PubMed and Scopus
databases were accessed between 6-13th December 2017 and the
literature included in the discussion were selected following two stages:
Database searching – involved searching
selected journals with key search terms.
Hand searching – involved refining the
search by manually removing a cohort of papers in relation to the research
title and exclusion criteria.
This was divided into a
primary and secondary stage. The primary
stage involved selecting key search terms and performing searches with no
inclusion or exclusion criteria applied.
The key terms selected were guided by the surgical words most frequently
encountered when reading related literature prior to the study. Synonyms, abbreviations and similar terms
such as ‘choice vs type’ and ‘ACL vs anterior cruciate ligament’ were used to
broaden the search and limit the likelihood of any papers being missed. Boolean operators and truncation methods (*)
were not used.
The primary stage
returned a large number of results (n = 489 ) and the secondary stage involved
the application of inclusion criteria to help refine the results.
There was a ten year
limit placed on selecting literature.
This was because the purpose of this review was to inform current
practice and any evidence, even level 1, published before the selected period
may have been superseded by findings of newer studies.
The results returned
from searches using the word ‘type’ as opposed to ‘choice’ and ‘anterior
cruciate ligament’ as opposed to ‘ACL’ were used for hand searching as there
was a greater number of results which reduced the chance of any relevant papers
Following the secondary
stage, 107 papers were identified and the literature was then subjected to a
manual hand searching stage. The study
design of each paper was firstly identified to eliminate any literature that
was not based on randomised control trials.
Next, a selection of RCT papers that remained were removed based on the
unsuitability of their abstracts relative to the research title and exclusion
Once completed, six
relevant papers were identified and were included in the appraisal. To help guide the appraising the “Critical
Appraisal Skills Programme” checklist for randomised control trials was used. 10
With more time, a
greater variety of MeSH headings sourced from the OVID database could have been
trialled and searching further databases, such as OVID, could have been undertaken. Four additional papers were sourced from
citations and references from the literature and included in the appraisal. These were subjected to the same
inclusion/exclusion criteria to ensure consistency between the papers being
included. Other references will include
other relevant reviews or meta-analyses that have used other study designs to
help contextualise the literature review.
Figures 5 and 6 below
outline the number of literature results included in the appraisal and the
number rejected from each database.