Is Wright Test an Appropriate Screening Test for Diagnosis of Brucellosis?

Problem statement: Diagnosis of brucellosis is generally based on cult ure, polymerase chain reaction and serology. The first two methods are not accessible in all parts of world and are expensive. The routine method for diagnosis of bruc ellosis is considering Wright test as the first screening test; if the results are Wright positive’ Wright would be the next choice otherwise 2ME would be requested. This method of laboratory data collection is not appropriate and it is probable to have some cases of brucellosis missed and in clinic a practice we observed that some cases of brucellosis are Wright negative but Coombs’ Wright positive. Approach: In this study we calculated sensitivity, specificity, positive predictive value , negative predictive value and likelihood ratio of Wright and Coombs’ Wright in brucellosis suspected patients. Results: 122 patients suspected to brucellosis were studied. 53.3% were female. Sensit ivity and specificity Positive Predictive Values (PPV) and Negative Predictive Values (NPV) of Wrigh t were 32.5% (CI 95%: 22.8-42.3), 96.4% (CI 95%: 89.5-100), 96.6% (CI 95%: 0.9-100) and 93.1% (CI 95%:83.8-100) respectively. Sensitivity, specificity, positive predictive value and negative predictive v alue for Coombs’ Wright were 97.7% (CI 95%: 94.6100), 100% (CI 95%: 100-100), 100% (CI 95%: 100-100) and 93.1% (CI 95%:83.8-100) respectively. Conclusion: Coombs’ Wright is more sensitive than Wright for di agnosis of brucellosis. Instead of considering Wright, Coombs’ Wright and 2ME (mercapt oethanol) tests and interpretation of these three test we can just apply Coombs’ Wright and 2ME to reduce the expenditures and use a more sensitive test for diagnosis of brucellosis.


INTRODUCTION
Brucellosis is one of the most common zoonotic diseases in many region of the world, especially in Iran and its incidence is increasing (Aliskan, 2008;Karami and Movassagh, 2010). Infection is transmitted by dairy products like milk, cheese and contact with infected animal and aerosol (Hatami et al., 2010;Rajaii et al., 2006). Signs and symptoms of disease are extremely various. It can mimic many infectious diseases and involve any organ in human body (Gomez et al., 2008;Rajaii et al., 2006). Laboratory tests used for diagnosis of brucellosis include: blood culture, bone marrow culture, Polymerase Chain Reaction (PCR), ELISA, agglutination test and Rose Bengal (Abdi-Liae, et al., 2007;Aliskan, 2008;Heydari et al., 2008). The most useful and common test which is used is the standard tube agglutination test called Wright. Two other complementary tests are 2-Mercaptoethanol (2ME) and Coombs' Wright (Abdi-Liae et al., 2007;Aliskan, 2008;Hatami et al., 2010;Heydari et al., 2008). Now the first screening test is Wright and many physicians request this test as the first step in diagnosis of brucellosis. If result of Wright is negative, Coombs' Wright would be requested. Some of the patients who are infected with brucellosis have negative result of Wright test. This is a problem in diagnosis of brucellosis and it makes it necessary to have more laboratory tests and more cost for the patient. Now we believe that if a patient with signs and symptoms of brucellosis never received any treatment for the disease, through a single Coombs' Wright test his/her brucellosis could be diagnosed; because this test is as effective as Wright and in cases of chronic disease, presence of incomplete antibody or blocking antibody the test could be positive (Abdi-Liae et al., 2007;Afsharpaiman and Mamishi, 2008;Heydari et al., 2008;Karami and Movassagh, 2010).
Thus we decided to do this study for finding the best screening test and to decrease the cost and number of unnecessary laboratory test through applying a single test and to prevent confusion in interpretation of tests.

MATERIALS AND METHODS
It is an analytical cross-sectional study. In this study 122 suspected patients to brucellosis who referred to infectious disease clinic were included. All patients were examined completely and results including past medical history, physical examination and laboratory data (Count Blood Cell, Wright, 2ME, Coombs' Wright, liver enzyme test and ESR) were recorded in a questionnaire. Definite diagnosis of brucellosis was achieved by clinical and laboratory findings. Leucopenia, anemia, thrombocytopenia were defined as WBC< 5000 cell/ µL, hemoglobin < 13g dL  in men or < 12 g dL 1 in women and platelet <150000, respectively. Wright test was considered positive if its titer was equal or greater than 1/160 and Coombs' Wright was considered positive if its titer was equal or greater than 1/40 as recommended by Iranian National center of diseases control. The data was entered to SPSS version 11.5 software. Chisquare, Fisher exact and Mann-Whitney U tests were used for analyzing the data. Also, Sensitivity, specificity, positive predictive value, negative predictive value and Likelihood ratio for Wright and Coombs' Wright were calculated as shown below (Table 1).

DISCUSSION
In our study 65 persons (53.3%) were female. In other study incidence of brucellosis in females were reported as 64% (Hatami et al., 2010), 88% (Gomez et al., 2008), 34.02% (Abdi-Liae et al., 2007) and 34.1% (Afsharpaiman and Mamishi, 2008). High frequency of brucellosis among women in our study is the result of higher frequency of women's contact with cattle and dairy products (production of dairies like cheese, butter and so on is done by women). 57.4% of our patients were urban and in other studies this rating was 12% (Gomez et al., 2008) and 41% (Afsharpaiman and Mamishi, 2008). Using non-pasteurized dairy products is a popular behavior among urban people in this region and some patients have double living sites (urban and rural), because of these factors brucellosis is a common zoonotic disease. The mean age of patients were 40.2(±17). In another study the mean age of patients was 41 (Gomez et al., 2008) and range of patient's age was somehow similar to our study. Myalgia was the most common symptom and was seen in 85.5% of patients and low backache was the second most common symptom (84.7%). In the other study fever 52.9% (Abdi-Liae et al., 2007), arthritis and arthralgia 79.5% (Afsharpaiman and Mamishi, 2008), fever 83.8% (Hajia et al., 2009) were the most common symptom.
In our study Wright test was considered positive if its titer was equal or greater than 1/160. Coombs' Wright was considered positive if its titer was equal or greater than 1/40. In other study titer 1/80 and 1/160 (Karami and Movassagh 2010), 1/80 and 1/80 (Hatami et al., 2010), 1/80 and both titers of 1/80 (Abdi-Liae et al., 2007) and 1/40 (Hajia et al., 2009) for Wright and Coombs' Wright respectively were considered positive. Like the other study and based on recommendation of CDC of Iran we selected Wright of 1/160 and Coombs' Wright of 1/40 as positive.
In our study sensitivity, specificity, positive predictive value and negative predictive value were 32.5, 96.4, 966 and 93.1%, respectively. Taleski et al. (2002) study the sensitivity of Wright and Coombs' Wright were 84 and 86%. The specificity of Wright and Coombs' Wright were 100% for both methods. In the other study Wright were positive in all cases except one and Coombs' Wrights were positive in all cases (Rajaii et al., 2006). In two other studies titers of Coombs' Wright were more than Wright (Gomez et al., 2008;Heydari et al., 2008). In one study sensitivity of Wright and Coombs' Wright were 97.7 and 100% respectively (Afsharpaiman and Mamishi, 2008).
Culture is the gold standard of diagnosis of brucellosis (Lucero et al., 2007;Araja and Awar, 1997;Hajia et al., 2007;Taleski et al., 2002), but this method is difficult and time-consuming (Parizadeh et al., 2009;Rajaii et al., 2006;Taleski et al., 2002) and its sensitivity is reported as 10-30% (Kazemi et al., 2008;Parizadeh et al., 2009). Therefore we consider clinical, serology and responses to treatment as definite definition. Now it can be concluded that sensitivity of Coombs' Wright is more than Wright. Coombs' Wright can be used as the first screening test and it is the best screening test for diagnosis of brucellosis. We propose Coombs' Wright for screening and we believe that 2ME test can be done on Coombs' Wright and Wright test can be omitted from panel of diagnostic tests of brucellosis. This strategy can also decrease cost of diagnosis and can reduce confusion about interpretation of tests. Therefore, we suggest Coombs' Wright test for all patients suspected to brucellosis and we may use 2ME in order to discriminate acute, chronic and exposure to antigen and 2ME on Coombs' Wright specimen is enough to follow up the disease.

CONCLUSION
Coombs' Wright is more sensitive than Wright for diagnosis of brucellosis. Instead of considering Wright, Coombs' Wright and 2ME (mercaptoethanol) tests and interpretation of these three test we can just apply Coombs' Wright and 2ME to reduce the expenditures and use a more sensitive test for diagnosis of brucellosis.