Postgrad Med J 2000;76:80–84 © The Fellowship of Postgraduate Medicine, 2000

Classic methods revisited

Postgrad Med J: first published as 10.1136/pmj.76.892.80 on 1 February 2000. Downloaded from http://pmj.bmj.com/ on 17 December 2018 by guest. Protected by copyright.
Widal agglutination test − 100 years later:
still plagued by controversy
Lateef A Olopoenia, Aprileona L King

Summary Agglutination is a classic serologic reaction that results in clumping of a cell sus-
We review the significance of the pension by a specific antibody, directed against a specific antigen. Such tests have
Widal agglutination test in the been widely used for detection of antibodies against various disease-producing
diagnosis of typhoid fever. Over micro-organisms in serum for a long time. The Widal agglutination test, devel-
100 years since its introduction as oped by F Widal in 18961 to aid in the diagnosis of typhoid fever, utilises a sus-
a serologic means of detecting the pension of killed Salmonella typhi as antigen, to detect typhoid fever in serum
presence of typhoid fever, the from suspected S typhi-infected patients who present with febrile illness. The
Widal test continues to be plagued value and clinical application of the Widal test in developed countries has
with controversies involving the diminished considerably in recent years2 and a large number of antigenically
quality of the antigens used and related determinants of both typhoid and non-typhoid Salmonella organisms are
interpretation of the result, par- now recognised. We therefore decided to review the significance of this
ticularly in endemic areas. Areas sero-diagnostic test for typhoid fever in modern medicine, and to discuss new
of concern with clinical and labo- and innovative alternative diagnostic tests. Hopefully, this review will oVer both
ratory significance discussed in the novice and the experienced physician the opportunity to appreciate the limi-
this review include: the tech- tations of the Widal test.
niques of test performance,
interpretation of results, limita- Widal agglutination
tion of the value of the test results
in endemic typhoid areas, the Widal agglutination was introduced as a serologic technique to aid in diagnosis
quality of the antigens used, and of typhoid fever. The test was based on demonstrating the presence of aggluti-
alternative diagnostic tests. nin (antibody) in the serum of an infected patient, against the H (flagellar) and
O (somatic) antigens of Salmonella typhi. While the definitive diagnosis of
Keywords: Widal agglutination test; typhoid typhoid fever depends on the isolation of S typhi from blood, stools, urine or
fever
other body fluids,3–5 the role of the Widal test had been to increase the index of
suspicion for the presence of typhoid fever by demonstrating a positive aggluti-
nation during the acute and convalescent period of infection with evidence of a
four-fold rise of antibody titre.6–8 In developed countries, the use of Widal agglu-
tination as a laboratory tool to aid in the diagnosis of typhoid fever during the
acute phase of the illness, has largely been abandoned,2 as the need for such a test
is minimal, especially in view of the low prevalence of typhoid fever. In addition,
adequate and improved sanitation, sewage systems, proper hygiene and better
means of isolating the organism from culture are available. Unfortunately, in
some developing countries, the situation is quite diVerent, and the Widal test
appears to be the only laboratory means employed in the diagnosis of typhoid
fever among suspected patients. As the test suVers from serious cross-reactivity
with other infectious agents, it may produce false-positive results, leading to an
over-diagnosis of typhoid fever. Reynolds et al 9 concluded that diagnosis of
typhoid fever based on serology (Widal agglutination) alone is frequently inac-
curate. Concomitant with this increase in diagnosis is the abuse of the first-line
drug of choice (chloramphenicol), which has led to the selection of resistant
strains of S typhi.

Performance technique

Liberty Medical & Research Center,

The Widal test reaction involves the use of bacterial suspensions of S typhi and
Lagos, Nigeria S paratyphi ‘A’ and ‘B’, treated to retain only the ‘O’ and ‘H’ antigens. These
L A Olopoenia antigens are employed to detect corresponding antibodies in the serum of a
patient suspected of having typhoid fever. The IgM somatic O antibody appears
Division of Allied Health, Howard first and represents the initial serologic response in acute typhoid fever, while the
University, Washington, DC, USA IgG flagella H antibody usually develops more slowly but persists for longer.3 8 10
L A Olopoenia
A L King
Two types of agglutination techniques are available: the slide test and the tube
test. The slide test is rapid and is used as a screening procedure. Using commer-
Correspondence to Dr LA Olopoenia, 422 cially available antigens of S typhi, a drop of the suspended antigen is added to
Butternut Street NW, Washington, DC 20012, an equal amount of previously prepared serum. An initial positive screening test
USA
requires the determination of the strength of the antibody. This is done by add-
Submitted 19 January 1999 ing together equal amounts of antigen suspension and serially diluted serum
Accepted 18 August 1999 from the suspected patient. Agglutinations are visualised as clumps. Weakly
Widal agglutination test 81

reactive agglutinations may require an adequate light source for proper visuali-
sation, while strongly reactive agglutinations are easily seen. The result of the
tests are scored from 0 to 4+, ie, 0 (no agglutination), 1+ (25% agglutination),

Postgrad Med J: first published as 10.1136/pmj.76.892.80 on 1 February 2000. Downloaded from http://pmj.bmj.com/ on 17 December 2018 by guest. Protected by copyright.
2+ (50% agglutination), 3+ (75% agglutination) or 4+ (100% agglutination).
The smallest quantity of serum that exhibits a 2+ or 50% agglutination is con-
sidered the end-point of serum activity or titre.
The tube agglutination test requires much more technical work than the rapid
slide test, and is a macroscopic test.11 12 It also serves as a means of confirming
the results of the slide test. A mixture of suspended antigen and antibody is
incubated for up to 20 h at 37°C in a water bath. Agglutinations are visualised in
the form of pellets, clumped together at the bottom of the test tube. Results are
scored from 0 to 4+ positive agglutination as described above for the slide test.
The tube test is useful to clarify erratic or equivocal agglutination reactions
obtained by the more rapid slide test.
Since the ultimate goal of the test is antigen–antibody complex reaction,
cross-reactions are encountered when antibody produced by non-typhoidal
antigens reacts with typhoid-specific antigens. Several other diseases caused by
non-Salmonella organisms (malaria, dengue, miliary tuberculosis, endocarditis,
chronic liver disease, brucellosis, etc) have been shown to exhibit this
cross-reactivity in typhoid endemic regions, and these cross-reactions increase
the error rate of the result of the Widal test.
Lack of standardisation of antigens also compromises the technique, as shown
by Devillier et al.13 The value of Widal test depends upon the standardisation and
maintenance of the antigens to produce consistent results, and it has become
evident from work done in recent years on standardisation of the Widal test and
interpretation of the results that both the O and H antigens are necessary for
proper serologic analysis of the suspected serum. However, according to Welch
in 1936,11 no Widal test, regardless of the composition and standardisation of the
antigens used, is infallible, and thus it is unlikely that any will be developed that
will lower the validity of the isolation of the aetiologic agent. Unfortunately, more
than 60 years after Welch published his paper, the problems of ambiguity, insen-
sitivity and non-specificity of Widal antigens continue. The widespread use of
typhoid–paratyphoid vaccine, as well as the large number of cases of repeated
exposure to Salmonella species, tend to lower the specificity of the Widal test. We
consider that serologic studies are helpful in typhoid fever cases in endemic
regions only if patients have four-fold or greater increases in O or H agglutinin
titres in serum specimens obtained 2–3 weeks apart.

Interpretation of the test results

Causes of negative Widal
While performance of the test may require some detailed technical work, inter-
agglutination tests
preting the test result is the more arduous task. Salmonella are divided into dis-
x absence of infection by S typhi tinct serologic groups (A through E) on the basis of their somatic O antigens.
x the carrier state While all group D organisms, such as S typhi possess O antigen 9, about 60 of
x an inadequate inoculum of bacterial the 78 group D serotypes including S typhi also have O antigen 12.14 Thus,
antigen in the host to induce antibody infection by any of the group D serotypes can produce antibodies that can react
production with the O antigen used in the Widal reaction. Also, since all groups A and B
x technical diYculty or errors in the
performance of the test
organisms possess O antigen 12, cross-reactions with O antibody of group D
x previous antibiotic treatment serotype can occur with any of the group A and B serotype O antigens. Depend-
x variability in the preparation of ing on the relative quality and quantity of antigenicity of the O antigens 9 and 12
commercial antigens contained in other common non-typhoidal Salmonella serotypes, cross-reaction
may occur frequently enough to lessen considerably the diagnostic specificity of
Box 1 the Widal reaction. A comparative study of S typhi O antigen obtained from dif-
ferent manufacturers tested against the same serum, which had previously been
shown to be positive by the slide agglutination test, revealed marked variability
associated with the Widal agglutination titre.13 A negative agglutination test may
Causes of positive Widal be for one of several reasons, given in box 1. A negative Widal test result does not
agglutination tests therefore necessarily rule out the absence of infection. Such results are best kept
x the patient being tested has typhoid
as a reference for subsequent comparative analysis.
fever A positive agglutination tests (on two successive occasions) on the other hand,
x previous immunisation with Salmonella may also be open to several diVerent interpretations (box 2).
antigen. Although there are controversies surrounding the increase in titre beyond the
x cross-reaction with non-typhoidal first week of illness in some endemic areas, it is generally accepted by clinicians
Salmonella. that, toward the end of the first week of illness, titres of either O or H antibody
x variability and poorly standardised
commercial antigen preparation
may rise to as high as 1:160. However, the lack of paired sera may lead to an
x infection with malaria or other erroneous interpretation of test results.10 15
enterobacteriaceae In endemic typhoid regions, a single testing of a serum specimen for Widal
x other diseases such as dengue agglutinin cannot provide a reliable diagnosis due to:
• repeated exposure to small inocula of S typhi or to other Salmonella spp that
Box 2 contain type 9 or 12 antigens
82 Olopoenia, King

Table Widal agglutinin titres in Nigerian patients with or without a positive malaria smear

Group Patients (n) 1:40 1:80 1:160

Postgrad Med J: first published as 10.1136/pmj.76.892.80 on 1 February 2000. Downloaded from http://pmj.bmj.com/ on 17 December 2018 by guest. Protected by copyright.
1 *(PMS, NSTC, LPTI) 45 85% 12% 3%
2 *(NMS, NSTC, LPTI) 69 45% 15% 10%

*PMS = positive malaria smear; NMS = negative malaria smear; NSTC = negative S typhi culture; LPTI =
lack of prior typhoid immunisation.

• previous typhoid fever immunisation

• other infectious agents such as malaria.
Although a number of reports from some developing countries have suggested
that a single Widal test is suYcient to make the diagnosis of typhoid fever,16–20
others have disputed the usefulness of such a single test result.9 10 14 15 21 22 In
some developing countries where the use of a single Widal test appears to be the
norm, there has been an increase in the rate of false-positive results. We have
studied Widal agglutinin in malaria infection in a Nigerian population and found
that 85% of patients with a negative S typhi culture but positive malaria smear
had Widal titres of 1:40, 12% had titres of 1:80, and 3% had titres of 1:160. In
contrast, 45% of patients with both S typhi cultures and malaria smears negative
had Widal titres of 1:40, 15% had titres of 1:80, and 10% had titres of 1:160
(table). Schroeder23 concluded in a review of clinical interpretation of serologic
tests for typhoid fever that the tests are nonspecific, poorly standardised, confus-
ing and diYcult to interpret. Erroneous interpretation of the test result may lead
to misdiagnosis and mismanagement of the patient, resulting in major morbidity
and mortality.
In interpreting Widal test results, it is important that there should be close
communication between the physician requesting the test and the laboratory,
since modifications of technique in individual laboratories may aVect the Widal
titres and some patients with bacteriologically confirmed typhoid fever may fail
to develop the usual rise of antibody titres. The results of the tests should be
reported as either ‘no agglutination’ or, if agglutination is present, in titres (1:20,
1:40 or 1:80) rather than in descriptive (negative or positive) terms, as the latter
may be misleading and contribute to the false interpretation of the test result by
the physician. The function of the laboratory is to perform and report the test
result to the requesting physician, who in turn will use the data to help make the
proper diagnosis. Unfortunately, in several areas of developing countries, the
laboratory performs the test, makes the diagnosis and prescribes the antibiotics.
It should be stressed that a single Widal agglutination test has no diagnostic
significance. According to HoVman et al,10 the results of a single Widal test, tube
dilution, micro-agglutination or slide agglutination are virtually un-interpretable
unless the sensitivity and specificity of the test for the specific laboratory and
patient population are known, as well as predictive values. Even in the extreme
case of a high titre in a single Widal agglutination test, the causative organism
may often be due to other species of Salmonella, rather than S typhi. Sansone et
al6 published a case report where the Widal reaction to typhoid O antigen on
admission for an unexposed patient was 1:320, with an increase in titre to 1:20
480 by the fourth day. While both blood and urine cultures were negative for S
typhi in this case, a non-typhoidal Salmonella sp was isolated from the stool of this
patient which was identified as S javiana. In an individual with no prior exposure
to S typhi infection (either lack of active infection or absence of passive immuni-
sation), a higher than 1:50 or 1:100 titre on an initial single test, usually corre-
lates fairly well with exposure to typhoid fever. However, even these single high-
value titres in an endemic area where repeated exposures to S typhi may have
occurred, do not have any clinical relevance in the absence of a positive isolate of
the causative organism or its antigen.11

Limitations of the Widal test

While the Widal test has played a major role in the diagnosis of typhoid fever in
the past, recent technical developments have revealed several pitfalls in its use
and interpretation of its result. Clinically, it is obvious that a single Widal test in
an unvaccinated or unexposed patient may have some diagnostic relevance.
However, the result of such a single test has no diagnostic significance in an
endemic region; in part due to diYculty in establishing a steady-state or baseline
titre of Widal agglutination, which limits the usefulness of the test as a reliable
diagnostic indicator of the disease process.
The results of studies done in Nigeria to evaluate the clinical value of a single
Widal test and the presence of Widal agglutinin in malaria infection are
Widal agglutination test 83

summarised in the table.15 24 The common denominator between the two groups
was the lack of prior immunisation against typhoid fever and absence of positive S
typhi culture. One would not therefore expect any patient to have any specific

Postgrad Med J: first published as 10.1136/pmj.76.892.80 on 1 February 2000. Downloaded from http://pmj.bmj.com/ on 17 December 2018 by guest. Protected by copyright.
Widal agglutinin in their serum, unless there are related, undetected, antigenic
determinants of S typhi present in the cells of other organisms. The presence of
Widal agglutinin under conditions of positive malaria smear, negative S typhi cul-
ture and negative prior typhoid immunisation (as seen in group 1), would suggest
that malaria parasite may have some undefined antigenic determinants similar to
S typhi which can induce antibody production. This could explain the febrile con-
dition seen in some of these patients. On the other hand, the presence of Widal
agglutinin under conditions of negative malaria smear, negative S typhi culture and
negative prior immunisation against typhoid fever (as seen in group 2) suggests
that other infectious agents, in addition to Salmonella and malaria parasite, may
also share common antigenic determinants with S typhi. These findings are in
agreement with other reports from India with similar environmental and disease
(malaria, typhoid) conditions,25–31 two cases from Canada,32 and a case from
Baltimore,33 all of which cast further doubt on the reliability and the use of Widal
test for the diagnosis of typhoid fever in endemic regions.
The use of the Widal test to diagnose typhoid fever should therefore be lim-
ited to situations in which there is no other confirmatory supportive test, such as
positive culture, available. Similarities between typhoidal and non-typhoidal
Salmonella antigens mean that a serological method of diagnosis is the least
accurate for typhoid fever. Due to the inexperience of some clinicians in typhoid
endemic countries, many cases of pyrexia of unknown origin receive the diagno-
sis of typhoid fever, based upon a false-positive Widal test result rather than a
positive culture of S typhi.

Antigen detection as an alternative to Widal agglutination

While bacteriological culture remains the gold standard for definitive diagnosis
of typhoid fever, lack of its immediate availability during the acute febrile illness
may limit its use. In an acute febrile illness in an endemic typhoid region where
the clinical picture is ambiguous, a rapid, accurate, specific and sensitive test
should be used to diVerentiate typhoidal from non-typhoidal febrile illnesses.
Clinicians usually elect to treat, rather than wait for blood or stool culture
results, which may take 3–5 days. While there might be some merit in this
approach, particularly in areas where culture facilities are either poor or not
available, and where Widal testing is the norm, the use of rapid antigen screen-
ing directly from the stool of the suspected patient would be more useful.
Khan et al34 35 have described a new rapid immuno-enzymatic dipstick test for
detection of Salmonella directly from the stool. The test which is non-invasive,
involves homogenisation of stool sample in a buVer solution and immersion of a
dipstick (previously coated with antibodies) in a tube containing the supernatant
from the homogenised stool samples. The contents of the tube (dipstick and
supernatant) are incubated at room temperature for 15 min and a second tube is
incubated for an additional 5 min for full development of colour. The dipstick is
air dried and the result is visualised as a horizontal mark on the dipstick. While
this test is new, Khan et al have reported a preliminary sensitivity of 94%, spe-
cificity of 98%, negative predictive value of 99.5%, and positive predictive value
of 74%. A large-scale field trial is underway to determine the true sensitivity,
specificity and the predictive values. It is hoped that such a direct stool testing
will be a useful discriminating test which can be used with confidence in areas
where both malaria and typhoid may have similar clinical presentations.

Conclusion

More than 100 years after the introduction of the Widal test for diagnosis of
typhoid fever, the controversy that surrounded the test has not abated. It has
become increasingly obvious that bacterial agglutination systems (particularly
Widal), while oVering a simple methodology, often result in misleading
information because of the polyvalent nature of the antigens involved. Whereas
cross-reacting antigens are widely distributed in the microbial world, the specifi-
city and sensitivity of bacterial agglutination is not suYcient when used in
human serum assays. We believe that Widal test cannot be expected to give a
reliable diagnostic result in endemic regions for the following reasons:
• the inherent variabilities of the test
• diYculty in establishing a steady-state baseline titre for the population
• repeated exposures to S typhi in endemic regions
• cross-reactivities with other non-Salmonella organisms
• lack of reproducibility of the test result.
84 Olopoenia, King

The use of Widal agglutination should not be encouraged, given all these
negative points. As cultures are time consuming, increased eVorts should be
made to find a better, more rapid, sensitive and specific test (such as antigen

Postgrad Med J: first published as 10.1136/pmj.76.892.80 on 1 February 2000. Downloaded from http://pmj.bmj.com/ on 17 December 2018 by guest. Protected by copyright.
screening) to supplement clinical and culture data.

1 Widal FM. Serodiagnostic de la fiévre typhoide 13 DeVillier AB, Deupree RH, Dickinson C, 24 Olopoenia L, Oyewole F, Onafowokan R, et al.
a-propos d’uve modification par MMC Nicolle Beeler MF. Comparative study of ‘O’ antigens. Widal agglutination in malaria infection. Med
et al. Halipie. Bull Soc Med Hop Paris Am J Clin Pathol 1965;44:410–2. Rev 1996;3:5–6.
1896;13:561–6. 14 Hook EW. Salmonella species (including ty- 25 Khubnani H, Phalke D, Khubnani AH, Jain RC.
2 Washington, JA II. In: Henry JB, ed, Medical phoid fever). In: Mandell GL, Douglas RG, Coexistence of anti-Salmonella agglutinins in
microbiology in clinical diagnosis and management Bennett JE, eds, Principle and practice of infectious falciparum malaria (letter; comment). J Assoc
by laboratory methods, 18th edn. Philadelphia, diseases, 2nd edn. New York: Wiley Medical Physicians India 1995;43:585–6.
PA: WB Saunders, 1991; pp 1055–8. Publications, 1985; pp 1256−67. 26 Jhaveri KN, Nandwani SK, Mehta PK, Surati
3 Manson-Bahr PEC, Bell DR. Manson’s tropical 15 Olopoenia L, Oyewole F, Onafowokan R, et al. RR, Parmar BD. False positive modified Widal
diseases, 19th edn. London: Bailliere-Tindall, Widal agglutination in malaria infection in test in acute malaria. J Assoc Physicians India
1987; pp 194–206. Nigeria (abstract). Clin Immunol Immunopathol 1995;43:754–5.
4 Gilman RH, Terminel M, Levine MM, 1995;75:314. 27 Kumar A, Katiyar GP. Mixed infection with
Hernandez-Mendoza P, Hornick R. Compari- 16 Mohammed I, Chikwem JO, Gashau W. Deter- plasmodium-vivax and Salmonella typhi in an
son of relative eYcacy of blood, stool, urine, mination by Widal agglutination of baseline titre infant (letter). Indian Pediatr 1995;32:243–4.
bone marrow and rose spot cultures for recovery for the diagnosis of typhoid fever in two 28 Singh PS: Two hundred false Widal positive
of Salmonella typhi in typhoid fever. Lancet Nigerian states. Scand J Immunol 1992; cases from Eastern UP (letter). J Assoc Physi-
1975;1:1211–5. 11(suppl):153–6. cians India 1996;44:842.
5 Geddes AM. Unexplained fever. BMJ 1974;4: 17 Rasaily R, Dutta P, Saha WR, et al. Value of a 29 Singh PS. Coexistence of enteric fever with
397–8. single Widal test in diagnosis of typhoid fever. malaria (letter, comment) J Assoc Physicians
6 Sansone P, Saslaw MS, Hennekens CH. High Indian J Med Res 1993;97:104–7. India 1995;43:71–2.
titer Widal reaction. JAMA 1972;220:1615–6. 18 Pang T, Puthucheary S. Significance of Widal 30 Samal KK, Sahu CS. Malaria and Widal
7 Somerville PC, Lewis M, KoornhoV HS, et al. test in the diagnosis of typhoid fever in endemic reaction. J Assoc Physicians India 1991;39:745–
The Widal test in the diagnosis of typhoid fever areas. J Clin Pathol 1983;36:471–5. 7.
in Transvaal. S Afr Med J 1981:59(24);851–4. 19 Choo KE, Razif AR, Oppenheimer SJ, et al. 31 Koul PA, Waheed A, Shah PA, Igbal Jand Aslam
8 Anon. Typhoid and its serology. BMJ 1978;18 Usefulness of Widal test in diagnosing child- K. Malaria and Widal reaction (letter). J Assoc
Feb:389–90. hood typhoid fever in endemic areas. J Paediatr Physicians India 1993;41:617–8.
9 Reynolds DW, Carpenter RL, Simon WH. Child Health 1993;29:36–9. 32 Gopinath R, Keystone JS, Kain KC. Concur-
Diagnostic specificity of Widal’s reaction for 20 Wicks ACB, Cruickshank JG, Museeve N. rent falciparum malaria and Salmonella bacter-
typhoid fever. JAMA 1970;214:2192–3. Observations on the diagnosis of typhoid fever emia in travelers. Report of 2 cases. Clin Infect
10 HoVman SL, Flanigan TP, Klancke D, et al. The in an endemic area. SAfr Med J 1974;48:1368– Dis 1995;20:706–8.
Widal slide agglutination test, a valuable rapid 70. 33 Zenilman JM. Typhoid fever (clinical confer-
diagnostic test in typhoid fever patients at the 21 Aquino RL, Lansong MAD, Quimpo VS, Sam- ence). JAMA 1997;278:847–50.
infectious disease hospital in Jakarta. Am J Epi- brero LT, Saniel MC. Evaluation of a single 34 Khan W, Maldonado S, Akron S, Rodriguez W.
demiol 1986;123:869–75. Widal test in the diagnosis of enteric fever. SE Rapid screening for Salmonella by dipstick,
11 Welch H, Mickle FL. A rapid slide test for the Asian J Trop Med Public Health 1991;22:375–9. ELISA test (abstract). Proc Am Soc Microbiol
serological diagnosis of typhoid and paraty- 22 Levine MM, Grados O, Gilman RH, et al. Diag- Ann Meet, Dallas TX: May 1991.
phoid fevers. Am J Public Health 1936;26:248– nostic value of Widal test in areas endemic for 35 Khan W, Maldonado S, Akron S, Gardner M,
55. typhoid fever. Am J Trop Med Hyg 1978;27:795– Rodriguez W. Rapid screening for Salmonella
12 Gaultney JB, Wende RD, Williams RP. Microag- 800. directly from stool by dipstick ELISA test
glutination procedures for febrile agglutination 23 Schroeder AS. Interpretation of serologic tests (abstract). Proc 31st ICCAC, Chicago, IL: Octo-
tests. Appl Microbiol 1971;22:635–40. for typhoid fever. JAMA 1968;206:839–40. ber 1991.