A study is feasible if it is:
- Able to be completed within the proposed timeframe and resources
- Supported by the organisation where it is being undertaken
- The methodology is able to be implemented (eg, sample size and outcome data collection).
A pilot study is the best way to clarify issues of feasibility, and provides very valuable data to help plan the statistical analysis. Trialling the whole research process can ensure that the final study is workable. Audits of the type of data that are to be used in a study are another process which may be valuable in preparing for a study.
Another important reason to do a pilot study as the first phase of a research process is that being able to provide data from a pilot study significantly improves the chances of attracting research funding; indeed, bodies like the NHMRC usually require pilot data before funding a larger project.
Implementing a research protocol involves working out in advance every step of the data collection process, how it will be documented, and how it will then be recorded and analysed. The exact processes will vary according to the nature of the study, but many issues are likely to be common to all research.
The consent process is crucial, and there are some special considerations in palliative care. How patients will be contacted and by whom, how the study will be promoted, and what written information about the study is required as part of the consent process, will all be part of the ethical approval for the study. The practical aspects should be carefully thought out. Legislation related to consent varies in different states.
Studies in which the participant may have impaired ability to consent, for instance studies relating to delirium, may need a proxy to consent as well as consent from the patient. Studies involving children usually require consent from the child and a parent/guardian. In other studies related to symptoms or care of patients who are close to death, for instance studies related to terminal secretions, participants may sometimes be asked to pre-consent before they enter the terminal phase.
Data collection processes usually involve designing a document to complete with the data collected for each subject. This is often called a Case Report Form or CRF. The CRF should be planned with the subsequent data analysis process in mind - to make the next steps of coding and data entry as efficient as possible. Piloting the CRF can be extremely useful.
If the research process uses interviews, the process for documenting and coding these must be considered as the protocol is being developed. If questionnaires are used, how they will be given to the participants, how and when they will be completed, and how they will be returned, must be detailed. The administrative burden for studies involving questionnaires (printing, mailing etc,) may be significant. Sometimes access to a list of addresses may be required for mailing of questionnaires, and there may be privacy and ethical barriers that make this difficult.
Having the written support of a supervisor or the endorsement of a professional body may make this easier to negotiate. Special thought needs to be given to open surveys such as those distributed through other bodies or online. Maximising response rates and reducing the danger of bias from a ‘volunteer effect’ should be considered in planning.
If participant assessments are to be performed – eg, related to problem severity or changes in symptoms, or effects of a treatment, or if the planned interviews are likely to cover sensitive or emotional issues - the skills and experience of the research staff need to be appropriate for these tasks. Sometimes researchers need to consider what they will do if a participant is very distressed or has symptoms that need followup (ie, who they may refer to in that situation, and whether they need the participant’s consent to do so) but also the possibility of interviewers being distressed or needing someone to debrief with may be important to consider as well, depending on the nature of the research question.
If access to patients’ medical records is required, the logistics of this will need to be carefully thought about and negotiated. State and health service requirements about confidentiality and access to casenotes are generally very stringent, and gaining access to some types of information may be difficult, especially if the researcher is not employed within that health service.
Coding and data entry are crucial steps in the study processes. Allowing enough time for these is essential. Considering how they will be done is also important. Generally when qualitative studies are coded, more than one coder is involved, and interviews are transcribed. This is an important cost to budget for and early identification of a research team to share the coding is necessary. If it is a quantitative study, how and by whom data will be entered, and the security of the data, are essential questions to answer.
Piloting the coding frame and documenting key definitions and coding decisions is invaluable to ensure that the quality of the data is reliable, and that problems can be identified early on. The CareSearch Research Data Management System provides valuable support to researchers. It allows online data entry with secure data storage, and can easily be used for multi-site studies. It is free for most palliative care researchers.
Storage of research data
There is a legal requirement to retain research documents, and also the medical records of patients who have participated in clinical trials. The length of time varies according to the type of study - but may be for as long as 20 years. Access to safe and secure archive facilities is therefore required. Electronic data requires specific consideration with regard to storage and security. Ethics committees may expect detailed information about how and where records will be stored, both during and after the study.
Last updated 11 January 2018