Skip Navigation

Conditions Associated with Leukocytosis in a Tertiary Care Hospital, with Particular Attention to the Role of Infection Caused by Clostridium difficile

  1. Anna Wanahita1,
  2. Elizabeth A. Goldsmith2, and
  3. Daniel M. Musher1,2
  1. 11Infectious Disease Section, Medical Service, Veterans Affairs Medical Center, Houston, Texas
  2. 2Departments of Medicine, Molecular Virology, and Microbiology, Baylor College of Medicine, Houston, Texas
  1. Reprints or correspondence: Dr. Daniel Musher, Infectious Disease Section, Veterans Affairs Medical Center, Houston Texas 77030 (daniel.musher{at}med.va.gov).
 
Next Section

Abstract

Few modern studies have enumerated the conditions associated with leukocytosis. Our clinical experience has implicated Clostridium difficile infection in a substantial proportion of patients with leukocytosis. In a prospective, observational study of 400 inpatients with WBC counts of ⩾15,000 cells/mm3, we documented ⩾1 infection in 207 patients (53%). Of these 207 patients, 97 (47%) had pneumonia, 60 (29%) had urinary tract infection, 34 (16%) had soft-tissue infection, and 34 (16%) had C. difficile infection. C. difficile infection was present in 25% of patients with WBC counts of µ30,000 cells/mm3 who did not have hematological malignancy. Other causes of leukocytosis in the 400 patients included physiological stress, in 152 patients (38%); medications or drugs, in 42 (11%); hematological disease, in 22 (6%); and necrosis or inflammation, in 22 (6%). C. difficile infection is a prominent cause of leukocytosis and this diagnosis should be considered for patients with WBC counts of ⩾15,000 cells/mm3, even in the absence of diarrheal symptoms.

The WBC count is one of the most widely used laboratory tests; at Texas Medical Center alone, µ1,000,000 such determinations are done each year (authors' unpublished observation). One might imagine that there have continued to be articles in the modern era to help clinicians evaluate the finding of an elevated WBC count. However, although the National Library of Medicine cites 7600 articles on the general subject of leukocytosis published since 1965, and Index Medicus lists hundreds of articles published between 1955 and 1964, only a small number have dealt with the specific causes of this condition [1–11], and even fewer have actually enumerated them [3, 6, 8, 9, 11].

Clostridium difficile is a well-recognized etiologic agent of nosocomial infection that causes diarrhea, fever, and inflammation of the colonic mucosa with pseudomembrane formation [12–14]. Recent studies have emphasized the importance of leukocytosis as a manifestation of C. difficile infection [13–16], but the frequency with which this agent contributes to leukocytosis in hospital-based practice has not been reported.

We believed that it would be useful for clinicians to have data on the medical conditions associated with leukocytosis, especially in light of modern diagnostic techniques and therapeutic modalities. As we became more interested in the clinical factors associated with leukocytosis, our clinical experience suggested that a substantial proportion of patients, especially those with WBC counts of ⩾30,000 cells/mm3, have C. difficile infection. Accordingly, at our tertiary care hospital, we undertook a prospective, observational study of inpatients and outpatients with WBC counts of µ15,000 cells/mm3, with a dual purpose: (1) to determine conditions associated with leukocytosis, and (2) to define the role of infection caused by C. difficile.

Previous SectionNext Section

Patients and Methods

Clinical settin. The Houston Veterans Affairs Medical Center is a large, tertiary care medical center with 520 active beds that provides primary medical care for ∼55,000 adults. There were an average of 1920 admissions and 56,652 outpatient clinic and urgent care visits per month during the time of the study. The complete medical record for every case patient is available electronically, including all progress, consultation, and discharge notes and all results of laboratory, radiographic, and pathologic studies, as these results are obtained.

Identification of patients with leukocytosi. In a prospective observational study that began on 1 January 2001, we reviewed laboratory records each day to identify patients with WBC counts of ⩾15,000 cells/mm3. Each patient was included only once and was stratified on the basis of the highest WBC count recorded. As patients were identified, an investigator studied the complete medical record of each, with attention paid to the date of admission; any and all condition(s) that might cause fever and/or leukocytosis; findings of physical examination, laboratory tests, and imaging studies; daily progress notes; diagnoses by the primary care team; and opinions of consultants. Medical records were reviewed again every few days until the patient was discharged from the hospital. Patients were assigned diagnoses on the basis of all available information in the medical record through the time of discharge (and after, if results of essential studies were outstanding).

We originally intended to study 200 inpatients with WBC counts of 15,000–19,999 cells/mm3 and 200 with WBC counts of ⩾20,000 cells/mm3. Once the study had begun, we discovered that approximately one-third of patients with WBC counts of ⩾15,000 cells/mm3 at our hospital were outpatients, and we decided to collect data from these individuals as well but to tabulate the results separately. We examined their medical records for the period preceding the relevant clinic visit and at the time of the visit and the records of other outpatient visits and/or hospital admissions for the subsequent 4 months. The study protocol was approved by the Institutional Review Board, Baylor College of Medicine (Houston).

Classification of potential causes of leukocytosi. Patients were assigned to ⩾1 of 6 groups corresponding to potential causes of leukocytosis, which were categorized according to the general classification of Holland and Gallin [17] (however, in reporting results, we changed the order of the categories of causes to reflect the actual frequency with which they were implicated). These causes were as follows: (1) infection, (2) physiological stress, (3) medications or drugs (including illicit drugs), (4) disease of the hematopoietic system, (5) necrosis or inflammation, and (6) unknown. The greatest care was used in categorizing patients. For example, patients with acute myocardial infarction were assigned to category 5 (necrosis/inflammation), whereas those with unstable angina were assigned to category 2 (physiological stress). Patients with acute pancreatitis (with or without a pseudocyst) were included in category 5, unless infection was also thought to be present, in which case the patient was listed in both categories 1 and 5. Multiple potential causes were recorded, when present; however, because 400 patients were inpatients, we did not include hospitalization itself as a form of physiological stress, even though it may well be one.

Laboratory tests fo. C. difficile infectio. Fecal samples were analyzed for C. difficile toxin by use of Premier A (Meridian Diagnostics), an EIA that detects toxin A. Fecal samples were refrigerated at 8°C overnight or at -80°C over the weekend; assays were performed 3 times weekly. The US Food and Drug Administration has approved an EIA kit for detection of C. difficile toxins A and B (Premier Toxins A and B; Meridian Diagnostics), but that kit was unavailable from the manufacturer for use in the United States throughout the time of our study.

Statistical analysis. The χ2 test was used to compare the prevalence of conditions for various groups of subjects. To determine whether there was a difference in the mean values between 3 groups, we used analysis of variance [18]. P values <.05 were considered significant.

Previous SectionNext Section

Results

Hospitalized patient. In accord with our intention, we accrued 200 patients with WBC counts of 15,000–19,999 cells/mm3 and 200 with WBC counts of ⩾20,000 cells/mm3. For the purpose of presenting our results, we stratified patients on the basis of WBC counts of 15,000–19,999 cells/mm3, 20,000–29,999 cells/mm3, and ⩾30,000 cells/mm3 (table 1). Because multiple factors that might cause leukocytosis were present, the total number of conditions exceeds the number of patients. Overall, the majority of patients (207 [52%]) had ⩾1 infections. Physiological stress was next-most common causal factor; it was identified in 152 patients (38%). Medications or drugs were implicated in 42 patients (11%), and hematological disease and necrosis or inflammation were each found in 22 patients (6%). Leukocytosis of unknown cause was present in only 16 (4%) of all 400 patients; for most of these patients, acute leukocytosis developed in the hospital, the patient died without a diagnosis, and no autopsy was performed; it is likely that most of these patients had infections.

Figure 1
Figure 1

Percentage of patients with clinical findings of Clostridium difficile infection, stratified by WBC count. Black denotes the proportions of patients with laboratory-proven diagnosis; hatching denotes the proportions of patients with some clinical or laboratory support for the diagnosis; white denotes the proportions of patients with clinical features consistent with the diagnosis but no supporting laboratory results, past history, or therapeutic response. For the purposes of this calculation, patients with hematological malignancy were excluded.

Table 1
Table 1

Summary of causal factors associated with leukocytosis in 400 hospitalized patients.

Infectio. Among inpatients with WBC counts of ⩾15,000 cells/mm3, infection was the most frequently implicated cause of leukocytosis (range, 48%–60% of patients; table 1). The percentage of patients with infection increased slightly but not significantly (P = .23) with the degree of leukocytosis. As shown in table 2, nearly all of these patients had acute infection attributable to bacterial disease, including 97 patients with pneumonia (47% of all patients), 60 (29%) with urinary tract infection, 34 (16%) with soft-tissue infection, 17 (8%) with intra-abdominal infection, 9 (4%) with vascular infection, and 7 (3%) with osteomyelitis. The highest proportion of patients with WBC counts of ⩾30,000 cells/mm3 were in the intravascular infection group (5 [56%] of 9 patients), and the lowest proportion was in the osteomyelitis group (0 [0%] of 7).

Table 2
Table 2

Infections associated with leukocytosis in hospitalized patients.

C. difficile coliti. A novel finding of this study was the importance of C. difficile colitis as a cause of leukocytosis, as demonstrated by documentation of the presence of C. difficile toxin A in the feces of 34 patients (9%) (table 2). On the basis of this unexpected finding, which required that the diagnosis be considered and an appropriate sample be sent for analysis, we undertook a more intensive study to determine whether additional patients might have clinical features suggesting this diagnosis. To do so, we reviewed the entire computerized record (which was available for the previous 4.5 years) for results of other studies for C. difficile toxin; for the admitting intern's note; for the pharmacy record (also stored in the computer system) of antibiotics prescribed in the preceding 6 weeks; and for all recorded notes (especially those of the nurses) with any mention of diarrhea.

Forty-two patients with leukocytosis had a clinical picture consistent with C. difficile colitis, including all of the following features: a history of having received antibiotic(s) in the preceding 6 weeks; fever, leukocytosis, and diarrhea; or another major manifestation of colitis, such as ileus, severe abdominal pain, or toxic megacolon. These 42 patients were stratified to 1 of 5 subcategories according to EIA test status and/or other clinical findings; these subcategories are listed in table 3 in decreasing order of likelihood that C. difficile infection was present. For 4 patients, EIA for C. difficile toxin A yielded a borderline (“indeterminate”) result. Seven patients either had no test done or had an EIA result negative for C. difficile toxin at the time leukocytosis was detected (for 4 patients, this test was only done once), but review of the medical record revealed C. difficile infection(s) in the past that had been documented by EIA. Eleven patients were treated with metronidazole in the absence of a positive diagnostic test result and responded well. The remaining 20 patients either had no study done (5 patients) or had ⩾1 EIA negative for C. difficile toxin (15 patients). Thus, of 400 inpatients with WBC counts of ⩾15,000 cells/mm3, 76 (19%) had a clinical picture consistent with C. difficile infection, 56 (14%) had such a clinical picture with some laboratory or clinical support, and 34 (9%) had confirmed infection. The proportion of patients who had a clinical syndrome consistent with C. difficile colitis or who had proven C. difficile colitis increased significantly as the WBC count increased (P < .01). When individuals with leukemia were excluded from analysis, 25% of all patients with WBC counts of ⩾30,000 cells/mm3 had proven C. difficile infection (figure 1).

Table 3
Table 3

Subcategories of patients with a clinical picture consistent with Clostridium difficile infection, according to results of EIA for C. difficile toxin A.

Physiological stres. Acute physiological stress was present in 152 (38%) of 400 patients, including the following conditions: surgery with general anesthesia (in 57 [41%] of the 152 patients), acute respiratory distress syndrome (ARDS) (in 40 patients [29%]), hypovolemia due to hemorrhage or dehydration (in 34 patients [25%]), and diabetic ketoacidosis (in 8 patients [6%]). Physiological stress was generally more likely to be associated with an elevated WBC count of 15,000–19,999 cells/mm3 than with a count of ⩾30,000 cells/mm3 (P = .04); for example, hemorrhage was associated with WBC counts of 15,000–19,999 cells/mm3 in 9 patients, 20,000–29,999 cells/mm3 in 3 patients, and ⩾30,000 cells/mm3 in 0 patients. ARDS was significantly more likely to be associated with WBC counts of ⩾30,000 cells/mm3 (P < .01) than with lower levels of leukocytosis, but many of the patients with ARDS were thought to have concurrent infection. In patients who had leukocytosis associated with physiological stress, when other factors were not also present, the leukocytosis promptly resolved after the stress was ended.

Medications or drug. The great majority of patients (37 [88%] of 42) for whom medications or drugs were implicated as a cause of elevated WBC counts were receiving prednisone (⩾40 mg/day) or its equivalent. This medication appeared to be more commonly responsible for WBC counts of 15,000–19,999 cells/mm3, whereas other drugs were implicated in patients with various levels of leukocytosis: granulocyte colony-stimulating factor (G-CSF; 2 patients), cocaine (2 patients), and lithium (1 patient). In no patients were medications or drugs solely responsible for WBC counts of ⩾30,000 cells/mm3.

Hematological conditio. The proportion of patients with an associated hematological condition increased with the WBC count (P < .01). Overall, 22 patients had a malignancy that involved the hematopoietic system: 15 patients had leukemia, 2 had lymphoma, and 3 had widespread metastatic cancer that involved the bone marrow. Two patients had stable elevations in WBC counts resulting from prior splenectomy.

Necrosis or inflammatio. In 22 patients, necrosis or inflammation was associated with leukocytosis. This group included 7 patients with myocardial infarction (patients who had unstable angina were assigned to the category “physiological stress”), 3 with pulmonary infarction, 2 with pancreatitis, and 2 with incarcerated umbilical hernia without necrosis. Of these 22 subjects, 4 had WBC counts of ⩾30,000 cells/mm3; only 1 patient (who had myocardial infarction) had no associated condition other than necrosis or inflammation.

Unknow. The majority of patients for whom no cause for leukocytosis was determined were in extremis and died shortly after the high WBC count was documented. Infection was probably responsible in most instances, but its presence was not documented, and autopsies were not performed.

Outpatient. A total of 133 outpatients were found to have WBC counts of ⩾15,000 cells/mm3. Because the total numbers were smaller, these patients were stratified into 2 categories: those with WBC counts of 15,000–19,999 cells/mm3 (n = 74) and those with WBC counts of ⩾20,000 cells/mm3 (n = 59) (table 4). As might have been predicted, infection was implicated as a cause of leukocytosis in a smaller proportion of outpatients than inpatients (24 [18%] of 133 vs. 207 [52%] of 400; P < .001). Only 1 (1%) of 133 outpatients was proven to have C. difficile colitis, compared with 34 (9%) of 400 hospitalized patients (P < .001). In contrast, 60 (45%) of 133 outpatients had a hematological condition (vs. 22 [11%] of 400 inpatients; P < .001). Three patients had WBC counts of µ20,000 cells/mm3 that were attributable to medications; 2 were receiving G-CSF, and the other was receiving risperidone. The proportion of outpatients with no identified cause for leukocytosis (33 [25%] of 133 patients) was much greater than the proportion of inpatients (16 [4%] of 200; P < .01).

Table 4
Table 4

Causal factors associated with leukocytosis in outpatients.

Presence of band form. Among inpatients with infection, ⩾3% band forms (a “left shift”) were detected in 20% of patients with WBC counts of 15,000–19,999 cells/mm3, 37% of patients with counts of 20,000–29,999 cells/mm3, and 53% of patients with counts of ⩾30,000 cells/mm3; the proportion of infected patients with a left shift was significantly greater for higher WBC counts (P = .01), although the mean percentage of bands was not necessarily greater for patients with higher WBC counts (7.0% for patients with WBC counts of 15,000–19,999 cells/mm3 vs. 9.3% for patients with WBC counts of ⩾30,000 cells/mm3; P = .45). An elevated percentage of band forms (⩾3%) was also observed in 12% of patients with physiological stress, 19% of patients with medication- or drug-associated leukocytosis, and 5% of patients with necrosis or inflammation. Band form percentages of µ12% were seen exclusively in patients with infection (19 patients) and those receiving G-CSF (2 patients).

Previous SectionNext Section

Discussion

At its inception, our study had 2 related goals: (1) to elucidate causes of leukocytosis in hospitalized patients and (2) to determine the role of C. difficile infection as a contributing cause of leukocytosis. The justification for the first goal was clear. Although the complete blood count is the most commonly ordered laboratory test, and nearly every textbook has a table showing the causes of leukocytosis, there is a remarkable paucity of studies that have actually tabulated factors associated with elevated WBC counts. The justification for the second goal was our clinical experience, which had suggested that a substantial proportion of patients with leukocytosis, especially those without a clearly recognized cause and with very high WBC counts, are likely to have C. difficile infection.

Leukocytosis is well described in many specific infectious states—for example, appendicitis [4, 19], cellulitis [20], and pneumonia [21]. Elevated WBC counts have also been reported in noninfectious conditions such as myocardial infarction [22–24], acute pulmonary embolism [25], alcoholic hepatitis [26], surgical stress [27] or other physiological stress [28, 29], as well as in association with medications, especially glucocorticosteroids [30–33]. In contrast, relatively few articles [2–11] have dealt generally with leukocytosis in human disease, and even fewer [3, 6, 9, 11] have enumerated potential causative conditions.

Shapiro and Greenfield [4] performed a meta-analysis to examine the sensitivity and specificity of a WBC count of µ11,000 cells/mm3 for the diagnosis of specified conditions. They concluded that the absence of leukocytosis does not enable the physician to exclude the diagnosis of sepsis, nor does a WBC count of µ11,000 cells/mm3 reliably distinguish appendicitis from inoperable causes of intra-abdominal pain. Their inclusion of all cases of WBC counts of µ11,000 cells/mm3 and the likelihood that physiological stress [28, 29] was, in part, responsible for these elevated counts probably explains this finding.

Two earlier studies, one that included 75 patients [6] and one that included 100 patients [9], tabulated factors associated with WBC counts of ⩾25,000 cells/mm3 among hospitalized subjects (table 5). Chang and Wong [6] implicated hematological disease and related medications, such as G-CSF, more frequently than did our study, which presumably reflects the population at an affiliated cancer hospital. The higher proportion of patients with physiological stress in Chang and Wong [6] and the lower proportion in Reding et al. [9] may reflect differences in the rate of surgical procedures. Otherwise, conditions associated with leukocytosis in these 2 series of cases generally were similar to ours.

Table 5
Table 5

Causal factors associated with leukocytosis in hospitalized patients in this and previous studies.

Our results showed a notable uniformity in the factors associated with WBC counts of 15,000–19,999 cells/mm3, counts of 20,000–29,999 cells/mm3, and counts of ⩾30,000 cells/mm3, although some tempering of the distinctions in the data for the patients in these groups may have resulted from our decision to stratify patients on the basis of the highest WBC count recorded. Physiological stress was less likely to be associated with WBC counts of ⩾30,000 cells/mm3, but the finding of a WBC count elevated to that level is not inconsistent with stress alone. In a prospective study of 18 patients who underwent extensive elective surgical procedures, which were often associated with hemorrhage and required transfusion, Pepper and Lindsay [27] found a mean WBC count of nearly 18,000 cells/mm3 on the first day after surgery and a peak WBC count of µ25,000 cells/mm3.

An increase in band forms (to ⩾3%) was seen with nearly the same frequency in patients with all categories of illness, but large percentages of band forms (⩾12%) were present only in patients who had infection. In a preliminary study of 99 patients, nearly one-half of whom had hematological disorders, Zubler [3] found no relation between the cause of leukocytosis and the presence of band forms; she did not, however, stratify patients on the basis of the size of band form counts. Although some authorities have suggested that exogenous steroids or stress mediated by endogenous steroids [31] cause demargination of granulocytes and are not accompanied by a left shift, extensive evidence summarized by Dale et al. [30] showed that hydrocortisone also mediates release of polymorphonuclear cells, including precursors, from bone marrow. Our study supports the observations of Shoenfeld et al. [31], who concluded that a band count of µ6% is unlikely to be attributable to glucocorticosteroid therapy.

Documenting an association between C. difficile colitis and leukocytosis is not novel. For example, in a retrospective study of 35 patients with proven C. difficile infection, Bulusu et al. [15] found that the mean peak WBC count was 15,800 cells/mm3. Seven patients (20%) had WBC of 20,000–30,000 cells/mm3, and 2 (6%) had WBC counts of µ30,000 cells/mm3. Other studies have noted leukocytosis (WBC counts of ⩾30,000 cells/mm3) in patients with this condition [34–38]. An important finding of our study, however, is the frequency with which C. difficile infection was found among all patients with leukocytosis. Of 400 inpatients with WBC counts of ⩾15,000 cells/mm3, 19% had a clinical picture consistent with C. difficile infection; 14% had such a clinical picture with some laboratory support, and 9% had confirmed infection. The proportion of cases attributable to C. difficile increased with higher WBC counts, such that nearly one-third of patients with WBC counts of µ30,000 cells/mm3 had a consistent clinical picture and 21% had proven C. difficile infection. Our results, together with those of a prospective interventional study now in progress (authors' unpublished data), confirm the observation [15] that an elevated WBC count may precede abdominal discomfort or diarrhea, which suggests that the diagnosis of C. difficile infection should be considered for all patients with leukocytosis, especially if some other cause is not immediately apparent.

Our documentation of C. difficile infection likely underestimates the magnitude of this problem. Manabe et al. [36] showed that 3 EIA tests detected C. difficile infection in 90% of patients who had evidence from tissue culture or EIA for the presence of C. difficile toxin, whereas a study confined to the first stool sample would have yielded the diagnosis for only 72% of patients. Gerding et al. [39] found that 10% of patients who had negative results of toxin assays but who had stool cultures positive for C. difficile had pseudomembranous colitis. False-negative assay results for C. difficile toxin may be due, in part, to the instability of the toxin in stored samples [16, 40]. Our hospital laboratory performs assays for C. difficile toxin only 3 times weekly, storing specimens at 4°C–8°C overnight and at -80°C over the weekend; the manufacturers of the EIA kit state that storage should be at 4°C–8°C but that samples should be stored frozen if µ72 h will elapse before testing. During the time of the present study, we assayed only for C. difficile toxin A, in part because that is a generally accepted procedure [14] and in part because the only US Food and Drug Administration-approved kit that measures levels of toxins A and B was unavailable from the manufacturer; an assay confined to toxin A may miss additional diagnoses [41, 42]. Taken together, these observations suggest that a single stool sample is likely to be positive for toxin A in no more than 60%–70% of cases of C. difficile colitis, as has been shown in prospective studies [43, 44]; the package insert, which claims that the assay has ⩾90% sensitivity [45, 13], needs to be considered in this context. Most patients in our study for whom there was clinical suspicion but no laboratory confirmation of C. difficile infection had only a single fecal specimen that was studied; in contrast, patients with positive test results may have had as many as 3 or 4 negative study results before a positive result was documented.

In summary, in the course of an observational study to examine all causes of leukocytosis in a tertiary care hospital, we discovered a surprisingly great contribution by C. difficile infection, especially among patients with WBC counts of ⩾30,000 cells/mm3. Accordingly, we have initiated a prospective interventional study to look at all patients with leukocytosis. Preliminary results suggest that, at our facility, if EIA for C. difficile toxin is done for every hospitalized patient who has leukocytosis without an apparent explanation, ∼50% of patients will have a positive assay result (authors' unpublished observations).

  • Received November 9, 2001.
  • Revision received January 30, 2002.
Previous Section
 

References

  1. ↵
    1. Wintrobe MM
    . Diagnostic significance of changes in leukocytes. Bull N Y Acad Med 1939;15:223-40.
    Medline
  2. ↵
    1. Hardison CS
    . Laboratory procedures: their clinical significance—the leukocyte count. JAMA 1968;204:85-6.
    FREE Full Text
  3. ↵
    1. Zubler MA
    . Leukocytosis in infection. Arch Intern Med 1987;147:197.
    Abstract/FREE Full Text
  4. ↵
    1. Shapiro MF,
    2. Greenfield S
    . The complete blood count and leukocyte differential count. Ann Intern Med 1987;106:65-74.
    Abstract/FREE Full Text
    1. Wasserman M,
    2. Levinstein M,
    3. Keller E,
    4. Lee S,
    5. Yoshikawa TT
    . Utility of fever, white blood cells, and differential count in predicting bacterial infections in the elderly. J Am Geriatr Soc 1989;37:537-43.
    MedlineWeb of Science
  6. ↵
    1. Chang R,
    2. Wong GY
    . Prognostic significance of marked leukocytosis in hospitalized patients. J Gen Intern Med 1991;6:199-203.
    CrossRefMedlineWeb of Science
    1. Gin-Shaw S,
    2. Moore GP
    . Selected white cell disorders. Emerg Med Clin North Am 1993;11:495-516.
    Medline
  8. ↵
    1. Seebach JD,
    2. Ruegg MR,
    3. Seifert B,
    4. Fehr J
    . The diagnostic value of the neutrophil left shift in predicting inflammatory and infectious disease. Am J Clin Pathol 1997;107:582-91.
    MedlineWeb of Science
  9. ↵
    1. Reding MT,
    2. Hibbs JR,
    3. Morrison VA,
    4. Swaim WR,
    5. Filice GA
    . Diagnosis and outcome of 100 consecutive patients with extreme granulocytic leukocytosis. Am J Med 1998;104:12-6.
    CrossRefMedlineWeb of Science
    1. Abramson N,
    2. Melton B
    . Leukocytosis: basics of clinical assessment. Am Fam Physician 2000;1:2053-60.
  11. ↵
    1. Crabtree TD,
    2. Pelletier SJ,
    3. Antevil JL,
    4. Gleason TG,
    5. Pruett TL,
    6. Sawyer RG
    . Cohort study of fever and leukocytosis as diagnostic and prognostic indicators in infected surgical patients. World J Surg 2001;25:739-44.
    CrossRefMedlineWeb of Science
  12. ↵
    1. Knoop FC,
    2. Owens M,
    3. Crocker IC
    . Clostridium difficile: clinical disease and diagnosis. Clin Microbiol Rev 1993;6:251-65.
    Abstract/FREE Full Text
  13. ↵
    1. Manabe YC,
    2. Vinetz JM,
    3. Moore RD,
    4. Merz C,
    5. Charache P,
    6. Bartlett JG
    . Clostridium difficile colitis: an efficient clinical approach to diagnosis. Ann Intern Med 1995;123:835-40.
    Abstract/FREE Full Text
  14. ↵
    1. Johnson S,
    2. Gerding DN
    . Clostridium difficile-associated diarrhea. Clin Infect Dis 1998;26:1027-36.
    FREE Full Text
  15. ↵
    1. Bulusu M,
    2. Narayan S,
    3. Shetler K,
    4. Triadafilopoulos G
    . Leukocytosis as a harbinger and surrogate marker of Clostridium difficile infection in hospitalized patients with diarrhea. Am J Gastroenterol 2000;95:3137-41.
    CrossRefMedlineWeb of Science
  16. ↵
    1. Klingler PJ,
    2. Metzger PP,
    3. Seelig MH,
    4. Pettit PDM,
    5. Knudsen JM,
    6. Alvarez S
    . Clostridium difficile infection: risk factors, medical and surgical management. Dig Dis 2000;18:147-60.
    CrossRefMedlineWeb of Science
  17. ↵
    1. Fauci AS
    1. Holland SM,
    2. Gallin JI
    . Disorders of granulocytes and monocytes. In: Fauci AS, editor. Harrison's principles of internal medicine. 15th ed. New York: McGraw-Hill; 2001. p. 366-74.
  18. ↵
    1. Brown BW Jr,
    2. Hollander M
    . Statistics: a biomedical introduction. New York: John Wiley & Sons; 1977. Chi-squared tests; p. 195-224.
  19. ↵
    1. Nase HW,
    2. Kovalcik PJ,
    3. Cross GH
    . The diagnosis of appendicitis. Am Surg 1980;46:504-7.
    MedlineWeb of Science
  20. ↵
    1. Fleisher G,
    2. Ludwig S,
    3. Henretig F,
    4. Ruddy R,
    5. Henry W
    . Cellulitis: initial management. Ann Emerg Med 1981;10:356-9.
    CrossRefMedlineWeb of Science
  21. ↵
    1. Musher DM,
    2. Alexandraki I,
    3. Graviss EA,
    4. et al
    . Bacteremic and nonbacteremic pneumococcal pneumonia: a prospective study. Medicine 2000;79:210-21.
    CrossRefMedline
  22. ↵
    1. Green SM,
    2. Vowels J,
    3. Waterman B,
    4. Rothrock SG,
    5. Kuniyoshi G
    . Leukocytosis: a new look at an old marker for acute myocardial infarction. Acad Emerg Med 1996;3:1034-41.
    MedlineWeb of Science
    1. Kyne L,
    2. Hausdorff JM,
    3. Knight E,
    4. Dukas L,
    5. Azhar G,
    6. Wei JY
    . Neutrophilia and congestive heart failure after acute myocardial infarction. Am Heart J 2000;139:94-100.
    MedlineWeb of Science
  24. ↵
    1. Barron HV,
    2. Cannon CP,
    3. Murphy SA,
    4. Braunwald E,
    5. Gibson CM
    . Association between white blood cell count, epicardial blood flow, myocardial perfusion, and clinical outcomes in the setting of acute myocardial infarction: a thrombolysis in myocardial infarction 10 substudy. Circulation 2000;102:2329-34.
    Abstract/FREE Full Text
  25. ↵
    1. Afzal A,
    2. Noor HA,
    3. Gill SA,
    4. Brawner C,
    5. Stein PD
    . Leukocytosis in acute pulmonary embolism. Chest 1999;115:1329-32.
    Abstract/FREE Full Text
  26. ↵
    1. Mitchell RG,
    2. Michael M,
    3. Sandidge D
    . High mortality among patients with the leukemoid reaction and alcoholic hepatitis. South Med J 1991;84:281-2.
    MedlineWeb of Science
  27. ↵
    1. Pepper H,
    2. Lindsay ST
    . Responses of platelets, eosinophils, and total leucocytes during and following surgical procedures. Surg Gynecol Obstet 1960;110:319-26.
    MedlineWeb of Science
  28. ↵
    1. Lloyd-Jones DM,
    2. Camargo CA Jr,
    3. Giugliano RP,
    4. O'Donnell CJ
    . Effect of leukocytosis at initial examiantion on prognosis in patients with primary unstable angina. Am Heart J 2000;139:867-73.
    MedlineWeb of Science
  29. ↵
    1. Rovlias A,
    2. Kotsou S
    . The blood leukocyte count and its prognostic significance in severe head injury. Surg Neurol 2001;55:190-6.
    CrossRefMedlineWeb of Science
  30. ↵
    1. Dale DC,
    2. Fauci AS,
    3. Guerry D,
    4. Wolff SM
    . Comparison of agents producing a neutrophilic leukocytosis in man. J Clin Invest 1975;56:808-13.
    MedlineWeb of Science
  31. ↵
    1. Shoenfeld Y,
    2. Gurewich YGLA,
    3. Pinkhas J
    . Prednisone-induced leukocytosis: influence of dosage, method and duration of administration on the degree of leukocytosis. Am J Med 1981;71:773-8.
    CrossRefMedlineWeb of Science
    1. Carmen J,
    2. Okafor K,
    3. Ike E
    . The effects of lithium therapy on leukocytes: a 1-year follow-up study. J Natl Med Assoc 1993;85:301-3.
    Medline
  33. ↵
    1. Brody SL,
    2. Wrenn KD,
    3. Wilber MM,
    4. Slovis CM
    . Predicting the severity of cocaine-associated rhabdomyolysis. Ann Emerg Med 1990;19:1137-43.
    CrossRefMedlineWeb of Science
  34. ↵
    1. Bartlett JG
    . Antibiotic-associated diarrhea. Clin Infect Dis 1992;15:573-81.
    Abstract/FREE Full Text
    1. Anand A,
    2. Bashey B,
    3. Mir T,
    4. Glatt AE
    . Epidemiology, clinical manifestations, and outcome of Clostridium difficile-assocated diarrhea. Am J Gastroenterol 1994;89:519-23.
    MedlineWeb of Science
  36. ↵
    1. Manabe Y,
    2. Vinetz JM,
    3. Moore RD,
    4. Merz C,
    5. Charache P,
    6. Bartlett JG
    . Clostridium difficile colitis: an efficient clinical approach to diagnosis. Ann Intern Med 1995;123:835-40.
    Abstract/FREE Full Text
    1. Lowenkron SE,
    2. Waxner J,
    3. Khullar PIJS,
    4. Niederman MS,
    5. Fein AM
    . Clostridium difficile infection as a cause of severe sepsis. Intensive Care Med 1996;22:990-4.
    MedlineWeb of Science
  38. ↵
    1. Ramaswamy R,
    2. Grover H,
    3. Corpuz M,
    4. Daniels P,
    5. Pitchumoni CS
    . Prognostic criteria in Clostridium difficile colitis. Am J Gastroenterol 1996;91:460-4.
    MedlineWeb of Science
  39. ↵
    1. Gerding DN,
    2. Olson MM,
    3. Peterson LR,
    4. et al
    . Clostridium difficile-associated diarrhea and colitis in adults: a prospective case-controlled epidemiologic study. Arch Intern Med 1986;146:95-100.
    Abstract/FREE Full Text
  40. ↵
    1. Fang FC,
    2. Madinger NE
    . Clostridium difficile colitis. N Engl J Med 1994;330:1754.
    CrossRefMedlineWeb of Science
  41. ↵
    1. Limaye AP,
    2. Turgeon DK,
    3. Cookson BT,
    4. Fritsche TR
    . Pseudomembranous colitis caused by a toxin A- B+ strain of Clostridium difficile. J Clin Microbiol 2000;38:1696-7.
    Abstract/FREE Full Text
  42. ↵
    1. Kader HA,
    2. Piccoli DA,
    3. Jawad AF,
    4. McGowan KL,
    5. Maller ES
    . Single toxin detection is inadequate to diagnose Clostridium difficile diarrhea in pediatric patients. Gastroenterology 1998;115:1329-34.
    CrossRefMedlineWeb of Science
  43. ↵
    1. Barbut F,
    2. Kajzer C,
    3. Planas N,
    4. Petit J-C
    . Comparison of three enzyme immunoassays, a cytotoxicity assay, and toxigenic culture for diagnosis of Clostridium difficile-associated diarrhea. J Clin Microbiol 1993;31:963-7.
    Abstract/FREE Full Text
  44. ↵
    1. Lozniewski A,
    2. Rabaud C,
    3. Dotto E,
    4. Weber M,
    5. Mory F
    . Laboratory diagnosis of Clostridium difficile-associated diarrhea and colitis: usefulness of premier cytoclone A + B enzyme immunoassay for combined detection of stool toxins and toxigenic C. difficile strains. J Clin Microbiol 2001;39:1996-8.
    Abstract/FREE Full Text
  45. ↵
    Premier Clostridium difficile toxin A: EIA for the detection of C. difficile toxin A in stool specimens. Cincinnatti, OH: Meridian Diagnostics;
| Table of Contents

This Article

  1. Clin Infect Dis. (2002) 34 (12): 1585-1592. doi: 10.1086/340536
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most