Most decisions in health care require a balancing of harms and benefits. Although people may make these decisions intuitively, informed decision making requires an explicit balancing of harms and benefits. Individuals making an informed decision therefore need to be aware of the harms and benefits relevant to their own situation. In situations with multiple outcomes, balancing harms and benefits for each of those outcomes to come to an overall decision may be challenging. Especially if individuals have varying risk profiles and differing preferences. The decision to participate in a nationwide screening program may be one of those situations, where different persons may weigh the potential harms and benefits of screening in different ways, and have a different risk profile for the screening condition.
The Dutch colorectal cancer (CRC) screening program started in 2013. The ‘wet bevolkingsonderzoek’ (law on population screening) states that possible participants are given the opportunity of a well informed choice. The decision to take part in a screening program is an individual choice, based on individual believes about screening and perceived risks. In contradiction to this, the information provided to those who are invited, is, till date, not individualized, but consists of general information. At the moment, the provided information is based on population risks and general pros and cons, and is not specified for individual risks, benefits and harms and preferences.
In this research project, we will use a recently developed quantitative benefit harm analysis model to weigh the benefits and harms according to individual preferences and an individual risk profile. This will allow us to move from overall guidance that leaves the potential participant insecure about how to weigh all received information to a personalized advice in which the harms and benefits are explicitly weighed and explained on a personal level. The upcoming of web based tools and better visualization options make it then possible to deliver this individual information in an informative way to the participant. Our end-product will therefore be a web-based personalized decision aid, which will enable invitees to make an informed decision about their screening participation.
For developing this web-based personalized decision aid, we take three steps.
(1) We will assess the benefit-harm balance for taking part in the current national CRC screening program versus not taking part in the screening program and whether this balance depends on certain personal characteristics and preferences.
(2) Based on these analyses, we will develop a web based decision aid, allowing potential participants to adjust the estimate of the benefit-harm balance of screening according to their personal preferences.
(3) We will compare the effects of this decision tool on the extent being informed, versus the usual way of informing potential participants.
In the first step, we will use the previously developed Adenoma and Serrated pathway to Colorectal CAncer (ASCCA) model to estimate the outcome risks within the screening program. Potential harmful outcomes are the different forms of CRC and subsequently risks of mortality, but also perforation due to colonoscopy, while having or not having CRC. Potential beneficial outcomes may be treatment success or reassurance when having no CRC. These outcomes are then weighed according to their importance for the participants. These preferences will be gathered through focus groups and in-depth interviews. In the final analysis, the outcome risks and weights are combined using the Gail/NCI approach. This approach was chosen because it can consider risks of multiple benefit and harm outcomes and the importance of these outcomes, summarized in one number, the benefit-harm index.
This benefit-harm index will be subsequently transferred to a color scheme, where different colors indicate that for this individual the harms of screening will exceed the benefits; benefits and harms are likely to be equal; or that the benefits exceed the harms. This color scheme will be seen on the screen, and participants will be able to change the schemes by adjusting preferences (see for example www.benefit-harm-balance.com/).
To assess the effect of the online decision aid on the level of user knowledge, the decision aid will be user-tested, adjusted and then compared against current practice in a randomized controlled trial, including representatives of the general public between 55 and 75 years of age. Finally, the decision aid will be promoted, providing potential participants a method to consciously make a decision about participating in the screening program.