Open Access Original Article

Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

DIAGNOSTIC ACCURACY OF X-RAY CHEST IN INTERSTITIAL LUNG DISEASE AS

CONFIRMED BY HIGH RESOLUTION COMPUTED TOMOGRAPHY (HRCT) CHEST
Farah Afzal, Saleem Raza, Mobeen Shafique*
Combined Military Hospital/National University of Medical Sciences (NUMS) Rawalpindi Pakistan, *Combined Military Hospital
Gujranwala/National University of Medical Sciences (NUMS) Pakistan
ABSTRACT
Objective: To determine the diagnostic accuracy of x-ray chest in interstitial lung disease as confirmed by high
resolution computed tomography (HRCT) chest.
Study Design: A cross-sectional validational study.
Place and Duration of Study: Department of Diagnostic Radiology, Combined Military Hospital Rawalpindi,
from Oct 2013 to Apr 2014.
Material and Method: A total of 137 patients with clinical suspicion of interstitial lung disease (ILD) aged 20-50
years of both genders were included in the study. Patients with h/o previous histopathological diagnosis, already
taking treatment and pregnant females were excluded. All the patients had chest x-ray and then HRCT. The x-ray
and HRCT findings were recorded as presence or absence of the ILD.
Results: Mean age was 40.21 ± 4.29 years. Out of 137 patients, 79 (57.66%) were males and 58 (42.34%) were
females with male to female ratio of 1.36:1. Chest x-ray detected ILD in 80 (58.39%) patients, out of which, 72 (true
positive) had ILD and 8 (false positive) had no ILD on HRCT. Overall sensitivity, specificity, positive predictive
value, negative predictive value and diagnostic accuracy of chest x-ray in diagnosing ILD was 80.0%, 82.98%,
90.0%, 68.42% and 81.02% respectively.
Conclusion: This study concluded that chest x-ray is simple, non-invasive, economical and readily available
alternative to HRCT with an acceptable diagnostic accuracy of 81% in the diagnosis of ILD.
Keywords: Chest x-ray, HRCT, Interstitial lung disease, Non-invasive.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

INTRODUCTION history, exposures, and drug history forming the

Interstitial lung diseases (ILDs) are a diverse clinical context3. More than 60% of cases are
group of pulmonary diseases which differ greatly idiopathic4. The remaining cases can be due to
in causes, presentation, clinical pressentation, various parenchymal lung disorders with a wide
imaging, pathologic features and natural history. range of occupational or environmental factors
ILD was once uncommon but now due to various including smoking, drugs, radiation therapy,
environmental factors it is quite common pulmonary aspiration, neoplasms and systemic
according to epidemiologic investigations1. diseases with pulmonary involvement2,4.
Diagnosis of ILD can be delayed when Although ILD primarily affects adults, it
clinicians neglect the initial symtoms or attribute may be seen in children. Few ILDs are seen in
them to more commonly encountered lung young age e.g., sarcoidosis, pulmonary
diseases such as chronic obstructive pulmonary Langerhan’s cell histiocytosis, and autoimmune
airway disease (COAD). In a study, prevalence of pulmonary disorders, whereas idiopathic
ILD was found to be 76% in all patients pulmonary fibrosis (IPF) usually presents
presenting with chronic pulmonary disease2. between 40 to 70 years of age. Familial IPF in
Diagnosis of ILD requires thorough workout of which two or more first-degree relatives are
clinical presentation, life style, occupational involved, the onset of fibrosis appears is in
comparatively younger age. Incidence and
Correspondence: Dr Farah Afzal, Radiology Department CMH mortality of interstitial lung disease are directly
Rawalpindi Pakistan (Email: [email protected])
Received: 08 Feb 2016; revised received: 30 Jun 2016; accepted: 04 Jul
proportional with age5,6.
2016

593
Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

Chest x-ray is the first line investigation of from October 2013 to April 2014. Sample size was
any pulmonary disease. During the recent years, 137 cases with 95% confidence level, 10.5%
many advances have been made in interpretation margin of error for sensitivity and 11% for
of plain chest radiographs so that interstitial lung specificity, expected prevalence of ILD as 76%
disease is diagnosed with more accuracy. and taking sensitivity and specificity of x-ray
However, due to limited spatial resolution and chest in interstitial lung disease as 47% and 82%
superimposition of various structures confident respectively.
diagnosis of ILD is hampered by inbuilt After approval from ethical review
limitations of the chest radiographs7. Chest committee, 137 patients admitted in other
radiograph (CXR) is a widely available, in departments of CMH, Rawalpindi and referred to
expensive, non invasive investigation but its the radiology department for HRCT scan with
sensitivity, specificity and accuracy has been clinical suspicion of ILD between the ages of
shown to be 47%, 82% and 77% respectively in 20-50 years of both genders were enrolled in the
diagnosing interstitial lung disease by Padley study through non-probability convenience
SPG et al8. sampling technique. After taking written,
The advent of high resolution computed informed consent and relevant history including
tomography (HRCT) has advanced our ability to h/o occupation, any drug intake and smoking,
detect and characterize ILDs. Structural changes patients had their CXR on Shimadzu x-ray
in the lungs can often be detected in patients with system run at 800mA, prior to HRCT
a normal chest radiograph. HRCT is indicated examination.
during the initial workup of a patient with strong Assessment of CXR and HRCT was double
clinical suspicion or CXR features suggestive of blinded. Each CXR was assessed by consultant
ILD9. It has higher sensitivity than the plain radiologist for presence or absence of interstitial
radiograph in identifying ILD and specific lung disease. Then the patients underwent HRCT
imaging patterns can help in characterization of imaging on whole body by Toshiba Astieon 16-
disease as well as assessment of disease activity9. slice CT scan system. Each HRCT was reviewed
As there is discrepancy between local and by consultant radiologist (at least 10 years of
international data so this study was planned to experience) unaware of the x-ray findings. Data
generate more local data. This study helped us to collected was recorded on a specially designed
determine the diagnostic accuracy of CXR in ILD proforma containing two parts. The 1st part
in our population. Although in previous studies included the demographic features of the patients
it was found to be very low but we have while 2nd part included the study variables.
determined its diagnostic accuracy in our local Collected data was analyzed through
population where most of the people belong to computer software SPSS version 16.0. Mean and
poor socioeconomic status and they would not be standard deviation was calculated for
able to afford HRCT. These patients could be quantitative variables i.e. age and duration of
provided with simple, non-invasive, economical disease. Frequency and percentage were
and readily available alternative to HRCT with calculated for qualitative variables i.e. gender,
the benefit of significant reduction in patient CXR and HRCT results i.e. presence or absence of
radiation dose and timely diagnosis and ILD. A 2×2 contingency table was used to
treatment. calculate sensitivity, specificity, positive
MATERIAL AND METHODS predictive value, negative predictive value and
This cross sectional vocational study was diagnostic accuracy of x-ray chest in diagnosing
conducted at Department of Radiology at interstitial lung disease taking HRCT as a gold
Combined Military Hospital (CMH), Rawalpindi standard.

594
Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

RESULTS limited spatial resolution and superimposition of

Total number of subjects was 137. Age range various structures confident diagnosis of ILD is
in this study was from 20-50 years with mean age hampered by inbuilt limitations of the chest
of 40.21 ± 4.29 years. Majority of the patients radiographs. Though there are quite a few studies
(48.17%) were between 41 to 50 years of age. A comparing plain chest radiography and HRCT in
total of 21.17% were between 21 to 30 years and specific interstitial diseases like systemic
30.66% between 31 to 40 years. Out of these sclerosis, ankylosing spondylitis and silicosis, the
59.85% had symptoms for 1 to 3 months while whole gamut of interstitial lung diseases has not
40.15% had it for more than 3 months. Out of 137 been studied as a group frequently enough10.
patients, 79 (57.66%) were males and 58 (42.34%) A study conducted on progressive systemic
were females with male to female ratio of 1.36:1. sclerosis inferred that HRCT was more sensitive
All the patients were subjected to CXR and ILD than chest radiography when assessing minimal
was detected in 80 (58.39%) patients. HRCT interstitial lung involvement in this group of
confirmed interstitial lung disease in 90 (65.69%) patients11. HRCT was reported to be superior to
cases where as 47 (34.31%) patients revealed no other imaging modalities with improved clarity

Table-I: Summary of results.

Positive result on X-Ray Negative result on X-Ray p-value
Positive result on HRCT 72 (TP)* 18 (FN)*** 0.213
Negative result on HRCT 08 (FP)** 39 (TN)****
*-TP=True positive **-FP=False positive ***-FN=False negative ****-TN=True negative
Table II: Stratification of male gender (n=79).
Positive result on X-Ray Negative result on X-Ray
Positive result on HRCT 39 (TP) 11 (FN)
Negative result on HRCT 03 (FP) 26 (TN)
Table III: Stratification of female gender (n=58).
Positive result on X-Ray Negative result on X-Ray
Positive result on HRCT 33 (TP) 07 (FN)
Negative result on HRCT 05 (FP) 13 (TN)

interstitial lung disease (table-I). In X-ray positive of parenchymal abnormalities, enabling a better
patients, 72 (true positive) had ILD and 8 (false and more confident characterization of
positive) had no ILD on HRCT. Among 57, X-ray pathologic processes in interstitial lung
negative patients, 18 (false negative) had ILD on diseases12. A study on patients of idiopathic
HRCT where as 39 (true negative) had no ILD on pulmonary fibrosis concluded that investigation
HRCT (p=0.213) (table-II). Overall sensitivity, by HRCT can lead to earlier detection and
specificity, positive predictive value, negative treatment as compared to conventional
predictive value and diagnostic accuracy of X-ray radiography13. It was also reported that
in diagnosing ILD was 80.0%, 82.98%, 90.0%, conventional radiography is not perfect in
68.42% and 81.02% respectively (table-III) detecting early parenchymal changes and small
(figure). opacities in patients of silicosis as compared to
DISCUSSION HRCT14.
During the recent years, many advances Plain CXR is considered a baseline
have been made in interpretation of plain chest investigation in patients with respiratory
radiographs so that interstitial lung disease is symptoms. Imaging features of early ILD are
diagnosed with more accuracy. However, due to often subtle and can be missed by less

595
Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

experienced radiologists. In cases of clinical male predominance was also reported in other
suspicion of ILD, HRCT chest scans should be previous publications16-18,22,23.
ordered, if available without specialist referral. CXR is a widely available, inexpensive, non
HRCT uses thin section CT images and high invasive investigation but it misses abnormalities
spatial frequency algorithm so it has superior in 20% of patients4. It can be normal in early
spatial and contrast resolution compared to disease and may be unable to characterize ILD. It
routine CT15. In our study, we have determined cannot be used as sole diagnostic investigation in
the diagnostic accuracy of CXR in diagnosing ILD such patients but is used as baseline investigation
taking HRCT as gold standard. in cases of clinical suspicion of ILD. Padley et al8
In this study, mean age was 40.21 ± 4.29 in his study has shown the sensitivity, specificity
years with majority of the patients (48.17%) and accuracy of chest x-ray as 47%, 82% and 77%
between 41 to 50 years of age. These findings are respectively in diagnosing ILD. In another study
very much comparable to the study of Gagiya et by Coutinho et al2, sensitivity, specificity, positive

Figure: Diagnostic accuracy of X-ray chest in diagnosing interstitial lung disease.

al16 and Shabbier et al17 who had found mean age and negative predictive values for the x-ray
of 43 & 41 years respectively. As majority of the diagnosis of interstitial lung disease was found to
ILDs have a long natural history that is why they be 70%, 90%, 62.3% and 93% respectively.
usually present in older adults in the sixth or Chest radiography usually demonstrates any
greater decade of life. Some types of ILDs eg of the following patterns: a reticular and netlike
sarcoidosis, connective-tissue disease–associated appearance of linear or curvilinear densities or
lung disease, and inherited forms of lung disease diffuse opacities with a predilection to the lower
present in younger adults18,19. Few ILDs show lobes. In advanced disease, the presence of a
sexual predilections. IPF mostly affects men (ratio coarse reticular pattern or multiple cystic or
of 1.5:1), whereas lymphangioleiomyomatosis honeycombed areas, and coarse reticular pattern
(LAM) and pulmonary tuberous sclerosis with translucencies are associated with poor
exclusively affect women20. The Bernalillo County prognosis. Pleural involvement is uncommon
study estimated an incidence of 31.5 cases per and its presence suggests another diagnosis24.
100,000/year in men and 26.1 cases per Tahbaz et al25 in a study on 49 patients found
100,000/year in women21. Similarly, in our study, only 3 patients (6.1%) with reticulonodular
79 (57.66%) were males and 58 (42.34%) were involvement on chest x-ray while HRCT showed
females with male to female ratio of 1.36:1. This pulmonary parenchymal involvement in 32 cases
(65.3%). He reported only 3 true positive, 29 false

596
Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

negative, 0 false positive and 17 true negative 2. Coutinho GF, Pancas R, Magalhaes E, Bernardo JE, Eugenio L,
Antunes MJ. Diagnostic value of surgical lung biopsy:
cases. On the whole he concluded very low comparison with clinical and radiological diagnosis. Eur J
sensitivity of chest x-ray of only 9.5% but in his Cardio-thoracic Surg 2008; 33: 781-85.
3. Santos J. Idiopathic pulmonary fibrosis: case report and clinical,
study specificity was found to be 100%.
histopathologic and epidemiologic attributes. Phil Heart Center J
CXR is a baseline invesitgaion of pulmonary 2008; 14(1): 67-75.
4. Sharma RP, Kaur G, Arora A, Khalasi Y, Vohra PV. High
diseases. But due to inbuilt limitation such as resolution PV, Interstitial lung disease in rheumatoid arthritis: a
poor spatial resolution and superimpostion of study of thirty cases. Chest 2006; 16: 835-9.
structures, subtle imaging features may not be 5. American Thoracic Society. Idiopathic pulmonary fibrosis:
diagnosis and treatment. International consensus statement.
visualized26. Grenier et al27 assessed the American Thoracic Society (ATS), and the European Respiratory
diagnostic value of chest radiography and high- Society (ERS). Am J Respir Crit Care Med 2000; 161: 646–64.
6. Olson AL, Swigris JJ, Lezotte DC, Norris JM, Wilson CG, Brown
resolution computed tomography (CT) in chronic
KK. Mortality from pulmonary fibrosis increased in the United
diffuse interstitial lung disease in 140 consecutive States from 1992 to 2003. Am J Respir Crit Care Med 2007; 176:
patients with diffuse infiltration of the lung 277–84.
7. Gross TJ, Hunninghake GW. Idiopathic pulmonary fibrosis. N
visible at radiography. Radiographs and CT scans Engl J Med 2001; 345: 517-25.
were separately read by three independent 8. Padley SPG, Hansell DM, Flower CDR, Jennings P. Comparative
observers without knowledge of clinical and accuracy of high resolution computed tomography and chest
radiography in the diagnosis of chronic diffuse infiltrative lung
pathologic data. First-choice diagnoses of all disease. Clin Radiol 1991; 44: 222-26.
three observers that were made with a high level 9. Nagapurkar AD, Gandage SG, Kulkarni DS, Marathe VN,
Kachewar SG. Role of high resolution computed tomography in
of confidence (probability, greater than or equal evaluation of pulmonary diseases. Intl J Health & Biomedical
to 75%) were more accurate with CT than with Res 2013; 1(3): 190-96.
radiography (p<0.001). The interobserver 10. Raghu G, Brown KK. Interstitial lung disease: Clinical
evaluation and keys to an accurate diagnosis. Clin Chest Med
agreement for the proposed diagnosis was 2004; 25: 409-19.
significantly better with high-resolution CT 11. Schurawitzki H, Stiglbauer R, Graninger W, Herold C,
(p<0.001). Polzleitner O, Burghuber OC, et al. Interstitial lung disease in
progressive systemic sclerosis: High resolution CT versus
CONCLUSION radiography. Radiology 1990; 176: 755-59.
12. Remy-Jardin M, Remy J, Deffontaines C, Duhamel A.
This study concluded that chest x-ray is Assessment of diffuse infiltrative lung disease: comparison of
simple, non-invasive, economical and readily conventional CT and high resolution CT. Radiology 1991; 181:
157-62.
available alternative to HRCT with an acceptable 13. Raniga S, Sharma S, Arora A, Khalasi Y, Vora PA. Utility of high
diagnostic accuracy of 81% in the diagnosis of resolution computed tomography (HRCT) in diagnosis and
ILD. management of Idiopathic Pulmonary Fibrosis- a study of 10
cases. Ind J Radiol Imag 2006; 16(4): 841-6.
RECOMMENDATION 14. Sun J, Weng D, Jin C, Yan B, Xu G, Jin B, et al. The value of
HRCT in the diagnostics of small opacities and complications of
We recommend that chest x-ray could be silicosis in mine machinery manufacturing workers, compared
used alone in our general practice for ILD to radiography. J Occup Health 2008; 50: 400-5.
15. Gulati M. Diagnostic assessment of patients with interstitial lung
diagnosis in our society where majority of people disease. Prim Care Respir J 2011; 20(2): 120-7.
belong to poor socioeconomic status and can’t 16. Gagiya AK, Suthar HN, Bhagat GR. Clinical profile of interstitial
afford the HRCT which results in their lung diseases cases. National J Med Res 2012; 2(1): 2-4.
17. Shabbier G, Amin S, Ullah F, Rehman S, Khan S. Tomographic
incomplete treatment as well as increase in their scan in diagnosis of interstitial lung diseases in local population.
morbidity and mortality. J Postgrad Med Inst 2012; 26(2): 149-52.
18. Colby TV, Swensen SJ. Anatomic distribution and histo-
CONFLICT OF INTEREST pathologic patterns in diffuse lung disease: correlation with
HRCT. J Thorac Imaging. Spring 1996;11:163]. J Thorac Imaging
This study has no conflict of interest to 1996; 11: 1-26.
declare by any author. 19. Ryu JH, Olson EJ, Midthun DE, Swensen SJ. Diagnostic
approach to the patient with diffuse lung disease. Mayo Clin
REFERENCES Proc 2002; 77: 1221-27.
1. Ryu JH, Daniels CE, Hartman TE, Yi ES. Diagnosis of interstitial 20. Ryu JH, Colby TV, Hartman TE. Idiopathic pulmonary fibrosis:
lung diseases. Mayo Clin Proc 2007; 82: 976-86. current concepts. Mayo Clin Proc 1998; 73: 1085-101.

597
Diagnostic Accuracy of X-Ray Chest in Interstitial Lung Disease Pak Armed Forces Med J 2017; 67 (4): 593-98

21. Gross TJ, Hunninghake GW. Idiopathic pulmonary fibrosis. N Fard S, Halvani A et al. HRCT in early diagnosis of asbestos
Engl J Med 2001; 345: 517-25. related pulmonary disease 2004; 3(12): 35-42.
22. Mahasur A. Diffuse fibrosing alveolitis in Indian experience. 26. Christopher B, Griffin CB, Steven L, Primack, MD. High
Lung India 1983; I: 171-9. resolution CT: Normal anatomy Techniques, and pitfalls of the
23. Turner M. Cryptogenic fibrosing alveolitis-clinical features and lung, Radiol Clin North Am 2001; 39: 1073-90.
their influence on survival. Thorax 1980; 35: 171-80. 27. Grenier P, Valeyre D, Cluzel P, Brauner MW, Lenoir S, Chastang
24. Reynolds HY. Diagnostic and management strategies for diffuse C. Chronic diffuse interstitial lung disease: diagnostic value of
interstitial lung disease. Chest 1998; 113; 192-202. chest radiography and high-resolution CT. Radiology 1991;
25. Tahbaz MO, Emam-Hadi MA, Nooraei M, Karam MB, Zahiri- 179(1): 123-32.

598