Methodological approach | Possible impact on validity of the review |
---|---|
Only including peer-reviewed journal publications, and excluding any literature that was not peer-reviewed such as congress presentations and government reports | This may have led to the exclusion of relevant evidence that has only been published in non-peer-reviewed formats. However, this is an accepted methodological adjustment for a rapid review and is unlikely to miss any pivotal studies, which would likely be published in peer-reviewed journals |
Only including English language publications | Given that this review focused on evidence relevant to the UK setting, this limitation should not have led to the exclusion of any pivotal studies |
Searches were run without date limits, and studies were initially considered for inclusion regardless of when they were conducted or published. However, given that a high volume of studies reporting epidemiology outcomes were identified, it was necessary to reduce the number of studies selected for extraction. Studies completed after 2000 and 2006 were ultimately included for VP and VCI respectively | Although some evidence from older studies was therefore excluded from the evidence synthesis, there is evidence that rates of VP have changed over time. The underlying risk factors for VP and VCI, particularly in vitro fertilisation(IVF), have also changed over time. The most recent estimates are therefore likely to provide the most relevant estimates of current epidemiology |
Articles were reviewed by a single reviewer in the first instance. A second reviewer examined all included articles, 20% of excluded articles and any articles where there was uncertainty about inclusion | Although a systematic review would require all articles to be reviewed in duplicate to reduce the risk of bias as much as possible, this pragmatic strategy would have ensured that any articles where the eligibility was unclear were reviewed twice. Furthermore, input from clinical experts and public consultation on the results acted as a safeguard to minimise the risk of critical studies being missed |
Searches for full-text articles were carried out at Cambridge University Library, but some articles were not freely available at this library and were therefore not reviewed | One article was included in the evidence synthesis on the basis of the abstract alone, but for the remainder of the articles, it was judged that the full-texts would not contain any additional pivotal data from relevant populations that would affect the conclusions of the review |
Not contacting authors of publications for further information | The anticipated impact of this was expected to be minimal, especially given the small proportion of timely responses typically received when undertaking this activity. Furthermore, none of the queries would have related to a matter with the potential to change the direction of the conclusions drawn in the review |
Use of published quality assessment checklists without adaptation, to minimise the time required for protocol development | This action reduced time required at the protocol development stage, but in retrospect may have increased the time taken to integrate the quality assessment results into the discussion. Nevertheless, this had no effect on the conclusions of the review |