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 University of Sheffield, School of Health and Related Research [ScHARR]. (See the "Availability of Companion Documents" field.)
Clinical Effectiveness
Search Strategy
Because of the range of interventions and comparators under review, the literature search aimed to identify all literature relating to the prevention and treatment of osteoporosis. The main searches were conducted in May and July 2002, and updated in September and October 2002. The utilities searches were performed in October and November 2002.
Sources Searched
Fourteen electronic bibliographic databases were searched, covering biomedical, science, social science, health economic, and grey literature. A list of the databases searched is provided in Appendix 2 of the Assessment Report (see "Availability of Companion Documents" field).
In addition, the reference lists of relevant articles and sponsor submissions were handsearched, and various health services research-related resources were consulted via the Internet. These resources included health economics and health technology assessment (HTA) organisations, guideline-producing agencies, registers of generic research and trials, and specialist sites. These additional sources are listed in Appendix 3 of the Assessment Report (see "Availability of Companion Documents" field).
Search Terms
A combination of free-text and thesaurus terms was used. General "population" search terms (e.g., osteoporosis, bone, density, diseases, fracture, etc) were used in order to identify all potentially relevant studies. "Intervention" terms were not used in the main searches since it was felt that these might restrict the results and cause possibly relevant articles to be missed. Utilities searches were performed for breast cancer and for osteoporosis fractures as part of the economic evaluation section of the report. Copies of the Medline search strategies are included in Appendix 4 of the Assessment Report (see "Availability of Companion Documents" field). Search strategies for the other databases are available on request.
Search Restrictions
No language, date or study-type restrictions were applied to the searches. However, the Biosciences Information Service (BIOSIS) search was performed as title only, and the Citation Indexes searches were limited with brief clinical trials, systematic reviews, guidelines, and economics filters, and to title only, in order to keep the number of hits to a sensible level. An randomized controlled trial (RCT) filter, an economics and quality of life evaluations filter, and a systematic reviews filter were used in the main searches performed in Medline and Embase to assist the identification of articles of these types (see Appendix 5 of the Assessment Report [see "Availability of Companion Documents" field]). After the searches were completed, because of the large number of references retrieved, only the articles identified using these specific filters, the articles from the databases that were not searched with filters (such as BIOSIS), and the papers found through handsearching etc, were reviewed.
Inclusion and Exclusion Criteria
Inclusion Criteria
Participants: Women with primary osteoporosis who were at least 6 months postmenopausal
Interventions:
- Bisphosphonates
- Alendronate
- Etidronate
- Risedronate
- Selective oestrogen receptor modulators (SERMs)
- Teriparatide (recombinant human parathyroid hormone (1-34))
Comparators:
- Vitamin D
- Calcitriol (a vitamin 1alpha-hydroxylated derivative)
- Pharmacological doses of calcium
- Oestrogens (opposed and unopposed)
- Exercise
- Placebo
- No treatment
Outcome Measures: Vertebral or nonvertebral fracture, associated effects, quality of life related to the study intervention, continuance and compliance
Study Design: Randomised controlled trials (RCTs). Trials were accepted as RCTs if the allocation of subjects to treatment groups was described by the authors as either randomised or double-blind.
A discussion of outcome measures is presented in section 3.1.2.1. of the Assessment Report (see "Availability of Companion Documents" field.)
Exclusion Criteria
Studies were excluded if they included participants with secondary osteoporosis (e.g., related to therapy with corticosteroids), or drew their participants exclusively from patients with specific diseases known to affect fracture rates (e.g., Parkinson's disease).
Only published studies (including those only available in abstract form) were included. As unpublished studies are more likely than published studies to demonstrate small or absent treatment effects, it is recognised that this approach is likely to overestimate the true effects of treatment. However, it was not possible in the time available to seek out unpublished studies.
It had originally been intended to include all relevant studies, whatever the language of publication. However, for practical reasons, it was in fact possible only to include those published in English, French, German, Italian or Spanish. This led to the exclusion of one possibly relevant study published only in Japanese.
Sifting
In principle, the references identified by the literature searches were sifted in two stages, being screened for relevance first by title and then by abstract. However, as it was not possible to identify all relevant studies with fracture outcomes from titles alone, the title sifting stage was used essentially to reject studies which were clearly irrelevant. Following this, the abstracts of all studies which used the relevant interventions in the relevant populations were screened (for studies which did not provide abstracts, the full studies were screened). Twenty-eight studies which had been identified by the literature searches were not identified as relevant at the abstract sifting stage, but were identified from other reviews as reporting fracture outcomes. The reason for this was that, as fracture was only a secondary outcome measure in many studies, it was therefore not reported in the abstract.
Economic Analysis
Identifying the Studies
The review has drawn on papers identified from a series of systematic searches undertaken for a Health Technology Assessment (HTA) review of treatment for osteoporosis. These include searches of papers reporting economic evaluation of the prevention and treatment of osteoporosis, and those reporting on quality of life associated with the main fracture states, breast cancer and coronary heart disease. Studies were identified through searches of electronic databases, hand searching, citation searching, reference list checking and those known to researchers involved in the HTA review (Appendix 8 of the Assessment Report [see "Availability of Companion Documents" field]).