Abdominal computed radiography for the diagnosis of enterolithiasis in horses: 142 cases (2003–2007)

Omar Maher William R. Pritchard Veterinary Medical Teaching Hospital, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by Omar Maher in
Current site
Google Scholar
PubMed
Close
 DVM, DACVS
,
Sarah M. Puchalski Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by Sarah M. Puchalski in
Current site
Google Scholar
PubMed
Close
 DVM, DACVR
,
Christiana Drake Department of Statistics, University of California-Davis, Davis, CA 95616.

Search for other papers by Christiana Drake in
Current site
Google Scholar
PubMed
Close
 PhD
, and
Sarah S. le Jeune Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA 95616.

Search for other papers by Sarah S. le Jeune in
Current site
Google Scholar
PubMed
Close
 DVM, DACVS

Click on author name to view affiliation information

Abstract

Objective—To evaluate the sensitivity and specificity of abdominal computed radiography (CR) for the diagnosis of enterolithiasis in horses and to examine how these parameters are affected by the number and anatomic location of enteroliths and by gas distension of the gastrointestinal tract.

Design—Retrospective case series.

Animals—Horses ≥ 1 year old that underwent abdominal CR and subsequent exploratory laparotomy or postmortem examination.

Procedures—3 reviewers blinded to signalment, history, clinical signs, and diagnoses separately evaluated abdominal computed radiographs of horses included in the study. Each set of radiographs was evaluated for the presence or absence of enteroliths, the amount of gas distention, and the image quality. Signalment, definitive diagnosis on the basis of findings on exploratory laparotomy or postmortem examination, and the number and location of enteroliths were obtained from medical records.

Results—Of the 142 cases reviewed, 58.4% (83/142) had confirmed enterolithiasis. For the 3 reviewers, overall sensitivity was 85% and specificity was 93%. Sensitivity was lower for small colon enteroliths than for large colon enteroliths (50% and 94.5%, respectively) and was significantly affected by gas distention. Sensitivity was not significantly affected by the number of enteroliths.

Conclusions and Clinical Relevance—Computed radiography provided high sensitivity and high specificity for the diagnosis of enterolithiasis in horses. Caution should be exercised when the radiographic results are negative, as the sensitivity for small colon enterolithiasis was relatively low and gas distension negatively affected detection of enteroliths. Abdominal CR is indicated as a diagnostic test in horses examined for colic in geographic regions in which enterolithiasis is endemic.

Abstract

Objective—To evaluate the sensitivity and specificity of abdominal computed radiography (CR) for the diagnosis of enterolithiasis in horses and to examine how these parameters are affected by the number and anatomic location of enteroliths and by gas distension of the gastrointestinal tract.

Design—Retrospective case series.

Animals—Horses ≥ 1 year old that underwent abdominal CR and subsequent exploratory laparotomy or postmortem examination.

Procedures—3 reviewers blinded to signalment, history, clinical signs, and diagnoses separately evaluated abdominal computed radiographs of horses included in the study. Each set of radiographs was evaluated for the presence or absence of enteroliths, the amount of gas distention, and the image quality. Signalment, definitive diagnosis on the basis of findings on exploratory laparotomy or postmortem examination, and the number and location of enteroliths were obtained from medical records.

Results—Of the 142 cases reviewed, 58.4% (83/142) had confirmed enterolithiasis. For the 3 reviewers, overall sensitivity was 85% and specificity was 93%. Sensitivity was lower for small colon enteroliths than for large colon enteroliths (50% and 94.5%, respectively) and was significantly affected by gas distention. Sensitivity was not significantly affected by the number of enteroliths.

Conclusions and Clinical Relevance—Computed radiography provided high sensitivity and high specificity for the diagnosis of enterolithiasis in horses. Caution should be exercised when the radiographic results are negative, as the sensitivity for small colon enterolithiasis was relatively low and gas distension negatively affected detection of enteroliths. Abdominal CR is indicated as a diagnostic test in horses examined for colic in geographic regions in which enterolithiasis is endemic.

Abdominal radiography is routinely used as part of the diagnostic workup of adult horses in geographic regions where enterolithiasis is a common cause of colic.1 The identification of enterolithias via radiography enables rapid and appropriate management of the patient, including surgical removal of the enteroliths, which is particularly helpful in horses with mild clinical signs where the need for surgical intervention may not be clear. In a 1994 study2 of horses evaluated at the William R. Pritchard Veterinary Medical Teaching Hospital, University of California-Davis, analogue radiography was reported to have a mean sensitivity of 76.9% and mean specificity of 94.4% for the diagnosis of enterolithiasis.

Computed radiography has become more common in veterinary diagnostic imaging in recent years and is generally assumed to improve the capabilities of radiography as a diagnostic technique, compared with analogue radiography. However, the effect of CR systems on the diagnostic accuracy of abdominal radiography for the specific identification of enterolithiasis is not known. The purpose of the study reported here was to evaluate the sensitivity and specificity of abdominal CR for identifying enteroliths in adult horses evaluated at the University of California's Veterinary Medical Teaching Hospital and to examine how these parameters would be affected by the number and anatomic location of enteroliths and by the presence of gas distension of the gastrointestinal tract. We hypothesized that mean sensitivity and specificity would be improved with the implementation of CR when compared with the previous study2 that used analogue radiography.

Materials and Methods

Case selection—Computerized medical records of all horses ≥ 1 year old evaluated at the William R. Pritchard Veterinary Medical Teaching Hospital that underwent abdominal CR between January 2003 and February 2007 with subsequent exploratory laparotomy or postmortem examination were reviewed. Signalment, definitive diagnosis on the basis of findings on exploratory laparotomy or postmortem examination, and number and location of enteroliths were recorded.

All abdominal radiographs were obtained by use of CR.a To obtain a complete study, at least 4 projections were taken with 14 × 17-inch cassettes as described previously.2 The radiographic cassettes were housed in an overhead rail system cassette holder with either a 10:1 or a 12:1 grid. Tube output parameters were set at 140 kVp and 160 mAs with a range of 100 to 200 mAs. All digital images were stored on the hospital computer picture-archiving and communications system.

The radiographs were evaluated on a dedicated image review workstation comprised of 3 liquid crystal display monitors. Radiographs were reviewed on 2 of the 3 monitors. The 2 monitorsb were 20.8-inch, 3-megapixel, monochrome, medical-grade monitors calibrated to the grayscale display function.

Radiographic evaluation—Radiographs of cases meeting inclusion criteria were evaluated separately by 3 reviewers (a board-certified surgeon, a board-certified radiologist, and an equine surgery resident). All 3 reviewers were blinded to the signalment, history, initial clinical signs, and diagnoses. Each set of radiographs was evaluated for the presence or absence of enteroliths, the amount of gas distension (score of 1 = mild, 2 = moderate, and 3 = severe), and the radiographic image quality (score of 1 = adequate and score of 2 = inadequate).

Statistical analysis—Sensitivity and specificity were calculated for each of 3 reviewers. A consensus measure for the 3 reviewers was formed by scoring a positive finding when at least 2 of the 3 reviewers had a positive finding, and a negative finding when at least 2 of the 3 reviewers had a negative finding. Sensitivity and specificity were also calculated for this consensus finding. The ability of each reviewer to correctly diagnose the presence or absence of enteroliths was calculated as the κ statistic,3 a measure of agreement between postmortem or surgical findings and radiographic diagnosis. Furthermore, sensitivity was also calculated by location of the enteroliths and as a function of the number of enteroliths determined by use of postmortem or surgical findings. Furthermore, dependence of sensitivity by location and number of enteroliths and the association between location of enteroliths and gas distension were assessed by use of χ2 tests of association or the Fisher exact test. All tests were considered significant if P < 0.05. The McNemar test was used to assess the nature of the disagreement between postmortem or surgical findings and the radiographic diagnosis.

Results

One hundred forty-two horses met the inclusion criteria. The most common breeds were Arabians (n = 48) and Quarter Horses (43) followed by Thoroughbreds (9), Appaloosas (8), Paints (8), Morgans (3), warmbloods (6), American Miniature Horses (5), Shetland Ponies (3), Tennessee Walkers (3), and American Saddlebred, Paso Fino, and Belgian (1 each). There were 19 sexually intact males, 62 geldings, and 61 females. Mean age was 8.2 years (median, 17 years; range, 1 to 34 years).

Final diagnosis on the basis of surgical or postmortem findings in the 142 horses included enterolithiasis (n = 83), large colon displacements (15), sand impactions (8), fecaliths (7), nonstrangulating lesions of the small intestine (5), colitis (3), large colon impactions (6), small colon impactions (5), strangulating lesions of the small intestine (4), cecal impactions (2), internal abscess (1), and undetermined cause (3).

Overall, 89% of radiographic images reviewed were deemed of adequate diagnostic quality. One enterolith was found in 50 patients, 2 to 5 enteroliths were found in 22 horses, and 6 or more enteroliths were found in 11 horses. Enteroliths were localized to the right dorsal colon in 54 patients, transverse colon in 8 patients, small colon in 12 patients, and more than 1 location in 9 patients. There was no association between location of the enterolith (large colon vs small colon) and severity of gas distension seen on radiographs by all 3 reviewers.

Sensitivity for diagnosis of enterolithiasis (Table 1) was high for all 3 reviewers (83%, 85%, and 92%), as was the specificity (91%, 88%, and 92%). There was good agreement among reviewers, with κ values of 0.86, 0.81, and 0.76, respectively, for the comparisons between reviewers 1 and 2, 1 and 3, and 2 and 3. Overall (consensus) sensitivity was 85%, and specificity was 93%. The McNemar test revealed values of P = 0.25, 0.04, and 0.99 for the 3 reviewers. Sensitivity was not affected by the number of enteroliths (P = 0.066). Sensitivity was lower for small colon enteroliths (P = 0.016) and was negatively affected by the amount of gas distention (P = 0.007).

Table 1—

Sensitivity of CR for the detection of enteroliths in 142 horses ≥ 1 year old that underwent abdominal CR and subsequent exploratory laparotomy or postmortem examination at the University of California William R. Pritchard Veterinary Medical Teaching Hospital between 2003 and 2007. Radiographs were evaluated separately by 3 reviewers (a board-certified surgeon, a board-certified radiologist, and an equine surgery resident). Sensitivity and specificity were calculated for each of the 3 reviewers. A consensus measure of sensitivity for the 3 reviewers was formed by scoring a positive finding when at least 2 of the 3 reviewers had a positive finding, and a negative finding when at least 2 of the 3 reviewers had a negative finding.

VariableSensitivity (%)P value
No. of enteroliths 0.066
  182.98 
  2–590.91 
  ≥ 681.82 
Location of enteroliths 0.001
  Large colon94.55 
  Large colon and small colon88.89 
  Small colon50.00 
  Transverse colon75.00 
Gas distention 0.007
  Mild95.00 
  Moderate85.71 
  Severe60.00 

Discussion

In the present study, CR provided high sensitivity and high specificity with excellent interobserver agreement for the diagnosis of enterolithiasis in adult horses at our hospital. Computed radiography sensitivity for detecting enteroliths was higher (85%) than that in a previously reported study2 of our hospital population that used conventional radiography (76.9%). Caution should be exercised when the radiographic results are negative for enterolithiasis as the sensitivity for detection of small colon enterolithiasis was relatively low and a large quantity of gas negatively affected detection of enteroliths. Results suggested that CR was indicated as a diagnostic test in horses examined for colic in geographic areas in which enteroliths are endemic.

Although it is difficult to directly compare the results of the present study with those of a prior study2 of analogue radiology conducted in our hospital because of differences in the dates of admission of the horses included, similarities in study design, inclusion criteria, and hospital population allow for some conclusions to be drawn. The present study found that the overall sensitivity of CR (85%) for diagnosing enterolithiasis was higher than the sensitivity for conventional radiology (76.9%).2 In the previous study,2 the most common error leading to false-negative diagnoses in enterolith detection was reported to be inadequate penetration.2 Inadequate penetration or underexposure in digital radiography manifests as a decrease in signal-to-noise ratio, giving digital radiographs an overall appearance of graininess or image mottle.4,5 Signal-to-noise ratio is a large factor in image quality; however, with equine abdominal radiography, even in the presence of underexposure, digital image display software may help to mitigate this problem. Although CR requires more radiation than conventional radiography to produce a diagnostic quality radiograph with a similar signal-to-noise ratio6–8 and adult equine abdominal radiographs are taken at near maximal capability of the x-ray generator, underexposure did not seem to have an impact on sensitivity of CR images, unlike for analogue radiographs.

In the group of horses included in this study, the breed distribution was similar to that in the previous study2 on abdominal radiography and a retrospective study9 on enterolithiasis in 900 horses at the same institution. Arabians represented 33.8% of the horses radiographed, with Quarter Horses being the second most common breed (30%). This might be a reflection of the referral of predisposed breeds in California for radiographic screening for enteroliths or the greater likelihood that predisposed horses examined with equivocal colic signs would be screened with radiography at our hospital.

As in the earlier study,2 detection of large colon enteroliths (94.5%) was superior to detection of small colon enteroliths (50%). From a radiographic standpoint, the ability to detect an enterolith is largely dependent on contrast resolution.10 Contrast resolution, defined as the ability to distinguish differences in opacity, is influenced by the radiography system and intrinsic patient contrast. Computed radiography as a system provides greater contrast resolution than does analogue radiography, which may explain the overall increase in sensitivity. However, the reasons for a greater ability to detect large colon versus small colon enterolithiasis are difficult to discern. Factors influencing intrinsic patient contrast would be possible reasons. The ventral and somewhat peripheral anatomic location of the large colon likely decreases the amount of superimposition of abdominal viscera, thereby increasing the intrinsic contrast when compared with the largely central location of the small colon. Furthermore, the ingesta of the large colon are generally less densely packed, providing more contrast around the margins of the enterolith.

Increased gas distension negatively affected sensitivity of enterolith detection overall in the present study. Typically, horses with small colon enterolithiasis have more severe clinical signs associated with a complete obstruction at this location.9 Orad intraluminal gas accumulation may influence the intra-abdominal position of the bowel and compress intestinal content around the periphery of small colon enteroliths, making them more difficult to detect. Gas distension may have a less profound effect on the position and intraluminal ingesta of the large colon, in that gas distention was identified in horses with both large and small colon enterolithiasis and no association could be made between anatomic location of the enterolith and degree of gas distension in the present study.

Analysis of the effect of the number of enteroliths on the overall sensitivity did not show a significant influence in this study. A greater number of enteroliths did not improve overall sensitivity for enterolithiasis with CR; however, the P value was 0.066. In general, enteroliths in horses with multiple stones were smaller than enteroliths in horses with 1 solitary enterolith, and smaller stones may be more difficult to detect.

Statistical analyses indicate that although the sensitivity values between the reviewers were different, the interreader agreement was excellent with κ values ranging between 0.76 and 0.86. The 3 reviewers included a surgery resident, a board-certified surgeon, and a board-certified radiologist, thereby providing a cross section of individuals who would likely be evaluating abdominal radiographs in a clinical setting. The excellent agreement between reviewers boosts confidence that other clinicians could use this diagnostic test appropriately. The McNemar test was used to assess the nature of the disagreement between the reviewer's findings and postmortem and surgical findings, and bias was found for the second reviewer, who was more likely to have false-negative results than false-positive results. The reason for this finding was unclear.

ABBREVIATION

CR

Computed radiography

a.

Fuji FCR5000 Plus, Fujifilm Medical Systems, Stamford, Conn.

b.

ME351i, Totoku, North America Inc, Irving, Tex.

References

  • 1.

    Rose JA, Sande RD. Radiographic diagnosis of equine enterolithiasis, in Proceedings. 26th Annu Meet Am Assoc Equine Pract 1980; 211.

  • 2.

    Yarbrough TB, Langer DL, Snyder JR, et al. Abdominal radiography for diagnosis of enterolithiasis in horses: 141 cases (1990–1992). J Am Vet Med Assoc 1994; 205: 592595.

    • Search Google Scholar
    • Export Citation
  • 3.

    Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: Wiley & Sons, 1981.

  • 4.

    Drost W, Reese D, Hornof W. Digital radiography artifacts. Vet Radiol Ultrasound 2008; 49 (suppl 1):S48S56.

  • 5.

    Widmer WR. Acquisition hardware for digital imaging. Vet Radiol Ultrasound 2008; 49 (suppl 1):S2S8.

  • 6.

    Bacarini L, Giacomich R, Saccavini C. The physical aspects of traditional radiology and of computed radiology compared [in Italian]. Radiol Med (Torino) 1995; 89: 506519.

    • Search Google Scholar
    • Export Citation
  • 7.

    Miettunen RH, Korhola OA. The effect of scatter reduction on the signal-to-noise ratio in computed radiography. Eur J Radiol 1991; 12: 167170.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Pongnapang N. Practical guidelines for radiographers to improve computed radiography image quality. Biomed Imaging Interv J 2005; 1:e12.

    • Search Google Scholar
    • Export Citation
  • 9.

    Hassel DM, Langer DL, Snyder JR, et al. Evaluation of enterolithiasis in equids: 900 cases (1973–1996). J Am Vet Med Assoc 1999; 214: 233237.

    • Search Google Scholar
    • Export Citation
  • 10.

    Bushberg JT, Leidholdt E, et al. The essential physics of medical imaging. 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2002.

All Time Past Year Past 30 Days
Abstract Views 228 0 0
Full Text Views 762 600 292
PDF Downloads 305 150 14
Advertisement