Sex and Gender-based Analyses in Norwegian Treatment Studies: a Scoping Review
Mapping review
|Published
The objectives of this scoping review were to investigate the use of sex-specific analyses in studies on outcome and effect of treatment in a Norwegian context.
Key message
Biological sex affects clinical outcome of treatment of many diseases, and knowledge about sex differences in effects of treatment can contribute to safer and more tailored treatment for the individual patient. The objectives of this scoping review were to investigate the use of sex-specific analyses in studies on outcome and effect of treatment in a Norwegian context.
We conducted a scoping review. A scoping review is a type of evidence synthesis that displays the available evidence relevant to a broader research question, and that narratively describes existing literature on a topic. The search was completed in August 2022 in the databases MEDLINE, PsycINFO, and Cochrane Central Register of Controlled Trials (CENTRAL).
- We found 133 studies from Norway with at least 50 patients in the time period 2017 – 2022 on outcome of treatment where there was at least one sex-based analysis.
- There were more (n= 178) publications that did not report any sex-based analyses on treatment outcome, compared to publications that did report at least one sex-based analysis.
- One of three publications had referred to the results in the discussion, and one of four had referred to the results in the conclusion.
- A little under half of the studies had an even sex distribution in the study samples with a proportion of women between 41% to 60%.
We found substantial differences across treatment studies in terms of the consideration of sex, and it is our conclusion that the effect of sex is not reported on in a systematic way. When sex-based analyses are conducted this should be made clear in methods, results and discussion.
Summary
Introduction
Biological sex affects clinical outcome of treatment of many diseases. Knowledge about sex differences in effect and outcome of treatment can contribute to safer and better tailored treatments for the individual patient. To ensure adequate sex-specific knowledge it is necessary that sex as a variable is considered in health and medical research. Important considerations are the inclusion of both men and women in the study sample, as well as use of sex stratified analyses. Studies also need to consider the results in the context of relevant existing knowledge.
There is no standardized way of applying sex in health and medical research, and the implementation of methods to get sex-specific knowledge is not well documented in a Norwegian context. Until now, no systematic reviews or reports have documented to what extent Norwegian health research conducts and reports findings from sex-based analyses, or whether studies have samples that are representative of the population they are studying in terms of sex.
Objectives
The objectives of this scoping review were to investigate the use of sex-specific analyses in studies and effects of treatment in a Norwegian context. The scoping review aimed to answer the following research questions:
- Whether sex-based analyses are conducted in studies on selected diseases, and how results from analyses are used and incorporated in the discussion and conclusion.
- Whether study samples have a sex-distribution that appropriately reflects incidence.
- Whether there is a marked difference between observational studies and randomized studies in terms of use of sex-based analyses and study samples.
- To what extent included studies report sex-specific differences in treatment results.
Methods
We included studies on treatment outcome on patients in Norway that were published from 2017 to 2022. Included studies had at least one author with affiliation with a Norwegian institution. We included studies with at least 50 participants, with no age requirements.
The search was conducted in July and August 2022 in the following databases: MEDLINE, PsycINFO, and Cochrane Central Register of Controlled Trials (CENTRAL). We searched for 62 predetermined diseases, and we used search words related to “women” and “gender/sex”. The search was limited to Norway by the use of the search word “Norway”. Selection and data charting was conducted in the digital tool EPPI Reviewer.
Results
The search produced 3233 unique references were checked by title and abstract, and 427 were read in full text. We found 133 studies from Norway with at least 50 patients published in the time period 2017 – 2022 on outcome of treatment where there was at least one sex-based analysis. Out of 62 selected diseases, there were most studies on cancer (n = 47), musculoskeletal diseases (n = 28), og cardiovascular diseases (n = 25). Twenty-one of the selected diseases were not represented in any studies, and 19 diseases were represented with one study.
There were more (n = 178) publications that did not report any sex-based analyses on treatment outcome, compared to publications that did report at least one sex-based analysis (n = 133). There were 36 and 33 analyses that showed significant findings, with better results for women and men respectively. Of these there were five studies that showed mixed results where women and men scored differently on separate measures. There were 67 studies that only found non-statistically significant results, and 19 studies that found mixed non-significant and significant results. One of three publications had referred to the results in the discussion, and one of four had referred to the results in the conclusion. There were no large differences between study design types in use of results in the discussion, while the mention of results in the conclusion varied from 12% to 20% across study designs.
Under half of the studies had an even sex distribution in the study samples with a proportion of women. There were 21 (15%) publications with an especially low proportion of women (11% - 30%) that were all conducted on patients with cardiovascular diseases or cancers.
There were few studies on children under the age of 18. Among studies that had conducted sex-stratified analyses, 6 (5%) were done on the age groups 0 – 18. Among studies that had not conducted a sex-stratified analysis, 7 (4%) were on children 0 – 12 years, and 6 (3%) were on youth 13 – 17 years.
Discussion
This scoping review presents an overview of the use of sex-based analyses and consideration of sex differences in treatment studies on a wide range of diseases that affect both men and women in Norway the last five years. Few studies had conducted separate analyses of treatment results for men and women, and a small selection reported having conducted a test of interaction of sex and treatment outcome. Researchers should be aware that relevant information and sex differences in treatment outcome can be missed if such analyses are not conducted and results reported. Of studies that did conduct a sex-based analysis, a minority had referred to the results in the discussion and considered the results in the context of relevant existing knowledge and research. To enable study results to be used in future research, findings from studies with sex-based analyses should be discussed and the relevance and implications of the results on research and clinical practice should be evaluated.
Conclusion
We found substantial differences across treatment studies in terms of the consideration of sex as an explanatory variable, and it is our conclusion that sex is not reported on in a systematic way in Norwegian treatment studies across the 62 diseases considered in this review. There are substantial differences across studies in terms of how results from analyses are applied in the discussions and conclusions, and this is done in a minority of publications. When sex-based analyses are conducted this should be made apparent in methods, results, and discussion. We have highlighted that there are few studies that reported a test of interaction of sex and treatment outcome, and a minority of studies conduct separate analyses for men and women. It is through these types of analyses that we develop knowledge about when men and women respond similarly or differently to treatment, and we would like to see more studies use such methods.
