+ MODEL

Asian Journal of Surgery (2015) xx, 1e5

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.e-asianjournalsurgery.com

ORIGINAL ARTICLE

A novel structural risk index for primary

spontaneous pneumothorax: Ankara
Numune Risk Index
Yucel Akkas a,*, Neslihan Gulay Peri a, Bulent Kocer a,
Tevfik Kaplan b, Aslihan Alhan c,d

a
Department of Thoracic Surgery, Ankara Numune Research and Training Hospital, Ankara, Turkey
b
Department of Thoracic Surgery, Ufuk University Faculty of Medicine, Ankara, Turkey
c
Department of Statistics, Ufuk University Faculty of Arts and Science, Ankara, Turkey

Received 30 July 2015; received in revised form 11 September 2015; accepted 16 September 2015

KEYWORDS Summary Background: In this study, we aimed to reveal a novel risk index as a structural risk
primary spontaneous marker for primary spontanoeus pneumothorax using body mass index and chest height, struc-
pneumothorax; tural risk factors for pneumothorax development.
Ankara Numune Risk Methods: Records of 86 cases admitted between February 2014 and January 2015 with or
Index; without primary spontaneous pneumothorax were analysed retrospectively. The patients were
body mass index; allocated to two groups as Group I and Group II. The patients were evaluated with regard to
chest height; age, gender, pneumothorax side, duration of hospital stay, treatment type, recurrence, chest
structural factors height and transverse diameter on posteroanterior chest graphy and body mass index. Body
mass index ratio per cm of chest height was calculated by dividing body mass index with chest
height. We named this risk index ratio which is defined first as ‘Ankara Numune Risk Index’.
Diagnostic value of Ankara Numune Risk Index value for prediction of primary spontaneous
pneumothorax development was analysed with Receiver Operating Characteristics curver.
Results: Of 86 patients, 69 (80.2%) were male and 17 (19.8%) were female. Each group was
composed of 43 (50%) patients. When Receiver Operating Characteristics curve analysis was
done for optimal limit value 0.74 of Ankara Numune Risk Index determined for prediction of
pneumothorax development risk, area under the curve was 0.925 (95% Cl, 0.872e0.977,
p < 0.001).

Conflicts of interest: The authors declare no conflicts of interest.

* Corresponding author. Department of Thoracic Surgery, Ankara Numune Research and Training Hospital, Ankara, Turkey.
E-mail address: y.akkas@yahoo.com (Y. Akkas).
d
Akkas Y: Conception and design of study, acquisition of data (clinical), data analysis and interpretation, drafting of manuscript and
critical revision, approval of final version of manuscript. Peri NG: Conception and design of study. Kocer B: Conception and design of study,
acquisition of data (clinical), approval of final version of manuscript. Kaplan T: Data analysis and interpretation. Alhan A: Data analysis and
interpretation.

http://dx.doi.org/10.1016/j.asjsur.2015.09.004
1015-9584/Copyright ª 2015, Asian Surgical Association. Published by Elsevier Taiwan LLC. All rights reserved.

Please cite this article in press as: Akkas Y, et al., A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune
Risk Index, Asian Journal of Surgery (2015), http://dx.doi.org/10.1016/j.asjsur.2015.09.004
 + MODEL
2 Y. Akkas et al.

Conclusions: Ankara Numune Risk Index is one of the structural risk factors for prediction of
primary spontaneous pneumothorax development however it is insufficient for determining
recurrence.
Copyright ª 2015, Asian Surgical Association. Published by Elsevier Taiwan LLC. All rights
reserved.

1. Introduction p value < 0.05 was accepted as statistically significant.

Diagnostic value of Ankara Numune Risk Index value for
Spontaneous pneumothorax is defined as air accumulation prediction of PSP development was analyzed with receiver
in the pleural space without the presence of trauma. operating characteristic (ROC) curve. Sensitivity, speci-
Spontaneous pneumothorax is classified as primary spon- ficity, and positive predictive and negative predictive
taneous pneumothorax (PSP) in the absence of an under- values of these limits were calculated in the presence of
lying pulmonary disease and secondary spontaneous valuable limit values. In assessment of the area under the
pneumothorax (SSP) in the presence of an underlying pul- curve, the diagnostic value of the test was interpreted as
monary disease.1,2 PSP is mostly seen as the result of statistically significant if the Type 1 error level was < 5%.
rupture of the subpleural bullae in the 2nd and 3rd de-
cades.3,4 The incidence of PSP is 7.4e18/100,000 in males
and 6/100,000 in females.5 There are many risk factors for 3. Results
the development of spontaneous pneumothorax, including
male sex, smoking, low body weight, tall height, low body Of 86 patients, 69 (80.2%) were male and 17 (19.8%) were
mass index (BMI), meteorological factors, and toxic female. Each group was composed of 43 patients (50%). In
metals.6e9 Group II, 27 patients (62.8%) were admitted with blunt
In this study, we aim to reveal a novel risk index as a trauma, 15 (34.9%) with penetrating trauma, and 1 (2.3%)
structural risk marker for PSP using BMI and chest height, with mediastinal mass lesion. Of the patients in Group I, 37
structural risk factors for pneumothorax development. (86%) were male and 6 (14%) were female. Mean age was
24.9  6.36 years (16e40 years) in Group I and 30.4  6.81
years (18e40 years) in Group II. Mean duration of hospital
2. Material and Methods stay was 5.98  1.74 (3e9) days in Group I and 5.58  6.08
(1e40) days in Group II. In Group I, while 11 (25.6%) pa-
Ethics committee approval was obtained from Ankara tients underwent video-assisted thoracoscopic surgery, 29
Numune Research and Training Hospital prior to the study (67.4%) underwent tube thoracostomy and 3 (7%) received
(date/number: 08.04.2015/E-15-468). Records of 86 cases observational therapy. In Group II, 17 patients (39.5%) un-
admitted between February 2014 and January 2015 with or derwent tube thoracostomy, 24 (55.8%) underwent obser-
without PSP were analyzed retrospectively. The patients vation, 1 (2.3%) underwent urgent thoracotomy, and 1
were allocated to two groups as Group I (admission due to (2.3%) received VATS. In Group I, pneumothorax was on the
PSP) and Group II (admission due to trauma or mediastinal right in 21 patients (48.8%), on the left in 21 (48.8%), and as
mass lesion). The patients were evaluated with regard to bilateral pneumothorax in 1 (2.3%). There was recurrence
age, sex, pneumothorax side, duration of hospital stay,
treatment type, recurrence, chest height (CH), and trans-
verse diameter on posteroanterior (PA) chest radiograph,
and BMI.
BMI was calculated as kg/m2 using weight and height
data in the files. PA radiographs of the chest were obtained
from the same distance on discharge of the patients and CH
was calculated by measuring the distance between apex
and costophrenic angle, transverse chest diameter was
calculated by measuring lateral distance of bilateral ribs 6,
and a picture archiving and communication system was
used for measurements (Figure 1). BMI ratio per cm of CH
was calculated by dividing BMI by CH. We named this value
the risk index ratio, which is defined as the Ankara Numune
Risk Index.
The SPSS, version 18.0 (SPSS Inc., Chicago, IL, USA) was
used for statistical analysis. Categorical variables were
evaluated with Chi-square test. Student t-test was used for
determining the difference between average values of the Figure 1 Chest height and transverse diameter measure-
variables between the two groups (Group I and Group II). A ment on posteroanterior chest radiography.

Please cite this article in press as: Akkas Y, et al., A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune
Risk Index, Asian Journal of Surgery (2015), http://dx.doi.org/10.1016/j.asjsur.2015.09.004
 + MODEL
Ankara Numune Risk Index 3

I, a statistically significant difference was found between

Table 1 Patient demographic characteristics.
recurrence and mean age (p Z 0.022) (p < 0.05). Recur-
Group I Group II All rence increased with age in Group I.
(n Z 43, (n Z 43, (n Z 86, The difference between Group I and Group II with regard
50%) 50%) 100%) to mean CH (p Z 0.000), mean BMI (p Z 0.000) and Ankara
Age 24.9  6.36 30.4  6.81 27.7  7.12 Numune Risk Index (p Z 0.000) was found to be statistically
(16e40) (18e40) (16e40) significant (p < 0.05). Although mean CH was higher in
Sex Group I compared with Group II, mean BMI and mean
Male 37 (86%) 32 (74.4%) 69 (80.2%) Ankara Numune Risk Index were lower in Group I compared
Female 6 (14%) 11 (25.6%) 17 (19.8%) with Group II. A statistically significant difference was not
Disease found between groups with regard to mean transverse
PSP 43 (100%) d 43 (50%) diameter (p Z 0.344) (p > 0.05) (Table 2).
Blunt trauma d 27 (62.8%) 27 (31.4%) When ROC curve analysis was carried out for optimal
Penetrating d 15 (34.9%) 15 (17.4%) limit value 0.74 of Ankara Numune Risk Index determined
trauma for prediction of pneumothorax development risk, the area
Mediastinal mass d 1 (2.3%) 1 (1.2%) under the curve was 0.925 (95% confidence interval [CI],
Hospital stay (d) 5.98  1.74 5.58  6.08 5.78  4.45 0.872e0.977, p < 0.001). 0.74 is the cut off value for
Treatment Ankara Numune Risk Index. Sensitivity and specificity values
Oxygen therapy 3 (7%) 24 (55.8%) 27 (31.4%) for Ankara Numune Risk Index were 0.9302; 95% CI
Chest tube 29 (67.4%) 17 (39.5%) 46 (53.5%) (0.8139e0.9760) and 0.7674; 95% CI (0.6226e0.8685) (k
VATS 11 (25.6%) 1 (2.3%) 12 (14%) 0.698; p < 0.001). Positive and negative predictive values
Thoracotomy d 1 (2.3%) 1 (1.1%) for limit value of Ankara Numune Risk Index were 0.8000;
Site NE 95% CI (0.6971e0.8756) and 0.9167; 95% CI (0.8318e0.9626)
Right 21 (48.8%) (Figure 2).
Left 21 (48.8%)
Bilateral 1 (2.3%)
Recurrence 11 (25.6%) NE
4. Discussion

Pneumothorax was first reported by Itard in 1803 and its

NE Z nonexamined; PSP Z primary spontaneous pneumo-
thorax; VATS Z video-assisted thoracoscopic surgery.
in 11 patients (25.6%) in Group I and 9 (81.8%) were male
(Table 1). Mean BMI was 19.9  1.66 (16.6e24.8) kg/m2 and
24.9  4.32 (16.7e35.1) kg/m2 in Group I and II, respec-
tively. Mean CH was 32.02  2.96 (26.73e37.34) and
28.65  3.08 (21.97e35.34) cm in Groups I and II, respec-
tively. Transverse thorax diameter was 28.90  2.20
(24.14e32.82) cm in Group I and 28.43  2.38
(23.71e33.63) cm in Group II. Mean BMI/CH was
0.627  0.08 (0.487e0.887) in Group I and 0.881  0.20
(0.593e1.498) in Group II (Table 2).
A statistically significant relationship could not be found
between Group I and Group II (p Z 0.279).
A statistically significant relationship was not found be-
tween recurrence and mean CH (p Z 0.767), mean trans-
verse diameter (p Z 0.311), mean BMI (p Z 0.404) and
Ankara Numune Risk Index (p Z 0.463) (p > 0.05). In Group
clinical findings were first described by Laennec in
1819.10 12 Kjaergard reported that PSP was different from
tuberculosis pneumothorax in 1931. Later, Devillier
explained the pathophysiology secondary to subpleural
bulla.13 Structural factors such as thorax measurements and
BMI also play a role in the development of subpleural
bullae.
Ample amounts of studies are available in the literature
investigating the role of structural factors in PSP develop-
ment. Peters et al14 reported that CH/transverse diameter
ratio was high in male patients with PSP, anteroposterior
diameter decreased in female patients, and the shape of
the thorax played a role in pneumothorax development
depending on stress distribution in the lungs. However,
Kawakami et al15 reported that stress distribution in the
lungs is not a major risk factor because they could not find a
difference between patient and controls with regard to
chest height in PSP patients. Casha et al16 reported that low
thoracic index leads to high apical stress and they increased
pleural pore and bulla development and although this stress

Table 2 Comparison of body mass index, chest height, transverse chest diameter and body mass index/chest height (Ankara
Numune Risk Index) of both groups.
Group I Group II p
BMI 19.9  1.66 (16.6e24.8) 24.9  4,32 (16.7e35.1) 0.000
CH 32.02  2.96 (26.73e37.34) 28.65  3.08 (21.97e35.34) 0.000
Transverse chest diameter 28.90  2.20 (24.14e32.82) 28.43  2.38 (23.71e33.63) 0.344
BMI/CH (Ankara Numune Risk Index) 0.627  0.08 (0.487e0.887) 0.881  0.20 (0.593e1.498) 0.000

BMI Z body mass index; CH Z chest height.

Please cite this article in press as: Akkas Y, et al., A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune
Risk Index, Asian Journal of Surgery (2015), http://dx.doi.org/10.1016/j.asjsur.2015.09.004
 + MODEL
4 Y. Akkas et al.
Figure 2 Receiver operating characteristic curve graph.
decreases in women and the elderly, it increases in the
young. Chang et al17 reported that CH and upper chest
width measurements increase in adolescents (12 years and
older) differently from the normal population and thereby
thorax measurements are a structural risk factor. Saita
et al18 reported that PSP patients had subclinical chest
deformity based on the tomography of thorax images, and
this condition was related to subpleural bleb formation
through leading to alveolar pressure heterogeneity.
BMI is the measurement of obesity found by dividing
body weight (kg) by the square of height (m2), which was
defined by Adolphe Quetelet in 1832.19 In the literature,
BMI of PSP patients is reportedly lower than that of
normal patients and this is a structural risk factor for
PSP.6e8 Case reports are available in the literature
reporting that PSP develops as the result of sudden
increased intra-alveolar pressure due to vomiting as lung
elasticity decreases as a result of impaired nutrition in
patients with a low BMI.20
In the literature, we can see that various radiological
sizes of the thorax are used for investigating the structural
properties of PSP. Peters et al14 investigated CH/transverse
diameter ratio and anteroposterior diameters on chest
graphs. Casha et al16 measured apical stress with apical
measurements of the lung using computed tomography of
the thorax and evaluated the relationship between them by
finding thoracic indices with Bellemare’s method. Saita
et al18 investigated the structural relationship by measuring
asymmetry ratios of anteroposterior and transverse di-
ameters on computed tomography of the thorax of PSP
patients. Chang et al17 evaluated CH by measuring from the
apex to the costodiaphragmatic sinus in the right lung on a
PA chest radiograph, and transverse diameter by measuring
from rib levels 2, 6, and 9 on PA chest radiograph. We
carried out our measurements in the manner described by
Chang et al,17 but we only used rib level 6 for transverse
measurement.
We did not use transverse diameter averages because we
did not find a statistically significant difference between
groups (p > 0.05). Instead, we used BMI and CH ratios that
showed a statistically significant difference between groups
(p < 0.05). In our study, BMI ratio per cm of CH was
Please cite this article in press as: Akkas Y, et al., A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune
Risk Index, Asian Journal of Surgery (2015), http://dx.doi.org/10.1016/j.asjsur.2015.09.004
calculated by dividing BMI/CH when determining the pneu-
mothorax risk index. This showed us that structural char-
acteristics in asthenic patients were directly proportional
with BMI and inversely proportional with CH. This formula,
which determines PSP risk, was first described by us and
named the Pneumothorax Risk Index (Ankara Numune Risk
Index). Ankara Numune Risk Index averages were found to
be significantly different between groups (p < 0.05). We
found a cut-off value of 0.74 for Ankara Numune Risk Index.
According to this information, we identified values < 0.74 as
a risk factor for PSP development.
In the literature, although some authors consider that
detection of bullous lesions on high-resolution computed
tomography obtained at the first exacerbation is a risk
factor for prediction of ipsilateral recurrence, some
others consider the opposite.12,21 We could not find a
significant difference between recurrence and CH, BMI,
and Ankara Numune Risk Index in Group I patients
(p > 0.05). Therefore, we considered that the Ankara
Numune Risk Index is insufficient for prediction of
recurrence.
Where to use the new structural index is an important
topic of discussion. This index is one of the primary struc-
tural risk factors that can predict the risk of spontaneous
pneumothorax. Therefore, it may be useful in the recruit-
ment medical examination used for professions with a high
risk of pneumothorax, such as pilots or divers.
5. Conclusion
There are many structural, congenital, and environmental
risk factors for PSP development. Ankara Numune Risk
Index is one of the structural risk factors for prediction of
PSP development. However, it is insufficient for deter-
mining recurrence; however, the Ankara Numune Risk Index
could give direction to the discovery of new structural
indices for the prediction of the recurrence of PSP in future
studies.
References
1. Fry WA, Paape K. Pneumothorax. In: Shields TW, LoCicero III J,
Ponn RB, Rusch VW, eds. General Thoracic Surgery. 6th ed.
Philadelphia: Lippincott Williams & Wilkins; 2005:794e805.
2. Beauchamp G, Ouellette D. Spontaneous pneumothorax and
pneumomediastinum. In: Pearson FG, Cooper JD, Deslauriers J,
eds. Thoracic Surgery. 2nd ed. Philadelphia: Churchill Living-
stone; 2002:1195e1213.
3. Sahin E, Kaptanoglu M, Nadir A, Akkas Y. Therapeutic ap-
proaches in spontaneous pneumothorax: a fourteen-year
experience. Turkiye Klinikleri Arch Lung. 2007;8:19e22.
4. Akkas Y, Katrancioglu O. A case of scoliosis and mental retar-
dation presenting with spontaneous pneumothorax. Cumhur-
iyet Medical J. 2012;34:103e106.
5. Sahn SA, Heffner JE. Spontaneous pneumothorax. N Eng J Med.
2000;342:868e874.
6. Ayed AK, Bazerbashi S, Ben-Nakhi M, et al. Risk factors of
spontaneous pneumothorax in Kuwait. Med Princ Pract. 2006;
15:338e342.
7. Celik B, Furtun K, Demir H, Yilmaz MA. Spontan pnömotorakslı
olgularımızın klinik özellikleri. Gulhane Tıp Dergisi. 2009;51:
71e74.
 + MODEL
Ankara Numune Risk Index 5

8. Chiu CY, Chen TP, Wang CJ, Tsai MH, Wong KS. Factors asso-
ciated with proceeding to surgical intervention and recurrence
of primary spontaneous pneumothorax in adolescent patients.
Eur J Pediatr. 2014;173:1483e1490.
9. Han S, Sakinci U, Kose SK, Yazkan R. The relationship between
aluminum and spontaneous pneumothorax; treatment, prog-
nosis, follow-up? Interact Cardiovasc Thorac Surg. 2004;3:79e82.
10. Itard JE, sur le pneumouthorax ou les congestions gazeuses qui
se forment dans la poitrine Thesis, Paris, 1803
11. Laennec RTH, Trait de l’auscultation mediate et des maladies
des poumons et du coeur. Tome Second Paris, 1819.
12. Celik B, Nadir A, Sahin E, Kaptanoglu M, Demir H, Furtun K. Risk
factors, clinical and radiological evaluation in patients with
recurrent spontaneous pneumothorax. Turkish J Thorac Car-
diovasc Surg. 2008;16:107e112.
13. Devilliers:“Du Pneumothorax Determine par la Rupture de la
Plevre et d’une vesicule Aerienne Emphysemateuse,” These,
Paris, 1826
14. Peters RM, Peters BA, Benirschke SK, Friedman PJ. Chest di-
mensions in young adults with spontaneous pneumothorax. Ann
Thorac Surg. 1978;25:193e196.
15. Kawakami Y, Irie T, Kamishima K. Stature, lung height, and
spontaneous pneumothorax. Respiration. 1982;43:35e40.
16. Casha AR, Manche A, Gatt R, et al. Is there a biomechanical
cause for spontaneous pneumothorax? Eur J Cardiothorac Surg.
2014;45:1011e1116.
17. Chang PY, Wong KS, Lai JY, et al. Rapid increase in the height
and width of the upper chest in adolescents with primary
spontaneous pneumothorax. Pediatr Neonatol. 2015;56:
53e57.
18. Saita K, Murakawa T, Kawano H, Sano A, Nagayama K,
Nakajima J. Chest wall deformity found in patients with pri-
mary spontaneous pneumothorax. Asian Cardiovasc Thorac
Ann. 2013;21:582e587.
19. Eknoyan G. Adolphe Queteled (1796-1874)- the average man
and indices of obesity. Nephrol Dial Transplant. 2008;23:
47e51.
20. Hochlehnert A, Löwe B, Bludau HB, Borst M, Zipfel S,
Herzog W. Spontaneous pneumomediastinum in anorexia
nervosa: a case report and review of the literature on pneu-
momediastinum and pneumothorax. Eur Eat Disord Rev. 2010;
18:107e115.
21. Young Choi S, Beom Park C, Wha Song S, et al. What factors
predict recurrence after an initial episode of primary sponta-
neous pneumothorax in children? Ann Thorac Cardiovasc Surg.
2014;20:961e967.

Please cite this article in press as: Akkas Y, et al., A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune
Risk Index, Asian Journal of Surgery (2015), http://dx.doi.org/10.1016/j.asjsur.2015.09.004