Perceived barriers to seeking cancer care in the Gaza Strip: a cross-sectional study

Background Several factors might contribute to the delay to get cancer care including poor cancer awareness and having barriers to seeking help. This study described these barriers in Gaza and their association with recalling and recognizing cancer symptoms and risk factors. Methods A cross-sectional study was conducted in Gaza. Adult visitors (≥18 years) to the largest three governmental hospitals and adolescent students (15–17 years) from ten high schools were recruited. A translated-into-Arabic version of the validated Cancer Awareness Measure (CAM) was used to collect data in face-to-face interviews. It described demographic data, barriers to seek cancer care as well as recall and recognition of cancer symptoms and risk factors. Responses were compared between adults and adolescents as well as males and females. Results Of 3033 participants approached, 2886 completed the CAM (response rate= 95.2%). Among them, 1429 (49.5%) were adult (702 females; 49.1%) and 1457 (50.5%) were adolescent (781 females; 53.6%). The mean age± standard deviation (SD) of adult and adolescent participants was 33.7±11.7 years and 16.3±0.8 years, respectively. Emotional barriers were the most common barriers with ‘feeling scared’ as the most reported barrier (n=1512, 52.4%). Females and adolescents were more likely to report ‘feeling scared’ as a barrier than males and adults, respectively. Higher recall scores for cancer symptoms were associated with lower likelihood to report ‘embarrassment’, ‘worry about wasting doctor’s time’ and ‘difficulty arranging transport’. This was also seen for recalling risk factors, where ‘embarrassment’ and all practical barriers showed significant inverse associations with higher scores. In addition, greater recognition scores of cancer risk factors were inversely associated with reporting ‘embarrassment’ and ‘feeling scared’. Conclusions The most commonly perceived barriers to seeking cancer care were ‘feeling scared’ and ‘feeling worried about what the doctor might find’, followed by practical and service barriers. Females and adolescents were more likely to report ‘fear’ as a barrier to seek medical advice. Having a higher recall of cancer symptoms and risk factors was inversely associated with reporting most barriers. To improve patient outcome, early presentation can be facilitated by targeting barriers specific to population groups. Supplementary Information The online version contains supplementary material available at 10.1186/s12913-020-06044-1.


Study design and population
This was a cross-sectional study conducted from the 1st of September to the 31st of October 2017 in the Gaza Strip, Palestine. It described people's challenges and barriers to seek medical advice for cancer symptoms and tested the association of recalling and recognizing some cancer signs, symptoms, and risk factors with the reported barriers.
A translated-into-Arabic version of the Cancer Awareness Measure (CAM) questionnaire was used as an assessment tool, which was developed and validated for such purposes in the United Kingdom [9]. The questionnaire consisted of four sections; the first described the demographic data. The second reported the barriers to seeking cancer care. The barriers were further categorized into emotional, practical, and service barriers, and were compared between adolescents and adults as well as between males and females. The third section comprised two open-ended (recall) questions to list cancer signs/symptoms and risk factors while the fourth section included 16 closed (recognition) questions to identify eight cancer signs/symptoms and eight risk factors. A 3point scale (no, I do not know, yes) was used to test the recognition of cancer signs/symptoms, while recognition questions of risk factors were based on a 5-point Likert scale (1= strongly disagree, 5= strongly agree).
Two bilingual experts translated the CAM from English to Arabic and another two bilingual experts backtranslated it from Arabic to English. Before starting data collection, a pilot study was conducted with 119 respondents to test the clarity of the questions of the Arabic version of the CAM. Assessment of internal consistency was carried out on the used scale comprising 16 items. Cronbach's alpha showed the questionnaire to reach acceptable reliability, α = 0.71.

Sampling methods
Cancer care in the Gaza Strip is provided by governmental, non-governmental organizations, and a few private providers. Governmental hospitals are the main entry point for healthcare services for people who have medical insurance. The Palestinian government's health insurance covers basic cancer therapies, while UNRWA sponsors the care of cancer patients by covering the cost of enrolling them in the government-funded health scheme [4]. Females represent 49.3% of the total population of the Gaza Strip, while adolescents represent about 23.0% [1]. There are 13 hospitals run by the Palestinian Ministry of Health in the Gaza Strip [1]. From these, the largest three hospitals, located in three geographical locations in the Gaza Strip, were selected to recruit participants by stratified sampling. Adults (≥ 18 years) admitted to or visiting those hospitals were the target population. Simple random sampling was used to recruit visitors or patients who were approached by data collectors on hospital grounds. Patients or visitors to the oncology departments at the participating hospitals were excluded from the study. A total of 1483 adult participants was approached in the included hospitals.
In parallel to this, stratified sampling was used to recruit adolescent participants from ten high schools (out of 147 high schools in Gaza [10]). Those high schools were in the same areas as the included hospitals to achieve uniformity of areas. Simple random sampling was then used to choose six classes in each high school. All students in the included 60 classes were invited to participate (n= 1550). Adolescents willing to participate were asked by data collectors to have a face-to face interview to fill out the CAM. The inclusion of both adults and adolescents gives the potential to have recommendations on future prevention interventions [5,6].
According to the Palestinian Ministry of Health report in 2017 [11], there was a total of 190,208 adolescents in the Gaza Strip. Among those, 92,850 (48.8%) were females and 97,358 (51.2%) were males. In addition, there was a total of 901,640 adults, of whom 500,207 (55.5%) were women and 401,433 (44.5%) were men. Based on the sample size calculator of the Australian Bureau of Statistics [12], with a confidence level of 95.0% and a margin of error of 3.0%, the calculated minimum sample sizes were 1062 and 1066 for the strata of adolescents and adults, respectively. The calculated minimum sample sizes were 1066 and 1065 for strata of females and males, respectively. A total sample size of 2131 was needed.
Data collectors were trained to recruit participants and facilitate the completion of the CAM. They were also trained to clarify the questionnaire to illiterate participants. Prior to completing the questionnaire, a detailed explanation of the study including its purposes was given to the participants. Written informed consent was obtained from the participants and from their legal guardian if they were under 16 years old. Ethical approval had been obtained from both the Palestinian Ministry of Health and the Ministry of Education and Higher Education prior to data collection.

Statistical analysis
Descriptive statistics were used for demographic data and the participants' reported barriers. For every correctly recalled sign/symptom or risk factor, the participant was given one point. Moreover, participants who answered symptom recognition questions correctly (yes) were given one point and incorrect answers (no or do not know) were given no point. Similarly, people who responded to recognition questions by 'agree' or 'strongly agree' were given one point for every cancer risk factor. Those who answered with 'disagree', 'strongly disagree' or 'not sure' were given no point. The total for each of recall and recognition scores for cancer symptoms and risk factors was then calculated.
The chi-square test was used to compare barriers in males vs females and adults vs adolescents and to test the association between belonging to any of these population subgroups with having the reported barriers. Univariable logistic regression was utilized to have an idea about the magnitude of that association followed by a multivariable logistic regression to adjust for other covariates. The multivariable logistic regression was also used to test the association between recall/recognition scores for cancer symptoms and risk factors with having barriers to seeking cancer care. Results of the univariable and multivariable logistic regression were reported as odds ratios (OR) with corresponding 95% confidence intervals (CI) and p-values. Data were analyzed using Stata software version 16.0 (StataCorp, College Station, TX).

Barrier to seeking cancer care
Overall, emotional barriers were the most reported, followed by practical and service barriers ( Table 2).  Difficulty making an appointment with the doctor was the most frequently reported service barrier (n=900, 31.2%). In the multivariable logistic regression, both females and adolescents were less likely than males and adults to report difficulty making an appointment with the doctor as a barrier to seek medical advice (OR= 0.83; 95% CI: 0.71-0.97; p=0.023 and OR= 0.75; 95% CI: 0.64-0.88; p< 0.001, respectively). However, men and adolescents showed more likelihood than women and adults to find difficulty talking to the doctor (OR=0.67; 95% CI: 0.56-0.80; p< 0.001 and OR= 1.48; 95% CI: 1.24-1.78; p< 0.001, respectively). Being busy to go to the doctor was the most observed practical barrier (n=991, 34.3%) followed by difficulty arranging transport (n=979, 33.9%). Females had a lower likelihood than males to be busy to go to the doctor (OR= 0.81; 95% CI: 0.69-0.94; p=0.008), while adolescents were more likely than adults to have this barrier (OR= 1.40; 95% CI: 1.19-1.64; p< 0.001).
Relationship between recalling/recognizing cancer symptoms and reporting barriers to seek cancer care After adjusting for age and gender, participants who had higher recall scores for cancer symptoms were found to have a decrease in the likelihood to report 'embarrassment' (OR= 0.90, 95% CI: 0.82-0.97, p=0.010), 'feeling worried about doctor's findings' (OR= 0.89, 95% CI: 0.80-0.98, p=0.024) and 'difficulty arranging transport' (OR= 0.91, 95% CI: 0.83-0.99, p=0.034) than those with lower scores (Table 4). No associations were noticed between recognition scores and reporting any barrier.

Discussion
'Feeling scared' and 'feeling worried about what the doctor might find' were the most common perceived barriers among Palestinians living in the Gaza Strip. Females and adolescents were more likely to report 'feeling scared' as a barrier than males and adults, respectively. On the contrary, females and adolescents were less likely than males and adults to report difficulty making an appointment. Adolescents were more likely to report 'busy to go to the doctor' as a practical barrier to seek medical advice. Significant associations between recall scores of cancer symptoms and reported barriers to early presentation were found for 'embarrassment', 'worry about doctor's findings' and 'difficulty arranging transports', where higher recall scores were associated with lower likelihood to report each barrier. On the other hand, recognition scores for cancer symptoms showed no significant associations. Reporting 'embarrassment' and all practical barriers showed significant associations, where higher recall of cancer risk factors was associated with less likelihood to have these barriers.

Impact of knowledge on perceived barriers to seek medical advice for cancer
Delay to cancer diagnosis can occur in any or all of three stages; presentation to the doctor (also called the patient interval), diagnosis, or access to treatment [13]. While diagnosis and treatment intervals were found to be comparable in low-and middle-income countries (LMICs) with those in high-income countries, the patient interval was found to be longer in LMICs [14,15]. A prolonged patient interval is a significant contributor to worse patient outcomes and a factor for 70% of global cancer mortality occurring in LMICs [16,17]. Poorer knowledge of cancer signs and symptoms was found to prolong the patient interval in LMICs as well as high-income countries [5,6,[18][19][20][21][22][23][24][25][26][27][28]. This might be at least partly caused by the fact that less knowledge may drive people to have more barriers to early presentation and vice versa. In this study, too, participants with greater knowledge of cancer symptoms were less likely to report barriers, especially 'embarrassment' and 'feeling worried about doctor's findings', illustrating the empowerment knowledge can give to people when seeking care for possible cancer symptoms. Furthermore, participants with better awareness of cancer risk factors also reported significantly less service barriers, demonstrating that greater knowledge of cancer risk factors has the potential to reduce barriers to seeking medical advice and, thus, promote early presentation.

'Fear' and 'worry'related barriers
Prevalent negative beliefs among the population around a cancer diagnosis could be one reason for the greater impact of 'fear' and 'worry' on reported barriers in LMICs compared to high-income countries, where practical barriers were most frequently reported [17,20,[29][30][31]. As in other low-income settings, stronger negative beliefs could be associated with a cancer diagnosis and could result from experiences of higher mortality and less chances of cure [32][33][34][35][36]. This assumption is underpinned by the fact that the most frequently reported barrier was 'feeling scared', illustrating the belief of cancer to be fatal [37][38][39]. This is difficult as late presentation prolongs the patient interval and, hence, also the time to diagnosis, exerting further negative impact on patient outcomes and reducing chances of cure [14,17,40]. In fact, in the Gaza Strip, 60% of the deaths due to breast cancer in 2016 (643 of 1072 deaths) were found to be avoidable, often due to late diagnosis [41], underlining the justified basis for negative beliefs to rise and disseminate within the population. Therefore, barriers can be reinforced by local experience and beliefs [42][43][44]. On the other hand, improved knowledge on cancer treatments and outcomes could address negative beliefs and remove barriers to presentation and by this promote early presentation. Participants with more knowledge about cancer risk factors reported less practical barriers to seek medical advice, such as 'difficulty arranging transport to see the doctor' or 'having other things to do', highlighting the fact that increased knowledge may play a role in mitigating barriers. Increased public knowledge of cancer symptoms, risk factors as well as management could be achieved by a combination of public awareness campaigns and individual interventions on school and community levels, which has been shown to be effective in highincome countries [45][46][47][48][49]. Funds are usually lacking for such interventions in LMICs; however, in the long-term, reduced morbidity and mortality might make such investments cost-effective also for low-income settings [50][51][52].
Differences and similarities related to gender and age Women and adolescents were more likely to report 'fear' as a barrier than men and adults, which has also been found by other studies [27,53]. Adolescents were also found in other studies to express 'fear' as an important barrier [54][55][56]. A school-based intervention program, aimed at increasing knowledge on cancer signs and symptoms was able to reduce such barriers, at least in the short-term [48]. Although adolescents were shown to retain the knowledge after six months, this knowledge had lost its positive effect on reducing barriers after that time. Therefore, effective intervention programs can work for adults and adolescents but must address knowledge as well as barriers to seeking help for greatest impact.
In this study, adolescents reported difficulty talking to the doctor, possibly due to a disinclination to express emotions to healthcare professionals or inexperience in doing so [57]. Men and adults identified difficulties in making an appointment with the doctor significantly more than women and adolescents. Women engage generally more with the healthcare system in Gaza than men, as women take up medical care regularly during pregnancies and childbirths and, thus, have more experience in making appointments with doctors. These differences illustrate that next to knowledge also culture influences health-seeking behaviour. Therefore, while good knowledge is associated with having less perceived barriers to help-seeking with cancer symptoms, it is not the only factor. In order to significantly reduce the patient interval, barriers to seeking medical advice have to also be addressed to promote early diagnosis and reduce excess mortality from cancer [17,41].

Future directions
In order to improve survival outcomes for cancer in low-income settings, like the Gaza Strip, it is essential to reduce the interval to diagnosis by promoting early presentation with symptoms. Essential interventions to achieve this have to focus on reducing barriers to presentation, such as 'embarrassment' or 'fear' or 'worry about what the doctor might find'. This can be achieved by nurturing greater public awareness and including more health content in school curricula, as greater knowledge was associated with less barriers in this study. Furthermore, reducing practical barriers by improving access to healthcare as well as the quality of care available to people will reduce barriers of 'fear' and 'difficulty making an appointment with a doctor', which were noticed in this study.

Strengths and limitations
This study had a large sample size and could be considered representative of the Gaza population. The study also had a high response rate and used a validated instrument, the CAM. In addition, the inclusion of both adults and adolescents provides the opportunity for additional recommendations on prevention interventions.
However, adult participants were mainly recruited from among the visitors to governmental hospitals (excluding oncology departments). This might have led to recruitment of people displaying a certain degree of help-seeking behaviour and, thus, leading to some degree of selection bias. On the other hand, hospitals are busy places in the Gaza Strip with two, three or more visitors present with patients at all times, in addition to relatives waiting outside for news. Therefore, visitors did not necessarily display help-seeking behaviour themselves. Furthermore, this study surveyed healthy participants, who had not experienced a cancer diagnosis and, hence, looked at 'anticipated' feelings and behaviours. This could differ from actual feelings, experiences and behaviours in presence of cancer symptoms. In order to look at this more closely, further research should examine potential delay in the three stages of presentation, diagnosis and treatment experienced by cancer patients.

Conclusion
The most commonly reported barriers to seeking cancer care were 'feeling scared' and 'feeling worried about what the doctor might find', followed by practical and service barriers. In this study, females and adolescents were more likely to report 'fear' as a barrier to seek medical advice. Having a higher recall of cancer symptoms and risk factors was inversely associated with reporting most barriers to seek cancer care. In order to improve cancer outcomes in LMICs, such as the Gaza Strip, reducing the patient interval is essential. This can be achieved by targeting interventions to population groups with view to their specific barriers.

Additional file 1.
Abbreviations UNRWA: United Nations Relief and Works Agency; CAM: cancer awareness measure; LMICs: low-and middle-income countries; CI: confidence interval; OR: Odds ratio