Abdominal and Chest Ultrasonography: A predictor for disease progression in nonsevere dengue

Abstract

Background

Dengue is one of the most rapidly spreading arboviral infections in the world. Ultrasound is well established in abdomino-thoracic evaluation of patients with dengue infection. The aim of this study was to explore the role of ultrasound in predicting occurrence of severe infection in dengue patients and in predicting deterioration in patients with nonsevere dengue.

Methods

The serologically proven dengue patients who reported to hospital during the study period were divided into three categories based on the dengue infection severity score. Ultrasound findings of abdomen and chest in these patients were noted in the initial, as well as follow-up scans and inferences drawn.

Results

61% belonged to the category of Dengue Fever, 35% were in Dengue Hemorrhagic Fever category, and 4% had Dengue Shock Syndrome. Positive ultrasound findings were seen to be significantly higher in patients with severe dengue. Logistic regression analysis revealed the presence of pericholecystic fluid to be significantly associated with the severe disease, while the presence of gall bladder wall edema, ascites, and any ultrasound finding were significantly associated with the disease progression. The odds of a patient with severe dengue having gall bladder wall edema, ascites, or any ultrasound finding were 2.74, 2.04, and 2.619 times, respectively.

Conclusion

Our study indicates that positive findings on ultrasound are significantly higher in severe dengue and also that ultrasound can be reliably used to identify the patients with nonsevere dengue who are likely to progress to severe dengue.

Introduction

Dengue is one of the most rapidly spreading arboviral infections in the world. Despite various measures taken for primary prevention of the disease, frequent outbreaks have become a norm and are straining the already limited resources of the public health system. Symptomatic dengue infections can present with a wide range of clinical manifestations ranging from a mild febrile illness to a life-threatening shock syndrome or multiorgan dysfunction. According to World Health Organization, the case fatality rate for Dengue Fever (DF) is roughly 5%.¹ On the other hand, the case fatality rate in patients with severe dengue infection, which consists of Dengue Haemorrhagic Fever (DHF) and Dengue Shock Syndrome (DSS), can be as high as 44%.¹ With the rising trend in dengue cases over the years, a dengue clinical severity score was introduced by Pongpan et al² in 2013.

Ultrasound is a reliable tool to detect organomegaly, gall bladder wall edema, and plasma leakage into serosal cavities. Several studies over the past decades have documented ultrasound findings in dengue patients.^{3, 4, 5} There is, however, still a paucity of data regarding the possible role of ultrasound in predicting deterioration before it manifests clinically. The aim of this study was to evaluate this relatively less investigated aspect of ultrasound in patients with dengue infection and also explore the possible role of ultrasound in the prediction of severe dengue.

Material and methods

This was a prospective study undertaken at a tertiary care government hospital in New Delhi, India. All patients presenting to the hospital with symptoms suggestive of dengue between 15 Oct 2016 and 15 Dec 2017 and subsequently proven to have dengue infection based on NS1Ag/IgM positivity were enrolled as the study participants. Clearance was obtained from the Institutional Ethics Committee before commencing the study. Written informed consent was taken from the participants.

Inclusion criteria

All patients who presented to hospital with symptoms of an acute febrile illness during the study period and were found to be dengue positive by NS1Ag Card/ELISA test or IgM positive by Card test.

Exclusion criteria

Patients who had any of the following pre-established findings/diagnosis before the onset of their acute febrile illness were excluded from the study:

• Pleural effusion

• Ascites

• GB wall thickening of any other etiology

• Cholelithiasis

• Hepatomegaly

• Splenomegaly

• Pericardial effusion

• Malaria/Chikungunya (Active/recent (<8 weeks) infection)

• Thrombocytopenia

Data were tabulated in Excel and analyzed in SPSS 20. Mean, SD, proportions, and percentages were calculated for various variables and logistic regression analysis was done on significant variables.

The dengue positive patients who were finally included in the study were categorized into three grades of severity as per the dengue infection severity score,² which has albeit been developed for and validated in pediatric population. Pongpan et al had developed this score based on patient's age, presence/absence of hepatomegaly, hematocrit percentage, systolic blood pressure, white blood count, and platelet count. Based on the sum of scores assigned to each predictor, the dengue patients are categorized into three subsets. Patients with a score of <2.5 were categorized as DF, a score of 2.5–11.5 placed the patients in the category of DHF, and a score of >11.5 defined the patients in the DSS category (Fig. 1). The odds ratio of ultrasound findings with the clinical severity score was done, and the inference was drawn regarding the role of ultrasound in predicting the severity of dengue infection. This study was conducted at a Government Services hospital that caters to a restricted clientele; hence, all patients who were detected to be suffering from dengue infection were subjected to regular follow-up as per study protocol regardless of disease severity.

Fig. 1.

Patients were divided into three categories based on dengue infection severity score² and subjected to abdominal and thoracic ultrasound at enrolment and on serial follow-up.

Data collection

Ultrasound examinations for all patients were performed on a GE LOGIQ P5 machine using a 3.5/5.0 MHz transducer. All patients were required to be fasting for a minimum of 8 h prior to the ultrasound scan, and all scans were performed by two Radiologists – one having one year of experience in Abdominal and Chest Ultrasonography and the other having eleven years of experience.

Scanning for the abdomen was done with the patient lying in the supine position. The liver was measured in the sagittal plane in the right midclavicular line and the maximum craniocaudal extent noted; hepatomegaly was defined as liver span >15 cm. The spleen was measured along its long axis and a maximum extent of spleen measuring >13 cm was taken as splenomegaly. Gall bladder (GB) wall edema was defined as a GB wall thickness measuring >3 mm with a classical layered appearance seen on ultrasound. Pericholecystic fluid and ascites were diagnosed based on their typical ultrasound appearances. The presence or absence of pleural effusion was determined by scanning bilateral posterior costophrenic angles with the patient sitting upright. The term “Any USG finding” refers to any one or more than one of the ultrasound findings of Hepatomegaly/Splenomegaly/GB wall edema/Pericholecystic fluid/Ascites/Pleural effusion, which were observed in the patients.

The findings of Pleural effusion, Ascites, Pericholecystic fluid, and GB wall edema were considered as a marker for plasma leakage. For eliminating/minimizing bias, the study protocol and the defining criteria of abnormal ultrasound findings were strictly adhered to, and the same equipment was used by both Radiologists.

The ultrasound of the abdomen and chest, as well as the laboratory tests (platelet count, hematocrit level and WBC count), were done for all the study participants within 24 h of each other at the time of enrolment, and the findings were recorded in each case. Ultrasound examinations were subsequently repeated for all patients every alternate day till clinical improvement and/or the resolution of ultrasound findings or patient's demise (Fig. 1).

Results

Based on the dengue infection severity score, the study population of 418 patients was divided into three groups, with 255 (61%) belonging to the DF group, 146 (35%) to the DHF group, and 17 (4%) to the DSS group.

Out of the 255 patients of DF, 173 were males, and 82 were females. In the DHF group, 99 patients were males, while 47 patients were females. Among the 17 patients of DSS, 11 were males and 6 were females. There were a total of 283 (68%) males and 135 (32%) females in the study population. The mean age of the study population was 30.5 ± 13.4 years with the age groups of 16–30 years and 31–45 years accounting for 168 (40%) and 153 (37%) cases, respectively. There were 50 patients aged <15 years, 35 patients in the 46–60 years age group, and 12 patients in the >60 years age group. We did not find any significant association between age category (less than and more than 60yrs) and severity of illness or progression of the disease.

Hepatomegaly on the abdominal ultrasound was found in 27/418 (6%) patients. While none of the patients in the DF group had liver enlargement, 10 patients of the DHF group and all 17 patients of the DSS group had hepatomegaly (Table 1).

Table 1.

Abdominal and chest Ultrasound findings in patients with nonsevere dengue (DF) and severe dengue (DHF and DSS).

Ultrasound findings	Categories of Patients based on Dengue Infection Severity Score of Pongpan et al2
	DF (n = 255)	DHF(n = 146)	DSS(n = 17)	Total (n = 418)
Hepatomegaly	0 (0%)	10 (7%)	17 (100%)	27 (7%)
Splenomegaly	8 (3%)	40 (27%)	16 (94%)	64 (15%)
Ascites	22 (9%)	54 (37%)	16 (94%)	92 (22%)
GB Wall edema	54 (21%)	67 (46%)	16 (94%)	137 (33%)
Pericholecystic fluid	2 (1%)	35 (24%)	16 (94%)	53 (13%)
Pleural effusion	9 (4%)	47 (32%)	16 (94%)	72 (17%)

Splenomegaly on ultrasound was seen in 64/418 (15%) patients wherein 8 patients were of the DF group, 40 patients were of the DHF group, and 16 patients were in the DSS group (Table 1).

Ultrasound revealed GB wall edema in 137/418 (33%) patients. Among these, 54 patients were in the DF group, 67 patients in the DHF group and 16 patients in the DSS group (Table 1).

Pericholecystic fluid was seen on ultrasound in 53/418 (13%) patients, of whom 2 patients had DF, 35 belonged to the DHF group, and 16 patients had DSS (Table 1).

Ultrasound was able to detect Ascites in 92/418 (22%) patients with 22 patients belonging to the DF category, 54 patients to the DHF category, and 16 patients belonging to the DSS category (Table 1).

Pleural effusion on ultrasound of the chest was seen in 72/418 (17%) patients, of whom 9 patients were of DF, 47 patients were of DHF, and 16 patients were of DSS (Table 1).

Ultrasound findings of splenomegaly, GB wall edema, pericholecystic fluid, ascites, and pleural effusion were found to be significantly higher in patients with severe dengue (DHF and DSS) with a p value of 0.000 (Table 2). Although hepatomegaly was also found to be significantly higher in patients with severe dengue, the p value for the same could not be calculated. Logistic regression done with the presence of disease severity as a dependent variable revealed the presence of pericholecystic fluid to be significantly associated with severe disease (Table 3).

Table 2.

Association between ultrasound findings and Severity of Dengue.

USG finding	Severity of Dengue		Total	p value
Hepatomegaly	Present	Absent
Present	27	0	27	NA
Absent	136	255	391
Splenomegaly
Present	56	8	64	0.000
Absent	107	247	354
Pericholecystic fluid
Present	51	2	53	0.000
Absent	112	253	365
GB wall edema
Present	83	54	137	0.000
Absent	80	201	281
Ascites
Present	70	22	92	0.000
Absent	93	233	326
Pleural effusion
Present	63	9	72	0.000
Absent	100	246	346
Any USG finding
Present	83	57	140	0.000
Absent	80	198	278

Table 3.

Logistic regression of ultrasound findings with presence of disease severity as dependent variable.

Variable	B	Std error	p value
Hepatomegaly	−20.028	0.000	NA
Splenomegaly	−1.206	0.617	0.05
Pericholecystic fluid	−1.779	0.890	0.04
GB wall edema	−1.884	1.158	0.10
Ascites	–0.550	0.521	0.29
Pleural effusion	–0.765	0.617	0.21
Any USG finding	2.222	1.174	0.05

All patients who had afore-mentioned positive ultrasound findings at the time of enrolment into the study and/or those patients who showed clinical deterioration subsequently underwent follow-up ultrasound every alternate day till resolution of findings/patient's demise. Out of the 255 patients with DF, 57 patients had sonographic evidence of plasma leakage and/or organomegaly at the time of enrolment, while 198 patients did not. On follow-up, 38/57 (67%) of these DF patients showed clinical deterioration and progressed to severe dengue (DHF/DSS), while only 11/198 (6%) of the remaining DF patients showed progression to severe dengue (DHF/DSS). Out of 146 DHF patients, 68 (47%) patients had sonographic evidence of plasma leakage and/or organomegaly at the time of enrolment, while 78 (53%) patients did not. On follow-up of these 78 DHF patients, 34/78 (43%) patients showed clinical deterioration with the appearance of sonographic evidence of plasma leakage and/or organomegaly. All 17 patients who were clinically diagnosed as DSS had sonographic evidence of plasma leakage and/or organomegaly at the time of enrolment, which was also noted in follow-up ultrasound scans.

Logistic regression was done for all significant deterioration of clinical condition, which revealed that the presence of GB wall edema, ascites, and presence of any of the afore-mentioned ultrasound findings were significantly associated with progression of the illness (Table 4). The odds of a patient having GB wall edema, ascites, or any ultrasound finding progressing to severe dengue are 2.74, 2.04, and 2.619 times, respectively (Table 5).

Table 4.

Association between ultrasound findings and progression of illness.

Ultrasound findings	Progression of disease		Total	p value
Hepatomegaly	Present	Absent
Present	4	16	20	0.735
Absent	91	300	391
Splenomegaly
Present	10	48	58	0.252
Absent	85	268	353
Pericholecystic fluid
Present	9	38	47	0.493
Absent	86	278	364
GB wall edema
Present	47	83	130	0.000
Absent	48	233	281
Ascites
Present	29	56	85	0.007
Absent	66	260	326
Pleural effusion
Present	17	48	65	0.526
Absent	78	268	346
Any USG finding
Present	47	86	133	0.000
Absent	48	230	278

Table 5.

Odds ratio for progression of disease and presence of ultrasound findings.

USG finding	Progression of disease		Total	Odds ratio	95%C I
Hepatomegaly	Present	Absent
Present	4	16	20	0.824	0.26–2.52
Absent	91	300	391
Splenomegaly
Present	10	48	58	0.657	0.31–1.34
Absent	85	268	353
Pericholecystic fluid
Present	9	38	47	0.766	0.35–1.64
Absent	86	278	364
GB wall edema
Present	47	83	130	2.74	1.71–4.41
Absent	48	233	281
Ascites
Present	29	56	85	2.04	1.20–3.43
Absent	66	260	326
Pleural effusion
Present	17	48	65	1.27	0.66–2.23
Absent	78	268	346
Any USG finding
Present	47	86	133	2.619	1.63–4.20
Absent	48	230	278

The mean time taken for resolution of ultrasound findings was 6.9 days in patients with DF, 10.5 days in patients with DHF, and 19.2 days in patients with DSS. Two patients in the DSS category succumbed to illness.

Discussion

This was a prospective study of serologically confirmed dengue patients, which was undertaken with the aim of evaluating the role of ultrasound in the prediction of severe dengue, predicting disease progression in patients with nonsevere dengue, and the possible role of ultrasound in secondary prevention of severe dengue.

A total of 142 patients had signs of plasma leakage and/or organomegaly on ultrasound at the time of enrolment into the study, which included 17/17 DSS patients, 68/146 DHF patients, and 57/255 DF patients. GB wall edema was the commonest finding in patients with nonsevere dengue, while patients with severe dengue had pleural effusion and ascites as the two commonest findings in addition to GB wall edema. Although we found abnormal ultrasound findings of hepatomegaly, splenomegaly, GB wall edema, pericholecystic fluid, ascites, and pleural effusion to be significantly higher in patients with severe dengue, logistic regression analysis done with the presence of severe disease as a dependent variable revealed that the presence of pericholecystic fluid was most significantly associated with severe disease.

Out of the 255 clinically nonsevere dengue patients, 57 were found to have signs of plasma leakage and/or organomegaly on ultrasound at the time of enrolment, while 198 patients did not have any such findings. On clinical and ultrasound follow-up, 38/57 (67%) of these DF patients showed clinical deterioration and progressed to severe dengue (DHF/DSS). On the other hand, only 11 of the remaining 198 DF patients, i.e. 6% showed progression to severe dengue (DHF/DSS). Thus, patients of nonsevere dengue with evidence of plasma leakage and/or organomegaly on their initial ultrasound had a 10-fold chance of disease progression than the patients who had a normal initial ultrasound.

Out of 146 patients who were clinically diagnosed as DHF, 68 (47%) patients had features of plasma leakage and/or organomegaly on ultrasound at the time of enrolment, while 78 (53%) patients had no abnormal ultrasound findings. On clinical and ultrasound follow-up of those DHF patients who had a normal initial ultrasound, 34/78 (43%) patients showed further clinical deterioration with the appearance of positive ultrasound findings in the form of sonographic evidence of plasma leakage and/or organomegaly. All 17 patients who were clinically diagnosed as DSS had evidence of plasma leakage and/or organomegaly on the initial ultrasound at the time of enrolment, and 2 patients eventually died due to multiorgan failure. Statistical analysis revealed that the presence of GB wall edema, ascites, and any ultrasound finding was significantly associated with the progression of illness, with the odds of a patient having GB wall edema, ascites, or any ultrasound finding progressing to severe dengue being 2.74, 2.04, and 2.619 times, respectively.

Michels et al.⁶ undertook a prospective study of 66 patients with proven dengue infection wherein they documented evidence of plasma leakage based on serial ultrasonography examinations. They found that 91% of patients in the severe dengue group developed ultrasonographic evidence of plasma leakage compared to 31% of patients in the nonsevere dengue group. They also found that the presence of subclinical plasma leakage at enrolment had a positive predictive value of 35% and a negative predictive value of 90% for identifying patients at risk of developing severe dengue. Similarly, in our study, we have found that more patients with severe dengue had evidence of plasma leakage and/or organomegaly at the time of enrolment (47% patients with DHF and 100% patients with DSS) compared to patients with nonsevere dengue (22% of DF patients). We also found that patients with nonsevere dengue who had gall bladder wall edema/ascites/any abnormal finding on ultrasound at enrolment were more likely to progress to severe dengue as compared to those who did not have these ultrasound findings.

In a retrospective study of 254 children with serologically confirmed dengue infection, Pothapregda et al⁷ found that on multivariate analysis, gall bladder thickening and hepatomegaly were associated with severe dengue infection. The authors also concluded that subclinical plasma leakage picked on ultrasonography was useful in predicting the increased risk for progression to shock, thus warranting careful monitoring of circulatory status in these patients. In our study, we have found that GB wall edema, ascites, pleural effusion, splenomegaly, and pericholecystic fluid were the commonest ultrasonography findings. All these ultrasound findings were seen more commonly in dengue infections of higher severity grades, and disease progression was linked to the presence of GB wall edema, ascites, and any abnormal ultrasound finding. Differences in demographic profile and study protocol could possibly account for some of the differences in observed findings; however, similar to the study by Pothapregada et al, our study also indicates that early detection of plasma leakage can serve as a predictor for progression in disease severity.

Setiawan et al⁸ conducted a prospective study of 148 pediatric DHF patients to examine the relationship between clinical severity of DHF and their ultrasonography findings. Their study revealed that while pleural effusions, ascites, and gall bladder wall thickening was seen in 30%, 34%, and 32% patients, respectively, in the mild group, these same findings were found in as many as 95% of patients in the severe group thus concluding that ultrasonography could be of use in early prediction of disease severity. In addition to the similar findings observed in our study, we have also found that patients with nonsevere dengue who had evidence of Gall bladder wall edema/ascites/any abnormal finding in initial ultrasound were more likely to progress to severe dengue as compared to those patients who did not have these findings.

Limitations

This study was conducted at a tertiary care referral hospital that caters to patients with more severe disease, and therefore, might not be truly reflective of actual disease demographics. This hospital is also a Government Services Hospital with a restricted clientele subset who are not necessarily a true representation of the general population. We have also used a severity score that was developed for a pediatric population, and hence, our findings would need to be validated in a bigger study. Since Ascites can itself be associated with passive GB wall thickening and pericholecystic fluid, it is possible that in some patients with Ascites, the findings of GB wall edema/pericholecystic fluid could have been erroneously noted as an independent finding and not attributed to Ascites per se. In addition to the limitations due to the inclusion/exclusion criteria used in our study protocol, there is also the inherent limitation of ultrasound being a user-dependent modality.

Conclusion

Ultrasound is a time-tested diagnostic modality with an unmatched safety profile that can reliably detect the presence of organomegaly and plasma leakage. In our study, we have established that there is a definite role of ultrasound in patients with nonsevere dengue wherein it can be used to identify that subset of DF patients who are more likely to deteriorate clinically, and thus, play an important role in the secondary prevention of severe dengue. The widespread availability of ultrasound machines provides us with this unique opportunity and ability to screen dengue patients at the initial point of care, thereby reducing morbidity and mortality associated with severe dengue. Early ultrasound examination in dengue patients can therefore be used to guide treating physicians toward a more proactive approach and timely management of these “at-risk” patients while at the same time ensuring optimal resource utilization.

Disclosure of competing interest

The authors have none to declare.