COMPARISON OF PERFORMANCE OF TWO WIDAL TEST TECHNIQUES WITH STOOL CULTURE IN THE DIAGNOSIS OF TYPHOID FEVER AT THE DSCHANG DISTRICT HOSPITAL, WEST CAMEROON

Objective: This study is aimed to compare the performance of two Widal test techniques with stool culture in the diagnosis of typhoid fever. 
Methods: A cross-sectional study was performed at the Dschang District Hospital in patients clinically suspected to have typhoid fever, whose stool was collected for stool culture. Furthermore, venous blood was collected and the serum was tested by both Widal slide agglutination test and Widal tube titration test. 
Results: The results showed that out of 750 participants include in the study, 325 (43.33%) were positive for Widal slide agglutination test, 174 (23.20%) for Widal tube titration test, and 159 (21.20%) for stool culture. The sensitivity, specificity, positive predictive value, and negative predictive value of Widal slide agglutination test with respect to stool culture were 97.48%, 71.23%, 47.69%, and 99.22%, respectively, but 100%, 97.46%, 91.37%, and 100% for the Widal tube titration test. With the stool culture, Widal slide agglutination test had a moderate agreement (kappa = 0.47), but Widal tube titration test had an absolute agreement (kappa = 0.94). 
Conclusion: Widal tube titration test should be used in place of Widal slide agglutination test in the diagnosis of typhoid fever in the case of limited access to stool culture test.


INTRODUCTION
Typhoid fever is an acute systemic infection caused by Salmonella enterica serotype Typhi. It is a worldwide public health problem and an important cause of morbidity and mortality with approximately 11-20 million people affected and 128,000-161,000 deaths each year [1]. This number is reduced by 1-4% in patients who received appropriate treatment. Laboratory results are very important to confirm the clinical diagnosis of this disease for effective treatment [2][3][4]. In the case of suspected typhoid fever, effective diagnosis requires the isolation and the identification of S. enterica serotype Typhi in many areas where the disease is endemic [5]. However, the situation is different in most developing countries as the Widal test (agglutination test) seems to be the most commonly used laboratory method for diagnosing the disease because it is relatively less expensive [6]. There are two types of agglutination technique, namely, slide agglutination test and the tube titration test. These serological tests are based on the detection of secreted circulating antibodies against Salmonella typhi O and H antigens. Notwithstanding the inconsistent results of Widal test in endemic areas, the two agglutination technics are still considered useful for typhoid fever screening [7]. The best diagnosis of typhoid fever is by the isolation of S. typhi from blood, bone marrow, or a specific anatomical lesion [7]. Advantageously, the causative agent of the typhoid fever can be cultured from the stool throughout the duration of the disease; although this is not as specific as blood culture. Nevertheless, feces taken from acute patients may be positive when the blood culture is negative. Widal test is simple to perform but has a low sensitivity and specificity that vary from one study to another [8][9][10][11]. However, Widal test continues to be accepted as an adjunct to the clinical diagnosis and isolation method. In Cameroon, there is very little information on typhoid fever and comparisons of diagnostics techniques in medical literature. Thus, this study aimed to compare the performance of Widal agglutinations tests commonly used at the Dschang District Hospital with stool culture in typhoid fever diagnosis.

Type and study area
This was a comparative cross-sectional study, conducted at the Dschang District Hospital, West Cameroon. A total of 750 participants were included in the study.

Inclusion criteria
Patients of all ages and both sexes whom typhoid fever diagnostic test was prescribed by the doctor following clinical diagnosis and who voluntarily agreed to participate were included in the study.

Exclusion criteria
Patients on antibiotics, patients who received typhoid vaccine, and patients who refused to give their consent were excluded from the study.

Natheu and Telefo
Collection of samples Approximately 20-30 g of stools were collected in wide-mouthed plastic container from each participant included in the study for stool culture. A 8-10 ml of venous blood were collected in the dry tubes from participants and centrifuged at 5000 rpm for 10 min to obtain sera that will be used for the Widal tests.

Widal slide agglutination test
The test was performed according to the manufacturer's instructions (Omega Diagnostic Ltd., 2015). Thus, one drop (50 μl) of each undiluted serum was placed on the circles of slides provided in the kit along with positive control serum followed by the addition of one drop (50 μl) of antigen solutions TO, AO, BO, CO, TH, AH, BH, and CH. The contents were mixed with separate applicator stick, and the slide was rocked gently for 1 min. If no agglutination was observed, the test was considered negative. If agglutination was visible within 1 min, the test was considered as positive.

Widal tube titration test
All the samples were subjected to Widal tube agglutination method to confirm the results of slide agglutination test. Following the manufacturer's instructions (MEDIFF, ONE FOR ALL, France), two groups of four rows containing six tubes in each row were set for each serum sample to be tested. Doubling dilutions of the test serum, i.e., 1:25, 1:50, 1:100, 1:200, 1:400, and 1:800 were prepared in all rows so that each tube contained 0.5 ml of the diluted serum. Then, in each dilution, 0.5 ml of the antigen solutions (TO, AO, BO, and CO in the first group of tubes; TH, AH, BH, and CH in the second group) were subsequently added. Hence, the final serum dilutions obtained for each antigen were 1:50, 1:100, 1:200, 1:400, 1:800, and 1:1600. Appropriate control tubes containing, for one, physiological saline and for the others, each antigen suspension were included to rule out auto agglutination of the reagent. All tubes were homogenized and incubated at 37°C for 24 h. By flicking on the bottom of each tube, the results were read as a granular appearance for the agglutinin O-positive reaction and as a fluffy appearance for the agglutinin H-positive reaction. The titer of the samples tested was for the last dilution showing a still clearly positive result. Titers between 50 and 100 were not taken into account.

Stool culture
Approximately 25 g of stools collected from each participant were diluted in 5 ml of NaCl solution to obtain the inoculum. A portion of each inoculum was used for microscopic examination (gram stain) to evaluate the percentage of Gram-positive and Gram-negative bacteria. Another part was seeded on Salmonella-Shigella (SS) medium and 1 ml was introduced into 9 ml of selenite broth for an enrichment culture. Both preparations were incubated at 37°C for 24 h. After 24 h, the incubated SS medium was observed. In case of the presence of bacterial growth (presence of colonies), subculture of the previously enriched selenite + stool mixture was made on a new SS medium and reincubated at 37°C for 24 h. After observation, colonies suspected to be Salmonella (black center colonies) were biochemically characterized using the Api 20E gallery to identify the type of Salmonella.

Analysis of collected data
Epi Info version 7.0 was used to analyze data collected. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated to compare Widal test with stool culture. Kappa test was used to assess the agreement between these diagnostic tests. The different calculation formulas were: Sensitivity (Se) = (true positives/true positives + false negatives) × 100, Specificity (SP) = (true negatives/true negatives + false positives) × 100, PPV = (true positives/ true positives + false positives) × 100, NPV = (true negatives/true negatives + false negatives) × 100, and Kappa = Pa-po/1-Pa. Compared to the stool culture, sensitivity, specificity, PPV, and NPV of the Widal slide agglutination test were 97.48%, 71.23%, 47.69%, and 99.22%, respectively ( Table 3). The agreement between these two tests was analyzed by the kappa test. However, moderate agreement was obtained between the Widal slide agglutination test and the stool culture (kappa = 0.49).

Of
We also compared the Widal tube titration test with the stool culture. Among 174 positive results of the Widal tube titration tests, 159 were positive for the stool culture. All negative results (576) of the Widal tube titration test were also negative with the stool culture ( Table 4). The overall sensitivity, specificity, PPV, and NPV of tube titration test were 100%, 97.46%, 91.37%, and 100%, respectively (Table 3). With respect to the concordance test, almost absolute agreement was obtained between the Widal tube titration test and the stool culture (kappa = 0.94).

DISCUSSION
In this study, 43.33%, 23.2%, and 21.2% of cases of typhoid fever were diagnosed by Widal slide agglutination test, Widal tube titration test, and stool culture, respectively. This indicates that there was a large variation among the tests. This very high positive result obtained in the case of Widal slide agglutination test compared to the other test shows the abundance of false positives for this test, which can be attributed to the endemicity of the disease in our area of study and probably the presence of cross-reactivity with antibodies from non-bacterial infections such as dengue fever, malaria, and hepatitis A [12,13]. Four cases out of the 425 negatives of the Widal slide agglutination test were positive with the stool culture. The result of the false-negative Widal slide agglutination test may be due to an early infection of the disease before the production of detectable antibodies or to a typhoid carrier [14].
Sensitivity of the Widal slide agglutination test and Widal tube titration test was 97.48% and 100%, respectively. This result shows the ability of the Widal test to detect truly positive results compared to the stool culture. Similar results were observed in a study conducted in India (91.2%) [15]. However, Widal sensitivity was lower in the study conducted by Olsen et al. (64%) [10] and Noorbakhsh et al. (2003) (75.86%) [11]. Specificity of the Widal test found in this study was   [8]. In contrast, other studies conducted in Nigeria on adults [16], Malaysia [9], Ethiopia [17], and Vietnamese hospitals showed that the Widal test has low specificity [10]. This may be due to geographical variations. In fact, the Widal test varies with geographical location, endemicity of typhoid fever, prevalence of non-typhoid Salmonella infection, and other infections that cross-react with antigen Salmonella [18].
Concerning the agreement between Widal slide agglutination test and stool culture or Widal tube titration test and stool culture, there was an absolute agreement (kappa = 0.94) between Widal tube titration test and stool culture, and a moderate agreement (kappa = 0.49) between Widal slide agglutination test and the stool culture. This indicates that the result obtained from the Widal slide agglutination test in the diagnosis of typhoid fever was less reliable as compared to the stool culture. Thus, Widal slide agglutination test cannot be used alone in the diagnosis of typhoid fever because a good number of patients will be falsely diagnosed positive and the doctor will submit them to antimicrobial treatment against typhoid fever without these patients suffering from the disease. This may consequently result in the development of resistance by these bacteria to antibiotics [20].

AUTHORS' CONTRIBUTIONS
CKN and BPT designed the study; CKN collected data. CKN and BPT write and approved the final version of the manuscript.  PPV has been shown to be the most important variable in a clinical diagnostic method, especially since it represents the proportion of patients with positive outcomes who are well diagnosed [19]. In this study, the PPV of Widal slide agglutination test was low (47.69%) but very high for Widal tube titration test (91.37%). This is the case of Gaikwad and Rajurkar works [20] and Jahan et al. (2016) [15] who also recorded a low PPV for Widal slide agglutination test, 28.91% and 53.8%, respectively. In addition, PNV value of Widal slide agglutination test was 99.22% and 100% for Widal tube titration test. This indicates that negative result in Widal test has a good predictive value for the absence of the disease, but positive result would have a low predictive value for the presence of typhoid fever, especially in the case of Widal slide agglutination test [16]. This clearly proves that although Widal slide agglutination test is very effective in the detection of negative samples, expressed positivity is not always useful in the correct diagnosis of the disease.