Abstract
BackgroundSeveral authors have developed and applied methods to routine data sets to identify the nature and rate of complications following interventional procedures. But, to date, there has been no systematic search for such methods. The objective of this article was to find, classify and appraise published methods, based on analysis of clinical codes, which used routine healthcare databases in a United Kingdom setting to identify complications resulting from interventional procedures.MethodsA literature search strategy was developed to identify published studies that referred, in the title or abstract, to the name or acronym of a known routine healthcare database and to complications from procedures or devices. The following data sources were searched in February and March 2013: Cochrane Methods Register, Conference Proceedings Citation Index – Science, Econlit, EMBASE, Health Management Information Consortium, Health Technology Assessment database, MathSciNet, MEDLINE, MEDLINE in-process, OAIster, OpenGrey, Science Citation Index Expanded and ScienceDirect. Of the eligible papers, those which reported methods using clinical coding were classified and summarised in tabular form using the following headings: routine healthcare database; medical speciality; method for identifying complications; length of follow-up; method of recording comorbidity. The benefits and limitations of each approach were assessed.ResultsFrom 3688 papers identified from the literature search, 44 reported the use of clinical codes to identify complications, from which four distinct methods were identified: 1) searching the index admission for specified clinical codes, 2) searching a sequence of admissions for specified clinical codes, 3) searching for specified clinical codes for complications from procedures and devices within the International Classification of Diseases 10th revision (ICD-10) coding scheme which is the methodology recommended by NHS Classification Service, and 4) conducting manual clinical review of diagnostic and procedure codes.ConclusionsThe four distinct methods identifying complication from codified data offer great potential in generating new evidence on the quality and safety of new procedures using routine data. However the most robust method, using the methodology recommended by the NHS Classification Service, was the least frequently used, highlighting that much valuable observational data is being ignored.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2288-14-126) contains supplementary material, which is available to authorized users.
Highlights
Several authors have developed and applied methods to routine data sets to identify the nature and rate of complications following interventional procedures
Eligible papers In total, 3688 records were assessed for relevance by two authors; 3180 were subsequently excluded using information from the title and abstract
Forty-four articles reported the identification of complications using methods based on clinical codes (59.1% of which used surrogate measures to identify complications) and were included in this study
Summary
Several authors have developed and applied methods to routine data sets to identify the nature and rate of complications following interventional procedures. The objective of this article was to find, classify and appraise published methods, based on analysis of clinical codes, which used routine healthcare databases in a United Kingdom setting to identify complications resulting from interventional procedures. An alternative is to use routine healthcare databases [11] that capture information from large populations across a broad range of interventions. Such data are available more readily and at a lower cost than bespoke patient registers [12,13] and have been used to assess outcomes of interventional procedures in clinical practice [14,15]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
