This exploratory study highlighted the main reasons for patients' refusal to comply with upper gastrointestinal endoscopy. Even though patients seem to accept the necessity and the importance of an EGD for their health, they expressed certain barriers and fears towards the examination as their main reasons for noncompliance. Various factors that may influence adherence to screening tests (such as breast and colorectal cancer (CRC) screening) have been published in the literature [17–19]. With the fear of pain and fear of the procedure itself noted as important barriers for performing a mammography [17, 19, 20]. Studies concerning compliance to endoscopy for CRC screening revealed that the fear of pain, concerns and anxiety towards the procedure served the role of negative predictors for any form of CRC screening [18, 21]. Similar fears and concerns were disclosed in this exploratory study.
The patient's level of education and knowledge of the procedure has been identified to play a critical role in adherence of the GP's instructions [22, 23]. However knowledge has not been proven to be the only crucial factor to impel a certain action [14, 24]. Almost 9% of the participating patients declared that additional data and explanation was needed, even though effort was made by the physicians (during consultation) to explain the need and the importance of endoscopy as well as describe the procedure itself. Many patients also reported that their own previous experience was either traumatic or not reassuring enough for them to proceed with a new endoscopy. Future intervention research should focus more on controlled beliefs and study their effectiveness on compliance rates to upper endoscopy.
Another interesting finding of our study was that the GP's recommendation for endoscopy (question 3) did not seem to facilitate patients to undertake the control of their behavior. It is known from the literature that advice, recommendation or encouragement from the personal physician can increase the likelihood of attendance of a certain examination [25–27]. Nevertheless in this study population, patients' perceptions seemed to play an important role as well, since 18% of the patients appeared to question their GPs' recommendation. Moreover, patients who undergo endoscopy generally report a larger number of, and more severe symptoms than those who do not . Accordingly, upper gastrointestinal symptoms do not seem to have been considered as severe enough to obligate the patients to undergo an EGD.
Although no direct question in the interview focused on exploring normative beliefs and particularly to assess the extent that the patients' behaviour depends on their friends, family and the society to perform the recommended behavior, nevertheless some patients reported unpleasant experiences of individuals in their close environment as a main reason for avoiding the endoscopy.
Health problems and current life demands were the most common external determinants causing non compliance to EGD; yet when patients were asked in specific, for a reason of inconvenience to endoscopy these factors were impressively diminished. It seems that although undoubtedly important, these external barriers that were indicated during interviews were often vague and other obscure reasons for non adherence were referred which would need further investigation .
Different health insurance coverage seems to have a negative effect on the use of CRC tests [18, 28]. However this was not the case in this study, since all EGDs were fully covered by the patients' insurance. The access to the referral centre and the long waiting times are other crucial factors of non adherence to recommendations but this factor cannot explain the high rate of non compliance in the study [17–19]. Strategies were utilized to overcome any waiting problems by scheduling the upper gastrointestinal endoscopy within a period of ten days. Access was not a main obstacle for the patients because there were no differences in compliance between the most and the less proximate to the hospital practices. In the TPB, demographic and personal characteristics of the patients may influence behaviour indirectly by affecting behavioural, normative, and control beliefs. The male gender was the only socio-demographic factor that predicted a tendency of non adherence in the population studied. This seems to be in controversy with other studies where female gender is the non compliance factor for CRC screening [18, 28].
This study followed mainly a qualitative approach and consequently its findings cannot be extended to wider populations. Also, the population studied poorly represented minorities and was not socioeconomically diverse. However, its low adherence suggests that this is likely to be an even greater problem among less socially advantaged patients. Patient-level influences on adherence have been determined, through interviews on a small sample of patients. Consequently the results are not strong enough to permit us to draw firm conclusions. The interviews reflect only the perceptions of patients who refused endoscopy and cannot be compared with the perceptions of patients who had undergone endoscopy.
Nevertheless, this exploratory approach was warranted because no prior studies have approached the causes of refusal of upper gastrointestinal endoscopy. The number of the interviewed subjects was higher than that in usual qualitative studies, due to the design of the study which was initially a part of quantitative research, and at a second stage we decided to focus more on an exploratory qualitative direction.
The TPB was successfully utilized to explain patient's answers. Compared to affective processing models, the theory overlooks emotional variables such as threat, fear, mood and negative or positive feeling and assesses them in a limited fashion. In the health related behavior situation, given that most individuals' health behaviors are influenced by personal emotion and affect, this could be a drawback for predicting health-related behaviors .