A comparative study of the performance of the widal slide agglutination test and the typhidot immunoassay for the diagnosis of typhoid fever in the West Region of Cameroon.

doi:10.21203/rs.3.rs-54260/v1

Background: The diagnosis of Typhoid fever, based on the Widal slide agglutination test, remains a major hurdle in developing countries like Cameroon due to varied perceptions of the value of the Widal test in determining clinical decision making. We undertook a study to evaluate the diagnostic performance of the Widal test and the typhidot immunoassay in patients suspected of having typhoid fever in the Menoua division, West Region of Cameroon.

Methods: Blood and stool samples were collected from 558 consenting febrile patients on the basis of suspicion of typhoid fever. These patients attended three district health services of the Menoua division between April 2018 and September 2019. These patients had clinical symptoms suggestive of typhoid fever as determined by their consultant. Serum from whole blood was used for the Widal slide agglutination test and for the Typhidot rapid immunoassay test based on manufacturer’s guidelines. A composite reference of fever plus positive coproculture for Salmonella enteric serovars typhi and paratyphi was used as reference. The sensitivity, specificity, predictive values of the positive and negative tests were calculated as well as the Cohen’s Kappa for agreement between the two tests.

Results: Of 558 patients, 12.90% tested positive for the reference method, 57.17% tested positive for the Widal slide agglutination test while 15.59% were positive for typhidot-IgM. The overall sensitivity, specificity, predictive values of the positive and negative tests were 80.56%, 94.03%, 66.6% and 97.03% respectively for typhidot-IgM; 94.44%, 48.35%, 21.32% and 98.33% for Widal slide agglutination test. The Cohen’s kappa estimates were 0.1660 (0.121-0.211), 0.386 (0.312-0.460) for Widal test and typhidot immunoassay respectively, with agreements of 53.76% and 76.16% respectively.

Conclusion: The Widal test was found to have a lower predictive value for the diagnosis of typhoid fever in our setting. However, the Typhidot test, although better, was not ideal. Diagnosis of typhoid fever should therefore rely on adequate clinical suspicion and a positive Typhidot test to improve the clinical management of Typhoid fever in our setting.

Infectious Diseases

Typhoid fever

Salmonella

enteric fever

typhidot-IgM

Widal test

Dschang

Widal test.

Enteric fever caused by Salmonella enterica serovars typhi and paratyphi remains a major burden in developing countries among which is Cameroon. The burden of typhoid fever worldwide, show that it causes 16.6 million new infections and about 600,000 deaths each year [1]. However, the incidence of typhoid fever has dropped to about 10/100,000 population/year in developed countries due improved leaving standard, proper hygiene and sanitation, and better health systems but the incidence is still higher 100/100,000population/year in less developed countries [2]. A key challenge to the effective control of typhoid fever is related to poor diagnosis. Diagnosis of typhoid fever in clinical settings is complicated because of overlapping symptoms with other common infections such as malaria, dengue and viral enteritis [3- 5]. For proper diagnosis a test with a good diagnostic performance especially in children with febrile diseases [6,7] is of importance. In addition, the misused of antibiotics via auto medication makes diagnosis difficult on clinical basis [8]. The gold standard for diagnosis of enteric fever is blood culture but this test not only has a poor sensitivity in clinical settings but it is also time consuming and expensive for patients and clinics in remote settings where culture facilities may not be always available and the population is poor [9]. The main diagnostic in such settings is based on the Widal slide agglutination test which is difficult to interpret for several reasons; cross-reactivities, time lag between infection and production of antibodies and persistence of target antibodies long after treatment with very low correlation with active disease [10, 11]. The tube dilution technique enables a quantification of specific antibodies and a change in titre can indicate active disease but also not very accessible. Many studies suggest that the Typhidot test, a plausible alternative based on detection of antibody production against the outer membrane preparation common to Salmonella typhi and S. paratyphi, has better performance characteristics. However, variations in sensitivity and specificity in diagnosis of enteric fever among adults and children have also been noted [12-14]. The assay gives good results during early infections with sensitivity of 68-95% and a specificity of 75-95% [15, 16]. An increase of negative productive value is important in endemic areas [17]. According to recommendations from World Health Organization for typhoid rapid antibody testing [18] and some studies that have evaluated the typhidot ability to detect antibodies have shown variations in sensitivity and specificity [15] in different settings. Although comparative studies have relied on the use of imperfect gold standard tests, a composite reference has been suggested to improve diagnostic value, but no agreement has been reached on which combinations can form a good composite reference standard Storey et al. (2015) [19]. In this study, we undertook to evaluate the diagnostic performance of the widal and typhidot tests against a composite reference made of a combination of fever within three days of attending the hospital and a positive stool culture test for Salmonella typh and/or paratyphi. The choice of this combination was based on the feasibility, independence of the tests, conditional to the disease, and although less specific, measures were taken to exclude other common febrile conditions such as malaria and respiratory tract infections. This was done in a bid to find a local strategy to better diagnose typhoid fever.

Study area and period

The Menoua Division, West Region of Cameroon has a surface area of 1380Km², it is divided into six subdivisions: Dschang, Santchou, Nkong-Ni, Penka Michel, Fokoue and Fongo-Tongo. According to altitude the division has Santchou at an altitude of 600m, Dschang at 1500m, its top at the plateau of Djuititsa at an altitude of 2200m. The Division has an average rainfall of 1717.7mm and temperature ranges from 13.6^oC to 25.35⁰C. 80% of inhabitants practice farming and the most important agricultural and horticultural plants include cabbage, carrot, onion, maize, banana, tomato, plantain, tea and beans. This study was conducted from April 2018 to September 2019. Menoua is a division of West Region of Cameroon. As of 2005 had a total population of 285, 764. The capital of this division is Dschang.

Study design and population

This was a prospective, hospital based, cross- sectional study among febrile patients suspected of having typhoid fever. Eligible patients were those who were suspected of typhoid fever, had fever (temperature >37.5 degrees at inclusion) or reported febrile episodes in the past three days. Eligible patients who provided consent were recruited in this study. A questionnaire containing demographic and clinical data, and known risk factors was administered to each participant.

Test methods

Blood sample collection

With the help of a sterile syringe and needle about 2-3ml of blood was collected from each participant consecutively as they arrive from the clinical room. This was done by well-trained laboratory technicians independent of the study team.

Test procedure for the Widal slide agglutination test

The qualitative slide agglutination test was performed using febrile patient antigen kits for Salmonella typhiand paratyphi (TYDAL Widal Antigens for Slide and Tube tests, Spain). This test was used to determine the presence of anti TO and TH antibodies in patient’s sera. The collected blood was centrifuged using a bacteriological laboratory centrifuge (Andres Hettich GmbH & Co. KG, 78532 Tuttigen, Germany), at 5000 turns for 5minutes at 4°C. For the determination of Salmonella antibodies by slide agglutination, a drop of Salmonella typhi O and H antigens were added on a drop of serum on the slide card and rotated for 100 rpm for one minute and recorded as either reactive or non-reactive.

Test procedure for the Typhidot (Immunoassay) test (typhidot^®Malaysian biodiagnostic)

The typhoid IgG and IgM rapid test device detects IgG and IgM antibodies to S. typhi and S. paratyphi through visual color development. Recombinant O and H antigens anti -human IgG and anti-human IgM antibodies used to detect the specific antibodies in human whole blood, serum or plasma samples. The patient’s sera sample was added to the sample well on the test panel, specific IgG and /or IgM antibodies, if present bind to the recombinant O and H antigens conjugated to colored particles on the sample pad. As the specimen migrates along the strip by capillary action and interacts with reagents on the membrane, the complex will be captured by anti-human IgG and /or anti-human IgM immobilized at the detection zone.

Venous blood was collected from patients suspected to have Salmonella infections (about 2-3ml). The blood specimens were then centrifuged at 5000r.p.m for 5 min using a laboratory centrifuge (Andres Hettich GmbH & Co.KG, 78532 Tuttigen, Germany. Using provided pipette, one drop of serum was added to the test well on the test cassette followed by one drop of buffer added to the sample well. The setup was allowed for 15 minutes for analyses to migrate across the membrane and colour development in the results windows.

Results interpretation

After 15 minutes, the test outcome was read as follows. A positive IgG and IgM was considered if one red band appeared on the control window (C), and two other red bands in both IgG and IgM windows. The shade of colour may vary from pink to purple but indicates a positive result even with a faint line; based on manufacturer’s recommendations. IgM was considered to be positive when one red band appeared on the control window (C), and on the IgM window; while a positive IgG was considered when one red band appeared on the control region (C), and on the IgG window. A negative result was indicated by appearance of the red colour only on the control window. A test assay was invalidated if the control line did not appear.

Stool sample collection and coproculture

Approximately 3 to 4g of fresh stool was collected by febrile patients for the Widal test, after brief training on how to collect the sample using a sterile wide- mouthed and transparent container. The assessors were provided to each of the participants. After collection, the stool was mixed in 3ml normal saline, and 1ml of this inoculum was diluted in to 9ml of freshly prepared Selenite F broth (Oxoid CM0395B&LP0121A), onto pre-labelled tubes and incubated at 37°C for 24 hours aerobically in bacteriological incubator (Laboratory Incubator, DNP-9052-1, Midfield Equipment & Scientific England), for amplification of bacterial growth. A loop-full of inoculum from the 3ml normal saline was streaked on SSA agar (L: S-BIOTECH, USA) and Mackonkey agar (Microxpress^®INDIA) and incubated for 24 hours for growth of colonies. After 24 hours a loop full of bacterial culture from incubated tubes containing the selenite culture was re-streaked into the Salmonella-Shigella agar (SSA) plates. The plates were examined carefully for the presence of characteristic colonies of Proteus spp, E. coli, Salmonella spp, Citrobacter spp Shigella spp, Kebsiella spp Serratia spp. Suspected colonies were re-streaked on Salmonella-Shigella agar for further identification and confirmation using Api 20E gallery (BioMerieux, France). The technicians performing coproculture were blinded to the procedure and results of the Widal and Typhidot tests.

Test outcome classification

The Widal test outcome was classified as positive or genitive based on the slide reactivity of patient serum. If, after the procedure the patient’s serum was reactive, it was noted as a positive test outcome. If the serum was non-reactive, it was noted as a negative test outcome. The procedure and classification were strictly followed for the total number of participants in the study.

The Typhidot immunoassay test outcome was determined by IgG and IgM positivity based on manufacturer’s guideline. Based on this guideline, a positive IgM and/or IgM+IgG were considered definite diagnosis of typhoid fever. For the purpose of this study, A positive IgG ‘only’ was not considered positive for the reference test as it may reflect only past exposure and also because the test has no quantitative value.

The reference test was a coproculture positive for Salmonella typhi/paratyphi and a history of fever within three days including the hospital visit day. Other non-typhoidal Salmonella although identified did not constitute a positive test. The were therefore considered negative by reference method. A positive Salmonella typhi/paratyphi coproculture and fever as determined by axillary temperature > 37.5 degrees or reported fever within the past three days was considered the referent. We included this criterion to avoid as much as possible spectrum bias, given the spectrum of disease severity that may present at the health facility and the likelihood of patients exposed to antibiotics before visiting the hospital.

Statistical analysis

We calculated the sample size based on the nomogram published by Carley et al, in 2003. Based on the nomogram for sensitivity plot at alpha=0.05,we estimated the sample size by using the following criteria: an estimated prevalence of 15%, a confidence limit of 0.05 we obtained an estimated sample of 558 patients.

The primary outcome of this study was the test accuracy and the predictive values of the positive and the negative tests of the Widal slide agglutination test and the typhidot immunnoassay. Secondary outcomes were a comparison of the performance characteristics such as sensitivity, specificity, likelihood ratios of positive and negative tests between the Widal and the Typhidot test in our setting. The area under the receiver-operating curve was used to estimate the diagnostic accuracy of the Widal and Typhidot tests while the sensitivity, specificity, predictive values were calculated based on standard formulae as indicated below.

Sensitivity =

Specificity =

Positive predictive value =

Negative predictive value =

Where TP=true positive, TN=true negative, FP=false positive and FN=false negative.

Regression analysis was done to model risk factors of true positive typhoid fever by reference test. All data were recorded in a pre-designed Epi-questionnaire before transferring to Graphpad Prism 8.0.2(Graphpad software 2019, California) for further processing. A p value of (<0.05) was considered significant for all comparisons except otherwise.

Socio-Demographic Characteristics of Respondents

A total of 558 patients presenting signs and symptoms suggestive of typhoid fever at our study sites (Table 1a). The majority of the participants were female, accounting for 67.38% (376/558). The majority of participants were in the age range [11–20] years old (219/558, 39.27%), were students. According to occupation, respondent that were student was more represented (n= 266) with 47.67% follow by participants working in the private sectors (144) with 25.80%. The respondent less represented were civil servant (n= 27) with 4.83% after house wife (n= 121) with 21.68%.

The distribution according to level of education shows that, the highest percentage was accounted for participant having secondary level of education (54.48%), followed by those of higher educational level (31.18%), primary education level (13.97%). The less represented were those that have never gone to school (0.17%). Distribution according to marital status shows that participants that are married were more represented (56.27%), followed by those who were single (37.27%), whereas widows were less represented (6.45%) (Table 1a)

Table 1a: Socio-Demographic Characteristics of Respondents

Characteristics	Number (n=558)	Frequency (%)
Gender
Female	376	67.38
Male	182	32.61
Age groups
≤ 10	37	6.63
11 – 20	73	13.08
21 – 30	219	39.24
31 – 40	48	8.60
41 – 50	57	10.21
51 – 60	54	9.67
61 – 70	48	8.60
≥ 71	22	3.94
Residence
Dschang	237	42.47
Fongo tongo	20	3.58
Nkong-gni	76	13.62
Penka Michel	12	2.15
Santchou	41	7.34
Occupation
Student	266	47.67
Civil servant	27	4.83
Private sector	144	25.80
Housewife	121	21.68
Level of education
None	1	0.17
Primary	78	13.97
Secondary	304	54.48
Higher education	174	31.18
Marital status
Married	314	56.27
Single	208	37.27
Widow	36	6.45

(n = number of patients);

Distribution of the prevalence of typhoid fever in the study population

The distribution of patients according to gender shows a significant difference (P= 0.0148) between male and female population, with male population been more infected (18.1%) compare to the female population (10.4%). Marital status also shows a significant difference (P = 0.0025), the prevalence of the typhoid fever in unmarried population is higher (19.2%). Neither age group, occupation nor the level of education affected the prevalence of typhoid fever (Table1b).

Table 1b: Distribution of the prevalence of typhoid fever in the study population

	Total (N=558)	Salmonella Typhi positive (n=72)	P-value (Chi²)
Age groups (years)
<10	37	6 (16.2)	0.8010 (3.814)
[11-20]	73	7 (9.6)
[21 - 30]	219	28 (12.8)
[31 - 40]	48	5 (10.4)
[41 - 50]	57	6 (10.5)
[51 - 60]	54	8 (14.8)
[61 - 70]	48	7 (14.6)
>71	22	5 (22.7)
Gender
Female	376	39 (10.4)	0.0148
Male	182	33 (18.1)	0.0148
Residence
Dschang	237	45 (19.0)	<0.0001 (29,16)
Fongo tongo	20	1 (5.0)
Nkong-ngni	76	10 (13.1)
Penka michel	12	9 (75.0)
Santchou	41	7 (17.1)
Occupation
Student	266	33 (12.4)	0.5105 (2.310)
Civil servant	27	6 (22.2)
Private sector	144	17 (11.8)
Housewife	121	16 (13.2)
Level of education
Non educated	1	0	0.9459 (0.3724)
Primary	78	11 (14.1)
Secondary	304	40 (13.15)
Higher education	174	21 (12.1)
Marital status
Married	314	28 (8.9)	0.0025 (11.95)
Single	208	40 (19.2)
Widow/widower	36	4 (11.1)

(N =total number of participants); (n= number of cases)

Clinical signs and symptoms of study participants.

In this study clinical signs and symptoms were identified and the frequency of these factors were noted and recorded in table 2 below. Abdominal pain and type of stool were associated with the typhoid fever at different degree.

Table 2: Clinical signs and symptoms distribution

	Total (N=558)	Typhoid fever positive	P-value	RR (95% CI
Fever	558	72	0.9999	N/A
Headache	363	48	0.7927	1.074 (0.6854 - 1.699)
Nausea	107	12	0.633	0.843 (0.4691 - 1.475)
Abdominal pains	457	66	0.021	2.431 (1.130 - 5.401)
Fatique	296	40	0.4444	0.827 (0.5365 - 1.279)
Vomiting	63	8	0.9999	0.9821 (0.4911 - 1.869)
Type of stool : Hard	93	4	<0.0001
Mucoid	254	23		N/A
Watery	170	44	<0.0001
Bloody	41	1

N: total number of participants; RR: Relative Risk; N/A: Not Available

Performance of Widal and Typhi-Dot test for typhoid fever diagnosis .

The comparison of Widal and Typhidot test for typhoid fever diagnosis (table 3) shows significance differences in their ability to diagnose typhoid relative to the reference test. The overall sensitivity, specificity, predictive values of the positive and negative tests were 80.56%, 94.03%, 66.6% and 97.03% respectively for typhidot-IgM; 94.44%, 48.35%, 21.32% and 98.33% for Widal slide agglutination test. The Cohen’s kappa estimates were 0.386 (0.312-0.460) and 0.1660 (0.121-0.211), for typhidot immunoassay and Widal test respectively, with agreements of 76.16% and 53.76% respectively (table 3).

Table 3: Performance of Widal and Typhidot test for typhoid fever diagnosis as referred to Typhoid stool culture test.

	Stool culture		P-value*	Sensitivity	Specificity	PPV	NPV	Likehood ratio	Accuracy	Kappa (95% CI)	Agreement (%)
	Positive	Negative
Widal
Positive	67	253	<0.0001	0.9444 (0.8657 – 0.9782)	0.4835 (0.4394 – 0.5279)	0.2132 (0.1718 – 0.2614)	0.9833 (0.9578 – 0.9935)	1.829	0.543	0.166 (0.121-0.211)	53.76
Negative	5	233
Typhidot IgM
Positive	67	128	<0.0001	0.8056	0.9403	0.6667	0.9703	13.5	0.941	0.386	76.16
Negative	5	358		(0.6997 – 0.8805)	(0.9156 – 0.9581)	(0.5624 – 0.7568)	(0.9507 – 0.9822)			(0.312-0.460)

PPV: Positive predictive value; NPV: negative predictive value;

Distribution of the prevalence according to type of test

Significant difference (P= 0.0001) was observed comparing the three diagnostic methods with low prevalence observed with stool culture. Out of the 558 participants in this study, 319 were positive for typhoid fever based on the Widal test with an antibody titer of 1:80 for both “O” and “H” antigens being taken as cut off point values indicative of a recent typhoid infection giving a prevalence of 57.17%, on the other hand 42.83% (n = 239) were negative for typhoid fever when Widal is used as the diagnostic tool. Interestingly, the prevalence of typhoid fever was 15.59% (n = 87) based on the findings from typhidot IgM. A total of 84.41% were proven negative for typhoid fever (n = 471) based on the same findings from typhidot and still from the same sample under study. More interestingly, the prevalence of typhoid fever infection was 12.90% (n = 72) base on the finding from stool culture and from the same sample under study (Table 4).

Table 4: Prevalence of typhoid fever according to Widal, Typhidot and stool culture tests

Diagnostic Test	Positive	Prevalence (%)	P-value
Stool culture	72	12.90	<0.0001
Widal	319	57.17
Typhi-Dot IgM	87	15.59

Typhoid fever constitutes one of the major causes of morbidity and mortality in developing countries among which is Cameroon. The clinical diagnosis of this disease remains difficult because its symptoms are similar to that of malaria and many other diseases [3, 21]. Laboratory confirmation of infection has relied on cheap slide agglutination tests which has its drawbacks leading to the appearance of false positive test results, the magnitude of which depends on the endemicity of the disease. Other tests such as the Typhidot immunoassay and the Tubex Widal dilution test, but have not been studied to understand their value in the diagnosis of typhoid fever in our setting. In the current study, we set out to compare the performance of the Widal Slide agglutination test and the typhidot test in patients suspected of typhoid fever using coproculture and fever/history of fever within three days as reference. We obtained a prevalence of 12.90% of typhoid fever by stool culture.

Our study revealed disparities in the performance of the Widal slide agglutination test and typhidot tests. Compared to the reference, the Widal test had a sensitivity of 0.9444 (0.8657 – 0.9782) and a specificity of 0.4835 (0.4394 – 0.5279) whereas the typhidot assay had a sensitivity of 0.8056 (0.6997 – 0.8805) and the specificity of 0.9403 (0.9156 – 0.9581). Our results indicate that the Widal test performs poorly compared to the typhidot test, taking stool culture and fever within three days preceding hospital attendance as referent. Indeed, the result indicate that Widal correctly identifies about half of patients who are uninfected with the typhoid fever bacteria, while only about 6% of uninfected patients will be wrongly classified as having typhoid fever meanwhile, they are uninfected. Similarly, the Widal test identifies 94% of febrile patients suspected of typhoid fever as positive, whereas only 80% of these patients were positive for typhidot immunoassay. However, only one in five febrile patients suspected for typhoid with a positive Widal test result truly had the disease (PPV=21.3%) while two in every three febrile patients suspected for typhoid fever who have a positive typhidot test truly had the disease (PPV=66.7%). This low accuracy of Widal test compared to typhidot could result from exposure to cross-reacting antigens in febrile infections other than Salmonella typhi. Indeed, Salmonella typhi shares the O and H antigens with other non-typhoidal salmonella, Enterobacteriaceae as well as some parasites such as Plasmodium falciparum [22, 23], contributing to increase the probability of false positive results. Despite disparities in some parameters, we obtained similar estimates of the predictive values of the negative test for both the Widal and typhidot tests (98.3% and 97.0% respectively). This indicates that both tests have the ability to rule out the disease in the population of suspected febrile patients.

The Widal test parameters obtained in the present study show some general patterns with some previous studies. High sensitivity, low predictive value of the positive test and high predictive value of the negative test. For example, in the study by Muthaiyan [24] and collaborators, they report a sensitivity of 88.5% [95%CI (69.82-97.42)], specificity 81.56% [95% (74.16-87.58)], positive predictive value 46.94% [95%CI (32.54-61.720)], negative predictive value 97.46% [95%CI (92.74-99.44)] for the Widal slide agglutination test. In a similar study in India in 2016 (24), the authors reported a sensitivity of 80.77% [95%CI (60.64-97.42)], a specificity 78.01% [95% (70.22-84.54)], a positive predictive value 40.38% [95%CI (27.01.-54.90)], and negative predictive value 95.65% [95%CI (90.14-98.56)]. Sherwal et al., (25), also reported a sensitivity of 74%, a specificity of 83%, a positive predictive value 87.5%. Rahman et al. (2007) [26], who had sensitivity of 81.8%, specificity of 69.2% and a positive predictive value of 45.6%. This pattern shows that generally, the Widal slide agglutination test, although highly sensitive, cannot be used to rule out the disease among those with positive results. Furthermore, the high predictive value of the negative test result indicate that this test can rule out the disease in patients that have a negative test outcome.

The prevalence has been shown to vary in different regions of Cameroon. For example, it was found to be 8.70% in the Bamenda (North West) in a similar study [27], 21% in Buea, South West region of Cameroon [28] and 2.5% obtained by Nsutebu et al. (2003) [29] in Tiko, Douala and Yaounde (Litorral and Centre regions respectively). Our estimate was also less than the prevalence obtained in a similar study in Nigeria [30]. Several reasons such as differences in sample size, risk factors and environmental conditions could account for these disparities in prevalence. Although disparate, it is not likely that the pattern of performance parameters of either typhoid test used in the present study may change from one region to another, although the magnitude of the performance parameters may be different.
In the present study five cases of confirmed typhoid fever by reference test had negative widal and typhidot-IgM tests. This observation, although few in number, may result from early sampling before antibodies become detectable in blood as was reported in one study [31]. Furthermore, Andualem et al. (2014) found two cases of this category in a study comparing the Widal slide agglutination test against blood culture as the referent [32]. Although not frequent, this observation may also account for false negative results obtained by the index test in comparative studies of typhoid immunodiagnostics. Delay in antibody production in early disease may be the main determinant of this observation. Similar observations have been made recently in serology tests used to detect exposure to the COVID-19 virus [33].

Our study has some limitations. Firstly, standard blood culture has been recognized as the reference test for evaluating diagnostic performance of typhoid fever tests. However, there are problems associated with the test. It is first of all an invasive procedure [34] and has been demonstrated to have low sensitivity even when there is no exposure to antibiotics [19]. In addition, obtaining adequate amounts of blood from groups like children is challenging. All these make the use of such a test challenging in our setting. We used a composite reference of positive stool culture and a fever greater or equal to 37.5 C or a history of fever in the past three days. This is not best reference either but represented the best combination under our local circumstances. The search for a perfect composite reference for diagnostic comparisons is definitely still a way ahead even with emerging technologies [34]. Secondly, we neither collected antibiotic exposure history nor clinical drug prescription data of these patients to quantify over-treatment and therefore the likely impact of treatment practices on our diagnostic accuracy estimations. This may likely have a small effect on our results which should be interpreted with this caution.

In the present study in which we compared the diagnostic performance of Widal slide agglutination test and typhidot-IgM test against a coproculture, the following diagnostic performance parameters were obtained: the sensitivity of typhidot immunoassay was 80.56% specificity 94.03%, positive predictive value was 66.67% and negative predictive value 97.03%. The sensitivity of the Widal slide agglutination test was 94.44%, specificity 48.35%, positive predictive value 21.32%, and negative predictive value 98.33%. Typhidot is a good alternative for Widal test in the diagnosis of typhoid fever as it is characterized by less false positive and negative. We note that the Widal test only rules in typhoid fever in one of five suspected patients that test positive, as opposed to two of three who test positive by the typhidot immunoassay. The Widal slide agglutination test in its present configurations should not be used to confirm typhoid fever diagnosis in Cameroon.

IgG: Immunoglobulin G, IgM: Immunoglobulin G, SSA: Salmonella-Shigella Agar

Ethics approval and consent to participate

Ethical clearance was obtained from the National Ethical Committee for Research involving Human Subjects (CNERSH). N^o 2019/11/56/CE/CNERSH/SP, Yaoundé Cameroon. Research authorizations were also obtained from all heads of districts in the Menoua division and from directors of the different hospitals where sampling and data collection was done. Research authorization was also granted by the University of Dschang. Data and sample were collected only after informed consent was obtained from each volunteer and guardian. Consent to participate in this study was written in a consent form for all participants.

Consent for publication

All authors consented for publication of this manuscript.

Availability of data and materials

The datasets and/or analyzed data, and materials used during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests

Funding

The authors received no funding for the present research and writing of this manuscript.

Authors’contributors

CBT and AIM designed the study. KO, LSF and MSN performed sampling and laboratory analysis,

KO and TFT performed statistical analyses. All authors participated in the write-up and approved the final version of the manuscript

Authors’ Information

OK is a PhD student in the Department of Biochemistry of the University of Dschang, Cameroon.
LSF is a PhD in the Department of Biochemistry of the University of Dschang, Cameroon.
IMA is a Lecturer in the Department of Biochemistry of the University of Dschang and a Researcher in the Laboratory for Public Health Research Biotechnologies at the Biotechnology Centre, University of Yaounde 1, Cameroon.
TFT is a PhD in the Department of Biochemistry of the University of Dschang,
CBT is an Associate Professor of Immunology and Molecular Biology in the Department of Biochemistry of the University of Dschang in Cameroon and Head of Department of Biochemistry of the University of Bamenda, Cameroun.
MSN is a principal laboratory technician of Microbiology Unit of the Dschang District Hospital

Acknowledgments

We wish to thank all participants who sacrificed their time and donated blood and stool sample for this study. We also thank particularly Doctors and laboratory technicians of the different hospitals for their contributions.

Pang T., Levine MM., Ivanoff B., Wain J. and Finlay BB. Typhoid fever - important issues still remain. Trends in Microbiology 1998; 6: 131-133.
Crump JA. and Mintz ED. Global trends in typhoid and paratyphoid fever. Clin Infect Dis. 2010; 50(2):241-6.
Khan NM., Shafee M., Hussain K., et al. Typhoid fever in paediatric patients in Quetta, Balochistan, Pakistan. Pak J Med Sci 2013; 29(4):929-32.
Malik AS. and Malik RH. Typhoid fever in Malaysian children. Med J Malaysia 2001; 56(4):478-90.
Siddiqui, F. J., Rabbani, F., Hasan, R., et al. Typhoid fever in children: some epidemiological considerations from Karachi, Pakistan. Int J Infect Dis 2006; 10 (3):215-22.
Beig FK., Ahmad F., Ekram M., et al. Typhidot M and Diazo test vis-à-vis blood culture and Widal test in the early diagnosis of typhoid fever in children in a resource poor setting. Braz J Infect Dis 2010:14(6):589-93.
Choo KE., Davis TM., Ismail A., et al. Rapid and reliable serological diagnosis of enteric fever: comparative sensitivity and specificity of Typhidot and Typhidot-M tests in febrile Malaysian children. Acta Trop 1999; 72(2):175-83.
Mehul, G. A. Study of clinical profile of multidrug resistant Typhoid fever in children. Indian Pediatr. 2010; 45(2):156-9.
Basnyat B., Maskey AP., Zimmerman MD. and Murdoch DR. Enteric (typhoid) fever in travelers. Clin Infect Dis. 2005;41(10):1467-72.
Crump J. A., Luby SP. et Mintz, ED. The global burden of typhoid fever. Bull World Health Organ. 2004; 82(5):346-53.
Parry CM., Hien TT., Dougan G., White NJ. and Farrar JJ. Typhoid fever. N Engl J Med. 2002; 347(22):1770-82.
Jyoti KD., Naresh K., et Chandrakishore. A comparative study of Widal test and Typhidot (IgM and IgG specific assay) test in the diagnosis of enteric fever. IAIM, 2017; 4(7): 126-130.
Olsen SJ., Pruckler J., Bibb W., et al. Evaluation of rapid diagnostic tests for typhoid fever. J Clin Microbiol 2004; 42(5):1885-9.
Pegues DA. and Miller SI. In: Longo DL., Fauci AS., KasperDL., Hauser SL., Jameson JL. and Loscalzo J., (editors). (2012). Harrison’s Principles of Internal Medicine. 18th ed. New York: McGraw Hill; 2012. p. 1274-81.
Bhutta ZA. Current concepts in the diagnosis and treatment of typhoid fever. BMJ. 2006; 333(7558):78-82.
Jesudason M., Esther E. and Mathai E. Typhidot test to detect IgG and IgM antibodies in typhoid fever. Indian J Med Res 2002: 116; 70-72.
Udayakumar S., Pushpalatha K., Naveen SHM., Swathi PM., Yoganand R. and Sushma C. Comparative study of TyphidotM with Widal and blood culture in diagnosis of enteric fever. Indian J Child Health. 2017; 4(1):64-67.
World Health Organization (WHO), Geneva (2003). Background Document: The Diagnosis, Treatment and Prevention of Typhoid Fever. Retrieved from https/extranet.who.int/iris on 15/12/2019 at 5:32 CAT.
Storey HL, Huang Y, Crudder C, Golden A, de Los Santos T, Hawkins K. A meta-analysis of typhoid diagnostic accuracy studies: A Recommendation to adopt a Standardized Composite reference. Plos one; 2015. 10(11): e0142364. doi: 10.1371/journal.pone.e0142364.
Temgoua E., Ntangmo TH., Njine T., Serve M-A. Vegetable Produc-tion System of Swam Zone in Urban Environment in West of Cameroon: Case of Dschang City. Universal Journal of Environmental research and Technology 2012; 2:83-92
Malik AS. et Malik RH. Typhoid fever in Malaysian children. Med J Malaysia 2001; 56(4):478-90.
Pradhan P. Co-infection of typhoid and malaria. Riew¨, Journal of Medical Laboratory and Diagno-sis,2 no. 3. 2011 pp. 22-26.
Cheesbrough M., District Laboratory Practice in Tropical Countries (part1), Cambridge University Press, Cambridge, UK 2nd edition,2006.
Muthaiyan J., Moomin MRA., Varadarajan P., et al. A comparative study of Typhidot test and Widal test with the culture-positive typhoid fever in children. Evolution Med. Dent. Sci. 2016; 5(87):6476-6479, DOI: 10.14260/jemds/2016/1464
Sherwal BL., Dhamija RK., Randhawa VS., Jais M., Kaintura A. and Kumar MA. Comparative study of Typhidot and Widal test in patients of Typhoid fever. J Indian AcadClinMed2004; 5:244-46.
Rahman M., Siddique AK., Tam FC., Sharmin S., Rashid H., Iqbal A., et al. Rapid detection of early typhoid fever in endemic community children by the TUBEX O9-antibody test. Diagn Microbiol Infect Dis. 2007;58(3) :275-81.
Nkeza A., Kengne M., Nwobegahay JM., Ter GD. et Njukeng PA. The Susceptibility Pattern of Salmonella Species to Commonly Used Antibiotics in the Bamenda District Health Area, Cameroon. The Cameroon Association for Medical Laboratory Science. 2014. 1-8
Njunda AL., Assob NJC., Nsaghac SD., Ndec FP., Kamga FHL. and Njimbia CA. Antibiotic susceptibility profile for salmonella in the Buea health district, Science Journal Microbiology; 2012; 1(4) 97-100.
Nsutebu EF., Martins P. and Adiogo D. Prevalence of typhoid fever in febrile patients with symptoms clinically compatible with typhoid fever in Cameroon. Trop Med Int Health. 2003; 8(6):575–578.
Ramyil MS., Ihuoma OJ., Ogundeko TO., et al. Comparative study on the use of Widal test and stool culture in the laboratory diagnosis of Salmonella infection in adult and children in Jos Metropolis, Plateau State, Nigeria. IJSR; 2013; 2(12):435–441.
Jesudason M., Esther E. and Mathai E. Typhidot test to detect IgG and IgM antibodies in typhoid fever. Indian J Med Res 2002: 116; 70-72.
Andualem G., Abebe T., Kebede N., Gebre-Selassie S., Mihret A. and Alemayehu HA. Comparative study of Widal test with blood culture in the diagnosis of typhoid fever in febrile patients. BMC Res Notes; 2014; 7:653.
Shen L, Wang C, Zhao J, Tang X, Shen Y, Lu M, Ding Z, Huang C, Zhang J, Li S, Lan J, Wong G and Zhu Y. Delayed specific IgM antibody responses observed among COVID-19 patients with severe progression, Emerging Microbes & Infections; 2020; 9:1, 1096-1101, DOI: 1080/22221751.2020.1766382
Narayanappa D, Sripathi R, Kumar JK, Rajani HS. Comparative study of Dot Enzyme Immunoassay (Typhidot-M) and Widal test in the diagnosis of Typhoid fever. Indian Pediatr; 2009; 47:33 1-33.

STROBEchecklistcrosssectionalKarimoetal.docx

A comparative study of the performance of the widal slide agglutination test and the typhidot immunoassay for the diagnosis of typhoid fever in the West Region of Cameroon.

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Supplementary Files

Status:

Version 1