To our knowledge, this is the first review on non-participation in vaccination programs as well as population-based screening programs embedded in General Practice.
In total, 28 studies in 24 original populations matched our search criteria regarding three program types: vaccination, early detection of disease, and identification of high risk of a disease. Non-participation levels ranged between 10% and 99% in all three program types. Median non-participation level was 38%. This review showed that program type does not seem to influence non-participation in population-based disease prevention programs in General Practice. In addition, no differences were found between the three program types with respect to factors related to non-participation. However, within each program type, and even within programs targeting the same disease, considerable differences were shown, which may be partly related to differences in recruitment strategies used.
Three groups were identified which might need specific attention, since they are less likely to participate in population-based disease prevention programs in General Practice; younger individuals, the people living in a socially deprived area, and former non-participants. In general, non-participants found participation not to be of personal relevance.
A previous review on vaccination programs also showed younger individuals to be less likely to participate, however, screening studies showed ambiguous outcomes of age
]. Living in socially deprived area was also shown to be related to non-participation in vaccination programs
] as well as in colorectal and cervical cancer screening
]. Previous non-participation, either in the same or other programs is a strong predictor of current non-participation
], which was also reflected in the lower likelihood of non-participation in individuals with a higher health care consumption in the diabetes screening programs
]. An explanation could be that people get familiar with the tests and procedures involved and overcome their barriers.
Lack of personal relevance and low perceived risk are important factors in non-participation
]. People may not respond since they feel and think they live healthy
]. However, risk of disease is also a difficult concept for people to understand and often underestimated, especially in diseases that may be preventable by behavioural change
The lowest non-participation levels were seen in programs applying more active and personal approaches, including recruitment in General Practice. A passive approach including recruitment by posters and leaflets in the waiting room of the General Practice practice, resulted in only little participation
]. Personal contact with a GP or nurse therefore seems to play a decisive role in increasing participation
]. Two other reviews showed that invitation, reminding, and counselling by telephone all seem to be effective interventions to increase uptake, however, they are much more intensive and expensive
]. Furthermore, the results of this review indicate that it might be important to more specifically address to the personal relevance of vaccination and screening. Recent studies showed that the concept of risk is better understood if information is presented in terms of natural frequencies rather than as an absolute or relative risk
]. In addition, low future time orientation, as often seen in the socially deprived, may be addressed to, since this is associated with a lower likeliness of changing (health) behaviour. Invitations should therefore stress the immediate benefits and remove barriers to e.g. participate in vaccination or screening and/or changing lifestyle
Noticeably, relatively few programs were aimed at identification of high risk of disease. For example, only one out of eight of programs focused on diabetes was aimed at identifying high risk of this disease. The relatively low participation level in this study may have been due to the fact that the invitation letter informed individuals that when they were at high risk of diabetes they would be offered a lifestyle intervention. The prospect of having to change ones lifestyle may have discouraged some people to participate
Our inclusion criteria are mainly related to health care systems in which the GP is a gatekeeper, which is the case in the Netherlands. Within this system GPs are able to pro-actively invite potentially high risk individuals within their patient population. This is reflected in the studies presented in this review. In countries with a non-gatekeeper system, the described programs would be performed in an opportunistic manner, which was an exclusion criteria in this review. Furthermore, population-based prevention programs in Europe are generally free of charge, while individuals in the USA have to pay or get reimbursement through their insurance. Generalizability of the results of this review, therefore, may be limited.
Some limitations of this systematic review can be noted. First, as with systematic reviews in general, despite of a thorough search strategy, relevant (un)published studies may have been omitted, as well as non-English or Dutch articles. Adding the concept ‘feasibility’ to the search strategy may have resulted in more included studies. Additionally, research on breast- and colon cancer screening was not retrieved since these programs need more specific equipment, which is usually not available in General Practice practices. Furthermore, although participation may have been studied, the results may not have been included in the final article. Next, the relatively small number of studies included, and the large heterogeneity in study characteristics and results require cautious interpretation of the results of this review.
Additionally, six included studies were part of the same international trial. Therefore, although we tried carefully to prevent presentation of duplicate data some bias may have been introduced. However, in the UK and Denmark arm of the study different target groups and recruitment strategies were used. In addition, the majority of the included studies were trials instead of real-world programs. This may have led to an overestimation of the participation levels, since these studies may be more likely to be performed in well-organized General Practice practices in less deprived areas. Finally, although we identified some factors that might be related to non-participation, it remains unclear from this review how these factors interrelate.
The findings of our systematic review provide guidance for future research and General Practice. Only 17% of the studies addressed both characteristics and reasons of non-participation. To gain more insight in who is not willing to participate in disease prevention programs and why, it is important to assess both.
Future research on willingness to participate should focus on programs aimed at identifying high risk of disease, since they may have great potential in improving population health but are also more demanding to people than the other two types. Additionally, research should be focused on the consequences this may have for recruitment in these programs.
Recruitment strategies should address to personal relevance, and to be tailored to the specific needs of subgroups of individuals. Those who are less likely to participate, such as individuals living in a deprived area, former non-participants and younger individuals, may need to be more actively recruited, in which contact with a health professional may be pivotal, especially when invited for the first time. For example, GPs and practice nurses may emphasize the importance of participation during regular encounters. In those who are more likely to engage in vaccination and/or screening, an invitation letter signed by the GP may be sufficient.