Vol.4, No.

1, 15-19 (2012)
http://dx.doi.org/10.4236/health.2012.41004

Health

Pleural effusion: Presentation, causes and treatment

outcome in a resource limited area, Ethiopia
Mekonnen Desalew1, Amare Amanuel2, Alemu Addis3, Hurissa Zewdu1, Ali Jemal4
1

Department of Internal Medicine, University of Gondar Ethiopia, Gondar, Ethiopia; *Corresponding Author: desalewm@yahoo.com
Department of Neurology, Addis Ababa University, Addis Ababa, Ethiopia, Ethiopia
3
CU-ICAP, Addis Ababa, Ethiopia
4
Department of Microbiology, Immunology and Parasitology, University of Gondar, Gondar, Ethiopia
2

Received 23 October 2011; revised 15 November 2011; accepted 12 December 2011

ABSTRACT
Background: Pleural effusion is a common clinical problem with different causes. Objective:
To demonstrate clinical features and outcome of
pleural effusion. Methods: Prospective descriptive study was conducted involving 110 patients
with pleural effusion admitted to a resource limited hospital in Ethiopia. Results: Males and
females were almost equally represented. Cough,
fever and weight loss were prominent presenting symptoms accounting 90, 77.3 and 77.3 percent respectively. Right side effusion was the
common presentation 50 (45.5%). Forty (37.4%)
patients had HIV infection among 107 tested.
Tuberculosis was the commonest cause 78 (70.9%)
followed by parapneumonic effusion 36 (32.7%)
and empyema 27 (24.5%). Malignant pleural effusion was detected only in one patient. Eighty
one (73.6%) improved from their illness and 7
(6.4%) died. Lympocytic pleural effusion found
to be associated with tuberculosis (OR = 3.942
(1.527 - 10.179), P = 0.005. There were no associations between HIV infection, anemia, elevated
ESR and side of pleural effusion with tuberculosis. Conclusion: Tuberculosis was the leading
cause of pleural effusion in our setup even
though etiologic diagnosis was difficult. Strengthening the laboratory and pathology services in
the area is strongly recommended.
Keywords: Pleural Effusion; Tuberculosis;
Parapneumonic Effusion; Empyema; Pleura

1. INTRODUCTION
Pleural effusion develops when more fluid enters the
pleural space than is removed. It is a common clinical
problem with different possible causes. It can be due to
local or systemic or infectious or non-infectious causes.
Copyright 2012 SciRes.

Patients with pleural effusion have higher morbidity and

mortality than patients without pleural effusion [1].
A definitive diagnosis, as is provided by the finding of
malignant cells or specific organism in the pleural fluid
can be established in approximately 25% of cases. However, a presumptive diagnosis, based on the prethoracentesis clinical impression and pleural effusion analysis,
clinical decision-making information can be gained in
over 90% of patients [2]. Diagnosis of the cause is often
difficult. The relative frequency of causes of pleural effusion is known to vary in different parts of the world [3].
However, in developing nations infectionsespecially
tuberculosis and parapneumonic effusions are more prevalent [4].
Pleural effusions are frequently encountered in hospitalized HIV/AIDS patients [5]. Among patients with tuberculosis pleural involvement appears to be more common when co-infection with HIV is present [6].
There were no similar studies published in the area
and this study was conducted to fill this gap. This prospective descriptive study demonstrates the presentation,
causes and treatment outcome of patients presented to
Gondar University Hospital with pleural effusion.

2. PATIENTS AND METHODS

The prospective study was conducted at Gondar University Hospital. It is a tertiary teaching hospital with
400 beds. The study include all patients with pleural effusion admitted to the medical wards of Gondar University Hospital from Jan 2006 to Nov 2007 who gave an
informed written consent and fulfill the inclusion criteria.

2.1. Inclusion Criteria

Adult patients (age 13 years) presented to Gondar
University Hospital with pleural effusion who required
diagnostic thoracentesis were included to the study.

2.2. Ethical Clearance

Research protocol was approved by the Research and
OPEN ACCESS

16

M. Desalew et al. / Health 4 (2012) 15-19

Publication Office of the University of Gondar. Permission to proceed the study was obtained from the medical
director of Gondar University Hospital. Confidentiality
was maintained while using the patients data. Informed
written consent was obtained from each patient and the
information was analyzed anonymously.

Table 1. Age and Sex distribution of 110 patients with pleural

effusion admitted to Gondar University Hospital from Jan. 2006
to Nov. 2007.
Sex
Age Group

Male
(n = 54)

Female
(n = 56)

15 - 20

11

Total Number
(n = 110)

Total Percent

18

16.4

2.3. Data Collection

21 - 30

13

23

36

32.7

The following data were collected using data sheet for

each patient who qualifies for pleural effusion: sociodemographic, clinical findings, underlying disease, duration of symptoms, duration of hospital stay, Chest X ray,
hematocrit, erythrocyte sedimentation rate, pleural fluid
analysis (cell count differential count, cytology and microbiologic analysis) and HIV serology. We also collected
data on treatments given including antibiotics, anti tuberculosis and tube thoracostomy. We also followed patients during their stay at the hospital and outcome variables and complications were noted. A pretested structured questionnaire was used to collect data. The questionnaire was prepared by the investigators for this study
purpose. It was pretested and several revisions were made
before its application.
Outcome variables compiled as improved were patients who got clinical improvement as a response to
treatment and same were patients who were discharged
against medical advice, disappeared from the hospital
and those referred to other center for further evaluation
and management. Those with deterioration were worsening clinical condition despite treatment. Deaths were
ascribed to hospital deaths while patients were on treatment.

31 - 40

10

18

16.4

41 - 50

11

17

15.5

51 - 60

8.2

61 - 70

7.3

71 - 78

3.6

2.4. Statistical Analysis

SPSS 12.1 computer software was used for data entry
and analysis. P value less than 0.05 was used for statistical significance. Associations were assessed using binary
and multinomial logistic regression.

3. RESULTS
In this prospective descriptive study, we assessed 110
patients with pleural effusion admitted to Gondar University Hospital from Jan 2006 to Nov 2007.
The sociodemographic characteristics are described in
Table 1. The number of male and female patients in the
study was comparable (54 and 56 respectively). The mean
age was 37 years with the range of 63 (15 - 78). Seventy
two (65.5%) patients were age under 40 years.
The mean duration of the clinical symptoms prior to
presentation was 66 days with a range of 362 days (3 to
365). The mean hospital stay was 25 days with a range of
100 days (3 to 103).
Copyright 2012 SciRes.

The presenting symptoms were represented with Table

2.
Twenty one (19.1%) patients had history of previous
treatment for tuberculosis. Previous history of pleural
effusion was reported in 4 (3.6%) patients.
On physical examination, body swelling was traced in
43 (39.1%) of patients which comprised 39 patients had
leg edema, 22 patients had ascites and 4 patients had peri
orbital edema. Clinically significant lymphadenopathy
was found in 8 (7.3%) of patients.
Comorbidity with underlying chronic illness was
found in 23 (20.9%) patients comprising 3 hypertensive,
8 congestive heart failure, 2 chronic liver disease, 2 renal
failure and 6 with other chronic illnesses. In this description, one patient might have more than one comorbid
chronic illness.
Right sided accounting 50 (45.5%) followed by left
sided 41 (37.3%) then bilateral 19 (17.3%).
Underlying HIV infection was found in 40 (37.4%)
patients out of 107 tested. Twelve patients were taking
antiretroviral therapy.
Chest X-ray findings of 110 patients reported were
described in Table 3.
Appearance of pleural fluid was documented in 100
patients and majority of them had straw colored 70
(66%), pus 24 (22.6%) and 6 (5.7%) had hemorrhagic
effusion.
On the pleural fluid analysis, lymphocytes were predominant in 57 (64.8%) and neutrophils predominate in
31 (35.2%) samples.
There were only three pleural fluid samples among 88
stained with gram stain showing one gram negative rods,
one gram positive cocci and one mixed bacteria of gram
positive cocci and gram negative rods. Culture of the pleural fluid among 88 samples grew bacteria in only three
of the cases which were S. aureus in two cases and S.
epidermidis in one case.
Pleural fluid cytology was done for 27 patients and
lymphocytic effusion reported in 16 patients followed by
OPEN ACCESS

M. Desalew et al. / Health 4 (2012) 15-19

Table 2. Presenting symptoms of 110 patients with pleural effusion admitted to Gondar University Hospital from Jan. 2006 to
Nov. 2007.
Symptoms

Number

Percent

Cough

99

90

Fever

85

77.3

Weight loss

85

77.3

Night sweating

80

72.7

Chest pain

77

70

Sputum production

73

66.4

Dyspnoea

64

58.2

Hemoptysis

16

14.5

Table 3. Chest X-ray findings in 110 patients with pleural effusion admitted to Gondar University Hospital from Jan. 2006 to
Nov. 2007.

Chest X-ray findings

Number

Percent

Right pleural effusion

50

45.5

Left pleural effusion

41

37.3

Bilateral pleural effusion

19

17.3

Hydropneumothorax

3.6

Loculated effusion

2.7

Consolidation

3.6

Parenchymal infiltrates

2.7

One chest x-ray may have more than one feature.

empyema in 5 patients. The others were 4 patients with

reactive effusion and 2 were non conclusive.
Hematocrit level was determined for 97 patients and
64 (66%) had anemia. Erythrocyte Sedimentation Rate
(ESR) was elevated in 52 patients (80%) among the 65
patients determined.
Table 4 represents final diagnosis of studied patients
with pleural effusion.
Of the 78 patients with tuberculosis, 38 had right sided
pleural effusion followed by 29 with left sided and 11
patients had bilateral pleural effusion.
Among the 40 HIV/AIDS patients reported 30 had tuberculosis, 8 had empyema and 2 had para pneumonic
effusion.
Anti-tuberculosis and antibiotics were given to 79
(71.8%) and 72 (65.5%) patients respectively. Chest tube
drainage was done in 25 (22.7%) patients. Two patients
with empyema died before chest tube was inserted.
Out come at hospital discharge was improved in 81
(73.6%), same in 17 (15.5%), deteriorated with out death
in 5 (4.5%) and 7 (6.4%) died while they were in the
hospital.
Of the 7 deaths encountered 4 had tuberculosis, 2 had
empyema and 1 had constrictive pericarditis. Three of
the deceased had HIV infection. Causes of death were
Copyright 2012 SciRes.

17

Table 4. Final diagnosis of 110 patients with pleural effusion in

patients admitted to Gondar University Hospital from Jan. 2006
to Nov. 2007*.
Diagnosis

Number

Percent

78

70.9

36

32.7

27

24.5

Tuberculosis

Parapneumonic effusion
Empyema

Visceral Leishmaniasis

10

9.1

Heart failure

7.3

Lymphoma

3.6

Chronic liver disease

1.8

Renal failure

1.8

Metastasis

0.9

Constrictive pericarditis

0.9

One patient may have more than one diagnosis; The diagnosis of TB was
based upon clinical assessment, chest X-ray findings, suggestive pleural
fluid analysis and responses to anti-TB chemotherapy; Empyema was
diagnosed based on the finding of gross pus appearance or pleural fluid
culture or Gram stain showing organisms; Among 27 patients with empyema 14 were due to tuberculosis and 13 were due to complicated pneumonia.

respiratory failure related to underlying pleural effusion

in 4 cases, one with sepsis related multi organ failure,
one with complicated fluid electrolyte imbalance related
to abdominal surgery and one patient with constrictive
pericarditis died of heart failure precipitated by hospital
acquired pneumonia. Two of the cases complicated by
respiratory failure were having empyema and they died
before chest tube was inserted.

4. DISCUSSION
In our study the single most important cause of pleural
effusion was tuberculosis accounting 78 (70.9%) of patients as compared to other Ethiopian studies assessing
causes of empyema demonstrated as 59.1% at Tikur Anbessa Specialized teaching hospital in Addis Ababa [7]
and 56% at Gondar University Hospital [8]. Among empyema cases in our study 14/27 (51.9%) had tuberculosis
which is comparable to the above empyema studies in
Ethiopia. The second cause of pleural effusion is complicated pneumonia with parapneumic effusion accounting 32.7% as comparable to a study conducted in Ghana
(21.2%) [4]. Infectious causes of pleural effusion are still
common causes of pleural effusion in developing countries as described also in other studies [4,9], unlike data
from developed countries where primary and secondary
malignancies outnumbered [10].
The majority of patients were ages below 40 years old
(65.5%) which is consistent with the above mentioned
Ethiopian studies 81.6% in Addis Ababa [7] and 82% of
patients below the age of 50 years at Gondar [8] and also
can be explained by the general population of Ethiopia
OPEN ACCESS

18

M. Desalew et al. / Health 4 (2012) 15-19

[11].
The commonest symptoms of patients with pleural effusion were cough (90%), fever (77.3%), weight loss
(77.3%) and chest pain (70%) which is in agreement
with other studies (4, 9, 10).
The magnitude of HIV infection in pleural effusion
due to tuberculosis patients were 38.5% (30/78) as compared to 55.5% in a study conducted at Ghana [4] and
61.4% in a study at New York [12]. Among 40 HIV positive patients with pleural effusion 30 had tuberculosis.
We didnt get statistically significant associations between HIV infection and tuberculosis as a cause of pleural effusion in our study.
Lympocytic pleural effusion accounted in 64.8% patients and was statistically associated with tuberculosis
(Table 5).
We didnt get statistically significant associations between tuberculosis and side of effusion, anemia, elevated
ESR and HIV infection with individual binary and multinomial logistic regression (Table 5).
There was a diagnostic constraint observed in our
study especially on pleural fluid protein, glucose, lactate
dehydrogenase (LDH) which has vital role in classifying
the fluid as exudative and transudative. Fluid cytology
reports were not also complete due to lack of sustained
pathology service. Pleural biopsy was not also done in
any patient. Advanced tests like adenosine diaminase
(ADA) and interferon (IFN) were not available in the
country as a whole. Due to this incomplete laboratory
data, classification of pleural fluid to transudative and
exudative was unattempted in this study [8].
Pleural fluid microbiologic studies were not satisfactorily providing evidences. There were only three specimens gram stain demonstrate bacteria and three specimens grew bacteria on culture. Acid fast stain didnt demonstrate organism among 88 samples examined.

In conclusion, tuberculosis is the leading cause of

pleural effusion in our study and it is wise to consider it
in a resource limited setups. Empirical therapy with anti
tuberculosis chemotherapy in patients with undiagnosed
pleural effusion should be taken as a last resort [15]. Laboratory and pathology services have vital contribution
for etiologic diagnosis of pleural effusion and it is recommended to strengthen the service in the area [8].

5. ACKNOWLEDGEMENTS
We acknowledge the Research and Publications Office of the University of Gondar for funding this study. Many thanks go to all patients
who participated in the study. We are also equally grateful to all staff
members of the department of internal medicine of Gondar University
Hospital for their multidisciplinary support.

REFERENCES
[1]

Khan, J. and Ellis, M.E. (1997) Anaerobic bacterial pneumonia, lung abscess, pleural effusion/empyema. In: Ellis
M. E. Ed., Infectious Disease of the Respiratory Tract.
Cambridge University Press, Cambridge, 358-373.

[2]

Collins, T.R. and Sahn, S.A. (1987) Thoracocentesis: Clinical value, complications, technical problems, and patient
experience. Chest, 91, 817-822.
doi:10.1378/chest.91.6.817

[3]

Storey, D.D., Dines, D.E. and Coles, D.T. (1976) Pleural

effusion, a diagnostic dilemma. JAMA, 236, 2183-2186.
doi:10.1001/jama.1976.03270200021022

[4]

Afful, B., Murphy, S., Antunes, G. and Dudzevicius, V.

(2008) The characteristics and causes of pleural effusions
in Kumasi Ghana, a prospective study. Tropical Doctor,
38, 219-220. doi:10.1258/td.2007.070275

[5]

Joseph, J., Strange, C. and Sahn, S.A. (1993) Pleural

effusion in hospitalized patients with AIDS. Annals of Internal Medicine, 118, 856-859.

[6]

Frye, M.D., Pozsik, C.J., Sahn S.A. (1997) Tuberculous

pleurisy is more common in AIDS than non-AIDS patients with tuberculosis. Chest, 112, 393-397.
doi:10.1378/chest.112.2.393

[7]

Ali, A., Biluts, H. and Gulilat, D. (2003) Managment of

empyema thoracis in Tikur Anbessa Hospital. A three
year experience. East and Central Journal of Surgery, 1,
47-50.

[8]

Amare, A., Ayele, B. and Mekonnen, D. (2010) Empyema

thoracis: Presentation and treatment outcome at gondar
university hospital, Northwest Ethiopia. East and Central
Journal of Surgery, 1, 119-123.

[9]

Al-Alusi F.A. (1986) Pleural effusion in Iraq: A prospective study of 100 cases. Thorax, 41, 492-493.

Table 5. The characteristics of tuberculous and non tuberculous

effusions in patients admitted to Gondar University Hospital
from Jan. 2006 to Nov. 2007.
Tuberculosis

Non tuberculosis

(n = 78)

(n = 32)

Right sided effusion

38 (48.7%)

12 (37.5%)

Left sided effusion

29 (37.2%)

12 (37.5%)

Bilateral effusion

11 (14%)

8 (25%)

Characteristics

HIV positive*

30

10

Anemia*

48

16

Elevated ESR*

40

12

Lymphocytic effusion*

48

Complete data for HIV status, hematocrit level, ESR level and pleural fluid
analysis for lymphocytes were not available to demonstrate comparisons
percentages; Hematocrit level below 36 for females and 39 for males were
taken as anemia [13]; Lymphocytic effusion was defined as lymphocyte
percentage greater than 50% [14].

Copyright 2012 SciRes.

[10] Porcel, J.M. and Vives, M. (2003) Ethiology and pleural

fluid characterstics of large and massive effusions. Chest,
124, 978-983.
[11] Central Statstical Agency. (2005) Ethiopian demographic
and health survey. Population pyramid, Addis Ababa, p.
OPEN ACCESS

M. Desalew et al. / Health 4 (2012) 15-19

14. doi:10.1136/thx.41.6.492
[12] Relkin, F. and Aranda, C.P. (1994) Pleural tuberculosis
and HIV infection. Chest, 105, 1338-1341.
doi:10.1378/chest.124.3.978
[13] World Health Organization. (1968) Nutritional anemias:
Report of a WHO scientific group. Switzerland. World
Health Organization, Geneva.

Copyright 2012 SciRes.

19

[14] Porcel, J.M. and Light, R.W. (2006) Diagnostic approach

to pleural effusion in adults. American Family Physician,
73, 1211-1220. doi:10.1378/chest.105.5.1338
[15] Sane, M. et al. (2007) Empiric antituberculous therapy in
patients with unexplained exsudative pleural effusion. Is
it valid in Senegal? Revue De Pneumologie Clinique, 63,
247-250. doi:10.1016/S0761-8417(07)92647-4

OPEN ACCESS