Number of participants
It is important to calculate the number of participants (ie sample size) you need in your study to be able to answer your research question. More participants than are needed to answer the question mean unnecessary patient burden, additional costs and time. Yet having too few participants may mean you are not able to observe clinically important differences. If this happens, a study is said to be under powered. Advice from a statistician on the required sample size can be critical.
Reaching the desired sample size can be a problem for some studies. There may be few suitable participants at a single site, or within a time frame. The requirement for adequate powering of studies may make multi-site research a necessity.
Early involvement of a statistician is crucial.
Trial participants
It is worth remembering that participants in palliative care studies may possibly be very ill and may only be able to participate for a limited time. This means there are several important research protocol considerations:
- Make the research protocol as short as possible to reduce expected withdrawal and burden
- Test and justify each data tool used to ensure appropriateness and burden in a palliative care population
- Include expected withdrawal rates in the initial sample size calculation.
When participants in studies are as ill as those in palliative care, adverse events may be more likely to occur. The research protocol should outline specific management approaches instigated to reduce and track adverse events:
- Investigators should define any expected potential adverse events in the protocol where possible; and these should be collected using validated measures if possible
- Specific data should be collected during the course of each participant's data collection to identify any adverse events
- Develop a standard operating procedure for adverse event reporting to the relevant research ethics committee(s) if one is not already available
- Track all adverse and seriously adverse events throughout the study
- Death may not necessarily be a serious adverse event in palliative care and this needs to be considered within the research setting.
Clinical service participation
Research activities may create a burden for the clinical service within which they are conducted. It may also be challenging if clinicians don't fully understand the need for research in this discipline. Help create understanding and reduce burden by:
- Meeting with the staff to understand perceptions of benefits and burdens of research for patients, staff and the unit
- Providing some level of infrastructure support for the research activities if possible (eg, administrative support to help with recruitment)
- Including the clinical staff in the research design where appropriate
- Holding regular meetings with clinical staff to monitor any burden and to let them know how the study is going
- Developing site and study specific recruitment strategies, eg, working with a respiratory clinic to recruit participants for a dyspnoea study
- Scripting and role playing how the research can be introduced to clinicians, patients and their caregivers
- Offering clinical staff the opportunity to participate in relevant research roles.
This page was created on 25 March 2008 and is due for review in March 2010
Last updated 26 May 2008