Why Do People Choose Not to Have Screening for Hemochromatosis?

Introduction

H FE-associated hereditary hemochromatosis (HH) is a common, inherited disorder of iron overload. About 1 in 150 individuals from the state of Victoria, Australia, who are of northern European descent, are homozygous for the HFE C282Y substitution and, in this population, a minimum of 28% of men and 1% of women have been shown to develop significant disease from iron overload (Delatycki et al., 2005; Allen et al., 2008). Untreated, HH can result in hepatic cirrhosis, hepatocellular carcinoma, cardiomyopathy, and diabetes mellitus, with the average age of onset being 41 years (McDonnell et al., 1999). Mutations in a number of genes can lead to HH; however, homozygosity for the HFE C282Y substitution accounts for the majority of disease with more than 95% of HH in Australia being due to this mutation (Jazwinska et al., 1996). Importantly, individuals at high genetic risk of HH who maintain their iron indices in the normal range by venesection are at very low risk of HH-related morbidity (Allen et al., 2008). The role of genetic screening for HH remains controversial. Issues relate to disease penetrance, potentially creating anxiety in people who would never develop disease and health economic considerations. We have previously shown that a workplace population-based genetic screening program, HaemScreen, is both practicable and acceptable to the community and did not result in significant morbidity for those identified as being at high risk of disease (Delatycki et al., 2002; Delatycki et al., 2005; Nisselle et al., 2006). Here, we present data from the same workplace study, examining why people did not attend information and screening sessions.

Methods

The methods of subject recruitment and data collection for the study have been described previously (Nisselle et al., 2004; Delatycki et al., 2005). Approval for this study was granted by the Ethics in Human Research Committee of the Department of Human Services, Victoria. Workplaces were recruited within four broad categories: corporate, medical/research, education, and government. This included a mix of white- and blue-collar businesses, central city-based, metropolitan, and regional workplaces. Advertising within workplaces was used to inform potential participants about information and in-house screening sessions, including posters, postcards, brochures, e-mails, newsletters, and intranet. Numerous strategies were employed to reduce possible barriers to participation. These included the use of multiple advertising strategies to maximize the receipt of information by employees, conducting multiple sessions at workplaces to facilitate attendance and avoid any travel requirements, ensuring that participation did not cost anything and collecting DNA samples via cheek brush not blood sample.

A self-administered questionnaire, which can be viewed at http://ironxs.com.au/HaemScreen-questionnaire.pdf, was used to assess attitudes of staff who did not attend an information and screening session (nonattenders). The questionnaire provided 15 reasons for not attending a session; individuals could tick as many as relevant and were provided space to add other reasons. Respondents were then asked questions about their health and demographic details. Convenience sampling was used to select workplaces that were representative of a broad range of workplace type and location, and staff age, ethnicity, and gender. The questionnaires were distributed immediately after the information and screening sessions via three, overlapping methods to optimize response rates: (1) staff in common areas were approached directly and asked to complete questionnaires; (2) questionnaires were left in common areas with reply-paid envelopes; (3) key contacts within each workplace distributed paper-based and online versions of the questionnaire. The method of subject recruitment meant that it was not possible to calculate a response rate for this study.

Results

In total, over 100,000 people from 49 workplaces were eligible to participate in the study, of whom 11,841 attended screening sessions (9.3%), and of these, 11,307 (95.5%) had screening for the HFE C282Y substitution. It is not known how many of those people who did not attend screening sessions were aware of the program, and therefore made a considered decision not to attend (Delatycki et al., 2005).

The nonattender questionnaire was distributed at 24 workplaces and completed by 872 individuals. Table 1 compares demographic information for the nonattenders with those who attended information and screening sessions. Using a significance level of p<0.005, the data show that in the nonattenders group, there were more women, more who had partners, and more with university degrees than in the attenders group. Importantly, there was no difference in the proportion reporting northern European ethnicity, proportion who had not heard of HH before or general health perception (using the SF36) between the two groups. While the difference in mean age between groups of 1 year is statistically significant, it is not clinically significant. Overall, the nonattenders sample are considered sufficiently similar to the attenders in relation to the measures recorded, so are considered representative of the wider target population for this program.

Practical constraints accounted for the majority (70.1%) of reasons provided for not attending information and screening sessions (Table 2); these included not being aware of the sessions or being unavailable when sessions were held. The next most common group of reasons given for nonattendance related to the belief that participation was unnecessary for the responding individuals (33.4% provided one of the following reasons for nonattendance; low iron levels/blood donor, healthy already, recent check up/blood test, already been tested for HH, or HH not important/have not heard of it). Importantly, as these individuals did not attend an information and screening session, their level of knowledge about their risk of HH is not known. Some commentators have questioned whether a concern about insurance discrimination is a factor in the choice not to have genetic screening; however, this was only cited by 1.0% of respondents as a reason contributing to their nonparticipation. Similarly, concerns about anxiety if found to be at an increased risk of developing HH, lack of trust in the results, or confidentiality concerns were uncommonly listed (together accounting for 5.1%) as a factor in the decision-making process.

Importantly, 53.4% of nonattenders indicated that they would attend an information and screening session if offered again.

Discussion

In setting up screening programs, it is as important to understand why people do not have screening as why people do have screening, because this will impact on many aspects of delivery and maintenance of programs. This study of 872 individuals who did not attend information and screening sessions that were part of a study of workplace genetic screening for HH found that by far the most common reasons were practical rather than issues related to the screening process or perceptions of harms related to having genetic information. This is supported by the fact that more than half of respondents indicated that they would attend screening sessions if the program were offered again. While intention does not always result in action, this reinforces the finding that any perceived harms relating to participation were not a major factor in the decision-making process for many of these people. A limitation of this study is that we cannot be certain that those who completed the questionnaire are fully representative of all nonattenders. However, multiple workplaces of various workplace types were included in the sampling process.

In the past, concerns about insurance discrimination were regarded as a major barrier to genetic screening (Burke et al., 1998; Barash, 2000). In 2005, findings from the HEIRS study (a North American primary care-based HH screening study) found that 40% of participants agreed genetic screening was not a good idea because of potential issues with insurance (Hall et al., 2005). Yet, later findings from that study showed no evidence of genetic discrimination for HFE C282Y homozygotes 1 year after diagnosis (Hall et al., 2007). The situation with insurance differs in Australia compared to the United States in that only life and disability insurance are potentially impacted by the results of genetic testing, as health insurance is community rated. We reached an agreement with the Australian Life and Disability Insurance Industry to ensure that individuals identified as asymptomatic C282Y homozygotes would not be discriminated against as long as they normalized their iron indices (Delatycki et al., 2002). Only 1% of respondents in the current study indicated concerns relating to insurance were a factor in their decision not to participate. In addition, we are not aware of any of the 47 C282Y homozygotes newly identified in this study having had any problems in obtaining insurance (Delatycki et al., 2005).

Our finding that practical issues were the main barrier to participation in screening is supported by findings from other studies, with similar reports of practical reasons outweighing concerns about insurance as a cause of nonparticipation in genetic screening for HH (Humphreys et al., 2000; Patch et al., 2005).

The Health Belief Model (HBM) is a theoretical framework used to explain health behaviors, including those that relate to avoidance of screening. The HBM includes internal constructs such as perceived benefits of, and barriers to, the health behavior in question. When looking at genetic screening, perceived severity of the disease being screened for and perceived risk of developing the disease also play a part in determining screening uptake behavior (Strecher and Rosenstock, 1997). Barriers to attending HaemScreen information and screening sessions were minimized where possible, such that the program was made readily accessible to those who saw the advertising and were available to attend. For example, sessions were held at multiple convenient times and locations tailored to the peculiarities and practices at each workplace. Of the people who attended a HaemScreen session, 95.5% went on to be screened for HH. The participation rate in the study was 5.8% in the first part of the program (Nisselle et al., 2004), but by the end was 9.3% (Delatycki et al., 2005). This increase may have been due to changes in how the program was delivered by removing as many barriers as possible. However, among those who did not attend a session, motivation to be screened may have been decreased through poor knowledge of HH risk and severity. Only 17.7% of the 11,841 who attended sessions had previously heard of HH. The nonattender data presented here support the notion that many of those who did not attend an information session had a poor understanding of HH and genotypic screening, so were not motivated to do so: 79.2% of nonattenders had not heard of HH before completing the questionnaire, and 33.4% indicated that they did not think screening was necessary for them (listing reasons such as “I am already healthy”, “I have had a recent check up/blood test”, “I have a diet very low in iron/am a blood donor”, “HH is unimportant/I have not heard of it,” and “I have already been tested for HH” as explanations for their absence). Most of these responses indicate an inaccurate understanding of HH and so highlight the knowledge gap in this population.

Conclusion

The nonattender data presented here indicate that concerns about insurance, anxiety, and use of genetic information are not major factors for why people did not attend workplace information and screening sessions for HH. Practical barriers were the major reasons identified. This highlights that when implementing screening programs, as many practical barriers as possible need to be overcome, so that a maximum number of people who would like to be informed about screening are given the opportunity to do so. Approaches to overcome some of the barriers identified include conducting the program on multiple occasions and at multiple venues, multiple methods, and instances of advertising the program, and education of the community about the condition that is being screened. Additionally, minimizing cost associated with participation in the program and collecting DNA samples via cheek brush sample rather than blood sample are very likely to improve uptake of genetic screening.