Material and Methods:
The present review and analysis followed the PRISMA
guidelines.9
Information Sources: Eligible studies were searched in
MEDLINE, Embase, Ovid, Scopus, Web of Science and Journal on web
databases, through EBSCO. All relevant articles were identified. Only
studies obtained in full, through electronic and other search methods,
were reviewed.
Search Strategy: MeSH phrases and open phrases were
entered into unique search strategies.
Keywords: Key terms employed for the search: 1)
Pneumothorax; 2) Chest ultrasound; 3) specificity; 4) sensitivity; 5)
diagnostic accuracy; 6) ultrasonography; and 7) Chest trauma.
Boolean Operators: The Boolean operator ‘OR’ was
employed to complement truncated synonyms in each search theme. The
Boolean operator ‘AND’ makes up the sum of each four main search themes
to specifically output papers that give at least one result for each
time.
Search Limits: The search limit for literature ranged
from the year 2000 up until 2020 as the concluding year, and the
literature published only in the English language with full text was
chosen.
Process of Study Identification: Endnote X8soft ware was
used to import the search data results and remove the duplicates.
Abstracts were screened for eligibility criteria, and the full text was
extracted when an article was selected. Following inclusion into the
review, data was extracted.
Data collection: All the titles and the extracts were
screened independently by the reviewers and upon a detailed review of
the full-text articles, the data were extracted and documented in a data
extraction table, elaborating data items evaluated for the review.
Eligibility criteria: Studies investigating the accuracy
of chest ultrasound in diagnosing pneumothorax, conforming to the
diagnostic criteria of pneumothorax and met quality criteria set by
QUADAS 2 was included.
Data items: The data extraction table included Study ID,
sample size, location, criteria of the condition, ultrasound type,
machine made, and personnel involved.
Risk of bias in individual studies: QUADAS – 2 criteria
evaluated risk of bias of included studies which is a tool for
evaluating diagnostic accuracy10. It includes patient
selection – random sequencing, index test – detection of the
condition, reference evaluation – comparison with flow and timing of
the study. However, if there is a non-agreement between the two
reviewers for quality evaluation, the mutual agreement was reached
through a discussion & consultation with a senior expert in the field.