To estimate the relative risk (RR) of connective tissue disease in general and specific diseases, such as scleroderma, in particular between women with and without breast implants.

MEDLINE, TOXLINE and Dissertation Abstracts were searched from mid-October 1994, and Current Contents from January to mid-October 1994. Printed abstracts from professional society national meetings (January 1992 to mid-October 1994), and references of retrieved studies, were also examined.

The keywords used to search MEDLINE and TOXLINE are stated, but it is unclear how they were combined.

One paper and one revision to a paper, both published in 1995, were also used to revise summary estimates in a note added in proof.

Study designs of evaluations included in the review

Studies that provided an estimate of the association, e.g. RR, between breast implants and connective tissue disease (general or specific) in women were included.

Cohort studies had to include both women with and without breast implants, or use rates from an external reference population. Case-control studies had to include both diseased and non-diseased women.

Specific interventions included in the review

Silicone gel breast implantation (reconstructive or cosmetic).

Participants included in the review

Women with or without breast implants, and with or without connective tissue disease, were included.

Outcomes assessed in the review

RR of connective tissue disease in general and specific types of connective tissue disease, such as scleroderma, rheumatoid arthritis and systemic lupus erythematosus.

How were decisions on the relevance of primary studies made?

Abstracts of potentially relevant studies were reviewed by one unblinded reviewer, and random samples were independently reviewed by a second reviewer for inclusion or exclusion; 100% agreement was achieved. For those studies which mentioned both breast implants and connective tissue disease, the complete manuscript was reviewed. Only studies reported in the English language were included.

The studies had to consider at least one connective tissue disease, provide an estimate of the association between breast implants and connective tissue disease in women, and have a control group. The authors do not state how the papers were assessed for validity, or how many of the authors performed the validity assessment.

The authors do not state how the data were extracted for the review, or how many of the authors performed the data extraction.

How were the studies combined?

The unadjusted (crude) odds ratios from each study were combined using exact confidence limits (CIs) to calculate summary RRs.

Results were also combined using an asymptotic method based on weighted, natural logarithmically-transformed, adjusted RR estimates and by using an asymptotic method based on a random-effects model. These all gave similar results.

How were differences between studies investigated?

The homogeneity test associated with the meta-analysis RR was calculated for all studies, for specific study designs and for specific diseases. Also, various groups of studies were included and excluded from the analysis.

There were 6 case-control studies (total of 2,375 women with connective tissue disease and over 15,000 controls), 6 cohort studies (total of 4,040 women with breast implants) and one cross-sectional study (150 women with breast implants and 4,079 without).

The individual study RRs for the relationship between breast implants and the general class of connective tissue disease ranged from 0 to 4.5, and all CIs included '1'. Data for exact analyses were not available from one study. For the remaining 12 studies, the summary RR for connective tissue disease in general, in women with implants compared to those without was 0.76 (95% CI: 0.55, 1.04). Irrespective of which studies were included in this meta-analysis, e.g. by restricting study designs, there was no statistically-significant increased risk for connective tissue disease in general.

The RR of scleroderma (based on 7 studies) was 0.98 (95% CI: 0.57, 1.64).

The RR of rheumatoid arthritis (based on 6 studies) was 0.79 (95% CI: 0.48, 1.26).

The RR of systemic lupus erythematosus (based on 2 studies) was 0.26 (95% CI: 0.04, 0.90).

The accumulating data show no significant association between breast implants and connective tissue diseases. These data should be reassuring to women with breast implants.

This review assessed the available evidence from studies which compared the frequency of connective tissue disease among women with and without breast implants, and from those which compared the frequency of breast implants among women with and without connective tissue disease.

The search strategy appeared to be comprehensive, but included only those papers published in the English language.

The methodological quality of included studies was not formally, systematically assessed. However, sensitivity analyses suggested that different study designs did not yield significantly different results.

The available research does not show any evidence for increased risk of connective tissue disease associated with silicone gel breast implants.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.