| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Symptomatic dengue virus infection can present with a wide range of clinical manifestations, from mild fever to life-threatening shock syndrome.5–9 Dengue is often classified into 2 nosological entities: dengue fever (DF) and dengue hemorrhagic fever (DHF). DHF is the most serious manifestation. Clinical presentations, laboratory findings, and serologic analysis of DHF have been widely reported; however, these do not include studies of chest radiographic (CXR) presentations.10–18
Many outbreaks of dengue virus infection have occurred in Taiwan. A major outbreak of DF occurred in the southern part of Taiwan from 1987 to 1988, and a small number of clustered cases of DF have been reported from this area since then. A large DF epidemic caused by the DEN-2 virus occurred from June to December 2002 in the same region where a DF outbreak caused by the DEN-1 virus occurred in 1987 and 1988.19,20 More than 5,000 cases of symptomatic DF infection occurred during this outbreak, about 700 of whom were admitted to the Kaohsiung Chang Gung Memorial Hospital, the largest tertiary-care medical center in southern Taiwan. We retrospectively reviewed the medical records of these DF patients.
The aim of the present study was to evaluate CXR presentations of DHF patients, identify relative risk factors in their abnormal CXR findings, evaluate the relationship between clinical course and abnormal CXR finding, and determine if CXR could be a useful prognostic tool in clinical management of DHF.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4-fold increase in dengue-specific hemagglutination inhibition titers in convalescent serum.21 A diagnosis of DHF was made in patients with laboratory confirmed DF in accordance with the World Health Organization definition (i.e., presence of fever, hemorrhagia, thrombocytopenia, and overt evidence of plasma leakage resulting from increased vascular permeability).22
There were 363 DHF patients with CXR examination. A total of 468 CXRs (posterior–anterior views) were performed; of these, 278 patients had only one CXR performed, 68 had two, 14 had three, and 3 had four.
To determine the clinical characteristics and risk factors of CXR presentation, the 468 CXRs were reviewed and interpreted by 3 experienced chest physicians and were divided into 2 groups: normal and abnormal CXR groups. The abnormal CXRs were reviewed for the presence of radiographic findings, such as infiltration only, small pleural effusion (PE) (if effusion was < 2 intercostal spaces), moderate PE (if effusion was between 2 and 4 intercostal spaces), and massive PE (if effusion was > 4 intercostal spaces), and if the lesions were unilateral or bilateral.
To evaluate the evolution of the CXR presentations, comparison of 2 different CXRs obtained for the same patient was also interpreted by 3 experienced chest physicians and were divided into 3 conditions: progressive change, no change, and improved.
D-day was defined as number of days after onset of fever at the time when the CXR was obtained. Levels of white blood cells (WBC), hematocrit, platelets, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and albumin (ALB) as well as prothrombin time (PT) and activated partial thromboplastin time (APTT) were checked on the same day as the CXRs were obtained, if available. Lastly, data regarding microorganisms from blood culture, as well as the duration of admission, were obtained from the medical records.
PE was defined as exudative if at least 1 of the following 3 criteria was present: 1) pleural fluid protein-to-serum protein ratio > 0.5; 2) pleural fluid lactic dehydrogenase (LDH)-to-serum LDH ratio > 0.6; and 3) pleural fluid LDH > two-thirds of the upper limit of normal for serum LDH. If none of the above criteria was present, the pleural effusion was defined as transudative.23
To evaluate the relationship between CXR findings and clinical course, the 363 DHF patients were divided into 2 groups: the normal CXR group, whose CXRs were without any abnormal findings, and the abnormal CXR group, who had at least one abnormal CXR finding. Also, patients in the abnormal CXR group were subdivided into the 1) infiltration group, whose CXRs were infiltration only without any PE; 2) small PE group, who had at least one CXR with small PE but without any moderate or massive PE; 3) moderate PE group, who had at least one CXR with moderate PE but without any massive PE; and 4) massive PE group, who had at least one CXR with massive PE.
Related clinical course (including renal insufficiency, liver function impairment, upper gastrointestinal [UGI] bleeding, acute respiratory failure, combination bacterial infection, and duration of admission) and outcome (survival or death) were evaluated. Renal insufficiency was defined as increased blood urea nitrogen and serum creatinine levels, while liver function impairment was defined as increased liver enzymes (AST and ALT). UGI bleeding was defined as the presence of hematemesis or tarry stool passage, and acute respiratory failure was defined when a patient was immediately intubated and required mechanical ventilation due to failure to respond to 40% oxygen via nasal cannula. Combination bacterial infection was defined as when bacterial microorganisms grew in culture. Other forms of bleeding—including oral, gum, nasal, conjunctival, and retinal bleeding as well as hemoptysis and vaginal bleeding—were also recorded.
The sequence relating the clinical course to the day of abnormal CXR was also evaluated and divided into 3 relationships: after abnormal CXR finding, before abnormal CXR finding, and concurrent with abnormal CXR, in the abnormal group of DHF patients.
Statistical analysis.
Data were collected and analyzed with software of SPSS for Windows version 13.0 (SPSS Inc., Chicago, IL). Much of the data provided was presented as summary statistics. Quantitative variables were characterized using mean ± SD. Statistical significance was determined by the Mann–Whitney U test for continuous variables and the 
2 test for dichotomous variables. P < 0.05 was considered statistically significant.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Distribution of CXRs.
The distribution of the chest X-rays on D-day is shown in Table 1
. Among the 468 CXRs, 256 (54.7%) had abnormal findings and 212 (45.3%) were within normal limits or unremarkable.
|
No CXRs had both infiltration and pleural effusion in our study. Infiltration only (109, 23.3%) and small PE (117, 25%) were the major findings among the abnormal CXRs. Infiltrations were predominately bilateral (49/109, 45.0%), while all PE lesions were predominately on the right side (70/117, 59.8% in small PE; 14/21, 66.7% in moderate PE; 8/9, 88.9% in massive PE; and 92/147, 62.6% in total). Echo-guided thoracentesis was performed in 7 DHF patients, all of whom had transudative effusion.
Evolution of CXRs.
A summary of CXR evolution is shown in Table 2. Most CXRs had changed progressively when compared with previous CXRs from Day 3 to Day 9. Improved CXR changes were observed starting on Day 6. No more progressive changes were observed after Day 14.
|
. Stratified analysis was done, and all of the CXRs were divided into 2 groups: Group 1 was obtained when D-day 
7, and Group 2 when D-day 8. There were no significant differences in Hct, PT, and ALT between the normal and abnormal CXRs in Group 1. However, there were significant differences in WBC, platelet, APTT, AST, and albumin levels in Group 1. On the other hand, there were no significant differences in most laboratory data variables in Group 2 except for a significant difference in albumin level.
|
Comparison of DHF patients between normal and abnormal CXR groups. Among the 363 DHF patients, 172 (47.2%) were in the normal group, 80 (22.0%) were in the infiltration group, 86 (23.7%) were in the small PE group, 20 (5.5%) were in the moderate PE group, and 5 (1.4%) were in the massive PE group.
A comparison of clinical course and outcome between normal and abnormal CXR groups of the 363 DHF patients is shown in Table 4 (part 1). There was no significant difference in mortality, but there were significant differences in renal insufficiency, liver function impairment, UGI bleeding, other bleeding, acute respiratory failure, concomitant bacterial infection, and duration of admission between the 2 groups.
|
(part 2) summarizes the comparison of the clinical course and outcomes between the normal and abnormal CXR groups of these 274 DHF patients. There were significant differences in renal insufficiency, liver function impairment, UGI bleeding, other bleeding, concomitant bacterial infection, and duration of admission between the two groups.
Sequence between abnormal CXR finding and clinical course.
Table 5 summarizes the sequence between abnormal CXR finding and clinical course in the DHF patients. More renal insufficiency and UGI bleeding were observed after abnormal CXR findings than before abnormal CXR findings. Almost all of the cases of acute respiratory failure were observed after abnormal CXR findings. Only slightly more liver function impairment and other bleeding were observed before abnormal CXR finding than after.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this retrospective study, CXR was not routine for DF (or DHF) patients when initially admitted. CXRs were most probably taken based on the judgment of the attending physician. Nonetheless, at least one CXR was performed in all of the DHF patients when the diagnosis of DHF was confirmed.
CXR presentations in DHF patients was seldom discussed in the past and had only been sporadically published in case reports.11–15 This has hindered the widespread dissemination of awareness of CXR presentation in DHF patients. As a result, clinicians have inevitably been less aware of the importance of CXR presentation when evaluating the clinical course of DHF patients.
In the few previous reports on this matter, PE was the most common finding.13–15 In our study, PE was also the most common finding (31.4% of total CXRs and 57.4% of abnormal CXRs), followed by infiltration only (23.3% of total CXRs and 42.6% of abnormal CXRs). Small PE was the predominate type among the CXRs with PE and in all abnormal CXRs. Furthermore, infiltration only and small PE were the major findings in DHF patients with abnormal CXR findings (Table 1).
In this study, 109 CXRs presented with infiltration only. Whether the infiltration is due to bacterial infection or not requires further evaluation. Among the 363 DHF patients, combination bacterial infection was observed in 19 patients (3 in the normal CXR group and 16 in the abnormal CXR group) (Table 4). Among the 19 DHF patients with combination bacterial infection, pneumonia (blood culture and sputum culture: Corynebacterium) was diagnosed in only 1. Except for this 1 patient, no evidence of pneumonia was observed in the other 362 DHF patients. We can conclude that pneumonia was not the cause of abnormal CXR findings for our DHF patients.
PE is characterized as either a transudate or an exudate. In our study, although diagnostic thoracentesis was performed in only 7 DHF patients, all were transudative PEs. Transudative PEs are largely due to imbalances in hydrostatic and oncotic pressures in the chest because transudates are an ultrafiltrate of plasma. Findings of a transudative PE generally imply that the pleural membranes are not diseased and that fluid accumulation is caused by systemic (nonlung, nonpleural) factors affecting the formation and absorption of pleural fluid.23,24
A cardinal feature of DHF that distinguishes it from classic DF is the increased vascular permeability (plasma leakage syndrome). Furthermore, plasma leakage syndrome is the most specific feature of DHF, which usually occurs between 3 and 7 days after onset of illness.22 In our study, the distributions of abnormal CXR were > 50% starting on the 4th day after onset of illness and > 60% from the 6th day (Table 1).
Plasma leakage also coincides with the most extreme depression of the platelet count and elevation of the liver function tests.22 In this study, aside from depression of the platelet count and elevation of liver function tests (AST), there were also significant differences in the elevation of WBC count and depression of ALB level in the group whose CXRs were obtained Day 7. Furthermore, most of the progressive changes were observed from Day 3 to Day 7 in DHF patients (Table 2). We could thus conclude that CXR presentation was consistent with the clinical course of plasma leakage syndrome.
No significant correlation in mortality was observed between patients with normal CXR and those with abnormal CXR, which may be due to the small sample size. However, significant correlations were observed in other clinical presentations (Table 4). The clinical course was also evaluated in DHF patients with chest radiography performed on or after the 3rd day after onset of illness (Table 5). Significant correlations were still observed in most clinical presentations except for acute respiratory failure and mortality, which might also be due to the small sample size. We could likewise reasonably conclude that abnormal CXR findings reflect the clinical course of DHF patients.
Table 1 shows the distribution of CXR presentation when D-day was after the 7th day. The distributions of abnormal CXR presentations were still > 50% between the Days 8 and 13. On the other hand, all of CXR presentations were normal after Day 14. Furthermore, improvement of CXR presentation was observed from Day 5 to Day 14 in DHF patients (Table 2), and no progressive change was observed after Day 14. Therefore, resolution of abnormal CXR findings occurred during the second week. Nonetheless, because of the small sample population, this requires further evaluation.
For DHF patients, plasma leakage is important and must be managed aggressively to prevent or reverse hypovolemic shock.25 Noncardiac pulmonary edema is a common complication of fluid replacement in patients with dengue and profound plasma leakage,25 but whether fluid replacement is a factor leading to or contributing to abnormal CXR findings in these dengue patients needs further evaluation. Among the 363 DHF patients, 213 (58.7%) had been treated with fluid replacement. Furthermore, only 26 had been treated with fluid replacement (at least 2,000 mL intravenous fluid supply) 24 hours before CXR was taken. Among these 26 dengue patients, 16 patients (16/212, 7.5%) were in the normal CXR group, 5 (5/109, 4.6%) in the infiltration CXR group, 3 (3/117, 2.6%) in the small PE CXR group, 2 (2/21, 9.5%) in the moderate PE CXR group, and none in massive PE CXR group. There was no significant correlation between DHF patients within different groups (P = 0.292), suggesting that fluid replacement was not a factor contributing to abnormal CXR finding in this study.
Abnormal CXR findings had led to changes in the clinical management in 17 DHF patients. Among these, the CXRs of 4 patients presented with infiltration, 9 with small PE, 1 with moderate PE, and 3 with massive PE. Regarding changes in clinical management, antibiotics in 10 patients were changed, 2 started using new antibiotics, 2 were started on diuretics, and 3 had intravenous fluid discontinued.
Usually, DF (or DHF) is suspected on the basis of clinical manifestations and laboratory results and a diagnosis is confirmed by serologic detection of the virus or antiviral antibodies. But results are often obtained too late to be of clinical benefit. Therefore, additional diagnostic tools for evaluating patients with suspected DF are needed. In this study, CXR is useful in guiding the clinical management of DHF patients and its presentations correlate with the clinical course and outcome. CXR may therefore be a complementary tool for evaluating DHF patients.
Because PE is the major CXR presentation in DHF patients, chest ultrasound can also be considered as another complementary tool to evaluate the clinical course. There have been only a few published reports of chest ultrasound studies in dengue patients.13,26 Further evaluation regarding the applicability of chest ultrasound in dengue patients and the difference between chest ultrasound and CXR in dengue patients is needed.
There are some limitations in our study. First, it was conducted at a single medical center, and the patient population may be biased by patient selection and referral patterns. Second, this study was a retrospective survey, which not only resulted in incomplete data but also did not control the laboratory examinations taken of all DF patients. Nonetheless, the study provides relatively rare information about CXR findings of DHF patients. Follow-up investigation or prospective evaluation should be done.
In summary, DHF is the most serious manifestation of dengue virus infection. The cardinal features of DHF distinguishing it from classic DF include increased vascular permeability (plasma leakage syndrome) marked by thrombocytopenia associated with a bleeding tendency and hepatomegaly and/or abnormal liver function. Infiltration only and small PE are the major abnormal CXR in DHF patients, which appear on or after the 3rd day after the onset of illness. Progressive change during first week and improvement during second week are observed in those with abnormal CXRs. There is a significant correlation among laboratory findings, clinical course, and outcome. CXR may be a complementary tool to evaluate the clinical course of DHF and CXR examination should be taken during the first week after the onset of illness for dengue patients.
Received July 25, 2006. Accepted for publication April 18, 2007.
* Address correspondence to Meng-Chih Lin, Division of Pulmonary & Critical Care Medicine, Department of Internal Medicine, Chang Gung Memorial Hospital—Kaohsiung Medical Center, Chang Gung University College of Medicine, 123 Dabi Road, Niaosung Shiang, Kaohsiung, Taiwan. E-mail: mengchih{at}adm.cgmh.org.tw 
Authors’ addresses: Chin-Chou Wang, Chao-Chien Wu, An-Shen Lin, Shih-Feng Liu, Yu-Hsiu Chung, Mao-Chang Su, and Meng-Chih Lin, Division of Pulmonary & Critical Care Medicine, Department of Internal Medicine, Chang Gung Memorial Hospital—Kaohsiung Medical Center, Chang Gung University College of Medicine, 123 Dabi Road, Niaosung Shiang, Kaohsiung, Taiwan. Jien-Wei Liu and Ing-Kit Lee, Division of Infectious Disease, Department of Internal Medicine, Chang Gung Memorial Hospital—Kaohsiung Medical Center, Chang Gung University College of Medicine, 123 Dabi Road, Niaosung Shiang, Kaohsiung, Taiwan.
Reprint requests: Meng-Chih Lin, Division of Pulmonary & Critical Care Medicine, Department of Internal Medicine, Chang Gung Memorial Hospital—Kaohsiung Medical Center, Chang Gung University College of Medicine, 123 Dabi Road, Niaosung Shiang, Kaohsi-ung, Taiwan, or Department of Respiratory Care, Chang Gung Institute of Technology, Chia-Yi, Taiwan, 123 Dabi Road, Niaosung Shiang, Kaohsiung, Taiwan, Telephone: +886-7-7317123, ext. 8199, Fax: +886-7-7322402, E-mail: mengchih{at}adm.cgmh.org.tw.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. E. E. Martina, P. Koraka, and A. D. M. E. Osterhaus Dengue Virus Pathogenesis: an Integrated View Clin. Microbiol. Rev., October 1, 2009; 22(4): 564 - 581. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
