Methods
The methodological approach adopted for this analysis was Qualitative Description. This is an inductive method that uses the naturalistic language of the informants to portray their 'perception and experience of the world and its phenomena' with an emphasis on precise low-inference description, rather than in-depth interpretation or development of an explanatory hypothesis. It is particularly suitable for 'mapping the terrain' of a poorly explored topic and gaining initial insight into informants' views. It can also incorporate quasi-statistical descriptive data, such as the number of times a posting has been viewed, to illustrate participant experience. Furthermore, when analysing web-forum postings it is not possible to probe participants for the very full and detailed narratives that would facilitate more in-depth qualitative analysis. For these reasons we considered Qualitative Description as an appropriate method for this study.
Text used for analysis was downloaded from the COBF website to Word and was then exported to the qualitative data management programme NVivo. The text was read and re-read multiple times, leading to the identification of individual codes that were identified inductively from the data. These were then grouped according to common ideas into related categories, and finally incorporated into five overarching main themes that are presented in this paper. One researcher led the analysis (AF). Coding and identification of themes were discussed in detail with a second researcher (FB) who was also familiar with the raw data and again among the whole research team to avoid idiosyncratic or selective interpretations and offer diverse perspectives. Using NVivo facilitated the maintenance of a detailed audit trail. Initially the 'Meet and Greet' section of the cystitis section of this web forum was analysed because it introduced many of the key issues developed throughout the site and provided background information on many of the women who use the forum. The remainder of the site, covering (at the point of analysis) 915 separate topics specifically relating to bladder infection comprising of 5,386 postings (see Table 1), was then analysed to provide a more detailed evaluation, looking for contradictory viewpoints, and to investigate areas not discussed in 'Meet and Greet'.
Permission was sought from the COBF and from Southampton University Ethics committee to analyse data that was publicly available on the website. This was granted in February 2012 (University of Southampton Ethics and Governance Online ID 946) and analysis of all material posted on the site from its inception in 2004 began in the same month and was completed in June 2012.
In recent years there has been an 'exponential' increase in the use of Internet websites to provide health related information and to operate as forums where site users can share experiences, discuss treatment options, and gain support. There has been a parallel increase in web-based research to analyse these data and many studies now report on the content, form, and function of these sites for a wide range of health related conditions. Web-based research into the experience of an illness such as RUTIs has the advantage of enabling access to many women from a wide geographical spread and diverse socio-economic backgrounds with 80% of UK households having internet access in 2012. Ecological validity is enhanced because material reflects women's spontaneous communication in their own words about their experiences. An internet forum also allows participants to preserve their anonymity, which may make participants more comfortable discussing personal, sensitive issues.
The main ethical issues related to this research revolved around the use of women's responses on a web forum without their explicit permission. We have adopted the recommendation for an inductive, casuistic approach to these issues that takes into account the specific context of the women involved and the site being investigated, and balances the potential harms and benefits that could result from this research. Consequently data has been drawn only from the publicly accessible forum and not from any password protected, private chat rooms that women could choose to use on the same site. As a result women participating on the site can be considered as providing their tacit consent for open access to these discussions. To help preserve their anonymity the names of women quoted in the paper have been changed. Finally, with the permission of the site managers, a version of this paper and an accompanying letter was posted on the COBF website explaining the rationale, methodology and purpose of the research. The letter clearly gave site users the option of withdrawing any comments they had posted from use in this paper. One site user did make contact and gave her permission for her material to be used as long as it was done so anonymously.
Legally the copyright of the text belongs to the COBF and it is clearly stated that 'The COB Foundation is the owner or the licensee of all intellectual property rights in our site and the material published on it'. Permission was requested from the COBF and granted prior to commencing this project.
We believe that this research conforms to the recent ethical recommendations for Internet mediated research outlined by the Association of Internet Researchers and the British Psychological Society. It is respectful of the women who contributed to the forum, and uses a scientifically valid methodology to present and analyse the findings. It is socially responsible, and may contribute to improving the future provision of care for this group of women. For these reasons we feel that this web-research project is ethically justified.