| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Division of Radiology/Hb 6, The Cleveland Clinic Foundation, 9500 Euclid Ave.,
Cleveland, OH 44195.
2 Department of Infectious Disease, The Cleveland Clinic Foundation, Cleveland,
OH 44195.
3 Department of Colorectal Surgery, The Cleveland Clinic Foundation, Cleveland,
OH 44195.
4 Department of Biostatistics and Epidemiology, The Cleveland Clinic Foundation,
Cleveland, OH 44195.
Received November 1, 2004;
accepted after revision March 14, 2005.
Presented at the 2004 annual meeting of the American Roentgen Ray Society,
Miami Beach, FL.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Over a 21-month period, 152 of 572 inpatients with C. difficile colitis were identified and had CT scans performed within 2 weeks of the diagnosis. These were independently and retrospectively reviewed by two reviewers. Those with colonic wall thickness greater than 4 mm were considered positive (CT-positive patients) and were further reviewed for specific findings in the colon. All 152 patients with CT scans were also retrospectively reviewed using the hospital information system for certain clinical parameters, admitting diagnoses, and reasons for scanning. The following were compared using several statistical tests: clinical parameters in CT-positive and CT-negative patients and surgical and nonsurgical groups to determine if positive scans or surgical treatment could be clinically predicted; specific CT findings in CT-positive patients to see if an association was found with clinical parameters or surgical treatment; and admitting diagnoses and reasons for scanning in scanned and unscanned populations to see which patients were more likely to undergo CT.
RESULTS. Seventy-six (50%) of 152 scanned hospitalized patients with C. difficile colitis were CT-positive. These patients most often had segmental involvement (50 [66%] of 76 patients), with the rectum (60 [82%] of 73 patients) and sigmoid colon (61 [82%] of 74 patients) most often affected. Positive scans were associated with increased WBC, abdominal pain, and diarrhea. Patients with signs and symptoms of infection or abdominal complaints were more likely to be scanned. No statistical correlation was found between specific CT findings and clinical parameters or clinical parameters and patients requiring surgery. There was no predictive value of specific CT findings for surgical treatment.
CONCLUSION. Half of the patients scanned had an abnormal CT, with segmental colonic disease more common than diffuse. Positive scans were more likely in patients with leukocytosis, abdominal pain, and fever. Specific CT findings did not correlate with clinical parameters and could not predict surgical treatment.
Keywords: abnormalities • Clostridium difficile colitis • colon • infectious diseases
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several studies have examined the colonic CT findings in patients with C. difficile colitis [3, 11-16]. In the majority of these studies, the findings were sensitive but nonspecific, limiting the diagnostic value of CT. Only one study in the literature has investigated the correlation between CT findings and severity of clinical disease [16]. Anecdotally, we have observed that patients with C. difficile colitis and obvious CT findings are often very ill. The purpose of this study was to determine the prevalence of an abnormal colon on CT in a large, adult, hospitalized population of C. difficile-positive patients and to compare the clinical presentation of these patients with other scanned C. difficile-positive patients without an abnormal colon. In addition, the clinical presentation in patients with C. difficile colitis who were not evaluated with CT was compared with scanned patients to determine whether differences were found in their presentation. Lastly, we also attempted to determine whether CT findings predicted surgical treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Of the 572 patients positive for toxins A and B, 171 (30%) had abdominal CT examinations within 2 weeks of their stool specimen collection date. Nineteen of these patients were eliminated from the study because of the following factors: six had underlying colonic disease (four diverticulitis, one cecal carcinoma, and one previous total colectomy), six had limited abdominal or pelvic CT examinations, five were under 18 years old, one study could not be retrieved, and one study was technically too poor to evaluate any portion of the colon. The remaining 152 patients formed the C. difficile colitis CT scanned population. The remaining 401 patients positive for toxins A and B did not have a CT examination and formed the unscanned study population.
CT Scan Parameters
All studies were performed on helical CT scanners and included both the
abdomen and pelvis. Seventy-two examinations were obtained with a
single-detector helical scanner (Somatom Plus-4, Siemens Medical Solutions);
66 of these studies were performed with 5-mm collimation and a pitch of 1.5,
and six were performed with 5-mm collimation and a pitch of 1. Forty-six
examinations were obtained with an MDCT scanner (Somatom Plus-4 Volume Zoom,
Siemens Medical Solutions); 44 of these studies were performed with 5-mm
collimation and pitch of 1, and two were performed with 3-mm collimation and a
pitch of 1. Twenty-five examinations were obtained with a single-detector
helical scanner (Emotion, Siemens Medical Solutions) with 5-mm collimation and
a pitch of 1.5. The remaining nine examinations were obtained with an MDCT
scanner (Sensation 16, Siemens Medical Solutions). Five studies were performed
with 3-mm collimation, and four studies were performed with 5-mm collimation.
All 16 had variable pitch depending on the z-axis coverage.
Because of the differing clinical indications for CT examination and underlying renal failure present in a portion of the patient population, the number of patients receiving IV, oral, and rectal contrast material varied greatly. One hundred forty-one (93%) of 152 patients received oral water-soluble contrast material (meglumine diatrizoate, Hypaque 25%; Amersham Health). One hundred eleven (73%) patients received 150 mL of 60% contrast material IV (ioversol, Optiray-300; Mallinckrodt), and 21 (14%) patients received water-soluble rectal contrast material (Hypaque 25%).
Image Review
The CT examinations were retrospectively reviewed separately by two
experienced abdominal imagers, one with 19 years of experience (reviewer A)
and the other with 13 years of experience (review B). These reviewers were
blinded to the prospective interpretation of the CT and clinical indicators
for the examination. The reviewers' findings were not compared with the final
radiologic report. One reviewer reviewed the CT examinations in alphabetical
order (reviewer A) and the other in reverse alphabetical order (reviewer B).
An electronic workstation (MagicView 1000, Siemens Medical Solutions) was used
to review all CT examinations. This allowed for precise measurements via the
electronic ruler and changes in window settings to better evaluate colonic
wall attenuation.
The reviewers determined if a scan was CT-positive or CT-negative based on colonic wall thickness. Based on the criteria from previously published reports, we defined abnormal colonic wall as greater than or equal to 4 mm [9, 11, 13, 14-16, 17]. If any segment of the colon had a wall thickness greater than 4 mm, the study was considered CT-positive. Abnormal colonic thickening was further categorized into mild (4-10 mm), moderate (11-15 mm), and severe (> 15 mm). The dependent portion of the colon was not used for wall thickness measurement as feces and debris within the colon often obscured the wall margin.
Once the scan was considered positive, more specific assessments were performed. First, we determined whether the colon was diffusely or segmentally involved. To do this, we divided the colon into six segments (rectum, sigmoid, descending, transverse, ascending, and cecum), and any affected segments were noted. Colon diameter (outer wall to outer wall) was also measured at each segment. We considered distention present if the bowel diameter was greater than 6 cm for any segment except the cecum (> 9 cm) [18]. Measurements for wall thickness and diameter were taken in the same plane and table location. For instance, if the measurement for wall thickness of the sigmoid colon was taken in the transverse plane at a certain table location, the diameter was also taken in the transverse plane at the same table location.
To avoid false-positive findings for colonic wall thickening, the most distended portion of the colon in each segment was used for wall thickness and diameter measurements. Also, the portion of the colonic segment used for these measurements had to have a diameter greater than 1.5 cm. The cecum was an exception, where greater than 2.5 cm was considered adequate distention. Certain segments of the colon were not evaluated because of previous resection, decompression of the entire segment prohibiting accurate measurement of wall thickness, or presence of a rectal tube obscuring the wall margins. In most cases, the entire segment was not fully distended. The rectum could not be evaluated in three cases because of previous resection in two cases and rectal tube placement in one case. The sigmoid, ascending, and transverse segments were resected in two cases and the cecum in five cases. The ascending segment was decompressed in one case.
In addition, the colonic folds were assessed for effacement and the presence of nodular thickening. Nodular fold thickening was considered to be present if a fold measured greater than 3 mm in width and was also further categorized as mild (> 3-9 mm), moderate (10-15 mm), and severe (> 15 mm). In those patients receiving IV contrast, the presence of a target sign (inner and outer layers of high attenuation surrounding a central area of decreased attenuation representing edema) was evaluated [17].
|
Clinical Presentation and Statistical Analyses
Using the hospital information system, all C. difficile-positive
patients (CT-positive, CT-negative, and unscanned patients) were reviewed for
clinical parameters, admitting diagnoses, and reasons for scanning when
applicable. The clinical parameters included age, sex, WBC, albumin, fever
(> 38.5°C), abdominal pain, nausea and vomiting, and diarrhea. For
purposes of statistical comparison, the admitting diagnoses and reasons for
scanning were broken down into five general categories: infectious signs and
symptoms including leukocytosis, fever, and hypotension; abdominal signs and
symptoms such as pain, distention, nausea and vomiting, and gastrointestinal
bleeding; neoplasia; vascular disease including abdominal aortic aneurysm,
aortic dissection, peripheral vascular disease; and miscellaneous diseases,
encompassing any admitting diagnosis or reason for scanning not falling under
the other four categories. In addition, the CT population was also evaluated
for cases in which surgical treatment was indicated or performed.
Statistical comparisons were then made between the different population groups (CT-positive, CT-negative, and unscanned patients) using the clinical information and specific CT findings obtained by the reviewers during image review, including the presence or absence of colonic thickening, segmental or diffuse involvement, presence or absence of colonic distention, preservation or effacement of colonic folds, nodular fold thickening, and target sign. Because of the relatively small CT population, ascites and pericolonic stranding were not used in the analysis.
To compare the clinical parameters of those patients who underwent scanning, chisquare tests were used for categoric indicators (or Fisher's exact test when the sample size was small). A two-sample Student's t test was used for age, and two-sample Wilcoxon's tests were used for WBC and albumin. Individual clinical indicators that reached statistical significance were included in a multivariable logistic model to assess the simultaneous effects of these variables in predicting which patients would have positive test results. Because of the retrospective nature of the study, not every variable could be accounted for and the significance of all clinical parameters could not be assessed in the multivariable model. A significance level of 0.05 was used for all analyses.
|
A comparison of the scanned (CT populations) and unscanned populations was also performed for the following variables: age, sex, admitting diagnoses, and reasons for scanning. For this comparison, 199 of the 401 patients who were not scanned were randomly selected and reviewed. These 199 patients represented the unscanned population. Several multivariable logistic regression models were fit and tested to assess the simultaneous effects on the various indications for CT and their interactions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
The reviewers disagreed on the presence of abnormal wall thickening in three cases. In two cases interpreted as CT-negative by reviewer A, reviewer B observed focal rectal thickening. In the third case, reviewer A observed focal sigmoid thickening in a case interpreted as CT-negative by reviewer B. Because the reviewers had nearly perfect agreement on the presence of wall thickening with a kappa statistic of 0.960 (95% confidence interval, 0.915-1.00), reviewer A's results were used to determine the frequency each colonic segment was thickened and the degree of wall thickening and nodularity for the affected segments.
The left colon was most often affected, with the rectum and sigmoid each involved in 82% of cases (60 of 73 cases and 61 of 74 cases, respectively) (Fig. 3). The descending colon was abnormally thickened in 57 (75%) of 76 cases. The right colon was affected to a lesser degree but was involved in well over 50% of cases (Fig. 4). The ascending colon was thickened in 42 (58%) of 73 cases and the cecum in 44 (62%) of 71 cases. The ascending colon had the greatest mean wall thickness (9.8 mm; range, 4-28 mm), and the cecum had the greatest mean diameter (47.3 mm; range, 22-71 mm). Evaluation of each segment was not possible in every case (Table 2).
|
Mild wall thickening (4-10 mm) was present in 55 cases. Fourteen had moderate thickening (11-15 mm) and seven severe (> 15 mm). Of the 25 patients with nodular fold thickening, nine were mild (> 3-9 mm), 11 moderate (10-15 mm), and five severe (> 15 mm). Ascites was present in 32 (42%) of 76 cases: 13 (41%) of 32 cases with a trace amount, eight (25%) of 32 with a small amount, nine (28%) of 32 with a moderate amount; and two (6%) of 32 cases with a large amount. Pericolonic stranding was present in the majority of cases (47 [62%] of 76 cases).
Population Comparisons
In the initial statistical comparison of clinical parameters in CT-positive
and CT-negative patients, positive scans were more likely to be found in
female patients and patients with higher WBC counts, abdominal pain, and
diarrhea. Forty (52.6%) of 76 of female patients had a positive scan versus 26
(34.2%) of 76 patients with a negative scan (p = 0.657). The mean WBC
was 19,400/µL for those patients with positive scans and 10,700/µL in
negative scans (p = 0.001). Abdominal pain was present in 47 (88.7%)
of 53 CT positive patients and 38 (65.5%) of 58 CT-negative patients
(p = 0.004). CT-positive patients had diarrhea in 38 (90.5%) of 42
cases and CT negative in 29 (53.7%) of 54 cases (p = 0.001). The mean
age, presence of fever, or nausea and vomiting were not statistically
significant clinical indicators of a positive CT scan.
In a multivariable regression analysis, the mean WBC, abdominal pain, and diarrhea were found to be statistically significant simultaneous predictors of a positive scan. Using this model, female sex was not a significant predictor as the initial statistical comparison indicated. Patients with abdominal pain were 9.3 times more likely to have a positive scan (p = 0.004) and patients with diarrhea, 9.7 times more likely to have a positive scan (p = 0.001) than those patients without these symptoms. For every increase in the WBC of 5,000/µL the chance of a positive scan doubled (p = 0.003) (Table 3).
|
No statistical correlation was found in the comparison of the specific CT findings and clinical parameters in CT-positive patients. However, some interesting trends were noted. There was a small degree of positive correlation (r = 0.21) between increasing mean WBC and wall thickness (unadjusted p = 0.021; adjusted p = 0.77). Twenty-three (64%) of 36 patients with abdominal pain had a target sign compared with very few patients without abdominal pain (unadjusted p = 0.026; adjusted p =0.91).
Similarly, no significant relationship or predictive value was found between the clinical parameters or specific CT findings in those CT-positive patients requiring or indicated for surgical treatment. Only six (4%) of 152 of patients had surgery or had signs and symptoms indicating surgery from the entire CT population. Five patients underwent surgery (two colectomy, one hemicolectomy, one ileostomy, and one exploratory laparotomy). One patient refused and chose instead a do-not-resuscitate status. Interestingly, three came from the CT-positive group and three from the CT-negative group.
In all surgical cases from the CT-positive population, the scan was performed 2 days before surgery or the decision for surgical treatment. Two CT-positive patients had known C. difficile colitis and sepsis refractory to medical therapy. One underwent subtotal colectomy and the other, a hemicolectomy. The third CT-positive patient had Crohn's disease and underwent ileostomy to divert her profuse fecal stream (believed to be secondary to C. difficile colitis) in the clinical setting of severe anal ulcers. The time interval between scanning and surgical treatment, however, varied from 2, 15, and 30 days in the CT-negative group. The patient indicated for surgery 2 days after her scan for refractory sepsis and severe abdominal pain refused surgical treatment. An exploratory laparotomy for colonic distention and suspected volvulus was performed 15 days after another CT-negative patient's scan. The last CT-negative patient indicated for surgery 30 days after developing a persistent ileus in the interval between scanning and surgery. He underwent total colectomy, which revealed a necrotic colon.
Comparison of the scanned and unscanned populations using the mean age,
sex, admitting diagnoses, and reasons for scanning found that patients with
infection (79 [52%] of 152 cases) (p 
0.001) and abdominal
complaints (93 [61%] of 152 cases) (p 0.001) were more likely to
undergo scanning, and those with neoplasm (41 [21%] of 199 cases) (p
= 0.046) or vascular disease (35 [18%] of 199 cases) (p = 0.001) were
less likely to be scanned. Sex and mean age were not statistically significant
predictors of a patient being or not being scanned
(Table 4).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Commonly described CT findings of C. difficile colitis include wall thickening, extent of disease, abnormal bowel wall attenuation (target sign), and nodular fold thickening. Wall thickening is the most common finding associated with C. difficile colitis and the least specific [17-19]. For our study, the presence of wall thickening was the only criterion for a positive CT study. The degree of wall thickening is generally thought to be greater than in other inflammatory or infectious forms of colitis (often > 10 mm) and is classically diffuse in distribution [12, 13, 17-20]. This contrasts our findings, in which the majority of CT-positive patients (55 of 76 patients) had mild thickening (4-10 mm), and the ascending colon was the most severely thickened segment, with a mean thickness of 9.8 mm. Also, only a third of the patients in our study had diffuse colonic involvement. Many cases isolated to a single colonic segment or the right colon and cases of rectal sparing were present in our patient population. Similar findings have been reported previously [9, 13, 21]. We also found a left colonic predominance with the rectum and sigmoid colon being affected in 82% of cases, similar to Boland et al. [12].
The target sign and nodular fold thickening are also common nonspecific findings associated with C. difficile colitis. The target sign was originally described in cases of inflammatory bowel disease and indicates mucosal hyperemia, submucosal edema, and inflammation in cases in which IV contrast was administered [3]. In our study, the target sign was present in 34 (61%) of 56 cases. Nodular fold thickening was present in 26 (34%) of 76 cases, which is similar to Boland et al. [12], who reported that it was present in 11 (28%) of 39 patients with wall thickening.
The only CT finding thought to be specific for C. difficile colitis is a type of nodular fold thickening called the "accordion sign," which results from oral contrast material being trapped between nodular, thickened bowel folds [3, 22-25]. This finding is thought to be the most specific for C. difficile colitis but is also the least common. The accordion sign was present in only two of 54 C. difficile positive patients in Kirkpatrick and Greenberg's [14] study, three of 64 patients in the study conducted by Boland et al. [12], and five of 26 patients in the study conducted by Fishman et al. [13]. Initially, we did not specifically assess for this finding as it was not our intention to compare colitis of different causes. However, in retrospect, it was seen in only four patients in our study.
Although the previously described CT findings can be used to determine the presence of C. difficile colitis, the fundamental problem in using CT as a diagnostic tool is that many, if not most, patients with this infection have no findings. In our experience, only 50% of scanned C. difficile-positive patients had abnormal bowel wall thickening (> 4 mm), a lower prevalence than in three relatively large studies. In these series, a thickened colon on CT was seen in 39 (61%) of 64 patients [12], 41 (76%) of 54 patients [14], and 23 (88%) of 26 patients [3]. But, as previous studies have stated, a negative scan does not preclude the presence of C. difficile colitis and "even in those who [do] show abnormalities, the findings [are] not always indicative of C. difficile colitis in particular" [14].
The prevalence of abnormal scans in our study is most likely an overestimation. Patients in our study were scanned at the discretion of the clinical team. Thus, symptomatic patients were more likely to be scanned, leading to an underlying selection bias. This is illustrated by a comparison of the clinical parameters in scanned versus unscanned patients in our study. We found that patients with signs and symptoms of infection and abdominal complaints were much more likely to be scanned than those with neoplastic or vascular disease. A smaller study composed of 25 patients also showed that CT was prompted by abdominal pain and fever in most cases [15]. In general, the majority of hospitalized patients with C. difficile colitis are not scanned, which likely reflects the fact that most patients with C. difficile are asymptomatic or only mildly to moderately ill [3, 16]. This was the foremost limitation in our study and is a common shortcoming in many other studies [12-14].
Ideally, to determine the true prevalence of an abnormal CT, all patients with positive immunoassay tests would undergo scanning, regardless of symptoms. One hundred seventy-one (30%) of the 572 C. difficile-positive patients in our study underwent scanning, versus 64 (10%) of 675 patients in the study by Boland et al. [12], and 54 (13%) of 430 patients in the study by Kirkpatrick and Greenberg [14]. Abnormal wall thickening was present in 39 (61%) of 64 patients, and 41 (76%) of 54 patients had colonic abnormalities on CT in these two studies, respectively. As a higher percentage of patients are scanned, the percentage of abnormal scans trends downward and more accurately reflects the prevalence of abnormal CT findings in patients with C. difficile colitis. We believe that the percentage of abnormal examinations would have fallen below 50% if all C. difficile-positive patients were scanned.
Because of the lack of specific CT findings and large number of normal CT examinations in patients with C. difficile colitis, most studies have concluded that CT has a limited diagnostic role. In the presence of abnormal CT findings with the proper clinical history, CT can suggest but not confirm the diagnosis of C. difficile colitis [14, 24]. Unfortunately, the clinical history is not always available, and not all patients present with classic symptoms. Many patients subsequently diagnosed with C. difficile colitis do not present with diarrhea but rather with generalized constitutional symptoms, obstipation, or even an acute abdomen [2, 9, 21, 23, 26]. In these cases, a positive CT scan can lead to the correct diagnosis and to appropriate management and therapy. Conversely, Kirkpatrick and Greenberg [14] found that a "patient with severe clinical features of C. difficile disease may show little or no CT evidence of colonic abnormality and that some patients with CT evidence of colitis may only suffer from mild symptoms." Our own results also show the disparity between clinical and CT findings in patients with C. difficile colitis, which is the major obstacle in using CT to guide therapy.
Mean WBC, abdominal pain, and diarrhea were found to be statistically significant simultaneous predictors of a positive scan in our study. Patients with abdominal pain and diarrhea were approximately nine times more likely to have a positive scan and for every increase in the WBC of 5,000/µL the chance of a positive scan doubled. Nonetheless, no statistical correlation was found between more specific CT findings and clinical parameters in CT-positive patients. In our experience, worsening clinical symptoms did not translate statistically into a greater number or degree of colonic abnormalities. A small degree of positive correlation, however, was noted between increasing bowel wall thickness and WBC. This relationship may have turned out to be statistically significant had the study population been larger. To our knowledge, only Boland et al. [16] compared clinical presentation with CT findings. Using fever, WBC count, and frequency and duration of diarrhea to estimate clinical disease severity, they found the clinical severity of disease did not statistically differ between patients with CT colitis and those without. The only CT finding they found that correlated with clinical severity of disease was nodular mucosal thickening, which was significantly more frequent in patients with a WBC greater than 11,000/µL [16].
Although C. difficile colitis is most often treated medically and not considered a surgical disease, CT in certain instances has played a dynamic role in determining the need for surgical intervention in severe cases [25-27]. Although the indication for surgery and operation of choice are still not well defined and the decision to operate is typically based on overall system deterioration or physical findings of peritonitis [28], some authors think CT can be used to monitor the efficacy of medical treatment and complications [27, 29]. This is based on the assumption that once medical therapy has started, symptoms should resolve or lessen within 1 to 4 days [27, 29]. If interval CT examinations show progression of the disease process or even rare complications such as perforation [25, 26, 29] on an initial scan, subtotal colectomy or hemicolectomy may be indicated [26-29].
In our experience, there was no significant relationship or predictive value between the clinical parameters in scanned patients or specific CT findings in those CT-positive patients requiring surgical treatment. Six (4%) of 152 patients were indicated for surgery from the entire CT population. Clinically, the mean WBC was 14,700/µL in patients indicated for surgery. This was not statistically different from the average WBC count found for CT positive (19,400/µL) and negative patients (10,700/µL). This also differed from smaller studies in which the average WBC was far greater than 20,000/µL [7, 28]. Of the clinical parameters, fever was the most common, having been present in all cases. Abdominal pain was present in four cases, diarrhea in two, and nausea and vomiting in none. Not surprisingly, abdominal pain is a common clinical finding in C. difficile colitis patients requiring surgery. Medich et al. [27] reported that all 12 patients with C. difficile colitis requiring surgical treatment presented with an acute abdomen. Likewise, in the study by Lipsett et al. [28] abdominal pain was present in 12 (92%) of 13 patients, and all six surgical patients in the study by Waddell et al. [7] had peritoneal signs.
Interestingly, only three of the six patients in our study indicated for surgery had bowel wall thickening. Similarly, Kawamoto et al. [30] found no significant difference in the CT findings of surgical and nonsurgical groups. However, all patients requiring surgery in their study had abnormal thickening of the colon and all patients in the study by Medich et al. [27] had colonic thickening and pancolitis. Of the three patients requiring surgery in our study, only one had diffuse colonic involvement. Also, colonic dilatation, which can be suggestive of an underlying toxic megacolon and was present in three (50%) of six patients in the study by Waddell et al. [7], was not seen in any of our surgical cases. Unfortunately, the small number of patients indicated for surgery may not have been large enough to produce a statistically significant relationship, a type II error. Of note, all three patients indicated for surgery with a negative CT examination were diabetics. This may suggest an underlying immunosuppressed state and could partially account for the lack of inflammatory colonic changes. Also, there was a greater time interval between the date of the CT and surgical treatment in the CT-negative group (2, 15, and 30 days) versus 2 days for each CT-positive patient. It may be that the colon became abnormal in the patients who underwent resection, especially the two with surgery 15 and 30 days after the CT.
Because of the retrospective nature of our study, there are many inherent limitations. Not all C. difficile-positive patients were scanned, and patients were scanned at the discretion of the clinical team. Thus, there was likely a selection bias. Patients were not always scanned with the same CT parameters and did not receive the same form of contrast in each case. Bowel wall thickening in some cases may have been caused by unrecognized coexistent diseases processes. Given the 2-week time interval between CT and collection of the stool specimen, it was possible that more acute forms of C. difficile colitis resolved before scanning or were not yet fulminant. Also, our C. difficile population was based on stool specimens positive by enzyme immunoassay for both toxins A and B. This type of assay has a reported sensitivity of 69% to 87% and specificity of 99% to 100% [5]. The high specificity confirms that C. difficile-positive patients in our population group were actually affected. However, the lower sensitivity of this assay test means that many C. difficile-positive patients scanned during the time frame of our study could have been excluded.
The small surgical population in our study prevents any evidence-based recommendations about the appropriate use and role of CT in C. difficile colitis. However, our clinical colleagues would suggest scanning patients in clinically suspected cases of toxic colitis (patients with diarrhea and two or more of the following: fever, tachycardia, leukocytosis, and hypoalbuminemia); or toxic megacolon (patients with severe abdominal distention, tenderness, and pain). The only way to determine the role of CT in C. difficile colitis would be to prospectively study every patient with disease, measuring multiple clinical and radiologic parameters. Practical considerations would preclude such a study.
Equally important is that C. difficile colitis was not considered clinically in several of our cases until the diagnosis was suggested by CT findings. Given the increasing rates of C. difficile colitis, clinicians should be aware of the diagnosis in hospitalized patients on antibiotics, including the postoperative period, before the disease process becomes severe [1].
In summary, it was our experience that 50% of scanned hospitalized patients with C. difficile colitis had positive findings on CT. This percentage is a probable overestimation as selection bias was highly likely. Patients were more likely to be scanned if they had signs and symptoms of infection or abdominal complaints. Patients with neoplastic and vascular disease were less likely to be scanned. Patients with abnormal scans most often had segmental colonic involvement, with the left colon being most often affected. Positive scans were associated with increased WBC, abdominal pain, and diarrhea. There was no statistical correlation between specific CT findings and clinical parameters. We found no predictive value of specific CT findings with regard to surgical treatment, and no statistical correlation was found between clinical parameters and patients requiring surgery. The absence of a significant statistical relationship may relate to the small surgical population producing a type II error. These findings point to the limitations of CT with regard to diagnosis and guidance of therapeutic treatment in cases of C. difficile colitis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Kenneally, J. M. Rosini, L. P. Skrupky, J. A. Doherty, J. M. Hollands, E. Martinez, W. McKenzie, T. Murphy, J. R. Smith, S. T. Micek, et al. Analysis of 30-Day Mortality for Clostridium difficile-Associated Disease in the ICU Setting Chest, August 1, 2007; 132(2): 418 - 424. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
