp, and Enterobacter sp.2

CDI is thought to predispose to bacterial translocation through the GI tract by alteration of mucosal indigenous microflora, overgrowth of certain pathogens, and presence of inflammation in the mucosa.3 Interestingly, C. difficile toxin A or B may play an active role in the bacterial adherence and penetration of the intestinal epithelial barrier.4  

Bonus pearl: Did you know that C. difficile may be found in the normal intestinal flora of 3% of healthy adults, 15-30% of hospitalized patients, and up to 50% of neonates? Why neonates seem immune to CDI is another fascinating story!



  1. Censullo A, Grein J, Madhusudhan M, et al. Bacteremia associated with Clostridium difficile colitis: incidence, risk factors, and outcomes. Open Forum Infectious Diseases, Volume 2, Issue suppl_1, 1 December 2015, 943,
  2. Kazanji N, Gjeorgjievski M, Yadav S, et al. Monomicrobial vs polymicrobial Clostridum difficile bacteremia: A case report and review of the literature. Am J Med 2015;128:e19-e26.
  3. Naaber P, Mikelsaar RH, Salminen S, et al. Bacterial translocation, intestinal microflora and morphological changes of intestinal mucosa in experimental models of Clostridium difficile infection. J Med Microbiol 1998; 47: 591-8. 
  4. Clostridium difficile toxins may augment bacterial penetration of intestinal epithelium. Arch Surg 1999;134: 1235-1242.
Why has my patient with Clostridium difficile diarrhea developed Klebsiella bacteremia?