Note from the National Guideline Clearinghouse (NGC): The National Institute for Health and Clinical Excellence (NICE) commissioned an independent academic centre to perform a systematic literature review on the technology considered in this appraisal and prepare an Evidence Review Group (ERG) report. The ERG report for this technology appraisal was prepared by Southampton Health Technology Assessment Centre (SHTAC), University of Southampton (see the "Availability of Companion Documents" field).
Clinical Effectiveness
Critique of Manufacturer's Approach
Description of Manufacturers Search Strategy
The manufacturer has replicated the search strategies used by SHTAC in the previous assessment report on adefovir and pegylated interferon alpha 2a which underpinned NICE's existing guidance (NICE Technology Appraisal 96; see the National Guideline Clearinghouse [NGC] summary Adefovir dipivoxil and peginterferon alfa-2a for the treatment of chronic hepatitis B). The manufacturer states that the full range of databases used by SHTAC were not searched for the submission due to difficulties in access. The minimum database search criteria specified by NICE were searched by the manufacturer (i.e., Medline, Embase, Medline in Progress (MEIP) and Cochrane). In addition, two of the Centre for Reviews and Dissemination (CRD) databases were also searched (Database of Reviews of Effectiveness [DARE]; health technology assessment [HTA] database). The host system used for the electronic bibliographic searching was not reported in the submission. The ERG requested clarification and the manufacturer reported that Dialog Datastar was used to search Embase, and that Ovid and Dialog Datastar had been used to search Medline (refer to Appendix 1 of the ERG report [see the "Availability of Companion Documents" field]).
The SHTAC strategy was extended by the manufacturer to incorporate entecavir, telbivudine, and lamivudine. The searches were limited to articles published in the English language. No time limits were applied to the clinical effectiveness searches, but the ERG requested clarification about the search dates of the various electronic bibliographic databases, as these vary according to which host system is used. The manufacturer responded with the information for each database (see Appendix 1 of the ERG report [see the "Availability of Companion Documents" field]). Each database was searched from its inception, up to approximately 21st September 2007.
The manufacturer also ran a 'simple search strategy' specifically to identify articles relating to entecavir. This was a bibliographic reference chasing exercise to check for any missed trials. It is stated that this strategy was also run for telbivudine, although terms for this drug are not presented in the actual strategy itself.
The search strategy, as adapted for each bibliographic database, was not presented in the submission. However, the strategy for Medline, Embase and the Cochrane Library was supplied on request to the ERG (see Appendix 1 of the ERG report [see the "Availability of Companion Documents" field]). The strategy contains a mixture of free text and index terms, although for the Embase search it is not explicit whether index terms were used. It is not clear from the search example given by the manufacturer if all the component databases of the Cochrane Library were searched or if the Cochrane Database of Systematic Reviews (CDSR) alone was used. The ERG noticed what appeared to be a few errors with the syntax used in the strategy and requested clarification from the manufacturer. The manufacturer confirmed that these were typographical errors in the submission, rather than errors in the strategies themselves (refer to Appendix 1 of the ERG report [see the "Availability of Companion Documents" field]). The strategies appear to be comprehensive although only the generic names of the drugs were included in the strategy, rather than including trade names and Chemical Abstract Service (CAS) registry numbers or applying field tags to search for these. It is not considered, however, that using these would have produced any additional references.
In terms of on-going trials the manufacturer reports searching clinicaltrials.gov (http://clinicaltrials.gov) and Current Controlled Trials (www.controlled-trials.com), as well as internal company databases. The National Research Register (NRR) is not reported as having been searched, although this is not a NICE pre-requisite. Conference proceedings have not been reported as individually searched, although the Cochrane Central Register of Controlled Trials (CCRCT) has been searched and this does include hand-searched conference proceedings.
In summary, the search process for clinical effectiveness studies reported by the manufacturer is generally comprehensive, with key databases searched using a combination of free-text and index terms. The search strategy is not, however, fully reproducible due to limitations in reporting.
Statement of the Inclusion/Exclusion Criteria Used in the Study Selection and Comment on whether They Were Appropriate.
Three different sets of inclusion criteria are presented in the manufacturer's submission (MS), all of which were applied to the same set of search results.
- The first set is for the clinical effectiveness systematic review of entecavir studies. This is the focus of the clinical effectiveness evidence for entecavir in the submission.
- The second set was for studies screened for possible inclusion in the mixed treatment comparison (MTC).
- The third set relates to a 'systematic review of licensed therapies for chronic hepatitis B', which incorporates adefovir, pegylated interferon alpha 2a, lamivudine, telbivudine, and entecavir.
Only fully published RCTs were eligible (see the "Number of Source Documents" field); however, observational extension studies were permitted. All other observational studies were excluded. Studies published in abstract form were excluded, and unpublished studies conducted by the manufacturer were only included where a clinical study report was available. Reviews were only analysed for bibliographic checking.
Cost-Effectiveness
Critique of Manufacturer's Approach
Description of Manufacturers Search Strategy
The cost-effectiveness searches have satisfied most of the minimum database criteria set by NICE (namely, Medline, Embase, and MEIP). The manufacturer has exceeded the criteria by searching internal company databases, The Cochrane Library, the HTA databases, the TRIP database (Turning Research into Practice), and websites of organisations including NICE, The Scottish Medicines Consortium (SMC), The European Association for the Study of Liver Diseases (EASL), The American Association for the Study of Liver Diseases (AASLD), as well as a Google internet search. It is not explicitly stated whether the NHS Economic Evaluation database (NHS EED) was searched, but it is assumed it was accessed via the CRD databases which were mentioned by the manufacturer as having been searched. It is not stated whether the Health Economic Evaluation Database (HEED), one of NICE's database criteria, was searched.
The date of the searches is recorded as "during September 5th and October 10th 2007". The host system used for Embase and Medline is reported as www.embase.com. It is stated that no time limits were applied, so presumably all databases were searched back to their inception.
It is reported that all search terms were mapped to EMTREE terms and exploded, as well as included as free-text terms. However, the strategy is not reproducible as the mapped terms are not recorded. It would have been preferable to record the exact search strategy that included the free text terms and subject headings, so that it could be reproduced, or at least have clearly defined which terms were free text and which were index terms.
The search strategy is not entirely transparent and therefore not easily reproducible because the list of free text terms is given, but they have not necessarily recorded the mapped index terms. The range of free-text terms looks sensible but there is no overt truncation of free text terms, although it is thought that the Datastar Dialog platform can be programmed to identify plurals and variations of endings of words. There is no indication in the search strategy as to which fields have been searched (title, abstract, subject headings etc.). However, it does say that the mapped headings have been exploded.
Refer to sections 3.1.1 and 3.1.2 of the ERG report (see the "Availability of Companion Documents" field) for more information on search strategies and details on inclusion/exclusion criteria used in the study selection.