Development of Web-based System for a Cohort Study
The Brazilian e-NutriHS
Luciana Dias Folchetti, Isis Tande da Silva, Bianca de Almeida Pititto and Sandra Roberta G Ferreira
Department of Nutrition, Faculty of Public Health, University of Sao Paulo, Av. Dr. Arnaldo, 715, Sao Paulo, Brazil
Keywords: Electronic Health Records, Health Information on the Web, Web-based System, Cohort Study, Nutrition.
Abstract: This study describes a web-based system developed to gather online information on health of college
students and graduates in nutrition, the e-NutriHS. The Nutrititionist Health Study – NutriHS is planned to
be a cohort study aiming to collect health-related data at a 3-year interval. The e-NutriHS consists of 6
questionnaires regarding demographic and socioeconomic data, dietary habits, physical activity, alcohol and
tobacco uses, anti-fat attitudes and personal and family histories. Validated and internationally recognized
lifestyle instruments were used. Our software and respective database are hosted in the School of Public
Health server; the software is based on free programming languages. e-NutriHS data obtained from
questionnaires can be transferred to excel format. An e-NutriHS prototype was created preceding online
attachment. An improved version of website was released based on 20 volunteers’ opinions. 503 users were
already registered. Our initiative of building a website designed for collecting data for epidemiological
studies, tailored to our local reality, is innovative under the perspective of the health informatics available in
the developing world. Considering that web-based systems produce reliable data, are easy to use, less costly
and less time-consuming, we conclude that our experience deserves to be shared, particularly with middle-
income economy countries.
1 INTRODUCTION
In recent years, the use of the Internet by the
populations, in particular for healthcare information
provision, is markedly increasing. This represents a
desirable scenario, but it is worrying that sometimes
there is a lack of scientific evidence which could
have harmful impact in individuals. In light of this
growth, researchers of health sciences have invested
in learning how to use the Web to spread high-
quality evidence-based data. Such initiatives have
developed their abilities to deal with electronic tools
for data collection and report, recognizing benefits
as well as pitfalls on using the Web. Since nutrition
is a key element for health, this issue represents one
of the most explored.
Research studies in epidemiology are often
conducted in hundreds or thousands participants,
followed for long periods of time. These studies
generate a significant amount of data, requiring
trained professionals with specialized skills for data
collection and processing. Health surveys have been
mostly based on paper questionnaires. Their quality
depends on accuracy in collecting, sorting, coding,
typing and checking of data to create a consistent
database before analysis. These steps are expensive
and time-consuming, prolong the study duration, but
are essential for the adequate data analysis and
results interpretation to achieve reliable conclusions.
Until recent years, relevant epidemiological studies
were based on paper-and-pencil surveys, particularly
those concerning lifestyle data (Colditz and
Hankinson 2005, Wilson et al. 2012, Hu et al. 1997).
Technological advances have allowed obtaining
standardized and reliable data in a faster and less
expensive way. The completion of data
electronically can ensure almost immediate
construction of database, reduce costs, errors and
biases and minimize duplicate information. This has
largely motivated the use of the Web, for instance in
nutritional epidemiology. Researchers in nutrition
commonly deal with analysis of dietary intake,
usually collected by questionnaires and/or food
recalls. Several softwares are available for data
analysis, tailored to dietary habits of a given
population. In contrast, a direct link between
questionnaires completion and data processing for
analysis is missing. Few softwares, specifically
designed to play this role, are found in literature and
none in developing countries. Web-based question-
574
D. Folchetti L., Tande da Silva I., de Almeida Pititto B. and R. G. Ferreira S..
Development of Web-based System for a Cohort Study - The Brazilian e-NutriHS.
DOI: 10.5220/0005277805740579
In Proceedings of the International Conference on Health Informatics (HEALTHINF-2015), pages 574-579
ISBN: 978-989-758-068-0
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
naires applied in the nutrition field would facilitate
the “dialogue” of data collected with software to
dietary reports, as well as to statistical procedures.
The development of web-based online self-
administered systems, employing validated health-
related surveys, sociodemographic status, dietary
habits and personal and family medical history, with
a link for online scheduling for face-to-face data
collection is warranted. Such system to obtain
health-related data would be particularly important
in middle-income economy countries where
resources for researches are limited. It is anticipated
that these methods will become a more feasible way
of implementing surveys, providing a number of
advantages over traditional methods, including
convenience for the participant, potentially large
cost savings for the researcher, efficiency in data
collection, higher data quality, a degree of perceived
anonymity for the participant, and high response
rates (Best et al. 2001, Hewson et al. 1996, Krantz et
al. 1997, Schaefer and Dillman 1998, Schmidt 1997,
Shettle and Mooney 1999).
In Brazil, initiatives for electronic collection of
data for monitoring health aspects are rare. A
national health survey, the Telephone-based
Surveillance of Risk and Protective Factors for
Chronic Diseases, under the coordination of the
Ministry of Health and supported by the University
of Sao Paulo and the Center for Disease Control,
uses desktop software to obtain data of the Brazilian
population but requires trained interviewers
(Azevedo e Silva et al. 2011). To our best
knowledge, self-administered online questionnaires
have not been reported.
The purpose of the current study was to describe
a web-based system developed to gather information
on health of a specific subset of the Brazilian
population, allowing participants to complete data
online. Characteristics of the participants were
established based on their potential skill to navigate
on the Internet and on the expected high quality of
data provided. The choice of nutrition college
students and graduates fulfills these criteria.
2 LITERATURE REVIEW
2.1 Epidemiological Studies of Health
Professionals
Monitoring of nutritional and lifestyle factors
associated with changes in health conditions in
epidemiological studies has contributed to identify
modifiable risk factors, to deepen the understanding
of the pathophysiology of diseases, especially non-
communicable chronic diseases (NCD) and to
propose interventions. In this sense, experiences
involving health professionals have brought
significant contributions to the knowledge about the
role of behavioral factors on health outcomes of
Americans (Colditz and Hankinson 2005, Wilson et
al. 2012, Hu et al. 1997). Particularly, cohort studies
due to their longitudinal design have provided the
most relevant evidence.
Approaches to gather information in cohort
studies have been mainly based on face-to-face
and/or telephone interviews, using paper-and-pencil
questionnaires. The Nurses’ Health Study I and II
and the Health Professionals’ Follow-up Study
including thousands of participants in North
America, are remarkable considering their scientific
contributions for identifying lifestyle risk factors for
NCD (Colditz and Hankinson 2005, Wilson et
al. 2012, Hu et al. 1997). Despite the use of paper-
and-pencil questionnaires, those studies had the
strength of involving professionals that guarantee a
high quality of data collected. Other epidemiological
studies with similar objectives were conducted in
America and Europe, including smaller samples and
non-restricted to health professionals (Turner et al.
2009, Mikkelsen et al. 2009).
In this century, researches using web-based
questionnaires have gained significant popularity
(Couper 2000). More frequently, they have been
used in psychological studies and marketing
research, but in the field of epidemiology, Internet
tools were employed in less than 1% of the reported
studies (van Gelder et al 2010). The Millennium
Cohort Study (Smith et al. 2007), the Nurses and
Midwives e-Cohort Study (Turner et al. 2009), and
the Danish Web-based Pregnancy Planning Study
(Mikkelsen et al. 2009) are examples of successful
large cohort studies where de recruitment and
follow-up occur over the Internet.
In developing countries, similar initiatives
involving health professionals are rare and,
considering the use of Internet as a tool for data
collection, are inexistent. In countries with a large
territory and limited resources like Brazil, the
development of a cohort study able to identify
environmental risk factors, based on web-based tools
would be highly desirable. Therefore, in 2014, our
research group, working for a recognized Brazilian
reference academic institution, decided to develop a
study with those characteristics, the Nutritionist
Health Study – NutriHS (www.fsp.usp.br⁄nutrihs).
DevelopmentofWeb-basedSystemforaCohortStudy-TheBraziliane-NutriHS
575
2.2 Web-based Data Collection
Epidemiologists conducting cohort studies often
work with huge amount of data to evaluate the
relationship between exposures and outcomes at
multiple points in time. Application of these
methods based on paper-and-pencil questionnaires
requires trained professionals for collecting, sorting,
coding, typing and checking data. These studies
frequently are relevant for actions in public health
demanding agility to be implemented. Therefore,
researchers have been abandoned traditional
methods, replacing by Web-based data collection.
Web-based questionnaires have several
advantages compared with traditional tools,
including data quality, reduction of costs and time
from the study initiation to the receipt of analysable
data (Schleyer and Forrest 2000, Wyatt 2000). The
interactive element of this system allows the
inclusion of visual aids; design issues can be
simplify for responders; and also the scripts of the
website can be immediately adjust to solve
unforeseen problems or to incorporate preliminary
results (van Gelder et al. 2010). Costs for printing,
postage, and data coding and entry are avoided. On
the other hand, the set-up costs, including
programming and design, may be too high in studies
with small sample sizes or in populations with low
response rates (Adams and White 2008, Rodriguez
et al. 2006, Dillman 2007).
During survey implementation on the Web, there
are unknown issues prompting some authors to
express concerns (Best et al. 2001, Schaefer and
Dillman 1998, Smith 1997). These concerns include
sampling problems, lack of participant access to
computers with Internet connections, security, and
response inconsistency across different media. The
possibility of bias associated with collecting
information over the Web previously raised is no
greater than that introduced by traditional paper
methods (Ekman et al. 2006).
In summary, it is notable that the majority of
respondents prefers Web-based version to postal
questionnaires or telephone interviews (Rankin et al.
2008, Touvier et al. 2010). This is particularly true
for young people who were born during the Internet
era. Being a more attractive method of data
collection, more Web-based systems tended to be
developed for epidemiologic studies.
3 MATERIAL & METHODS
This chapter includes details on the users of the e-
NutriHS, as well as the instrument (e-NutriHS) and
its prototype for pre-testing. The NutriHS was
conceived during the year 2013 by our research
group from the School of Public Health of the
University of São Paulo, Brazil. The NutriHS is
planned to be a cohort study mainly involving
students of Nutrition Graduation Courses from
Universities located in the State of Sao Paulo, but
also graduates. For its first phase, launched in 2014,
they are being invited to participate in the study, in
which health-related data will be collected at a 3-
year interval.
A web-based system was developed to gather
information online of this specific subset of the
Brazilian population, the e-NutriHS. Our software is
unique, tailored to the researchers’ purposes and to
local requirements of the university server.
3.1 e-NutriHS Users
The choice of nutrition college students and
graduates was based on four presumed conditions: a)
potential skill to navigate on the Internet; b)
availability of computer and Internet connections; c)
expected high quality of data provided; d) local
facilities in our University.
Nutrition students and graduates are being
invited to this first phase of NutriHS. The ethical
committee at FSP-USP approved this study. All
student or graduate who accepts participation has
signed Statement of Informed Consent electronically
prior the fulfilment of questionnaires.
The ones who agree to participate are being
investigated about the role of lifestyle habits on
health conditions. Instruments used to collect
feedback from users were e-mail and Facebook.
3.2 Survey Instruments
The e-NutriHS includes of 6 questionnaires
regarding demographic and socioeconomic data,
dietary habits, physical activity level, alcohol and
tobacco use, anti-fat attitudes and personal and
family medical histories. Contact information was
obtained to help the participant tracking during the
follow-up. Validated and internationally recognized
lifestyle tools were used, since standardization is
necessary for comparisons with other populations.
Main tools employed in the NutriHS include the
International Physical Activity Questionnaire
(IPAQ) (Craig at al. 2003), Food Frequency
Questionnaire (FFQ) (Fisberg et al. 2008) and the
Antifat Attitudes Questionnaire (Lewis et al. 1997).
HEALTHINF2015-InternationalConferenceonHealthInformatics
576
3.3 Online Social Network and Website
Colors for the website were carefully selected; green
and orange are related to health, nutrition and youth.
During the software building process, ease of use
was frequently reviewed until become a friendly
version. All the questionnaires were inserted in the
website exactly as the printed version, except for
essential instructions for completion. The length of
each questionnaire page was defined based on a
comfortable time for filling. Our logotype was
created reinforcing the impact of increasing
knowledge about nutrition on health.
In order to increase awareness of the NutriHS a
social page in the Internet was created before
starting the data collection. The most popular social
network – the Facebook – was chosen to publicize
the study (Figure 1).
Figure 1: Social online network of NutriHS
(https://www.facebook.com/nutrihs).
Our page in Facebook includes a link to the e-
NutriHS homepage. This place has also represented
an important vehicle to disseminate scientific
knowledge on nutrition for Internet users and to
support the dialogue among researchers and users.
With the University of São Paulo Informatics
Center support, the e-NutriHS website (Figure 2)
was developed to expedite and further analyze data
generated from the first phase of the NutriHS.
As our software and respective database would
be hosted in the School of Public Health server, our
team had to build the web system which was based
on free programming languages, such as HyperText
Markup Language (HTML), Cascading Style Sheets
(CSS), JavaScript, Hypertext P reprocessor (PHP)
and Structured Query Language (SQL), in database
management system MySQL.
Before fulfilling the questionnaires, users are
required to create a unique login and password.
Afterwards, their access to questionnaires is
released. When users enter incomplete or
implausible answers, prompts alert them to complete
Figure 2: The e-NutriHS website (http://www.fsp.usp.br/
nutrihs/index.html).
or to check information. Each page must be saved at
the end in order to feed database. User workflow is
depicted in Figure 3. e-NutriHS data obtained from
any questionnaire can be transferred to excel format
and provide print reports. Dietary data are
immediately processed permitting estimates of daily
energy and nutrient intakes, based on the nutritional
composition data of the USDA National Nutrient
Database for Standard Reference.
Figure 3: User online workflow.
3.4 The e-NutriHS Prototype
A prototype of e-NutriHS was created preceding
online attachment for NutriHS users and hosted on a
temporary url in order to test the system and to
check its ease-of-use, data entry and website design.
Twenty volunteers, graduates at the School of
Public Health who have been worked in the health
area, aged 25 to 60 years old, were invited to test the
prototype. They did the online registration and
answered to all questionnaires.
To check if answers were accurate, one
questionnaire per person was randomly chosen for
paper responses, sent by e-mail. The volunteers also
answered four simple questions about the design,
method preferred and ease-of-use of the website:
1. Which method do you prefer?
DevelopmentofWeb-basedSystemforaCohortStudy-TheBraziliane-NutriHS
577
2. Is attractive the design of the website?
Would you change anything?
3. Were the questionnaires easy to answer?
4. Did you have any kind of problem when
navigating the site?
4 INITIAL FINDINGS
4.1 Prototype Results
In average, volunteers fulfilled the web-based
questionnaires 30 minutes after they had received
the temporary url, meaning that these data were
already ready to be analyzed. Regarding the paper
version, they returned it at least one day after e-mail
delivery. It is important to take into consideration
that those paper questionnaires still needed to be
coded to enter the database.
Based on the comparative analyses of the
prototype with paper questionnaires, some technical
problems were detected in the FFQ. These script
errors were properly adjusted and solved.
Users preferred web-based method over the
traditional one, considered the website attractive,
made favourable comments about the colours used
and the interactive elements and suggested no
change. The questions were considered simple and
the prompts convenient, facilitating fulfilment. No
problem was mentioned related to navigation. An
improved version of website was released based on
the volunteers’ opinions.
4.2 NutriHS Preliminary Results
In March 2014, NutriHS participants have started to
use the online web-based system. No complain on e-
NutriHS access and use has been reported. In
general, users’ satisfaction seems to be positive
since only favourable comments have been received.
A total of 503 users were registered, being 459
college students and 44 graduates. Predominance of
female sex was detected (93%) which was expected
since the majority of nutrition students are women.
Mean age of the sample is 23.8 (SD 6.6) years;
63.5% are Caucasian and 84.6% single.
Preliminary data on energy and nutrient intakes,
physical activity levels, alcoholic beverages and
tobacco uses and others are available allowing
comparisons of students’ profile at the beginning
and end of the course. These conditions are in
contrast with the nutritional epidemiological studies,
which require a long period to get results.
Summaries of scientific papers on nutrition
recently published should be responsible – at least in
part – to the high access to our social network.
Taken together, these preliminary results should
bring relevant information not only for the student,
graduate and researcher but also to the academic
institutions’ staff.
5 LIMITATIONS & STRENGHT
There is a lack of originality in using web-based
system to collect and disseminate health
information. However, its use in cohort studies,
conducted in emerging countries, exploring the
lifestyle impact on health conditions, is rare. As far
as we know, in Brazil, this kind of web-based
research in nutritional epidemiology is a novelty.
The applicability of the current instrument is
limited since this was specifically created for a
subset of professionals. However, similar web-based
systems could be adapted for other practices.
Particularities of our design site limit comparisons
with other similar web systems available.
6 CONCLUSIONS
Our initiative of building a website designed for
collecting data for epidemiological studies, tailored
to our local reality, is innovative under the
perspective of the health informatics available in
developing world.
Preliminary findings suggested that e-NutriHS
may be useful to obtain data with potential to
identify gaps and/or problems in graduation process
as well as health disorders. In parallel, we speculate
that this website might encourage students to search
for e-learning on nutrition. Also, it is expected that
college authorities could obtain some subsidies to
review aspects of the academic curriculum.
Considering that web-based systems produce
reliable data, are easy to use, less costly and less
time-consuming, we conclude that our experience
deserves to be shared, particularly with middle-
income economy countries.
ACKNOWLEDGEMENTS
The authors have no conflict of interest to disclose.
This study was supported by a scholarship grant
from Coordination for the Improvement of Higher
Education Personnel (CAPES).
HEALTHINF2015-InternationalConferenceonHealthInformatics
578
REFERENCES
Adams, J. & White, M. (2008) Health behaviours in
people who respond to a web-based survey advertised
on regional news media. European Journal of Public
Health. 18 (3). p. 335–338.
Azevedo e Silva, G., Valente, J. G. & Malta, D. C. (2011)
[Tendências do tabagismo na população adulta das
capitais brasileiras: uma análise dos dados de
inquéritos telefônicos de 2006 a 2009] Revista
Brasileira de Epidemiologia. 14 (1). p.103-114.
Best, S. J. et al. (2001) An assessment of the
generalizability of Internet surveys. Social Science
Computer Review. 19. p.131–45.
Cade, J. E. et al. (2004) Food-frequency questionnaires: a
review of their design, validation and utilisation.
Nutrition Research Reviews. 17 (1). p. 5-22.
Colditz, G. A. & Hankinson, S. E. (2005) The Nurses'
Health Study: lifestyle and health among women.
Nature Reviews Cancer. 5. p. 388-96.
Couper, M. P. (2000) Web surveys – A review of issues
and approaches. Public Opinion Quarterly. 64. p. 464–
494.
Craig, C. L. et al. (2003) International Physical Activity
Questionnaire: 12-country reliability and validity.
Medicine & Science in Sports & Exercise. 35 (8). p.
1381-95.
Dillman, D. A. (2007) Mail and Internet Surveys: The
Tailored Design Method. 2nd ed. Hoboken, NJ: John
Wiley & Sons, Inc.
Ekman, A. et al. (2006) Feasibility of using web-based
questionnaires in large population-based epidemio-
logical studies. European Journal of Public Health.21.
p. 103–11.
Fisberg, R.M, et al. (2008) [Questionário de freqüência
alimentar para adultos com base em estudo
populacional]. Revista de Saúde Pública, 42. p. 550-
554.
Hewson, C. M., Laurent, D. & Vogel, C. M. (1996) Proper
methodologies for psychological and sociological
studies conducted via the Internet. Behaviour
Research Methods, Instruments, & Computers. 28. p.
186–91.
Hu, F. B. et al. (1997) Dietary fat intake and the risk of
coronary heart disease in women. New England
Journal of Medicine. 337. p. 1491-9.
Krantz, J. H., Ballard, J. & Scher, J. (1997) Comparing the
results of laboratory and World Wide Web samples of
the determinants of female attractiveness. Behaviour
Research Methods, Instruments, & Computers. 29. p.
264–9.
Lewis, R. J. et al. (1997) Prejudice toward fat people: the
development and validation of the antifat attitudes test.
Obesity Research. 5 (4). p. 297-307.
Mikkelsen, E. M. et al. (2009) Cohort profile: the Danish
Web-based Pregnancy Planning Study-‘Snart-Gravid’.
International Journal of Epidemiology. 38 (4). p. 938–
943.
Rankin, K. M. et al. (2008) Comparing the reliability of
responses to telephone-administered versus
selfadministered web-based surveys in a case-control
study of adult malignant brain cancer. Cancer
Epidemiology, Biomarkers
& Prevention. 17 (10). p.
2639–2646.
Rodriguez, H. P. et al. (2006) Evaluating patients’
experiences with individual physicians: a randomized
trial of mail, Internet, and interactive voice response
telephone administration of surveys. Medical Care. 44
(2). p. 167–174.
Schaefer, D. R. & Dillman, D. A. (1998) Development of
standard e-mail methodology: results of an
experiment. Public Opinion Quarterly. 62. p. 378–97.
Schleyer, T. K. & Forrest, J. L. (2000) Methods for the
design and administration of web-based surveys.
Journal of the American Medical Informatics
Association. 7 (4). p. 416–425.
Schmidt, W. C. (1997) World-Wide Web survey research:
benefits, potential problems, and solutions. Behaviour
Research Methods, Instruments, & Computers. 29.
p.274–9.
Shettle, C. & Mooney, G. (1999) Monetary incentives in
U.S. government surveys. Journal of Official
Statistics. 15. p. 231–50.
Smith, B. et al. (2007) When epidemiology meets the
Internet: Web-based surveys in the Millennium Cohort
Study. American Journal of Epidemiology. 166 (11).
p. 1345–1354.
Touvier, M. et al. (2010) Comparison between web-based
and paper versions of a self-administered
anthropometric questionnaire. European Journal of
Epidemiology. 25 (5). p. 287–296.
Turner, C. et al. (2009) Cohort profile: the Nurses and
Midwives e-Cohort Study - a novel electronic
longitudinal study. International Journal of
Epidemiology. 38 (1). p. 53–60.
van Gelder, M. M., Bretveld, R. W. & Roeleveld, N.
(2010) Web-based questionnaires: the future in
epidemiology? American Journal of Epidemiology.
172 (11). p. 1292-8.
Wilson, K. M. et al. (2012) Dietary acrylamide and risk of
prostate cancer. International Journal of Cancer. 131
(2). p. 479-87.
Wyatt, J. C. (2000) When to use web-based surveys.
Journal of the American Medical Informatic
Association. 7 (4). p. 426–429.
DevelopmentofWeb-basedSystemforaCohortStudy-TheBraziliane-NutriHS
579
