Secondary Logo

Journal Logo

Original Studies

Differential Diagnosis of Chikungunya, Dengue Viral Infection and Other Acute Febrile Illnesses in Children

Laoprasopwattana, Kamolwish MD*; Kaewjungwad, Lamy*; Jarumanokul, Roongrueng MSc; Geater, Alan PhD

Author Information
The Pediatric Infectious Disease Journal: May 2012 - Volume 31 - Issue 5 - p 459-463
doi: 10.1097/INF.0b013e31824bb06d

Abstract

There have been many outbreaks of chikungunya (CHIK) in Africa and Asia since it was first observed in 1952.1 An outbreak around the Indian Ocean and India between 2005 and 2006 infected more than a third of the population on Reunion Island, east of Madagascar, and more than a million people in India, before the outbreak subsided.2 The same strain of chikungunya virus (CHIKV) spread to Singapore and Malaysia in 2008,3,4 and then to southern Thailand in 2009.5

Previous studies found that ~90% of adult patients with CHIK presented with fever and arthralgia/arthritis of small joints such as in the wrist, ankle and around the metacarpophalangeal and metatarsophalangeal bones.6,7 However, arthralgia in children with CHIK was milder or more subtle than in adults,8–10 and because this symptom can be indicative of various diseases, and accurate history-taking from young children concerning the level or even presence of nonvisible symptoms such as arthralgia is difficult, differentiating CHIK from other similar infections in children is problematic.

Both CHIKV and dengue virus are transmitted to humans by Aedes mosquitoes, and both viruses can cause hemorrhagic fever. In Thailand, dengue virus is endemic year round, and the clinical manifestations of dengue fever such as high-grade fever, myalgia, arthralgia and rash are similar to those of CHIK.11 Kularatne et al studied adult patients admitted to a hospital suspected of having either dengue or CHIK during a concurrent outbreak of both viruses and found that rash and arthralgia could not differentiate CHIK from dengue, but arthritis was found only in CHIK. They also found that a low platelet count after 3 days of fever could help to differentiate dengue from CHIK.12 Hochedez et al studied adult travelers who had returned from CHIK and dengue outbreak areas who had fever with rash and found that arthralgia was found only in those who had CHIK, whereas leukopenia and thrombocytopenia were common in dengue cases.13

Due to this similarity in symptoms, during a CHIK outbreak in a dengue-endemic area there is a danger that a physician might overdiagnose CHIK in children who present with fever, rash or arthralgia, with the concomitant misdiagnosis of other serious infections such as dengue hemorrhagic fever (DHF) or bacterial infection. If the physician then prescribed a nonsteroidal antiinflammatory drug (NSAID) to relieve the high-grade fever or arthralgia, a normal procedure in treating CHIK, without further investigations, this could lead to severe bleeding in patients with thrombocytopenia or DHF. Therefore, being able to differentiate dengue from CHIK with some confidence during a CHIK outbreak is important.

To our knowledge, there is only 1 study that has compared the occurrence of the major symptoms of dengue viral infection (DVI) and CHIK in children in areas where both are occurring at once,11 as is the situation in this study from southern Thailand of a CHIK outbreak in a dengue epidemic area (CHIK is not endemic to any part of Thailand). The current study aimed to determine whether there are any clinical manifestations that can help to reliably differentiate dengue and other acute febrile illnesses (OFIs) from CHIK, and to examine and compare the clinical outcomes and complications of dengue and CHIK.

PATIENTS AND METHODS

A prospective cohort study, approved by the Ethics Committee of the Faculty of Medicine, Prince of Songkla University, was conducted during a CHIK outbreak between April and July 2009, in the outpatient pediatric clinic and pediatric ward at Songklanagarind Hospital, the major tertiary-care hospital in southern Thailand. Children who were from 1 month to 15 years of age, who lived in the CHIK outbreak area, who had a history of fever <7 days, and in whom the body temperature measured at presentation was ≥38.0°C, and who presented with arthralgia/arthritis, myalgia or rash were enrolled in the study. A complete blood count (CBC) was performed at the first visit of each patient. During the first week after their presentation and diagnosis, all patients were evaluated daily by a physician if hospitalized and by phone if not hospitalized, and then weekly to determine the duration of fever, arthralgia/arthritis, skin lesions and complications. Real-time reverse transcriptase polymerase chain reaction (real-time RT-PCR) was used to detect CHIKV in serum.14 A commercial indirect immunofluorescence assay (IFA; EUROIMMUN AG, Luebeck, Germany) was used to detect CHIK IgM in serum.15 CHIK was diagnosed if real-time RT-PCR or the IFA IgM was positive. DVI was diagnosed by using the capture enzyme-linked immunosorbent assay to detect dengue IgM and the hemagglutination inhibition test to detect total dengue antibodies. Acute DVI was diagnosed if the dengue IgM test was positive and indicated a 4-fold rise in hemagglutination inhibition.16

The initial serology test was performed at the first visit, and the second test 2 weeks later. Patients with negative results from both the CHIK and DVI tests were classified as OFI. Leukopenia was defined if the white blood cell count (WBC) was <5000 cells/mm3; thrombocytopenia was defined if the platelet count was <150,000 platelets/mm3 and anemia was defined if the hematocrit was <36%.

STATISTICAL ANALYSIS

Data were evaluated using descriptive statistics (mean and SD, median, and interquartile range [IQR] or frequency and percentage, as appropriate). Comparisons between patients with CHIK, dengue and OFI were made using analysis of variance or the Kruskal-Wallis test for continuous data and χ2 test for categorical data. Combinations of diagnostic parameters were tabulated against the diagnosis of CHIK (versus not-CHIK) and sensitivity and specificity, as well as positive predictive value (PPV) and negative predictive value, with exact confidence intervals calculated. Diagnostic indices showing some evidence (P < 0.2) of differing across the 3 diagnostic groups were included in an initial multinomial logistic regression model to identify independent associations with these diagnoses. The model was refined by sequential backward elimination of covariates not contributing significantly to the fit of the model (ie, P > 0.5) using the likelihood ratio test. Stata version 7 (StataCorp, College Station, TX) was used for statistical analysis.

RESULTS

Causes of Acute Febrile Illness in Patients Suspected of Having Chikungunya

During the study period, there were 50 patients whose symptoms indicated the possibility of CHIK. Laboratory tests confirmed CHIK alone in 32 (64%) of these patients: real-time RT-PCR in 21 cases and IFA IgM in 25 cases (Table 1). Of the remaining 18 patients, 10 were confirmed to have dengue, 7 an OFI and 1 had both dengue and CHIK (confirmed by a 4-fold rise in hemagglutination inhibition titers for dengue and IFA IgM positive for CHIK [this patient was excluded from the comparison analysis]).

TABLE 1
TABLE 1:
Confirmation Testing of Thirty-three Patients With Chikungunya

There were 4 patients who had previously been treated with an NSAID before their enrollment, of whom 1 and 3 patients were finally diagnosed with OFI and CHIK, respectively; all except 1 patient with CHIK who had severe and persistent arthritis stopped the NSAID at their enrollment. During the study period, 49 of 50 patients were treated with acetaminophen to relieve fever, arthralgia or myalgia.

Of the 49 patients included in the study, 22 (44.9%) were hospitalized; the hospitalization rates among the 3 groups were not significantly different at 37.5%, 60.0% and 57.1% in the CHIK, dengue and OFI patients, respectively.

Of the 32 confirmed CHIK patients, 5 (15.6%) patients had complications, of which 3 were serious, 1 each of fatal encephalitis, acute disseminated encephalomyelitis and chronic arthritis. The patient who developed chronic arthritis responded well to NSAID treatment and the arthritis cleared up within 1 month, and the patient with acute disseminated encephalomyelitis fully recovered after treatment with pulse methylprednisolone. The other 2 complicated patients had less serious problems, 1 each of febrile convulsion and confusion.

Of the 10 confirmed dengue cases, 8 patients had secondary DVI. Dengue fever, DHF grade I and DHF grade II were diagnosed in 3, 3 and 4 patients, respectively. The bleeding in the 4 patients with DHF grade II was only mild, such as petechia and epistaxis.

Clinical Characteristics in the CHIK, DVI and Other AFI Patients

Sex, age and family member living in the same home having CHIK were not different among the 3 groups. Children with CHIK saw a doctor sooner; 30 of 32 (93.8%) patients visited their doctor within 2 days after the symptoms began whereas 60% and 57% of the patients with dengue and OFI, respectively, visited the doctor only after the second day of fever. Patients with CHIK had a shorter total duration of fever, 29 of 31 (93.5%) patients (excluding the one who had fatal encephalitis) having total duration of fever less than 4 days, whereas 5 of 10 and 3 of 7 patients with dengue or OFI had fever for more than 4 days; however, this difference was not significant (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/B97). Of the 2 CHIK patients who had fever for more than 4 days, 1 patient, with underlying acute lymphoblastic leukemia, had fever for 12 days without other defining infection, and the other had fever for 6 days, with no underlying disease.

The mean days of fever before going to see a doctor were not significantly different between those who had and those who did not have a family member with CHIK (1.5 ± 1.0 versus 2.1 ± 1.6, respectively, P = 0.19).

Of the 50 patients, there were 11 younger than 5 years who could not reliably describe their pain symptoms. Of the remaining 39 patients, 30 (76.9%) indicated arthralgia, and more than 80% of the patients in all groups had myalgia. Arthralgia was found in a significantly higher proportion in patients with CHIK than in those with dengue or OFI (95.8%, 44.4% and 50.0%, respectively; Table 1). Most of the arthralgia was mild, although 1 patient with CHIK had chronic arthritis at the shoulder joint. The median durations of arthralgia among the CHIK, dengue and OFI groups were 4.0 (IQR 2.8–6.5), 3.0 (IQR 1.5–8.5) and 3.5 (IQR 1.2–8.0) days, respectively, which was not significantly different among the groups.

There were 6 patients who had myalgia without arthritis or rash. These 6 patients comprised 3 of the 9 dengue and 3 of the 6 OFI patients. None of the patients with CHIK had myalgia without arthritis or rash.

Children with CHIK presented with headache in a higher proportion than the dengue and OFI groups, but chills and general respiratory symptoms such as cough and rhinorrhea were not different among the 3 groups. Gastrointestinal symptoms, especially anorexia and vomiting, were more commonly found in the dengue group than the other 2, but abdominal pain and diarrhea were not different among the groups. Hemorrhagic manifestations were found only in 4 of 10 children with dengue (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/B97).

Physical examination results showed that patients with CHIK and dengue had higher body temperatures than patients with an OFI, but the differences were not significant. Hepatomegaly was found in 4 of 10 patients with DVI, and in no CHIK or OFI patients. Skin rash during fever was significantly more common in children with CHIK. Children with CHIK developed rash during the febrile stage of the illness, which subsided after defervescence. Acrocyanosis was found in 4 of 5 infants with CHIK, and 2 infants had skin blebs.

The CBC taken during the first visit showed that none of the children with dengue had anemia, but higher proportions had leukopenia or thrombocytopenia than did children with CHIK (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/B97).

Using Fever and/or (Arthralgia, Rash and WBC) to Diagnose CHIK

The specificity and PPV of the standard clinical triad (fever, arthralgia and rash) to diagnose CHIK in the study were 70.6% and 83.3%, respectively, whereas the specificity and PPV of the combination of fever and arthralgia to diagnose CHIK were only 47.1% and 74.2%, respectively. The specificity and PPV using only fever and skin rash during fever to diagnose CHIK were higher than the combination of fever and arthralgia, at 64.7% and 82.4%, respectively. Higher specificity and PPV were found if the patient had seen a doctor within 2 days of fever development (including patients who had total duration of fever <2 days; Table 2).

TABLE 2
TABLE 2:
Sensitivity, Specificity, Positive Predictive Value and Negative Predictive Value of Various Combinations of Diagnostic Factors in Diagnosing Chikungunya

Variables included in the initial multinomial model were vomiting, duration of fever at presentation (≤2 versus >2 days), joint pain and/or limited activity, skin rash during fever, WBC (≥5000 versus <5000 cells/mm3) and platelet count (≥150,000 versus <150,000 cells/mm3). (Low hematocrit was not included in the model as none of the patients with DVI had hematocrit lower than 36%). After model refinement, fever ≤2 days, skin rash during fever, and WBC ≥5000 cells/mm3 were identified as being independently and significantly associated with CHIK in comparison with dengue and OFI, with relative risk ratios of 10.4 (0.9–116) and 13.7 (1.3–145), 13.8 (1.2–164) and 14.8 (1.6–138), and 18.3 (1.7–194) and 1.8 (0.1–20.6), respectively.

DISCUSSION

During the 2009 CHIK outbreak in southern Thailand on which this study was based, slightly over one-third of the children suspected of having CHIK and enrolled in the study actually had dengue or an OFI.

Although arthralgia is the major symptom of CHIK and most of our patients with CHIK who were old enough to reliably report physical symptoms indicated arthralgia, the specificity and PPV of a diagnosis of CHIK based on only fever and arthralgia were only 47.1% and 74.2%, respectively. These relatively low numbers could be explained by noting that arthralgia was also common in children who had dengue or an OFI in our study. A study in adult patients by Kularatne et al also found that arthralgia could not differentiate CHIK from dengue.12 It has also been noted that arthralgia in children who have CHIK is usually milder or more subtle than in adults,8–10 and as a result, using a mild, subjective factor such as arthralgia to try to differentiate CHIK from other infections, especially dengue, in children is more difficult than in adults. In addition, in the youngest children, who could not complain directly of pain, we could not use arthralgia as an accurate indicator of the presence of CHIK, as getting an accurate history from young children concerning the presence of arthralgia or any other nonobvious symptom is difficult and uncertain at best.

Kularatne et al found that a low platelet count after day 3 of fever could help to differentiate dengue from CHIK, but the presence/absence of skin rash was not useful in such differentiation.12 This last finding was, however, different from our study and from that of Nimmannitya et al, in which skin rash during fever was helpful in differentiating CHIK from dengue.11 Although thrombocytopenia was more commonly found in children with dengue than those with CHIK, half of the children had thrombocytopenia, whereas 80% had leukopenia after day 3 of fever. Therefore, in our study, using a WBC ≥5000 cells/mm3 was more helpful than thrombocytopenia to differentiate CHIK from dengue. However, 2 patients with dengue who had CBCs performed on days 1–2 of fever had WBC ≥5000 cells/mm3, suggesting that if a patient sees a doctor on the 1st or 2nd day of symptoms, the WBC count may not differentiate dengue from CHIK, and a patient should have a follow-up CBC if their fever does not decrease or their clinical symptoms progress. In addition, during the febrile stage, when both CHIKV and dengue virus can be detected in the bloodstream, a more specific laboratory investigation such as antigen detection should be used to differentiate dengue from CHIK.

Although the specificity and PPV were both 100% when using the combination of fever ≤2 days, rash and WBC ≥5000 cells/mm3 to diagnose CHIK, we think it important to note that based on these symptoms alone, certain other serious infections (such as, for example, staphylococcus scaled skin syndrome) cannot be entirely excluded in this young age group. To prevent the misdiagnosis of other serious infections due to overdiagnosis of CHIK during a CHIK outbreak, we suggest that a CBC and septic workup be performed in young children with these symptoms, just as they would be in a time without a CHIK outbreak.

We found rash or arthralgia in nearly 90% of our children with CHIK, although previous studies have found rash or arthralgia in only 40–50% of children with CHIK.10,11 This difference could be explained by noting that in our study we included only patients who had myalgia, arthralgia or rash, so the number of patients who had rash or arthralgia would be higher than in previous studies with broader inclusion criteria. We found that none of the patients with CHIK had myalgia without arthralgia or rash, unlike patients with dengue or an OFI, who had myalgia without arthralgia or rash in 33.3% and 50.0% of cases, respectively, suggesting that patients who complain of myalgia without arthralgia or rash are unlikely to have CHIK.

Our patients with CHIK had shorter duration of fever than the DVI cases, with more than 90% of them having total duration of fever not more than 4 days. The earlier study referenced also found that patients with CHIK had, in general, a shorter duration of fever than DVI patients.11

Four of the 5 infants in our study had acrocyanosis, and skin blebs were found in 2 infants, 1 of whom had febrile seizure. These results were similar to a previous study in infants with CHIK that found acrocyanosis in 75% of the patients, seizure in 40% and blebs to be common.17

Previous studies have found mild hemorrhage rates of 3.5% and 6.4% in infants17 and adults,9 respectively; however, none of the children with CHIK in our study had hemorrhage. This could be explained by noting that in the 2 patients in our study who had thrombocytopenia, the platelet counts were more than 100,000 platelets/mm3, and also none of our patients (except the one who had chronic arthritis) were treated with an NSAID, which can cause gastrointestinal bleeding. Hemorrhage was found only in 4 patients with DVI, a finding compatible with an earlier study that found that thrombocytopenia and hemorrhage were more common in patients with dengue.11

Neurologic complications were noted in 4 children with CHIK (12.5%) in our study, whereas an earlier study in adult patients with severe CHIK found neurologic complications in 12.1%, most of whom presented with delirium and dizziness.9 A previous study found that during a CHIK outbreak, CHIK was the most common cause of encephalitis.18 Adult studies have found mortality rates of CHIK patients who had encephalitis of ~50%; however, most of the fatal cases had underlying diseases.19,20 Our patient who died from encephalitis had no underlying disease except obesity. We should also note, in discussing our various rates, that Songklanagarind Hospital is the major medical tertiary-care referral center in southern Thailand; thus, the proportion of patients we treat with complications in any situation will generally be higher than in the general population, in which the incidence of severe CHIK is less than 0.02%.2

Of our 33 confirmed CHIK patients, 1 was coinfected with DVI, which was about the same rate as noted in an earlier study that examined a CHIK outbreak in a dengue-endemic area.21 This coinfection did not increase the disease severity of either the dengue or the CHIK.

We found, as in previous studies, that CHIKV will be detected only during the febrile stage, and CHIK IFA IgM will be positive only after day 4 of illness.2 However, most of the patients suspected of CHIK lived in a dengue-endemic area, and the chance of misdiagnosis using only the generally accepted clinical profile of fever, arthralgia or rash in our study was nearly 40%. Thus in a similar setting, a dengue-endemic area with an apparent CHIK outbreak, we suggest that children suspected of having CHIK who have no complications should be scheduled to see the doctor if their fever has not decreased by day 3 to evaluate their clinical condition. A CBC should be done at this time, because we found the mean duration of fever in CHIK was 2.8 days and the earliest day that plasma leakage or shock develops in dengue is day 316 and a CBC performed on or later than day 3 was more helpful to differentiate CHIK from DVI than a CBC performed during the first 2 days of fever. During the first 3 days of fever, an NSAID should not be prescribed, even though the patient may have arthralgia and normal platelet count because thrombocytopenia commonly develops after day 3 of fever in patients with dengue and even some patients with CHIK.7,12

To summarize, we found that in our dengue-endemic area experiencing a CHIK outbreak, a clinical diagnosis of CHIK based only on fever, arthralgia and/or rash could lead to overdiagnosis of this disease, and therefore in areas known to be endemic for dengue, when dealing with children during a CHIK outbreak who have no skin rash during fever, or a low WBC count, the physician should not forget to include dengue in the differential diagnosis, and should not prescribe any treatment before definite diagnosis that would be contraindicated for dengue, such as an NSAID. In addition, most children with CHIK have only mild arthralgia, a symptom that can and usually does resolve spontaneously, and some patients also have thrombocytopenia. Even in patients with a normal platelet count, an NSAID should not be prescribed routinely during the first 3 days of fever. NSAIDs should be prescribed only in patients with CHIK who have severe or chronic arthritis with normal platelet count.

ACKNOWLEDGEMENTS

The authors thank Walailuk Jitpiboon for assistance with data analysis, and David Patterson, an English teacher and writing consultant, for help with the English, both of whom work with the Faculty of Medicine, Prince of Songkla University.

REFERENCES

1. Lahariya C, Pradhan SK. Emergence of chikungunya virus in Indian subcontinent after 32 years: A review. J Vector Borne Dis. 2006;43:151–160
2. Pialoux G, Gaüzère BA, Jauréguiberry S, et al. Chikungunya, an epidemic arbovirosis. Lancet Infect Dis. 2007;7:319–327
3. AbuBakar S, Sam IC, Wong PF, et al. Reemergence of endemic Chikungunya, Malaysia. Emerging Infect Dis. 2007;13:147–149
4. Leo YS, Chow AL, Tan LK, et al. Chikungunya outbreak, Singapore, 2008. Emerging Infect Dis. 2009;15:836–837
5. Theamboonlers A, Rianthavorn P, Praianantathavorn K, et al. Clinical and molecular characterization of chikungunya virus in South Thailand. Jpn J Infect Dis. 2009;62:303–305
6. Staikowsky F, Le Roux K, Schuffenecker I, et al. Retrospective survey of Chikungunya disease in Réunion Island hospital staff. Epidemiol Infect. 2008;136:196–206
7. Staikowsky F, Talarmin F, Grivard P, PLoS ONE, et al. Prospective study of Chikungunya virus acute infection in the Island of La Réunion during the 2005-2006 outbreak. 2009;4:e7603
8. Rezza G, Nicoletti L, Angelini R, et al. CHIKV study groupInfection with chikungunya virus in Italy: an outbreak in a temperate region. Lancet. 2007;370:1840–1846
9. Borgherini G, Poubeau P, Staikowsky F, et al. Outbreak of chikungunya on Reunion Island: early clinical and laboratory features in 157 adult patients. Clin Infect Dis. 2007;44:1401–1407
10. Sebastian MR, Lodha R, Kabra SK. Chikungunya infection in children. Indian J Pediatr. 2009;76:185–189
11. Nimmannitya S, Halstead SB, Cohen SN, et al. Dengue and chikungunya virus infection in man in Thailand, 1962–1964IObservations on hospitalized patients with hemorrhagic fever. Am J Trop Med Hyg.. 1969;18:954–971
12. Kularatne SA, Gihan MC, Weerasinghe SC, Med J, et al. Concurrent outbreaks of chikungunya and dengue fever in Kandy, Sri Lanka, 2006–07: a comparative analysis of clinical and laboratory features. Postgrad. 2009;85:342–346
13. Hochedez P, Canestri A, Guihot A, et al. Management of travelers with fever and exanthema, notably dengue and chikungunya infections. Am J Trop Med Hyg. 2008;78:710–713
14. Pastorino B, Bessaud M, Grandadam M, et al. Development of a TaqMan RT-PCR assay without RNA extraction step for the detection and quantification of African Chikungunya viruses. J Virol Methods. 2005;124:65–71
15. Litzba N, Schuffenecker I, Zeller H, et al. Evaluation of the first commercial chikungunya virus indirect immunofluorescence test. J Virol Methods. 2008;149:175–179
16. Anonymous. Dengue Hemorrhagic Fever: Diagnosis, Treatment and Control. 1997 Geneva, Switzerland World Health Organization
17. Valamparampil JJ, Chirakkarot S, Letha S, et al. Clinical profile of Chikungunya in infants. Indian J Pediatr. 2009;76:151–155
18. Lewthwaite P, Vasanthapuram R, Osborne JC, et al. Chikungunya virus and central nervous system infections in children, India. Emerging Infect Dis. 2009;15:329–331
19. Farnon EC, Sejvar JJ, Staples JE. Severe disease manifestations associated with acute chikungunya virus infection. Crit Care Med. 2008;36:2682–2683
20. Economopoulou A, Dominguez M, Helynck B, et al. Atypical Chikungunya virus infections: clinical manifestations, mortality and risk factors for severe disease during the 2005-2006 outbreak on Réunion. Epidemiol Infect. 2009;137:534–541
21. Leroy EM, Nkoghe D, Ollomo B, et al. Concurrent chikungunya and dengue virus infections during simultaneous outbreaks, Gabon, 2007. Emerging Infect Dis. 2009;15:591–593
Keywords:

chikungunya; dengue viral infection

Supplemental Digital Content

© 2012 Lippincott Williams & Wilkins, Inc.