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 assessment report. The assessment report for this technology appraisal was prepared by the Centre for Reviews & Dissemination (CRD)/Centre for Health Economics (CHE) University of York (see the "Companion Documents" field).
Assessment of Clinical Effectiveness
Search Strategy
The following sources were searched for studies relating to the clinical effectiveness of newer antiepileptic drugs (AEDs); lamotrigine, gabapentin, levetiracetam, oxcarbazepine, tiagabine, topiramate and vigabatrin. This first set of literature searches were designed to retrieve systematic reviews and randomised controlled trials only. However, some databases cannot be reliably restricted by study type and in these cases the search was not limited by study design, and the results of the searches were screened by hand. A range of free text terms and subject headings were used as appropriate. Further details of the search strategies are reported in Appendix 2 of the assessment report.
CRD Internal Administration Databases (searched 20.03.02)
- Database of Abstracts of Reviews of Effective (DARE)
- Health Technology Assessment Database (HTA)
Internet Resources and Databases (searched 02.04.02)
CD-ROM Resources
- Cochrane Controlled Trials Register (CCTR) (2002: Issue 1) (searched: 02/04/02)
- Cochrane Database of Systematic Reviews (CDSR) (2002: Issue 1) (searched: 02/04/02)
- EMBASE (1980 - 2002/02) (searched: 27/03/02)
- MEDLINE (1966 - 2002/03) (searched: 26/03/02)
- National Research Register (NRR) (2002: Issue 1) (searched: 02/04/02)
- PREMEDLINE (up to 22.03.02) (searched: 26/03/02)
- PsycINFO (1967 - 2002/07 Week 3) (searched: 03/09/02)
Online Resources (searched: 08/04/02)
- Conference Papers Index (CPI) (1973 onwards)
Paper Resources
- Clinical Evidence: A compendium of the best available evidence for effective health care. Issue 6, 2001. London: BMJ Publishing Group.
No date or language restrictions were placed on any of the literature searches. Due to financial and logistical constraints non-English publications were not included in the review. However, not limiting the literature searches by language enabled an estimate of the size of the non-English literature to be obtained. In addition, search strategies were not limited by age although the review only included data relating to adults. This was due to the fact that many records do not mention the appropriate patient group within the title, abstract or indexing. The bibliographies of all included studies were reviewed in order to identify any further relevant studies. A list of studies found from bibliographies and industry submissions, but not meeting the inclusion criteria for this review, are listed in Appendix 3 of the assessment report.
Inclusion and Exclusion Criteria
Two reviewers independently screened all titles and abstracts in order to determine relevance. Full paper manuscripts of potentially relevant titles and abstracts were obtained where possible and the eligibility of the study for inclusion in the review was assessed by two authors independently, according to the four criteria outlined below. Any discrepancies were resolved by consensus and if necessary a third reviewer was consulted. Studies that did not fulfill all of the criteria were excluded. Due to time and financial constraints only studies reported in English were included in the analysis section of this review. Eligible studies in other languages were identified but only brief details tabulated.
Study Design
The following study designs were included in the review:
- Single-blinded, double-blinded or unblinded randomised controlled trials (RCTs) using a parallel or crossover design, designed to assess the equivalence, non-inferiority or superiority of comparators
- Systematic reviews meeting fulfilling the criteria for inclusion in the Database of Abstracts of Review of Effect (DARE) (http://www.york.ac.uk/inst/crd/darehp.htm
Participants
Studies recruiting adults (i.e., individuals aged 18 years or over) with either newly diagnosed or refractory epilepsy were included. Seizure types included partial onset (with or without secondary generalisation) and generalised onset seizures. Trials enrolling only patients with single seizures, status epilepticus, seizures following neurosurgery or head injury, and trigeminal neuralgia were excluded. Studies that enrolled participants with excluded indications were evaluated to determine whether: 1) the study results reported data for the excluded indications groups of participants separately, or 2) the numbers of excluded indications participants was small. In either case the relevant data was included in this review.
Studies with mixed age groups were identified during the inclusion/exclusion process. The data reported in these studies was discussed and divided accordingly in coordination with the Birmingham review team responsible for reviewing the evidence for the treatment of children. The discussion determined whether 1) the study results reported data for the different age groups of participants separately, or 2) the numbers of younger or older participants were small. Data were only extracted if relevant to age group under consideration.
Interventions
Newer antiepileptic drugs (AEDs) (gabapentin, lamotrigine, levetiracetam, oxcarbazepine, tiagabine, topiramate, and vigabatrin) used either, as monotherapy and/or adjunctive therapy were included. Comparators included older AEDs, newer AEDs, or placebo. Trials where epilepsy surgery was the comparator were excluded. Older AEDs included acetazolamide, benzodiazepines, carbamazepine, ethosuximide, phenobarbital and other barbiturates, phenytoin, and valproate.
Assessment of Serious, Rare and Long-Term Adverse Events Studies
Search Strategy
Literature searches were carried out to identify serious, rare and long-term adverse events not likely to have been found by the clinical effectiveness RCT search strategies. The searches aimed to find all possible side effects of these seven drugs irrespective of the condition treated. Therefore, no epilepsy terms were added. It is well reported in the literature that conducting electronic database searching for adverse events is problematic. The procedure for tracing papers of adverse events is not as well established as in other areas of research such as RCTs and systematic reviews. A broad experimental search strategy was therefore adopted using textwords, and thesaurus terms for each drug limited to the appropriate subheadings and known serious or rare side effects as both textwords and thesaurus terms. Those adverse events deemed serious fell into one or more of the following categories; death, life threatening, hospitalisation, disability (including vision), congenital abnormality, cancer, and overdose.
Databases were searched from the date of inception to the most recent date available.
Internet Resources and Databases (all searched 09/09/02)
CD-ROM Resources (searched: 10/09/02)
- EMBASE (1980-2002 Week 36)
- MEDLINE (1996-August Week 4 2002)
Paper Resources (searched: 04/09/02)
- ABPI Medicines Compendium. Datapharm Communications Ltd. 2002.
- AHFSFirst professional edition version 2.71. American Society of Health-System Pharmacists. 2002.
- British National Formulary (BNF). London: British Medical Association/ Royal Pharmaceutical Society of Great Britain. Issue 43 March 2002,
- Dukes, M.N.G. and Aronson, J.K. (eds.). Meylers's Side Effects Of Drugs: An Encyclopedia of Adverse Reaction and Interactions. 14th edition. Oxford: Elsevier. 2000.
- Sweetman, S.C. (ed.) Martindale: the complete drug reference. 33rd edition. London: Pharmaceutical Press. 2002.
Further details of the full search strategy are reported in Appendix 2 of the assessment report.
Inclusion and Exclusion Criteria
In this review, non-randomised experimental studies and observational studies were included to enhance retrieval of information about serious, rare and long-term adverse events. Reporting of safety data in RCTs is largely inadequate and most systematic reviews of RCTs only include safety data as reported in the primary studies. Furthermore, RCTs are often too small and of insufficient duration to detect rare and delayed events. Consequently, evaluation of the safety of therapeutic interventions needs to go beyond RCTs.
Two reviewers independently screened all titles and abstracts against pre-defined inclusion criteria. Differences were resolved by discussion and full papers were obtained for all studies potentially eligible for inclusion. Two reviewers then independently applied the inclusion criteria to all full papers and differences were again resolved by discussion.
Three categories of studies were included:
- Studies that investigated the effects of newer AEDs, including safety and/or tolerability. Study designs eligible for inclusion were uncontrolled trials, open label extension phases of controlled trials, cohort studies (controlled or uncontrolled) and case-control studies. These studies, RCTs of newer AEDs in diseases other than epilepsy, and RCTs of dose or titration comparisons in epilepsy, were included only if more than 300 participants were exposed to the newer AED or if follow up exceeded 6 months. These limits were based on the duration and size of effectiveness RCTs. Combination therapies and dose comparisons were included within the aforementioned parameters. Case series, case reports, cross-sectional studies, audits and surveys were excluded.
- Studies that investigated a specific adverse effect (such as visual field defects). Study designs eligible for inclusion were as described above but without the restriction on study size or duration.
- Reports of prescription event monitoring studies and prospective post-marketing surveillance studies. Spontaneous case reports of suspected adverse drug reactions such as those collated by the Medicines Control Agency and other bodies were not included.
Assessment of Cost Effectiveness
Search Strategy
Those databases restricted by study design in the clinical effectiveness searches were searched again using a search strategy designed to retrieve cost effectiveness studies or economic models. Two specialist databases were also searched, the National Health Service (NHS) Economic Evaluation Database (NHS EED) and the Health Economic Evaluation Database (HEED). No economic filter was necessary for these databases.
CRD Internal Administration Databases (searched 20.03.02)
NHS Economic Evaluation Database (NHS EED)
CD-ROM Resources
- EMBASE (1980 - 2002/02) (searched: 27/03/02)
- Health Economic Evaluation Database (HEED) (March 2002) (searched: 28/03/02)
- MEDLINE (1966 - 2002/03) (searched: 27/03/02)
- PREMEDLINE (up to 22.03.02) (searched: 27/03/02)
Further details of the search strategies used are reported in Appendix 2.
Inclusion and Exclusion Criteria
Three reviewers independently screened all of the titles and abstracts of the retrieved references according to following inclusion criteria. Any disagreements were resolved by consensus.
Study Design
Only full economic evaluations were included. Types of designs included:
- Cost-effectiveness analyses including cost-minimisation analyses and cost consequences analyses
- Cost-benefit analyses
- Cost-utility analyses
Participants
Studies recruiting adults (i.e., individuals aged 18 years or over) with either newly diagnosed or refractory epilepsy were included. Seizure types included both partial onset (with or without secondary generalisation) and generalised onset. Trials enrolling only patients with single seizures, status epilepticus, seizures following neurosurgery or head injury, and trigeminal neuralgia were excluded. Studies that enrolled participants with excluded indications were evaluated to determine whether: 1) the study results reported data for the excluded indications groups of participants separately, or 2) the numbers of excluded indications participants was small. Any relevant data were included.
Interventions
Newer AEDs (gabapentin, lamotrigine, levetiracetam, oxcarbazepine, tiagabine, topiramate, and vigabatrin) used either, as monotherapy and/or adjunctive therapy were included. Comparators included older AEDs, newer AEDs, or placebo. Trials where epilepsy surgery was the comparator were excluded. Older AEDs included acetazolamide, benzodiazepines, carbamazepine, ethosuximide, phenobarbital and other barbiturates, phenytoin, and valproate.
Outcomes
In order to be included in the review of cost effectiveness evaluations had to report both costs and clinical effectiveness.