• Abdominocentesis;
  • Cattle;
  • Diagnosis


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References

Background: Peritoneal fluid analysis in cattle traditionally includes the classic parameters despite the fact that they have only moderate diagnostic accuracy and often fail to identify the pathogenesis or etiological factors. Therefore additional parameters recently have been established to improve diagnostic precision. In a recent study, reference ranges for several of these parameters have been proposed in dairy cows.

Hypothesis/Objectives: The aim of this observational study was to assess the diagnostic value of D-Dimer and other measurements of peritoneal fluid analysis in dairy cows with peritonitis.

Animals: The study included 110 Holstein-Friesian cows grouped into cows with peritonitis (n = 47) and cows without peritonitis (n = 63).

Methods: Peritoneal fluid was obtained by abdominocentesis. Total protein, albumin, glucose, cholesterol, fibrinogen, l-lactate, D-Dimer, lactate dehydrogenase (LDH), alkaline phosphatase, creatine phosphokinase, white blood cell, and red blood cell were determined in peritoneal fluid and venous blood. Serum-ascites albumin gradient (SAAG) and ratios of peritoneal fluid-venous blood were calculated. Sensitivity (SN) and specificity (SP) were calculated and receiver operating characteristic curve analysis performed.

Results: Peritoneal fluid D-Dimer was most accurate in diagnosing peritonitis in cows (SN and SP>95.0%). Total protein concentration, LDH and LDH ratio, and SAAG had sensitivities between 49.0 and 67.1%, and specificities between 88.4 and 95.5%. A low-peritoneal fluid glucose concentration was found to be highly indicative of septic peritonitis.

Conclusions and Clinical Importance: Measurement of the recently introduced parameters may increase the diagnostic value of peritoneal fluid analysis and provide additional specific information. Therefore these measurements should be included in the routine procedure.


alkaline phosphatase


confidence interval


creatine phosphokinase


lactate dehydrogenase


nicotinamide adenine dinucleotide


red blood cell


receiver operating characteristic


serum-ascites albumin gradient






white blood cell

Peritoneal fluid analysis is a useful diagnostic method in gastroenterology because this fluid generally reflects conditions in the peritoneal cavity.1–3 Traditionally, peritoneal fluid analysis in cattle has included only a few parameters that are used by the classic transudate-exudate classification.3 However, in some circumstances, these measurements fail to identify and evaluate the pathogenesis or etiological factors of the disease, and thus they have only moderate diagnostic accuracy.4 Furthermore, the cut-off value of peritoneal fluid protein concentration might not be appropriate for cattle.5 Additional measurements and calculations (Light's criteria, serum-ascites albumin gradient [SAAG], fibrinogen, C-reactive protein, glucose, d-lactate, l-lactate, hemoglobin, D-Dimer, inorganic phosphate, lysosomal enzymes, alkaline phosphatase [ALP], lactate dehydrogenase [LDH], creatine phosphokinase [CPK], and cytology) have been established in human and veterinary medicine to improve diagnostic precision, to allow early diagnosis, and to identify certain pathological conditions such as peritoneal bacterial infection, intestinal ischemia, or cell damage.6–13 In a recent study, reference ranges of several peritoneal fluid parameters have been proposed in dairy cows.5 However, the diagnostic accuracy of these parameters needs to be evaluated.

The first objective of this study was to examine peritoneal fluid in cows with clinically diagnosed peritonitis for changes that could indicate intra-abdominal inflammatory processes (albumin, fibrinogen, D-Dimer), cell damage (CPK, ALP, LDH), bacterial infection (glucose), and intestinal ischemic conditions (l-lactate, D-Dimer). Light's criteria, SAAG, and ratios between peritoneal fluid and blood were calculated. Secondarily, the diagnostic value of these parameters was assessed according to the recently proposed peritoneal fluid reference ranges,5 and receiver operating characteristic (ROC) curve analysis was performed. The accuracy of these values also was compared with parameters of the classic transudate-exudate system.3

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References

The study included 110 dairy cows from 38 different herds that were referred to the veterinary hospital by practitioners between December 2006 and May 2008 for further diagnosis and therapy. The cows comprised Holstein-Friesian or crossbreeds (Holstein-Friesian/Deutsche Schwarzbunte [German Black Pied]) with an average age of 4.4 ± 1.5 years. The mean duration of lactation was 62 days (13 days [1st quartile], 91 days [3rd quartile]) and the body condition score14 was 2.75 (2.50 [1st quartile], 3.50 [3rd quartile]). The study was approved by the Institutional Animal Use and Protection Committee. A thorough clinical examination was performed in each cow at admission. Cows were assigned to 2 groups (Group P, peritonitis; Group N, nonperitonitis) according to findings of the clinical examination. Almost all cows of both groups had been treated with different antimicrobial and anti-inflammatory drugs by owners or local veterinarians before admission.

Cows with at least 4 of the following clinical signs of peritonitis were included in Group P (n = 47): (1) increased abdominal wall tension, (2) kyphosis, (3) response to pain provocation tests (withers test, percussion of cranial abdomen), (4) rough peritoneal surface or adhesions found during careful rectal examination, (5) pneumoperitoneum, and (6) increased peritoneal fluid volume, fibrin formation, or both seen on transcutaneous abdominal ultrasonography. Additional pathological conditions (mastitis [n = 13], endometritis [n = 12], lameness [n = 9], injuries [n = 2], and skin lesions [n = 1]) were diagnosed in 32 of these cows.

The second Group N (n = 63) consisted of cows that did not show any of the above described clinical findings and served as control group. Cows included in this group had conditions that did not primarily affect the abdominal cavity (lameness [n = 31], mastitis [n = 22], respiratory tract diseases [n = 10]), injuries [n = 4], skin diseases [n = 3], dental problems [n = 2], and neoplasia [n = 1]). Most of these conditions include inflammatory reactions in their pathology. However, cows with a history of diseases that could potentially be associated with an inflammatory condition in the abdominal cavity (eg, abomasal displacement, abdominal surgery, endometritis) were excluded to avoid bias, even if no clinical signs of peritonitis were present. Results of the thorough clinical examination were used as reference diagnosis but the true prevalence was unknown. Additionally, tests to confirm the diagnosis (eg, postmortem pathological examination) were not possible in these cows that had been referred for treatment.

Respiratory rate, pulse rate, body temperature, and rumen motility were recorded twice daily. After physical examination and before any treatment, peritoneal fluid and venous blood samples were taken. Blood samples were obtained from the jugular vein for cytologic and biochemical analyses. As described,5 abdominocentesis was performed at the right ventro-caudal abdominal quadrant with a disposable needle (1.1 mm diameter, 50 mm length). The peritoneal fluid was collected in serum tubes (biochemical analyses), ethylenediamine tetra-acetic acid tubes (cytology), sodium citrate tubes (D-Dimer assay), and sodium fluoride tubes (l-lactate assay).

Peritoneal fluid was macroscopically assessed (color, turbidity); volume and density were measured with a small measuring cylinder and a hand refractometer. The numbers of white blood cells (WBC) and red blood cells (RBC) were counted with an automated analyzer.a All laboratory measurements were obtained by an experienced technician who was not aware of the clinical condition.

Routine differentiation of leukocytes (neutrophils, lymphocytes, eosinophils, basophils, and monocytes) was performed (QuickDiff staining) using venous blood and peritoneal fluid. Peritoneal fluid also was examined by oil immersion microscopy and Gram staining to identify the presence of bacteria and toxic neutrophils. The cytologic examinations were performed by one author (A.G.) to maintain consistency.

Biochemical parameters of peritoneal fluid and jugular venous blood (total protein, albumin, glucose, cholesterol, fibrinogen, CPK, ALP, and LDH) were measured with an automated analyzer.b Plasma and peritoneal fluid L-lactate concentrations were determined using a spectrophotometric method based on the reaction of L-lactate and NAD+ to pyruvate and NADH catalyzed by LDH.c The concentration of the fibrin degradation product, D-Dimer, was measured by a commercial available immunoassay.d The coefficient of variation of the D-Dimer assay in the present study was found to be 8.5% based on 10 repeated measurements of 1 sample. All biochemical analyses except D-Dimer and l-lactate assays were performed immediately after sampling. The samples for D-Dimer and l-lactate assays were stored (−21°C) for up to 2 months.

The peritoneal fluid-plasma/serum ratios of total protein, albumin, fibrinogen, D-Dimer, glucose, l-lactate, creatine kinase, ALP, and LDH were calculated.7 The SAAG was calculated by subtracting the peritoneal fluid albumin concentration from the concentration in serum.

Normality was tested by the Kolmogorov Smirnov test. The majority of data were not normally distributed and therefore given as median, 1st quartile, and 3rd quartile. The clinical examination is the diagnostic tool that was used as reference test. Although it is reasonable to believe that a thorough clinical examination is an appropriate diagnostic tool to diagnose peritonitis it must be considered as imperfect marker and cannot be used as a “gold standard.” Therefore, the classic calculation of sensitivity (SN) and specificity (SP) would overestimate the true diagnostic value.15 As a consequence, SN and SP were calculated applying a method that has been described for comparison to imperfect tests and adjusts the values for SN and SP accordingly.16,17 The calculations use the recently proposed reference ranges for dairy cows5 and the transudate-exudate classification.3 The 95% confidence intervals (CIs) for SN and SP were calculated by the Likelihood ratio test. As described, ROC analyses were performed and area under the curve calculated for assessment of diagnostic value of certain parameters.18,19 The Mann-Whitney U-test was used to compare the groups and a P value <.05 was considered to indicate statistical significant differences. A statistical programe was used for statistical analyses.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References

Cows of both groups were similar in age, lactation status, and body condition score. All data are presented as median, 1st quartile, and 3rd quartile. Heart rate (Group P: 74, 64, 88 beats per minute; Group N: 68, 62, 82 beats per minute [median, 1st quartile, 3rd quartile]), respiratory rate (Group P: 22, 20, 28 breaths per minute; Group N: 20, 18, 26 breaths per minute), body temperature (Group P: 38.9, 38.2, 39.6°C; Group N: 38.6, 38.1, 38.9°C), and rumen contraction (Group P: 1.0, 0.3, 2.6 per 3 minutes; Group N: 1.3, 0.6, 4.6 per 3 minutes) also did not differ between the groups. Numbers of RBC and WBC and values of biochemical analyses of blood (serum or plasma) varied considerably, but did not differ between the 2 groups. The median numbers of blood erythrocytes and leukocytes were within reference ranges.

The peritoneal fluid volume ranged from 4.3 to 213 mL in Group P (12.4, 4.9, 21.9 mL) and from 1.2 to 7.2 mL in Group N (3.7, 2.5, 4.8 mL), and differed significantly between the groups (P < .01). The biochemical measurements of peritoneal fluid analysis, the peritoneal fluid/blood ratios, SAAG, their SN and SP, and 95% CIs are shown in Tables 1 and 2. Using the cut-off values of the classic transudate-exudate system, the SN in predicting peritonitis in cows based on the peritoneal fluid protein concentration and peritoneal fluid leukocyte number was 64.5% (95% CI: 49.7–77.2%) and the SP was 51.0% (95% CI: 42.3–62.0%). Results of ROC curve analyses for the parameters that were statistically different between Groups N and P are shown in Table 3 and Figures 1–4.

Table 1.   Parameters of peritoneal fluid analysis (median, 1st, 3rd quartiles) in cows with peritonitis (Group P) and cows without peritonitis (Group N), and their reference ranges (Wittek et al5).
Parameter in Peritoneal FluidGroup N (n = 63)Group P (n = 47)Reference RangeSN % (95% CI)SP % (95% CI)
  • ALP, alkaline phosphatase; LDH, lactate dehydrogenase; CPK, creatine phosphokinase.

  • *

    Significant differences (P < .05) between groups; sensitivity (SN) and specificity (SP) and 95% confidence intervals (95% CI).

D-Dimer (mg/L)0.30 0.20, 0.40*1.15 0.90, 1.60*0.00–0.6095.7 (88.6–99.2)98.6 (90.4–99.7)
Total protein (g/dL)2.91 2.21, 3.09*4.09 3.41, 4.55*0.56–4.1849.0 (35.8–62.6)95.5 (87.7–99.0)
Albumin (g/dL)1.32 1.11, 1.48*1.60 1.33, 2.12*0.27–2.3918.2 (13.4–29.9)91.8 (90.4–99.7)
Fibrinogen (mg/dL)66 32, 9990 48, 1505–24117.4 (12.6–38.3)86.6 (84.9–93.4)
Cholesterol (mg/dL)32.2 24.6, 41.028.2 25.6, 36.712.8–77.311.1 (4.65–22.6)92.8 (82.7–97.1)
l-lactate (mg/dL)5.03 3.60, 8.3412.6 7.11, 42.91.71–11.7955.6 (42.2–70.9)88.1 (84.5–98.7)
Glucose (mg/dL)77.1 69.4, 100.771.8 1.80, 93.641.9–117.847.1 (33.3–60.8)90.2 (80.4–96.5)
CPK (U/L)41.6 23.4, 58.9*100 63.0, 196*12.0–167.430.9 (19.1–46.0)94.8 (85.5–98.8)
ALP (U/L)19.1 9.6, 25.232.0 17.0, 1945.0–61.132.7 (24.2–53.0)95.0 (86.0–98.9)
LDH (U/L)405 349, 568*888 613, 1,157*233–96061.5 (38.1–65.9)88.4 (82.5–94.1)
Leukocytes (103/μL)2.84 1.58, 3.90*7.44 3.93, 9.38*0.67–4.953.1 (39.0–66.3)91.3 (82.7–97.1)
Table 2.   Peritoneal fluid-blood ratios of biochemical parameters and SAAG (median, 1st, 3rd quartiles) in comparison with their reference ranges (Wittek et al5).
Peritoneal Fluid-Blood RatioGroup N (n = 63)Group P (n = 47)Reference RangeSN %SP %
  • ALP, alkaline phosphatase; LDH, lactate dehydrogenase; CPK, creatine phosphokinase; SAAG, serum-ascites albumin gradient.

  • *

    Significant differences (P < .05) between groups; sensitivity (SN) and specificity (SP) and 95% confidence intervals.

D-Dimer0.63 0.55, 0.80*2.50 1.88, 3.33*0.00–1.0996.2 (91.1–100)94.1 (85.1–98.8)
Total protein0.33 0.30, 0.39*0.59 0.51, 0.71*0.06–0.5154.5 (47.5–65.8)91.9 (85.1–98.1)
Albumin0.39 0.30, 0.44*0.60 0.49, 0.72*0.05–0.6445.0 (33.3–60.8)93.4 (87.7–99.0)
SAAG (g/L)20.3 16.4, 23.6*12.0 6.1, 17.8*12.3–30.267.1 (49.5–76.0)94.7 (87.7–99.0)
Fibrinogen0.18 0.04, 0.290.23 0.09, 0.330.00–0.944.3 (1.43–16.9)96.4 (88.4–96.6)
Cholesterol0.20 0.16, 0.32*0.49 0.40, 0.69*0.12–0.7714.2 (8.9–30.1)95.6 (87.5–99.0)
l-lactate0.61 0.49, 0.701.01 0.41, 1.730.26–0.9140.9 (31.6–57.4)90.2 (79.4–93.6)
Glucose1.01 0.96, 1.170.76 0.02, 0.980.68–1.3844.5 (33.3–60.8)94.8 (86,3–98.9)
CPK0.24 0.11, 0.33*0.39 0.32, 0.76*0.01–1.3312,7 (6.00–25.2)91.7 (87.7–99.0)
ALP0.28 0.10, 0.300.46 0.22, 1.650.01–0.8030.3 (20.5–47.5)94.2 (87.0–99.0)
LDH0.29 0.23, 0.40*0.68 0.43, 0.91*0.02–0.7956.4 (34.6–66.7)90.0 (85.2–97.1)
Table 3.   AUC and SEM calculated from ROC curves of parameters that were different between Groups P and N in peritoneal fluid analysis, different indices indicate significant differences between AUC (P < .05).
Parameter in Peritoneal FluidAUCSEMPeritoneal Fluid-Blood RatioAUCSEM
  1. LDH, lactate dehydrogenase; CPK, creatine phosphokinase; SAAG, serum-ascites albumin gradient; AUC, area under the curve; SEM, standard error of mean; ROC, receiver operating characteristic.

Total protein0.82c0.05Total protein ratio0.82c0.04
Albumin0.67d,e0.09Albumin ratio0.72d0.09
LDH0.84c0.04LDH ratio0.87b0.03
CPK0.82c0.06CPK ratio0.83c0.04
SAAG0.84c0.04Cholesterol ratio0.63e0.11

Figure 1.  Receiver operating characteristic curve of peritoneal fluid D-Dimer concentrations of 110 cows with (n = 47) and without peritonitis (n = 63).

Download figure to PowerPoint


Figure 2.  Receiver operating characteristic curve of peritoneal fluid leukocyte number of 110 cows with (n = 47) and without peritonitis (n = 63).

Download figure to PowerPoint


Figure 3.  Receiver operating characteristic curve of peritoneal fluid total protein and albumin concentrations and serum-ascites albumin gradient (SAAG) of 110 cows with (n = 47) and without peritonitis (n = 63).

Download figure to PowerPoint


Figure 4.  Receiver operating characteristic curve of peritoneal fluid lactate dehydrogenase (LDH) and creatine kinase (CK) activities and peritoneal fluid/serum LDH ratio of 110 cows with (n = 47) and without peritonitis (n = 63).

Download figure to PowerPoint

In peritoneal fluid, a low number of erythrocytes was found in both groups (Group P: 0.04, 0.02, 0.08 × 108/μL; Group N: 0.03, 0.01, 0.13 × 108/μL). The peritoneal fluid leukocyte number was different (P < .01) between Group N (2,500, 1,450, 3,700/μL) and Group P (6,400, 3,850, 17,750/μL), but varied in wide ranges especially in Group P.

Bacteria and toxic neutrophils were not found in any cow of Group N but were found in 16 cows of Group P. According to these findings, Group P was subdivided into Group PS comprising 16 cows with septic peritonitis, and Group PN comprising 31 cows with nonseptic peritonitis. Groups PN and PS differed significantly in rumen contraction rate (Group PS: 0.0, 0.0, 1.0 per 3 minutes; Group PN: 1.0, 1.0, 4.0 per 3 minutes), in serum total protein concentration (Group PS: 6.01, 4.74, 6.70 g/dL; Group PN: 7.39, 6.70, 8.18 g/dL), and in peritoneal fluid glucose concentration (Group PS: 0.90, 0.18, 1.80 mg/dL [0.05, 0.01, 0.10 mmol/L]; Group PN: 93.6, 72.2, 112.1 mg/dL [5.20, 4.01, 6.23 mmol/L]), which also resulted in a significantly lower glucose ratio in Group PS (0.006, 0.002, 0.018) compared with Group PN (0.96, 0.81, 1.12). In contrast to rumen contraction rate and serum protein (SN and SP below 60%), glucose concentration and glucose ratio had SN and SP of 100%. Additionally, peritoneal fluid LDH activity was significantly increased in cows with septic peritonitis compared with nonseptic peritonitis (Group PS: 1,216, 1,024, 2,008 U/L; Group PN: 755, 570, 965 U/L); however, the SN and SP of LDH did not exceed 70%. In contrast, activities of ALP and CPK did not differ between these 2 conditions.

The results of differentiation of peritoneal fluid cells are presented in Table 4. The main cell types were polymorphnuclear neutrophils (PMN) and lymphocytes, whereas monocytes, eosinophils, and basophils were found only in very low numbers (<2%). Epithelial cells were present in percentages between 0 and 10% without differences between the groups. Peritoneal fluid PMN were different between Groups N and P (P= .04) and also between Groups PN and PS (P= .03).

Table 4.   Percentages of polymorphnuclear neutrophils (PMN) and lymphocytes in peritoneal fluid of Group N, Group P, Group PN, and Group PS (subdivisions of Group P); data given as median, 1st, 3rd quartiles; comparison between Groups N and P, and between Groups PN and PS.
GroupPMN (%)PLymphocytes (%)P
Group N (n = 55)58.8, 51.4, 74.0.0422.7, 10.6, 30.5.07
Group P (n = 47)74.0, 65.0, 84.014.2, 8.0, 22.0
Group PN (n = 31)69.0, 56.5, 81.5.0315.0, 7.0, 22.2.18
Group PS (n = 16)84.0, 74.5, 91.511.5, 5.5, 20.0


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References

The aim of the study was to describe the potential for specific peritoneal fluid measurements to diagnose peritonitis in dairy cows. A major finding was that several parameters were potentially suitable for diagnostic purposes, because significant differences were found between cows with peritonitis and without peritonitis. The present study used the results of clinical examination as reference diagnosis for peritonitis; the true prevalence of peritonitis, however, is not known. This is a typical limitation of clinical studies because the reference tests are frequently less than perfect15,17 and may result in overestimation of diagnostic value. However, the model16,17 used for calculation of adjusted values for SN and SP should have minimized these effects and allowed more accurate estimations. Another limitation of the present clinical study is that the animals referred to the veterinary hospital might not completely reflect the population in field practice. The majority of cows in both groups had received anti-inflammatory and antimicrobial treatment before sampling which may have had some effect on peritoneal fluid characteristics. However, we believe this limitation would not have severely confounded the objectives and results of the present study. Additionally, this reflects the current situation in bovine practice because most cows have already been treated when presented to the veterinarian.

Applying the recently proposed reference ranges,5 most measurements had a high SP but only a low to moderate SN. Based on the reference ranges for peritoneal fluid protein concentration and leukocyte number, these 2 measurements have only moderate diagnostic value to predict peritonitis in dairy cows. Because cows with peritonitis can have normal peritoneal fluid leukocyte numbers and protein concentrations, normal values for these measurements cannot be used to exclude peritonitis in cows. The previously described low to moderate diagnostic accuracy of the standard transudate-exudate classification20,21 was confirmed by the results of the present study.

Increased peritoneal fluid albumin concentration results in a decrease of SAAG, a parameter with high-diagnostic importance in human patients,20 and indicates inflammatory conditions in the peritoneal cavity. However, in the present study, although no statistical comparison could be performed, SAAG did not seem to be superior in detecting peritonitis in cows compared with total protein concentration alone.

In contrast to the situation in human beings, cholesterol concentration and its ratio were not useful in diagnosing peritonitis in cattle. Possibly, these measurements allow more precise distinction between ascitic transudates from hepatic cirrhosis and exudates caused by malignant neoplasia or abdominal tuberculosis,7 2 clinical conditions that rarely occur in cattle. In addition, serum cholesterol concentration in cows is mainly dependent on the animal's food intake,22 and thus its ratio could result in inconclusive values.

Peritoneal fluid fibrinogen and the peritoneal fluid-serum ratio of fibrinogen also were not useful to exclude or diagnose intra-abdominal inflammatory conditions in cows. In contrast, the fibrin degradation product D-Dimer showed excellent accuracy in positive and negative prediction of peritonitis. However, D-Dimer ratio did not exceed the diagnostic value of the absolute peritoneal fluid D-Dimer concentration alone. Furthermore, peripheral blood D-Dimer might even be misleading because of the possible influence by modified blood coagulation during DIC, inflammation, thrombosis, or pulmonary embolism.23 Blood and peritoneal fluid concentrations of fibrin degradation products and D-Dimer have been described as accurate indicators of intestinal ischemia and peritonitis in laboratory animals and in human patients.12,24,25 The favorable diagnostic value in cattle might be a result of high fibrin formation activity during inflammation in cattle compared with other species,26 and the resulting high D-Dimer concentrations.

l-Lactate concentrations did not differ between cows with and without peritonitis, which is likely related to the source of this substrate. l-Lactate is the product of anaerobic tissue metabolism during ischemic episodes. In several species, intestinal ischemic conditions result in an increase of l-lactate in peritoneal fluid and plasma.6,27,28 Probably none of the cows with peritonitis in the present study suffered from any intestinal ischemic processes. However, results of a previous study showed that during abomasal volvulus and partly during LDA the abomasal wall is ischemic which results in l-lactate synthesis.29 Such ischemic conditions should result in increased peritoneal fluid l-lactate concentration in cattle similar to what has been described in other species.

In horses and dogs, decreased peritoneal fluid glucose concentrations are caused by bacterial glucose consumption.9,30 Therefore, low glucose concentrations can indicate the presence of bacteria in the peritoneal cavity (ie, septic peritonitis). Although there were a limited number of cows with septic peritonitis in this study, the results corroborate the findings in horses and in dogs. The peritoneal fluid glucose concentration was an excellent marker for septic peritonitis in cows of the present study, but this should be confirmed in a larger number of cattle.

Peritoneal fluid enzyme activities (ALP, CPK, and LDH) varied considerably in this study. Their SN to diagnose peritonitis in cows was only moderate, but their SP was higher. The diagnostic value of LDH ratio was found to be superior to LDH, ALP, and CPK which supports the inclusion of LDH activity and LDH ratio in Light's criteria. Light et al31 proposed 3 criteria for differentiating transudate and exudate in the pleural fluid of human patient: (1) pleural fluid-serum protein ratio >0.5, (2) pleural fluid-serum LDH ratio >0.6, and (3) pleural fluid LDH activity >200 U/L. These criteria later were adapted and widely applied to peritoneal fluid in human patients and horses.13,32–34 The cut-off values for protein and LDH ratios are similar in cows. In contrast, the peritoneal fluid LDH activity in cows was higher than that of humans, and therefore a cut-off value of 960 U/L has been calculated for dairy cows.5

Low numbers of leukocytes in peritoneal fluid were typical for Group N, which corresponded with the classic transudate-exudate classification and with reports on cattle peritoneal fluid analysis.3,5,35,36 Although significantly higher in Group P, the leukocyte number had only moderate diagnostic value because some cows with peritonitis in the present study had low leukocyte numbers. High percentages of PMN (>80%) indicate acute and severe inflammation, whereas percentages between 50 and 70% are typical for chronic inflammation.3,37 In the present study, cows with peritonitis (especially septic peritonitis) had a higher percentage of PMN in peritoneal fluid (Table 4), but there was wide overlap with animals without peritonitis, and in comparison with values reported in the literature.5,37

Generally, the inclusion of additional parameters in routine peritoneal fluid analysis results in improved diagnosis. The ratios between measurements of a marker in peritoneal fluid and blood slightly increase the diagnostic benefit in only a few of these parameters. Light's criteria have been shown to be useful in peritoneal fluid to detect intra-abdominal inflammation in cattle, but peritoneal fluid D-Dimer concentrations seem to be superior to other parameters. Parameters such as D-Dimer and glucose may provide detailed information on intra-abdominal conditions, which can be very beneficial for adequate and early treatment.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References

aTechnicon H 1, Bayer Diagnostics, Leverkusen, Germany

bHitachi 912, Boehringer Mannheim, Mannheim, Germany

cSigma-Aldrich Deutschland GmbH, Taufkirchen, Germany

dAxis Shield AS, Oslo, Norway

eSPSS 12.0, SPSS Inc, Chicago, IL


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. References