Introduction
Delivering health care services to remote populations is a challenge to public health
systems worldwide. Data from Australia show that people living in remote and rural
areas tend to have lower life expectancies, higher rates of injuries and diseases,
poorer access to health care services, and lower rates of bowel screening [1]. Regarding participation in colorectal screening programs, disparities between rural
and urban zones as well as indigenous and non-indigenous populations are observed
in many countries [2]
[3]
[4]
[5]
[6]
[7]. Cancer outcomes are worse for rural patients, and more advanced cancer stages at
diagnosis may explain this difference [8]. It is difficult to measure the impact of remoteness itself, as there are other
health determinants underpinning this scenario such as lower socioeconomic status
and medical data underreporting in these locations.
In an attempt to overcome this issue, the Belterra Project (BP) was set up as a pilot
initiative for gastrointestinal cancer screening in an Amazon rainforest village.
It addressed population awareness of colorectal cancer screening, healthcare assistance,
and medical capacity expansion for gastrointestinal cancer screening.
The objective of this study was to describe the implementation and present the results
of the BP, which was created to provide gastrointestinal cancer screening for a small
population in the Amazon rainforest.
Patients and methods
Study design and ethical concerns
This descriptive study was conducted at the Hospital Sírio-Libanês. The description
of the BP was approved by the Local Research Ethical Committee (register number 1.273.137)
on October 9, 2015.
Setting
Belterra is a small town in the Amazon rainforest, on the banks of the Tapajos River.
It was founded in 1934 by Henry Ford as a center for the extraction and sale of latex
for the American automobile industry. The municipality area is 2,640.699 km2, with over 80 % of forest preservation area. Its population was estimated at 18,000
inhabitants in 2014, most of whom lived in rural areas. Of this, the subpopulation
aged 50 to 70 years has been estimated at 2,359 people. The main economic activities
include subsistence agriculture and fishing. The town's health infrastructure is poor,
consisting of only six small health units with five nurses and 40 community health
agents. There is only one hospital where six doctors work, and there are no endoscopic
or surgical units. The nearest secondary health facility is located 34 km from Belterra.
Data collection
Data were collected using standard forms filled out by doctors and medical students
enrolled in the program. Data included general and demographic information on each
patient, endoscopic exam results, and anatomopathological results. All data were then
entered into the REDCap (Research Electronic Data Capture) platform [9] and available to researchers only. Only clinical information specifically concerning
an individual participant was sent to local health providers.
BP description and implementation
As previously stated, the BP encompasses three key goals: address population awareness
of colorectal cancer screening, medical capacity expansion, and healthcare assistance.
The first step was to raise awareness of the importance of screening for gastrointestinal
cancer among the population and healthcare agents. To this end, marketing material
was developed, and was strategically placed in different regional locations and delivered
to the residents.
The second step was to increase the theoretical knowledge and improve the technical
skills of local health professionals. Medical students, residents, and endoscopists
from neighboring cities were invited to participate in training under the supervision
of experienced endoscopists.
The third step consisted of offering gastrointestinal cancer screening to every individual
between 50 and 70 years of age based on their medical history (risk factor, lifestyle,
social determinants), physical examination, endoscopic examinations (including Helicobacter pylori testing), and stool tests. From October 2014 to December 2017, 19 expeditions to
Belterra took place. Each expedition lasted four days, during which endoscopies (EGD)
and colonoscopies were performed, and fecal samples were analyzed within 1 week from
collection for fecal occult blood, using a Biopix qualitative fecal immunochemical
test (FIT) and parasitological tests. Whenever a lesion was diagnosed, an endoscopic
treatment with polypectomy or endoscopic mucosal resection was attempted if possible.
In case of advanced lesion biopsies were taken and sent to histopathological exam.
Bowel preparation regimen was constituted of clear liquid diet on the day before the
exam and 500 milliliters of 20 % mannitol solution orally 6 hours previous to the
expected time of the exam. A standard video endoscope (EG530 WR, Fujifilm Corporation,
Tokyo, Japan) was used for upper gastrointestinal endoscopies (EGD) and adult video
colonoscope (EC 530 WL, Fujifilm Corporation, Tokyo, Japan) for colonoscopy using
an imaging processor FUJIFILM – EPX 2500.
In every expedition, three experienced endoscopists, one endoscopy resident, two medical
students, two nurses, one nurse technician, and four or five other support members
from different regions of the country comprised the team responsible for delivering
screening to the target population.
Community health care agents were responsible for explaining the procedures, providing
drugs for bowel preparation, and providing general information to the population.
Any patient with conditions discovered during the examinations that could not be managed
endoscopically was referred to a secondary hospital located in Santarem city, 34 km
away from Belterra.
Barriers and facilitators for implementation.
Public-private partnerships were established in order to secure not only financial
funding, but also to guarantee adequate treatment to patients whose lesions could
not be managed endoscopically, to collect and analyze data, and to provide medical
material and expertise.
Due to the demographic characteristics of the region, the project utilized not only
the local hospital but also a boat ([Fig. 1]), where two endoscopy units were built ([Fig. 2]) in order to deliver screening to the riverside and hard-to-reach populations. Both
facilities had very limited resources, and therefore all medical materials including
image processors, screens, gastroscopes and colonoscopes, anesthetics, syringes, needles,
sharps disposals, endoscopic tweezers, electrosurgery equipment, oxygen tanks, emergency
carts, endoscopes cleaning products, and other basic items needed for the procedures
had to be transported to these health units.
Fig. 1 The Abaré boat, a floating hospital where two endoscopy units were built to increase
access to gastrointestinal screening among remote riverside populations.
Fig. 2 View of the inside of an endoscopy unit in Abaré.
Further, the BP relied on backup support from the Hospital Regional do Baixo Amazonas,
which was the local secondary hospital that received referred patients who needed
further treatment in the event of unexpected complications. This hospital was located
45 kilometers away from Belterra. Patients with adenocarcinoma were referred to this
hospital which was then held responsible for staging and treatment of these patients.
Patients diagnosed with adenomatous polyps and those with positive H. pylori tests were then referred to the Hospital Municipal de Belterra to receive adequate
treatment and follow-up. [Fig. 3] is a flowchart of the algorithm use for this project.
Fig. 3 Flowchart of the algorithm.
Results
Screening results
During the 19 expeditions to Belterra, 2,022 individuals were enrolled in the program
and a mean of 80.9 patients were screened per campaign. The characteristics of the
participants are shown on [Table 1]. Among the 1,903 participants whose fecal occult blood test (FOBT) was available,
64 patients (3.4 %) had a positive result. Of these 64 positive results, one patient
(1.6 %) was diagnosed with colorectal adenocarcinoma and 13 had at least one polyp
diagnosed and removed. A total of 31 polyps were removed from these patients as a
few had more than one polyp removed. Low-grade dysplasia was diagnosed in nine patients,
another two patients had adenomas with high-grade dysplasia, and two had synchronic
polyps with low- and high-grade dysplasia. From the remaining patients with negative
FOBT, 38.2 % had at least one abnormal finding, with polyps accounting for 65.4 %
of the cases. From the 459 patients with negative FOBT, but with polyps diagnosed
on colonoscopy, 41 had high-risk adenomas and four had adenocarcinomas. Other abnormal
findings were diverticulum (50.4 %), flat lesion (3.8 %), colitis (3.3 %) and arterial-venous
malformation (1.6 %). Regarding the upper endoscopy findings within patients with
negative FOBT, four patients had gastric cancer diagnosed and one had duodenal cancer.
Other abnormal findings were esophagitis (7.7 %) and gastric (26.7 %) and duodenal
(25.6 %) peptic ulcers.
Table 1
Demographic characteristics and risk factors in the studied population.
|
Characteristic
|
Number
|
Percentage
|
|
Gender
|
Male
|
974
|
48.2
|
|
Female
|
1048
|
51.8
|
|
Age
|
< 60
|
1260
|
62.3
|
|
≥ 60
|
762
|
37.7
|
|
Body mass index
|
< 27.5
|
1107
|
57.3
|
|
≥ 27.5
|
824
|
42.7
|
|
Previous colonoscopy
|
Yes
|
160
|
8.1
|
|
No
|
1816
|
91.9
|
|
Alcohol consumption
|
No
|
1353
|
69.0
|
|
< once a month
|
306
|
15.6
|
|
< once a week
|
210
|
10.7
|
|
> once a week
|
91
|
4.7
|
|
Smoking history
(> 100 cigarettes in lifetime)
|
Yes
|
973
|
50.5
|
|
No
|
955
|
49.5
|
Overall, 1,952 (96.5 %) participants underwent colonoscopy, the remaining patients
either did not show up on the scheduled day (n = 64) or did not have adequate bowel
preparation (n = 8). The colonoscopy was considered complete in 98.1 % when the scope
reached the cecum. The main reasons for an incomplete exam were bad quality of bowel
preparation (n = 31) and anatomical difficulties (n = 6). This study had 514 patients
with colonoscopic findings. Of the 471 polyps found, the majority were adenomas and
were subdivided into high-risk (defined as a lesion with any of the following characteristics:
size larger than 10 mm, more than 50 % of villous component or high-grade dysplasia)
and low-risk adenomas. High-risk adenomas represented 14.0 % of the total number of
adenomas. the adenoma detection rate (ADR) was 16.5 %. The main findings from this
test are summarized in [Table 2].
Table 2
Main colonoscopic and histopathological findings (n = 1,952).
|
Colonoscopic findings
|
Number (%)
|
Histopathological finding
|
Number (%)
|
|
Polyps
|
471 (24.12)
|
Low-risk adenoma
|
277 (14.20)
|
|
High-risk adenoma
|
45 (2.31)
|
|
Hyperplastic polyp
|
149 (7.63)
|
|
Tumors
|
7 (0.61)
|
Adenocarcinoma
|
5 (0.26)
|
|
Neuroendocrine tumor
|
2 (0.10)
|
|
Flat lesions
|
31 (1.59)
|
Low-risk adenoma
|
31 (1.59)
|
Colorectal polyps were diagnosed in 472 patients. The majority of the patients presented
with adenomas (n = 284), of which 267 (92.4 %) had tubular adenomas, while 22 had
tubulovillous adenomas (7.6 %). Regarding the histologic grade, 276 (95.5 %) had a
total of 370 low-grade dysplasia, while 13 (4.5 %) had high-grade dysplasia. Another
27 patients had sessile serrated lesions (SSL), of which two had high-grade dysplasia
and 20 had low-grade dysplasia. Data were not available for five patients with SSL.
The distribution of the polyps was 637 (57.5 %) located in the right colon (cecum,
ascending and transverse colon) and 440 (40.1 %) in the left colon (descending colon,
sigmoid, and rectum). There were 21 polyps with unknown localization. The characteristics
and location of the colorectal polyps are shown in [Table 3]. The average size of polyps removed was 5 mm (range 1 to 40 mm; median 4 mm).
Table 3
Polyps characteristics and location.
|
Characteristic or location
|
Number of patients
|
Percentage of patients with polyps (%)
|
Number of polyps
|
Percentage of polyps (%)
|
|
Sessile
|
429
|
86.3
|
672
|
84.3
|
|
Pedunculated
|
57
|
11.5
|
113
|
14.2
|
|
No data available
|
10
|
2.1
|
11
|
1.4
|
|
Transverse colon
|
137
|
23.1
|
271
|
24.7
|
|
Sigmoid
|
130
|
22.0
|
216
|
19.7
|
|
Ascending colon
|
128
|
21.6
|
259
|
23.6
|
|
Descending colon
|
77
|
13.0
|
136
|
12.4
|
|
Cecum
|
52
|
8.8
|
107
|
9.2
|
|
Rectum
|
52
|
8.8
|
88
|
8.0
|
|
No data available
|
16
|
2.7
|
21
|
1.9
|
Among the five patients diagnosed with colorectal adenocarcinoma, one had a positive
FOBT result. Moreover, of 45 patients with high-risk adenoma, 41 tested negative for
fecal occult blood.
Overall, 1,958 participants underwent upper endoscopy (96.83 %) and the remainder
did not show up on the scheduled day (64 patients). The main findings are detailed
in [Table 4].
Table 4
Main upper endoscopic and histopathological findings (n = 1,958).
|
Endoscopic findings
|
Number (%)
|
Histopathological finding
|
Number (%)
|
|
Suspected gastroduodenal lesions
|
9 (0.46)
|
Adenocarcinoma
|
6 (0.31)
|
|
Lymphoma
|
1 (0.05)
|
|
Neuroendocrine tumor
|
1 (0.05)
|
|
High-grade dysplasia adenoma
|
1 (0.05)
|
No adverse events requiring medical intervention were observed during the program.
Discussion
To the best of our knowledge, this is the first report of a colorectal cancer screening
program within an Amazonian population and the findings are key to future interventions
in this region.
Providing gastrointestinal screening to remote and underserved populations has proved
to be a challenge. Notwithstanding the geographical, cultural, and financial barriers,
we suggest that through the participation of volunteers and the establishment of partnerships
with the public and private sector, the implementation of a cancer screening program
for remote Brazilian populations is feasible. Of the several barriers that have been
identified, we highlighted difficulties in transporting all medical material and professionals
to an unequipped hospital, which was required to reach riverside populations that
are not accessible by roads. We also highlighted difficulties in providing adequate
training to health professionals who are not familiar with population-based screening.
In addition, we presented main results from the project, which included diagnosis
of lesions in 26 % of participants undergoing colonoscopy, and nine patients diagnosed
with malignancies undergoing upper gastrointestinal endoscopy.
We observed over 95 % adherence to the screening project, which is considered very
high. Other studies from different countries have shown much lower adherence rates,
such as 10.5 % and 45.4 % [10]
[11]
[12]. This surprisingly high adherence is likely due to a conjunction of factor, such
as the investment in marketing and promotion of the screening program through outdoor
posters, flyers, and the work of community health care agents who promoted the campaign
for population awareness. Further, the implementation of such a program in a small
and restricted population facilitates the spread of the information among all inhabitants.
Moreover, in this study, the ADR was 16.5 %, below that recommended by the American
Society for Gastrointestinal Endoscopy, which considers the target ADR of 20 % to
be a quality indicator [13]. The low ADR observed in this study may be explained by the characteristics of the
population screened within this study, or by the lack of imaging magnification modes
or high-resolution endoscopic devices used, which may have compromised our ability
to diagnose small lesions [14]
[15]
[16]
[17]
[18].
Five patients (0.26 %) were diagnosed with colorectal adenocarcinoma. The adenocarcinoma
detection rate varies widely in the literature, from 0.2 % to 3.1 % [19]
[20]
[21]
[22]. Therefore, our results fall within the variability described in other studies.
Unlike many colorectal screening projects [11]
[12]
[23]
[24]
[25]
[26]
[27]
[28], this study provided every participant with both FOBT and colonoscopy, making it
possible to analyze the efficacy of these examinations under the circumstances of
remote populations. The positivity of FOBTs observed in this study was similar to
that in other studies, with a positivity rate of 2.4 % to 7.7 % [25]
[26]
[27]
[28]. We found a surprisingly low sensitivity of FIT for adenocarcinoma in our study,
among the five patients diagnosed with adenocarcinoma, only one (20 %) had a positive
FOBT. Other studies that have also performed colonoscopy and FOBTs in all patients
have found positive FOBT results in patients diagnosed with colorectal adenocarcinoma
at higher rates than those observed in this study [19]
[20]. The low sensitivity of the FOBT in this study raises concerns about its use in
remote regions, where the storage and transportation of stool samples is not optimal
and may lead to decreased efficacy.
There are controversies regarding gastric cancer screening efficacy, and recommendations
are based on its incidence in each country or region. Countries with high gastric
cancer incidence, such as Japan, South Korea, Venezuela, and Chile, have implemented
screening, each of them with their own protocol regarding time interval and modality
of screening [29]
[30]
[31].
Nevertheless, our project had a series of limitations that must be pointed out. First,
given the voluntary nature of this work and the multiple sources of funding, an accurate
measurement of the actual costs is underestimated and out of the scope of this paper.
However, this factor plays a major role when considering the possibility of reproducing
this model on a nationwide scale. Further, data are lacking regarding colon preparation
and scope withdrawal time, which are considered important aspects when analyzing the
colonoscopy quality. Finally, we did not follow up the patients with adenomatous polyps
and invasive carcinomas; therefore, we lack information about their clinical stages
and further treatments.
Clinical studies comparing the benefits and risks of these interventions and further
cost-effectiveness analyses addressing this topic are needed. Another remaining question
concerns the feasibility of adapting this model for larger populations, within the
context of national public health systems.
Conclusions
Despite the geographical, cultural, and financial barriers, this study suggests that
through volunteer participation and the establishment of partnerships with the public
and private sector, the implementation of a gastrointestinal cancer screening program
for remote Brazilian populations is feasible. Our results showed that nearly 26 %
of Belterra inhabitants who underwent colonoscopy screening had some type of colonic
lesion.
