Diagnostic utility of a single Widal test in the diagnosis of typhoid fever at Aga Khan University Hospital (AKUH),Nairobi,Kenya

Abstract

Typhoid fever is a systemic illness caused by Salmonella enterica serovar typhi (S. typhi) and is endemic in Kenya, where a single Widal test done on an acute phase serum is the most commonly used test. Using a cut-off value of 1:320 for both O and H agglutinins as being diagnostic of S. typhi, only 18 (26%) of our patients had diagnostic titres; 37 (53.6%) had O and H titres less than 1:40. Our study showed that Widal testing done on an acute phase serum of a patient suspected to have typhoid fever had limited diagnostic capability given its low sensitivity.

Introduction

The signs and symptoms of typhoid fever are non-specific, making it extremely difficult to make accurate diagnoses based only on clinical presentations.¹ In Kenya, often the Widal agglutination test is the only available means of diagnosing typhoid fever in suspected cases.²

Diagnosis by Widal test classically relies on the demonstration of a fourfold increase in antibody titres against the flagella (H) and somatic (O) antigens of Salmonella enterica serovar typhi (S. typhi) in paired sera taken 10–14 days apart. It is therefore essential to make a rapid therapeutic decision based on the results of a single acute phase serum Widal test.

This retrospective study was carried out in order to examine the sensitivity of a single Widal test done on acute phase sera in patients confirmed as having typhoid fever at the Aga Khan University Hospital, Nairobi, Kenya.

Materials and methods

The Aga Khan University Hospital, Nairobi, is a 254-bed private referral hospital that serves Nairobi and its environs. All patients whose blood or bone marrow cultures were positive for S. typhi between January 2000 to May 2008 were included in the study.

Relevant clinical data was collected in addition to the results for Widal, stool cultures, blood slide for malaria and ELISA for HIV, where available.

Blood and bone marrow cultures were performed by the BACTEC 9120 and BACTEC 9050 automated culture systems (Becton Dickinson, Maryland, USA). Standard microbiological techniques were used for the identification of S. typhi.³

Widal tests were performed with standardized S. typhi O and H antigens (Remel Europe Ltd, Kent, UK).

An analysis of the statistics was performed using SPSS 15.0.

Results

S. typhi was isolated from the blood and bone marrow cultures of 64 (73.6%) male patients and 23 (26.4%) female patients from January 2000 to May 2008 64 (73.6%). The mean age at diagnosis was 25.6 years. Eighty-six patients' files were reviewed – 79 (91.9%) inpatients and seven (8.1%) outpatients. Of these, 54.4% were admitted for 7 days or less, 36.7% for 8–14 days and 8.9% for more than 14 days.

Of the 87 patients with positive blood cultures, only 69 had been given a Widal test. Using a cut-off value of 1:320 for both O and H agglutinins as being diagnostic of S. typhi, only 18 (26%) had diagnostic titres; 37 (53.6%) had O and H agglutinins of less than 1:40. The sensitivity of O and H agglutinins at various cut-off values is shown in Table 1.

Table 1

Sensitivity of a Widal test based on different cut-offs for O and H agglutinins

Titres	Sensitivity (%)
O and H ≥ 320	26.0
O ≥ 320	29.0
H ≥ 320	40.6
O and H ≥ 160	37.7
O ≥ 160	39.1
H ≥ 160	42.0

The duration of the symptoms influenced the Widal results (χ² = 4.257, P = 0.039) with more positive results seen in patients with symptoms of more than one weeks' duration. There was no significant difference in the Widal results between paediatric patients and those older than 12 years of age.

Fifty-one (59.3%) patients reported a history of prior antibiotic use. However, there was no association between the prior use of antibiotics and the Widal result (χ²= 0.013, P = 0.91).

Discussion

The interpretation of a Widal test is largely determined by the endemic titres of a region. Unfortunately, there have been few studies of endemic Widal titres in Kenya.⁴

In this study, the Widal test was found to have a sensitivity of 26% when both O and H agglutinins were considered at a cut-off value of 1:320. A previous Kenyan study reported a sensitivity of 81.3% using a similar cut-off value.⁵ Lowering the cut-off value to 1:160 for both agglutinins only raised the sensitivity to 37.7%. The disparities might be explained by:

The difference in the prevalence of typhoid fever in the regions;

Differences in diagnostic titres;

The timing of the Widal test during the course of illness;

Variability in the reagents;

And prior antibiotic use.

Thirty-seven (53.6%) patients had undetectable levels of O and H agglutinins, despite typhoid fever having been confirmed by a blood or bone marrow culture. Unspecified host or bacterial factors, as well as prior antibiotic use, have been given as possible explanations for this observation.⁶

No consensus has been reached as to which agglutinin between O and H is of greater diagnostic value or whether they should be considered separately or together.^6–9 In this study, H was more sensitive than O in diagnosing typhoid fever with sensitivities of 40.6% and 29%, respectively, at a cut-off value of 1:320.

Conclusion

In conclusion, our study showed that, given its low sensitivity, a Widal test made on an acute phase serum of a suspected typhoid patient has limited diagnostic capabilities. Clinicians should, therefore, utilize blood or bone marrow cultures where available. In areas where facilities for culture are unavailable, antigen-based rapid tests should be evaluated as they are more specific and sensitive than antibody-based tests.¹⁰

Footnotes

Acknowledgements

The excellent technical assistance of the team of microbiology division is gratefully acknowledged.

References

Hoa

, Diep

, Wain

, Community-acquired septicaemia in southern Viet Nam: the importance of multidrug-resistant Salmonella typhi. Trans R Soc Trop Med Hyg 1998;92:503–8

Kakai

Laboratory diagnostic services in rural health centres, western Kenya. East Afr Med J 2001;78 (7 Suppl):S34–5

Murray

, Baron

, Jorgenson

, Pfaller

, Yolken

. Manual of Clinical Microbiology, 8th edn. Washington: ASM Press, 2003

Jumba

, Mirza

, Mwaura

FB.

Salmonella typhi and Salmonella paratyphi antibodies in Kenya. East Afr Med J 1995;72:755–7

Kariuki

, Mwituria

, Munyalo

, Revathi

, Onsongo

Typhoid is over-reported in Embu and Nairobi, Kenya. Afr J Health Sci 2004;11:103–10

Schroeder

SA.

Interpretation of serologic tests for typhoid fever. Jama 1968;206:839–40

House

, Wain

, Ho

, Serology of typhoid fever in an area of endemicity and its relevance to diagnosis. J Clin Microbiol 2001;39:1002–7

Coovadia

, Singh

, Bhana

, Moodley

Comparison of passive haemagglutination test with Widal agglutination test for serological diagnosis of typhoid fever in an endemic area. J Clin Pathol 1986;39:680–3

Pang

, Puthucheary

SD.

Significance and value of the Widal test in the diagnosis of typhoid fever in an endemic area. J Clin Pathol 1983;36:471–5

10.

Pandya

, Pillai

, Deb

Rapid diagnosis of typhoid fever by detection of Barber protein and Vi antigen of Salmonella serotype typhi. J Med Microbiol 1995;43:185–8