Symptom Awareness Measures for Breast and Cervical Cancer in Sub-Saharan Africa: A Scoping Review Tweet for journal Twitter profile: A scoping review of symptom awareness measures for breast and cervical cancer in Sub-Saharan Africa

Background:  In sub-Saharan Africa (SSA), breast cancer is the most commonly diagnosed cancer among women, while cervical cancer remains the leading cause of cancer death. Women often fail to recognise or misinterpret possible symptoms, so breast and cervical cancer symptom awareness information can promote timely help-seeking behaviour, diagnosis and start of treatment. Aim:  To identify tools that have been utilised to measure breast and cervical cancer symptom awareness in SSA and to establish if such tools have been validated in SSA populations. Methods:  A scoping review of articles published between January 1997 and February 2017, written in English and describing primary research in breast and/or cervical cancer symptom awareness-related topics in SSA contexts, was undertaken across five databases. The approach was supported by Colquhoun et al.’s methodological framework for scoping reviews. Results:  A total of 41 studies were included from 11 SSA countries. Almost half (20/41) used breast and/or cervical cancer symptom awareness tools but did not report on tool validation processes. The rest (21/41) made reference to some tool validation, yet only two reported a detailed account of their tool validation processes. One explored lay perceptions of breast cancer, while the other sought to establish the validity and reliability of a UK tool in a Kenyan context. Conclusion:  The findings point to the dearth of comprehensively validated and culturally relevant tools to measure breast and cervical cancer symptom awareness in the SSA context. They have informed the development and validation of an African Women Awareness of CANcer (AWACAN) tool, which can support the development and evaluation of interventions relevant to the SSA context.


Introduction
Breast and cervical cancer are the leading cause of cancer morbidity and mortality in women.
In Sub-Saharan Africa (SSA) breast cancer is the most commonly diagnosed cancer among women, whilst cervical cancer remains the leading cause of cancer death. 1 Yet, if diagnosed early, both cancers are treatable with curative intent.Worldwide, most cancers (85-90%) are diagnosed following symptomatic presentation. 2Consequently, understanding processes related to cancer symptom epidemiology, symptom awareness and responses to symptoms are important in developing interventions to promote timely cancer diagnosis.For both breast and cervical cancer, women often fail to recognise or misinterpret these symptoms or wait until symptoms (and disease) progress before they seek medical attention. 3,4For example, patients in Nepal have been shown to be more likely to recognise vaginal bleeding and seek care for symptoms, in contrast to recognising foul smelling vaginal discharge. 5For breast cancer, women are more likely to recognise and seek care for breast lumps in contrast to nonlump symptoms. 6There is evidence that interventions to increase awareness lead to better outcomes. 7For instance, a Malaysian study reported earlier help-seeking for breast and cervical cancers following an intervention targeted at raising public awareness of symptoms Background: In Sub-Saharan Africa (SSA) breast cancer is the most commonly diagnosed cancer among women, whilst cervical cancer remains the leading cause of cancer death.Women often fail to recognise or misinterpret possible symptoms, so breast and cervical cancer symptom awareness information can promote timely help-seeking behaviour, diagnosis and start of treatment.

Aim:
To identify tools which have been utilised to measure breast and cervical cancer symptom awareness in SSA.
Methods: A scoping review of articles published between January 1997 and February 2017, written in English, and describing primary research in breast and/or cervical cancer symptom awareness-related topics in SSA contexts, was undertaken across five databases.The approach was supported by Colquhoun et al's methodological framework for scoping reviews.
Results: 41 studies were included from 11 SSA countries.Almost half (20/41) used breast and/or cervical cancer symptom awareness tools but did not report on tool validation processes.The rest (21/41) made reference to some tool validation, yet only two reported a detailed account of their tool validation processes.One explored lay perceptions of breast cancer, while the other sought to establish the validity and reliability of a UK tool in a Kenyan context.

Conclusion:
The findings point to the dearth of comprehensively validated and culturally relevant tools to measure breast and cervical cancer symptom awareness in the SSA context.They have informed the development and validation of an African Women Awareness of CANcer [AWACAN] tool, which can support development and evaluation of interventions relevant to the SSA context. of these cancers. 8This suggests that breast and cervical cancer symptom awareness and interpretation can promote timely help-seeking behaviour, diagnosis and start of treatment.
Accurate measurement of cancer symptom awareness, knowledge and beliefs using validated tools would enable precise measurement of the impact of cancer awareness interventions.
Against this background, we conducted a scoping review to identify tools which have been utilised to measure breast and cervical cancer symptom awareness in SSA and evidence of their validity.0][11] Scoping reviews cover a broader scope than systematic reviews which have a more clearly defined, narrow scope.This renders scoping reviews useful when attempting to encapsulate and disseminate findings from a diverse body of knowledge.Inclusion and exclusion criteria in a scoping study are developed during study selection unlike in a systematic review where such criteria are pre-determined prior to study selection, during protocol development.Scoping reviews present evidence based on key themes but do not delve in assessment of the quality of such evidence, as is the case in systematic reviews.
Scoping reviews may be conducted as a preamble -to determine if it is necessary to conduct a full systematic reviewor as an autonomous exercise, as we did in this review.

Methods
The review was guided by two questions: (1) Which tools have been used to measure symptom awareness in SSA? (2) Have these tools been validated in any SSA populations?Following Colquhoun, et al., 12 the review process incorporated the first five stages of Arksey and O'Malley's six-stage framework for conducting scoping reviews 9 with Levac, et al. enhancements of each stage. 10The five stages are: stage 1: identifying the research question, purpose and objectives of the scoping study; stages 2-4: identifying relevant studies, study selection and charting the data respectively, and; stage 5: collating, summarising and reporting the results.
Identification of relevant studies commenced with a basic search for the terms breast cancer, cervical cancer, awareness, beliefs, measures, Africa on Web of Science, Scopus & Ebscohost electronic databases.We then sifted through key studies for potential broader search terms, which generated a list of terms (see Appendix 1).A scoping review of literature was then undertaken in PubMed, Web of Science, Ebscohost, Scopus and Cochrane databases.This process entailed refining the search strategy with the aim of generating as many relevant articles as possible.The final search strings contained the terms 'breast; cervix*; cancer; neoplasm; symptom*; sign; knowledge; perception*; appraisal; understand*; beliefs; attitudes; behavio$r; tool; scale; measure*]' in various combinations, with the Boolean phrases AND/OR.The database search was supplemented by articles identified by searching through reference lists of key articles.
We included peer reviewed journal articles published between January 1997 -February 2017, written in English and describing primary research in breast and/or cervical cancer symptom awareness-related topics in SSA contexts.Given the focus on tool validation which involves evaluating 'if the measurement tool employed actually measures the intended research concept or construct…or if the measurement tools used to quantify the variables provide stable or consistent responses', 13 that is, validity and reliability respectively, the research team limited articles to quantitative research studies.The exclusion criteria were: articles in languages other than English, non-peer reviewed articles and book chapters, grey literature, qualitative studies and studies from low-and-middle-income countries outside SSA.
Selection of studies involved consultations among the authors who met as a team to establish consensus.

Results
The PRISMA Flow Diagram and checklist 14 demonstrates our search strategy and included studies.

Kenya
Psychometric assessment of BCAM for assessment of BC knowledge and barriers to screening in Kenya Cross-sectional survey BCAM: Assessment of face and psychometric validity of BCAM using cognitive testing (and adaptation of some items), factor analysis of survey data and correlations these questions be rephrased for clarity.The study observed variances in cultural interpretation of some items as well as nuances around translation of BCAM items into Kiswahili.For example, participants requested for clarification on the BCAM question "do you think that redness of your breast skin could be a sign of breast cancer", noting that unless the top surface of the skin peels off, dark skin would not usually turn red.This resulted in rephrasing of the question to read "change in skin colour" in lieu of redness for contextual relevance.

Discussion
Our scoping review provides a synthesis of research 12 on tools used to measure symptom awareness in SSA and determining if such tools have been validated in SSA settings.In this regard, our review identified gaps in SSA breast and cervical cancer tool validation literature by presenting the extent of research and how this research was conducted 11 in the period between 1997 and 2017.Scoping reviews are useful in various research aspects, including: examining the extent, range and nature of research activity; establishing the relevance of conducting a full systematic review; succinctly presenting and circulating research results, and; identifying gaps in literature. 9In addressing how the research was conducted, the scoping review was limited to a narrative descriptive account of studies included and not a detailed analysis of the quality of these studies, which would be possible in a systematic review. 9r findings demonstrate that several tools have been utilised to measure breast and cervical cancer in SSA contexts.6][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] The other 21 studies identified tools used and offered information of varying detail regarding validation processes.  withnly two studies demonstrating detailed and systematic tool validation processes. 52,53Even so, these two studies are limited in that they show evidence of thorough validation focused on (a) breast cancer but not cervical cancer (b) only two of the BCAM domains namely, knowledge and/or perceptions of breast cancer and barriers to screening.
Our results point to the dearth of comprehensively validated and culturally relevant tools to measure breast and cervical cancer symptom awareness in the SSA context, and the need for systematic efforts to develop and validate such tools.These findings are consistent with a study on understanding low cervical cancer screening in Uganda, which noted lack of validated tools in African settings. 37Using consistent and reliable measures in cancer studies contributes to the quality of research results 54 which, in turn, will inform oncology practice in far as 'standardised, valid measurement is essential for monitoring levels of cancer awareness, examining its risk factors and consequences, and evaluating interventions to promote it'. 55rther, validated tools should be culturally relevant for their intended study populations. 52,53,56

Table 3
53,53 the two articles which gave a detailed account of their validation processes52,53.Both articles reported a study conducted by a research team from the Academic Model Providing Access to Healthcare [AMPATH] Oncology Institute, based in Eldoret, Kenya.One explored lay perceptions of breast cancer52, while the other sought to establish the validity and reliability of the UK Breast Cancer Awareness Measure (BCAM)55in a Kenyan context.53

Table 3 : Symptom awareness tools identified; thorough validation
53chira, et al.,53focused on validation of the BCAM in a Kenyan setting, based their work on 1061 women and an additional 48 women who participated in six cognitive focus group discussions.The authors opted to use BCAM in the absence of locally validated tools applicable in the Kenyan context.Of specific interest to this research team were two BCAM domains, breast awareness and perceived barriers to screening, which the Kenyan researchers associated with delays in timely interventions to enhance early breast cancer detection.
52anyu, et al.,52modified BCAM items by adding open-format questions relating to symptoms, severity and treatment of breast cancer, to suit their Kenyan audience.The specific questions were not detailed in the article.Further, two open-ended questions were added to the tool as follows: (1) "What are some beliefs, opinions and traditions that you have heard from others about breast cancer?"(2)"In your opinion, what are some of the early warning signs of breast cancer, the ways in which one may know first that s/he has this condition?"(p.149)The questionnaire was translated into Kiswahili, Kenya's national language, and this version subjected to content validation via three focus group discussions consisting of men and women without cancer aged 18 years and above, attending non-cancer outpatient clinics.Cognitive focus group interviews revealed that compound questions in the BCAM knowledge domain [for example, "do you think discharge or bleeding [italics added] from your nipple could be a sign of breast cancer?"] were confusing to participants, who recommended that