Handel et al.

Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8

DOI 10.1186/s40794-016-0024-y

RESEARCH Open Access

Knowledge, attitudes, and practices

regarding dengue infection among public
sector healthcare providers in Machala,
Ecuador
Andrew S. Handel1, Efraín Beltrán Ayala2,3, Mercy J. Borbor-Cordova4, Abigail G. Fessler5, Julia L. Finkelstein5,
Roberto Xavier Robalino Espinoza3, Sadie J. Ryan6,7,8 and Anna M. Stewart-Ibarra7*

Abstract
Background: Dengue fever is a rapidly emerging infection throughout the tropics and subtropics with extensive
public health burden. Adequate training of healthcare providers is crucial to reducing infection incidence through
patient education and collaboration with public health authorities. We examined how public sector healthcare
providers in a dengue-endemic region of Ecuador view and manage dengue infections, with a focus on the 2009
World Health Organization (WHO) Dengue Guidelines.
Methods: A 37-item questionnaire of dengue knowledge, attitudes, and practices was developed and administered
to dengue healthcare providers in Machala, Ecuador. Survey focus areas included: “Demographics,” “Infection and
Prevention of Dengue,” “Dengue Diagnosis and the WHO Dengue Guide,” “Laboratory Testing,” “Treatment of Dengue,”
and “Opinions Regarding Dengue.”
Results: A total of 76 healthcare providers participated in this study, of which 82 % were medical doctors and
14 % were nurses. Fifty-eight percent of healthcare professionals practiced in ambulatory clinics and 34 % worked
in a hospital. Eighty-nine percent of respondents were familiar with the 2009 WHO Dengue Guidelines, and, within that
group, 97 % reported that the WHO Dengue Guide was helpful in dengue diagnosis and clinical management.
Knowledge gaps identified included Aedes aegypti mosquito feeding habits and dengue epidemiology. Individuals
with greater dengue-related knowledge were more likely to consider dengue a major health problem. Only 22 % of
respondents correctly reported that patients with comorbidities and dengue without warning signs require hospital
admission, and 25 % of providers reported never admitting patients with dengue to the hospital. Twenty percent of
providers reported rarely (≤25 % of cases) obtaining laboratory confirmation of dengue infection. Providers reported
patient presumptive self-medication as an ongoing problem. Thirty-one percent of healthcare providers reported
inadequate access to resources needed to diagnose and treat dengue.
Conclusion: Participants demonstrated a high level of knowledge of dengue symptoms and treatment, but additional
training regarding prevention, diagnosis, and admission criteria is needed. Interventions should not only focus on
increasing knowledge, but also encourage review of the WHO Dengue Guidelines, avoidance of presumptive
self-medication, and recognition of dengue as a major health problem. This study provided an assessment tool
that effectively captured healthcare providers’ knowledge and identified critical gaps in practice.
Keywords: Dengue fever, KAP survey, Ecuador, Medical practitioners

* Correspondence: stewarta@upstate.edu
7
Emerging Pathogens Institute, University of Florida, Gainesvillee, fl, USA
Full list of author information is available at the end of the article

© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 2 of 10

Background could not be classified using the 1997 WHO classifica-

Dengue virus infection is a major cause of morbidity, tion system, compared to 1.6 % using the 2009 WHO
mortality, and economic hardship in the tropics and classification system. As the debate over dengue classifi-
subtropics [1, 2]. Infection occurs when one of four cation continues, it is critical to understand how clini-
dengue virus serotypes (DENV 1–4) are transmitted to cians interpret and apply the guidelines in clinical
humans by Aedes sp. (primarily Aedes aegypti) mosqui- practice.
toes [1]. Dengue infection may cause fever, headache, ab- Previous research has focused on how community
dominal pain, rash, muscle aches, and bone pain (hence members view dengue infections; however, there have
‘break-bone fever’). Infection with additional dengue se- been few attempts to date to better understand the per-
rotypes increases the risk of hemorrhagic disease, result- spectives of clinicians. These studies have been con-
ing in severe mucosal and gastrointestinal bleeding, ducted mainly in Asia [9, 11–15], with a single study
hypovolemia, and potentially death [1]. It is crucial that performed in Puerto Rico [16]; to our knowledge, no
healthcare professionals are able to accurately diagnose, studies to date have been conducted in Central or South
monitor, treat, and hospitalize patients infected with America. Results have varied considerably across these
dengue fever. studies. In a study of Sri Lankan practitioners [11],
Latin America has seen a surge of dengue infections Kularatne et al. report significant disagreement among
since the 1980s, increasing the need for physicians physicians over the utility of treating dengue with ste-
skilled in managing dengue. From 2010 to 2014, an aver- roids, antibiotics, and platelet transfusions. Lee et al.
age of 1.5 million cases per year were reported in the [12] noted that clinical practice varied significantly by
Americas [3], although total case estimates are higher practice setting, as physicians practicing in private prac-
due to underreporting [2]. Díaz-Quijano et al. [4] esti- tice were more likely to refer patients with dengue to
mated that dengue-related mortality rates have tripled the hospital and to utilize dengue PCR testing (vs. ser-
every decade in Latin America since dengue became en- ology), compared to physicians practicing at public
demic in the 1980s. The economic burden of dengue clinics. Thaver et al. [13] conducted a knowledge-based
fever is also tremendous: the estimated median cost of assessment in Pakistan and found that practitioners had
dengue treatment in the Americas is US$472 per ambu- a stronger understanding of dengue pathophysiology
latory case (72.9 % of cases) and US$1,227 per hospital- than clinical diagnosis and treatment. Together, these
ized case [5]. The total economic impact of dengue in studies provide evidence that clinical practice varies by
the Americas was estimated at US$2.1 billion per year region and over time, making it crucial to understand
(2000–2007 estimate; range US$1–4 billion) [5], under- local, current practices for dengue management when
scoring the significant economic burden of dengue fever identifying areas of potential improvement.
infection and the need to improve interventions. As the epidemiology of dengue has evolved over the
Understanding how clinicians manage suspected cases past century, so have healthcare systems’ strategies to re-
of dengue is crucial to improving patient outcomes. In duce infection rates. Healthcare providers who interact
2009, the World Health Organization (WHO) revised its directly with patients have an important role in both
classification system of dengue severity [6]. The central treating and preventing the spread of dengue. This study
aim of the new scheme is to improve clinical outcomes was conducted to assess the knowledge, attitudes, and
by identifying patients at highest risk of mortality who practices regarding dengue infection among healthcare
may require therapeutic interventions. However, accept- providers in a dengue-endemic city in Ecuador. We also
ance and incorporation of these recommendations has assessed familiarity with the 2010 Pan American Health
varied considerably since publication, with ongoing de- Organization’s (PAHO) Spanish translation [17] of the
bate regarding the utility of each classification scheme 2009 WHO Dengue Guidelines, and how these guide-
[7, 8]. Recent investigations of the WHO Dengue Guide- lines influenced their clinical practice, providing import-
lines are promising. Prasad et al. [9] compared the ant information to help guide future interventions.
sensitivity of the 2009 and 1997 WHO guidelines in
identifying the severity of dengue infection among 56 Methods
patients who tested positive for dengue infection in Study site and study population
northern India. The study found that, when compared We conducted a study of the knowledge, attitudes, and
to the ‘gold standard’ of actual level of medical interven- practices associated with dengue infection among
tion provided (i.e. outpatient versus inpatient treatment), healthcare providers practicing in Machala, Ecuador,
the 2009 WHO classification system had 98.0 % sensitiv- from December, 2013 through December, 2014. Machala
ity, compared to 24.8 % sensitivity using the 1997 sys- is an urban coastal city located in El Oro Province,
tem. In an analysis of 1,962 cases reviewed from 18 Ecuador (3.2667°S, 79.9667°W, altitude 6 m, population
countries, Barniol et al. [10] found that 13.7 % of cases 245,972), and has been well-described as hyper-endemic
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 3 of 10

for dengue fever (DENV 1–4) [18, 19]. Over a five year Dengue, with a subset of questions on local dengue
period (2010 to 2014), 72,060 cases of dengue were re- epidemiology based on peer-reviewed sources [3, 13,
ported in Ecuador, with an annual average of 14,412 18]. The questionnaire comprised of the following sec-
cases [20]. This study is part of an ongoing collaboration tions: “Demographics,” “Infection and Prevention of
with the Ministry of Health to strengthen dengue surveil- Dengue,” “Dengue Diagnosis and the WHO Guide,”
lance capacities, with the aim of studying public sector “Laboratory Testing,” “Treatment of Dengue,” and
healthcare providers; private physicians were therefore not “Opinions Regarding Dengue” (See Additional file 1:
included in our study. The Ecuadorian Ministry of Health Appendice A1 for English and Additional file 2:
previously collaborated with the Pan American Health Appendice A2 for Spanish versions of the survey in-
Organization (PAHO) to translate the 2009 WHO Dengue strument). The survey was piloted through face-to-face
Guidelines into a 2010 Spanish version of the guidelines interviews with physicians in Machala prior to con-
[17], which was distributed throughout Machala and ducting the full study.
serves as a focal point of our study.
Physicians and nurses were recruited as the study Data analysis
population because they serve as the frontline healthcare Survey responses were analyzed using R (Version 3.1.2).
workers for diagnosis and treatment of dengue and other Descriptive statistics (e.g. means, medians, frequency
febrile illnesses. Healthcare providers in Machala include distributions) were calculated. A Cumulative Knowledge
primary care providers working in local healthcare Score (CKS) was calculated as an aggregate of all
clinics (Centros de Salud) and tertiary care providers knowledge-based questions (See questions in Tables 2
practicing in public and private hospitals, including and 3). Correct answers received one point and incorrect
emergency care physicians, hospitalists, and subspecial- answers received zero points, for a maximum possible
ists. The public health system requires that individuals score of 14 points. Questions requiring participants to
visit a single assigned Centro de Salud prior to referral select multiple correct answer choices were given one
to hospital subspecialists. These clinics provide care free point per correct answer selected. A Clinical Scenario
of charge. Private clinics were not included in this study. Score (CSS) was similarly developed from three clinical
It is common for Ecuadorians to view hospital care as questions, with a maximum score of three points (See
superior to ambulatory clinics, leading some patients to Table 3). These same clinical questions were included in
seek primary care in the Emergency Department. the CKS. Bivariate Pearson Correlations (r) were con-
ducted to assess whether the CSS and CKS were associ-
Participant recruitment ated with awareness and/or support of WHO clinical
Two methods of recruitment were utilized in this study. guidelines, prior training, years of experience or number
Participants from the public health sector were recruited of patients treated, and region of medical practice. We
at dengue management training conferences in Machala, also examined whether dengue risk perceptions were as-
with survey distribution prior to the educational session. sociated with support for the WHO dengue guidelines,
These trainings were conducted in collaboration with and the proportion of patients referred for dengue la-
and sponsored by the Ecuadorian Ministry of Health and boratory testing or hospital admission. The questions
the Global Emerging Infections Surveillance and Re- were grouped by dependent variable, and a Bonferroni
sponse System (GEIS, a division of the United States correction was used for multiple comparisons. The alpha
Armed Forces Health Surveillance Center), with the goal level was set at 0.05 (i.e., values of p <0.05 were consid-
of improving recognition of dengue infection and aware- ered statistically significant).
ness of the World Health Organization’s Clinical Manual Closed-ended questions using a Likert scale and open-
of Dengue. These individuals were recruited for the ended questions were used to assess doctor and patient
training sessions as they play key roles in dengue man- perceptions of dengue (See Tables 5, 6 and 7). The fre-
agement. The second form of recruitment involved visits quencies of these themes were tabulated, and for each
to the Ministry of Health public health clinics and to the theme, and the average scores from the Likert scale were
Teófilo Dávila Hospital, the reference hospital for the used to identify themes that associated with greater risk
province of El Oro. perceptions.

Questionnaire development Results/Discussion

We developed a 37-item questionnaire, with the goal of In this study, several common themes emerged; health-
evaluating the knowledge, attitudes, and practices associ- care providers reported:
ated with dengue infection among healthcare providers.
Information regarding dengue infection was based on 1. High use and awareness of the 2009 WHO Dengue
the World Health Organization’s Clinical Manual of Management Guidelines.
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 4 of 10

2. High level of knowledge regarding dengue signs and Table 1 Characteristics of Study Participants (n = 76)
symptoms, but demonstrated significant knowledge Category Response selected n (%)
gaps regarding dengue epidemiology and prevention. Gender Male 31 (41 %)
3. Limited knowledge of WHO-recommended criteria Female 45 (59 %)
for dengue hospital admission, and under-utilization
Age (years) <30 24 (32 %)
of confirmatory laboratory tests.
4. High level of concern regarding the burden of 31–40 10 (14 %)
dengue in Machala; and a lack of training and basic 41–50 17 (23 %)
tools needed to adequately diagnose and manage 51–60 17 (23 %)
dengue infections. 61–70 6 (8 %)
5. High levels of presumptive self-medication and delay >70 0 (0 %)
in seeking medical attention among patients with
Medical role Doctor 62 (82 %)
dengue fever.
Nurse 11 (14 %)
This study assessed the knowledge, attitudes and prac- Other 3 (4 %)
tices of local healthcare providers in dengue manage- Years of Medical experience <1 6 (8 %)
ment. This study was restricted to one group of 1–4 26 (34 %)
healthcare practitioners in Machala at one point in time, 5–9 5 (7 %)
and accordingly, the small sample size of available pro-
10–14 7 (9 %)
viders within Machala may limit generalizability of
findings. Additionally, data collected were self-reported, 15–19 10 (13 %)
limiting our ability to assess healthcare practices and >19 22 (29 %)
causal inference. However, this study captures useful in- Practice setting (n = 74) Community Health Center 43 (58 %)
formation from a community with a high burden of den- (Subcentro de Salud)
gue, and this assessment framework can inform dengue Hospital 25 (34 %)
management in other settings. Diagnostic Laboratory 2 (3 %)
Other 4 (5 %)
Demographics
A total of 76 healthcare providers involved in dengue
care and treatment in Machala, Ecuador, participated in virus is endemic to the region and has the potential to
the study. Demographic information is presented in cause high morbidity. One participant reported that den-
Table 1. Surveys were administered to participants dur- gue may cause “the deterioration of the [individual],
ing visits to their offices or at training events, resulting family and community health”. The majority of respon-
in a 100 % response rate. Forty-one percent of partici- dents (78 %) also agreed with the statement, “My pa-
pants were male and 59 % were female. Ages ranged tients feel that dengue infection is a major problem for
from under 30 years old to less than 70 years old, with a their health,” with 34 % citing health complications and
median age group of 41 to 50 years old and a median of
10 to 14 years of healthcare experience. Participants Table 2 Physician responses to the statement “I think that
consisted of physicians (82 %), nurses (14 %), and other dengue is a major problem for my patient population” (n = 71)
healthcare professionals (4 %). There are 93 physicians Categorical responses Open-ended responses
working in the public health sector in Machala including Agree or Strongly Agree The region is an endemic zone
63 doctors at 17 public health clinics, 15 in the central (n = 63, 88 %)
Dengue has a high morbidity
hospital, and 15 in the social security hospital, giving an
There is a lack of preventative measures
inclusion rate of 67 % of all potential physician subjects
Patients self-medicate
in the city. Healthcare providers worked primarily in
public health clinics (58 %) and the Teófilo Dávila There exists poor infrastructure
Hospital (34 %). Dengue poses a high risk to others
There is a lack of education about dengue
Healthcare provider views of dengue burden There is a lack of social consciousness
As seen in Table 2, healthcare providers in Machala were regarding dengue
concerned with dengue infections, with 89 % of partici- Neutral (n = 4, 6 %) There is adequate education about dengue
pants agreeing that it is a “major problem for my patient Disagree or Strongly Good preventative measures are in place
population”. Of those in agreement, the majority re- Disagree (n = 4, 6 %)
Good medical attention is available
ported that dengue is a significant threat because the
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 5 of 10

mortality as the major concerns in the general population. Table 4 Physician responses to the statement “In my experience,
Seventeen percent of participants suggested that public a member of the community who exhibits dengue symptoms will
health measures within the city, including disease preven- seek medical attention” (n = 71)
tion efforts, local infrastructure, and education were Categorical responses Open-ended responses
inadequate for controlling disease transmission. A small Agree or Strongly Agree Patients believe that dengue has a high
subset (6 %) of providers reported that dengue is not a (n = 54, 76 %) morbidity if untreated
major problem because the region already has effective Dengue symptoms are severe
prevention and treatment interventions in place. A similar Patients want to prevent complications
proportion of providers also felt that a dengue diagnosis There is adequate education about dengue
creates an unnecessary sense of fear among patients.
There is easy access to medical attention
Neutral (n = 7, 10 %) Some patients will seek attention while others
Providers’ views of the community response to dengue self-medicate
and self-medication
Disagree or Strongly Patients do not seek medical attention until
The majority (76 %) of healthcare providers perceived Disagree (n = 10, 14 %) complications develop
that patients exhibiting symptoms of dengue would seek
Medical care is delayed by self-treatment
attention at a healthcare facility. An equal proportion
also reported that patients are aware of the steps needed
to prevent dengue infection (See Tables 2, 3 and 4 for Clinical Scenario subset of questions). The mean Cumu-
physician attitudes toward dengue); of those who agreed lative Knowledge Score was 10.5 of 14 possible points
with this statement, 30 % suggested that public health (SD ± 1.73). Using a Bonferroni correction, the statistical
awareness campaigns were successful. For example, one significance level for CKS was determined to be p <0.01.
participant reported, “due to constant [educational] cam- The Cumulative Knowledge Score correlated positively
paigns, [patients] know to seek out medical help before with: 1) reporting familiarity with the WHO Dengue
they develop alarms signs”. Guide (r = 0.427, p <0.01), 2) agreeing with the statement
Seventeen percent of providers reported that upon “I believe that dengue is a major problem for my patient
symptom onset, many patients “turn to self-medication population” (r = 0.433, p <0.01), and 3) agreeing to the
and do not seek out professional help”. Previous studies statement “My patients feel that dengue infection is a
have reported that communities in the urban periphery, major problem for their health” (r = 0.282, p <0.01). Not-
and particularly men, report self-medicating to treat ably, having previous dengue training was not significantly
dengue [21]. This tendency to self-medicate can result in correlated with the CKS (p = 0.225). These associations
greater dengue morbidity and mortality due to lack of provide evidence of the interrelatedness of a practitioner’s
clinical management, and has the potential to increase knowledge, patient care, and concern for dengue infection.
community susceptibility to other diseases by promoting Clinician education must not only focus on basic know-
bacterial resistance to over-prescribed antibiotics. ledge, but also emphasize dengue’s burden on individual
health and communities.
Clinical scenario scores and cumulative knowledge scores The Clinical Scenario Score analysis results are pre-
The Cumulative Knowledge Score analysis results are sented in Table 6. The mean Clinical Scenario Score was
presented in Tables 5 and 6 (Table 6 consists of the 2.1 of 3 potential points (Table 6). Using a Bonferroni
correction, the statistical significance level for CSS was
Table 3 Physician responses to the statement “My patients feel
determined to be p <0.0125. A higher CSS was corre-
that dengue infection is a major problem for their health” (n = 68) lated with the following responses: 1) reporting familiar-
Categorical responses Open-ended responses
ity with WHO Dengue Guidelines (r = 0.326, p <0.01), 2)
agreeing with the statement “I believe that dengue is a
Agree or Strongly Agree Dengue has a high morbidity
(n = 53, 78 %) major problem for my patient” (r = 0.37, p <0.01), 3)
Dengue is considered an alarming diagnosis
agreeing with the statement “I am fully trained to man-
Dengue decreases economic productivity age a patient with an infection of dengue without warn-
Dengue is difficult to detect ing signs,” (r = 0.383, p <0.01), and 4) agreeing with the
The environment is conducive to disease statement “In my experience, a community member who
transmission has dengue symptoms will seek medical attention” (r =
Dengue poses a high risk to others 0.453, p <0.01). Higher CSS was also associated with
Neutral (n = 7, 10 %) Adequate medical attention is available reporting that the WHO Guidelines are helpful, although
Disagree or Strongly Patients believe self-medication is adequate
this was not statistically significant after a Bonferroni cor-
Disagree (n = 8, 12 %) rection was applied (r = 0.245, p <0.05). These findings
There is a lack of education about dengue
emphasize the importance of practitioner ‘buy-in’ of
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 6 of 10

Table 5 Knowledge-Based Questions (n = 76)

Question Correct response n (%) with correct response
1. How is dengue spread? Aedes mosquito 75 (99 %)
2. At what time of day are people most likely to be Any answer other than “Night” 57 (75 %)
infected by dengue?
3. Which of the dengue serotypes have been found in DENV 1–4 are all present
Ecuador?
- Note: 0.25 point given per correct answer, with a total of DENV 1 54 (71 %)
1 point available
DENV 2 54 (71 %)
DENV 3 48 (63 %)
DENV 4 34 (45 %)
4. What advice do you give your patients to prevent 1. Frequently change the water in flower vases 66 (87 %)
dengue infection?
2. Remove containers that accumulate clean water 69 (91 %)
- Note: Question is worth a total of 2 points
- 0.25 point given per correct answer, with a maximum of 3. Eliminate tanks or puddles with stagnant water 55 (72 %)
1 point. Column to the right indicates the n (%) of
respondents receiving 4. Keep drinking water containers (cisterns, tanks) tightly 67 (88 %)
- 1 point given for not selecting “Take Paracetamol” closed
Did not select “Take Paracetamol” 54 (71 %)
5. Which group of patients should be hospitalized? 1. Dengue without warning signs (F) 76 (100 %)
- Note: Each response is worth 1 point. Question is worth
a total of 4 points 2. Dengue without warning signs but with comorbidities (T) 17 (22 %)
- If the answer is correctly selected, the respondent gains 3. Dengue with warning signs (T) 59 (78 %)
1 point
- If the answer is correctly left blank, the respondent gains 4. Severe dengue (T) 58 (76 %)
1 point Percent answering all 4 correctly 15 (20 %)
- Responses that are correct are marked here as (T) and if
incorrect are marked as (F)
6. According to the WHO’s 2010 Clinical Management of Correct Responses n (%) selecting response
Dengue guidebook, what signs and symptoms can be
used to identify an infection of dengue without alarm
signs?
- Note: Question is worth 1 point Headache 59 (78 %)
- Each response is worth 1/19 point, which is given for
either correctly selecting a true response or correctly Muscle pain 60 (79 %)
leaving a false response blank Retro-orbital pain 62 (82 %)
Positive tourniquet test 45 (59 %)
Fever/subjective warmth 64 (84 %)
Petechial rash 33 (43 %)
Vomit 25 (33 %)
Incorrect
Ascites 1 (1 %)
Constipation 5 (7 %)
Diarrhea 10 (13 %)
Dyspnea 3 (4 %)
Dysuria 2 (3 %)
Chest pain 1 (1 %)
Edema 2 (3 %)
Icterus 1 (1 %)
Lymphadenitis 3 (4 %)
Nasal secretions 11 (14 %)
Persistent cough 3 (4 %)
Thrombocytopenia 15 (22 %)
Oral Hydration 70 (92 %)
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 7 of 10

Table 5 Knowledge-Based Questions (n = 76) (Continued)

7. Select any the treatments you could use in a patient
suspected to have dengue
- Note: Question is worth 1 point IV Hydration 12 (16 %)
- 0.5 points given for hydration (either oral and/or IV) and
Paracetamol 71 (93 %)
0.5 points given for paracetamol. Recipient is given 0
points if anti-bacterial or anti-viral medication is selected Anti-bacterial 1 (1 %)
Anti-viral 1 (1 %)
Any of the following (listed individually in survey): Aspirin, 0 (0 %)
NSAIDs/Steroids/Immunosuppressants (methotrexate,
cyclosporine, etc.)/Opioids/Platelets/Plasma/Whole blood
transfusion
Note: One point given per question, unless otherwise specified

dengue’s detrimental impact, as clinical knowledge and this question as asking which medications may help man-
concern for dengue infections are strongly associated. Of age dengue, the survey clearly asked how dengue may be
note, the clinical scenarios comprised a small component prevented, indicating a misconception of prevention strat-
(3 of 14 points) of the above-mentioned Cumulative egies. In addition, 25 % incorrectly selected “night time” as
Knowledge Score. the most likely feeding time for Aedes mosquitoes. Similar
findings have been documented elsewhere: Huang et al.
Providing patients with accurate dengue prevention and [14] found that only 14.4 % of Taiwanese providers cor-
treatment guidance rectly identified Aedes mosquito feeding habits, compared
Although study participants demonstrated a high level to 82.8 % who correctly identified Anopheles mosquito
of understanding of dengue infection signs, symptoms, feeding habits. When participants were asked which dengue
and treatment, we identified specific gaps in knowledge virus serotypes are found in Ecuador, only 38 % correctly
of dengue prevention and epidemiology. A total of 29 % answered all four serotypes (DENV 1–4). Ho [15] also
of participants incorrectly selected “take paracetamol” as found limited knowledge of dengue epidemiology among
a method for preventing dengue infection (Table 5). Al- healthcare providers in Taiwan, with only 47.7 % correctly
though it is possible that some participants misinterpreted responding that dengue is endemic in that country. These
Table 6 Clinical Knowledge Questions
Question Response Selected n (%)
1. An 8-year old male patient presents to your office with a (n = 73)
4 day history of fever, nausea, vomiting three times per day,
Dengue fever 0 (0 %)
and joint aches. He is accompanied by his mother, who
reports that he has been less active over the past few days Dengue hemorrhagic fever 5 (7 %)
and seems to be getting more uncomfortable. You note the
following abnormalities on physical exam: The patient has Dengue shock syndrome 0 (0 %)
bleeding of the oral mucosa, a palpable mass on the right side Dengue without warning signs 2 (3 %)
2 cm below the ribs, and winces when you palpate his
abdomen. You do not observe fluid in the abdomen or Dengue with warning signs (T) 61 (83 %)
difficulty breathing. Based on current WHO guidelines, this Severe dengue 5 (7 %)
patient is best classified as:
2. A 5-year-old girl patient presents to your office with a few (n = 73)
days of fever and a distended, painful abdomen. Her mother
states that she has been less active over the past 3 days. It is Order dengue lab tests, tell the patient to get rest at home, 10 (14 %)
currently February and you have seen six patients in the past and ask the patient to return to your office in 24 h
3 weeks with dengue infections. The best course of action in Order dengue lab tests and admit the patient to the hospital 54 (74 %)
managing this patient is to: for 24 h of observation (T)
Order dengue lab tests and admit the patient to the Intensive 9 (12 %)
Care Unit for close monitoring and access to emergency care
3. A 27-year-old male patient presents to your office in February (n = 71)
with two days of fever and complaints of muscle aches. He
notes that he has had three episodes of non-bloody vomiting Order dengue lab tests, tell the patient to get rest at home, 52 (73 %)
in the past two days. The patient notes that his younger sister and ask the patient to return to your office in 24 h (T)
has similar symptoms. You recall hearing numerous reports of Order dengue lab tests and admit the patient to the hospital 19 (27 %)
dengue infection during the last month. The best course of for 24 h of observation
action in managing this patient is to:
Order dengue lab tests and admit the patient to the ICU for 0 (0 %)
close monitoring and access to emergency care
Note: Each question is worth 1 point. (T) if placed next to the correct response
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 8 of 10

misunderstandings may lead healthcare providers to give Table 7 Practice-Based Questions

patients incorrect, clinically significant advice. It is critical Question Response selected n (%)
to target specific local misconceptions of dengue preven- Approximately how many patients do 0 3 (5 %)
tion and transmission through training of medical profes- you see per week? (n = 60)
1–49 6 (10 %)
sionals, in order to reduce the burden of dengue.
50–99 28
(46.7 %)
Confusion regarding hospital admission criteria
100–149 16
Clinicians indicated confusion when developing appropriate
(26.7 %)
dengue treatment plans for their patients. When healthcare
>150 7 (11.7 %)
professionals were asked which groups of patients with
dengue require hospital admission, only 22 % correctly Are you familiar with the WHO’s 2010 Yes 67 (89 %)
stated that patients with “dengue without warning signs but Clinical Management of Dengue
No 8 (11 %)
guidelines?
with comorbidities” require hospital admission (Table 5).
Do you feel that the WHO’s Dengue Yes 64 (97 %)
The 2009 WHO Dengue Guidelines provide specific rec-
guidelines help in managing dengue?
ommendations for appropriate clinical observation based No 2 (3 %)
on a patient’s risk of significant morbidity [7]. These guide- Of those patients who you suspect 0 % of patients 1 (1 %)
lines state that any patient with a comorbidity (e.g. diabetes have dengue fever, approximately
10 % of patients 10 (15 %)
what percentage do you refer to a lab
mellitus, obesity, risk of hemorrhage such as peptic ulcer
for diagnostic testing? 25 % of patients 3 (4 %)
disease) should be admitted to a hospital during a dengue
infection, regardless of the severity of infection. Addition- 50 % of patients 8 (12 %)

ally, only 45 % of participants correctly responded to all 75 % of patients 5 (7 %)

three clinical scenarios (Table 6), demonstrating knowledge 100 % of patients 40 (61 %)
gaps of patient admission criteria. Do your patients ever use a private lab Yes 31 (47 %)
Hospital admission rates for dengue infection vary con- without a referral?
No 35 (53 %)
siderably between regions globally. For example, Toma-
shek et al. [16] found that only 31 % of Puerto Rican Approximately what percentage of 0 14 (26 %)
patients with dengue fever do you
medical providers used hospital admission criteria consist- refer to the hospital for additional
<10 % 31 (57 %)
ent with the 1997 WHO Dengue Guidelines. Conversely, medical treatment? 25 % 1 (2 %)
Lee et al. [12] reported that one-third of providers in 50 % 6 (11 %)
Singapore “always” or “often” admitted patients with sus-
75 % 1 (2 %)
pected dengue, regardless of infection severity. Globally, it
is estimated that less than 5 % of patients infected with 100 % 1 (2 %)
dengue will develop severe disease [22], and WHO recom- Do you feel you have adequate Yes 48 (69 %)
mends that patients who do not meet criteria for resources to treat your patients when
No 22 (31 %)
they have dengue?
hospitalization have frequent office follow-up [7]. This is
particularly important in resource-limited settings. Pa- If you said ‘No’ to the previous Sufficient training 7 (32 %)
question, what are you lacking?
tients with comorbidities who are not admitted to hospi- - Note: Percentages given as n/22,
Medication needed 9 (41 %)
tals may have worse clinical outcomes, underscoring the to treat
based on previous question
need for close monitoring of this patient population. - Note: Subjects may select multiple Instruments needed 8 (36 %)
options to treat
Diagnostic testing: under-utilization and inadequate Access to lab tools 10 (45 %)
resources Note: Percentages given do not include respondents who did not answer
Study participants indicated suboptimal use of confirma- the question

tory diagnostic laboratory tests when dengue infection

was suspected (Table 7). As appropriate in a region with inform appropriate medical interventions. It is important
many acute febrile illnesses with similar clinical presen- to note that 14 % of providers reported inadequate ac-
tations as dengue, 61 % of healthcare providers reported cess to diagnostic testing for dengue (Table 7). Addition-
referring all patients with suspected dengue infection for ally, providers who agreed with the statement “I am fully
laboratory test confirmation. However, 20 % of partici- trained to manage a patient with an infection of dengue
pants reported referring patients for confirmatory la- without warning signs” referred a higher percentage of
boratory tests 25 % of the time or less. As these patients their patients for laboratory testing (r = 0.345, p <0.01),
may actually be infected with other febrile illnesses such compared to those who disagreed with this statement.
as leptospirosis, malaria, or chikungunya, laboratory This may signal one of two possibilities: clinical confi-
confirmation is crucial for differential diagnosis and to dence is increased with better access to diagnostic
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 9 of 10

testing, or providers who report greater confidence in associated with dengue fever in a resource-limited en-
their clinical training refer more patients for confirma- demic region. These findings highlight several strategies
tory laboratory tests. Access to dengue diagnostic testing to improve diagnosis and clinical management of dengue
remains a key issue in this context. infections in this region. A strong healthcare policy be-
Healthcare providers were asked about availability and gins with accurate information, which can best be ob-
access to a variety of resources for dengue diagnosis and tained and disseminated through close collaboration
treatment. A total of 31 % of providers reported having in- between the public, primary healthcare providers, health
adequate resources (See Table 7 for specific resources). educators, and the public health sector.
There were no significant correlations between reported
lack of resources for dengue diagnosis and treatment and  Healthcare providers should receive continuous
insufficient training, Cumulative Knowledge Scores, treat- education about dengue prevention, transmission,
ments used, or other items from this survey. Identifying and high-risk patient populations.
the impact of resource deficiencies is difficult to assess  Providers’ needs should be assessed in future studies,
from the data collected, as no discernible differences in as nearly one-third of participants reported inadequate
knowledge, attitudes, or practices were identified in this access to crucial healthcare resources.
study. Further investigation of availability and access to re-  Health providers should educate their patient
sources for dengue diagnosis and treatment, and how they population about the harms of self-diagnosis and
influence daily clinical practice is needed. presumptive self-medication.
 Findings demonstrated that those providers who
Awareness and Implementation of the WHO dengue showed the greatest concern of dengue infections
guidelines were also the most knowledgeable and provided
Awareness of the 2009 WHO Dengue Guidelines was clinical care that more closely aligned with WHO
high, with 89 % of participants reporting previous know- recommendations. Future interventions should
ledge of the guidelines. Of these respondents, 97 % re- therefore provide core dengue information while
ported that these guidelines were helpful. This finding is emphasizing dengue’s impact on health and
in contrast to Kularatne’s study of Sri Lankan practi- development.
tioners [13], in which only 45 % of practitioners reported  Periodic reassessment of the local knowledge,
using the WHO Dengue Guidelines. However, Kular- attitudes, and clinical practices will be instrumental
atne’s study was conducted prior to the current version to reduce the burden of dengue fever and improve
of the WHO Dengue Guidelines, and may be more re- clinical management in high-burden settings.
lated to local medical practices and training.
Additional files
Impact of practice setting
In this study, there were no significant differences in re- Additional file 1: Appendix A1: Knowledge, Attitudes, and Practices of
ported knowledge, attitudes, and practice, between health- Dengue Survey – English Version. (DOC 49 kb)

care providers practicing in a hospital versus ambulatory Additional file 2: Appendix A2: Knowledge, Attitudes and Practices of
Dengue Survey – Spanish Version. (DOC 55 kb)
settings, including familiarity with the 2009 WHO Dengue
Guidelines, reporting that these guidelines were helpful,
Abbreviations
or overall dengue knowledge (p >0.05). Previous studies CKS, Cumulative Knowledge Score; CSS, Clinical Scenario Score; PAHO, Pan
have indicated that practice settings can influence clinical American Health Organization; WHO, World Health Organization
management of dengue fever. Ho et al. [15] found that
Acknowledgments
healthcare providers practicing at Taiwanese medical cen- Thank you to the healthcare providers of Machala for their participation and
ters (i.e. medical school-affiliated hospitals at the highest to SNEM Machala, MSP Ecuador, and the Center for Global Health and
accreditation level) had significantly different levels of Translational Science at SUNY Upstate Medical University for their support.
knowledge, compared to providers at non-medical centers.
Funding
In order for dengue interventions to be most effective in The IDSA Medical Scholars Program provided a stipend to ASH to support
hyper-endemic regions, healthcare providers of all types this study.
and at all settings must receive adequate training and
Availability of data materials
guidance, and differences in knowledge, attitudes, and Survey instruments in English and Spanish are included as supplements. Raw
practice by setting should continue to be assessed. survey data available upon request.

Conclusion Authors’ contributions

ASH was involved in study design, creation of the questionnaire, data
Findings from this study provide important insights into collection, statistical analysis, interpretation of results, and coordinating and
medical practitioner knowledge, attitudes, and practices drafting the manuscript. EBA was involved in study design, creation of the
Handel et al. Tropical Diseases, Travel Medicine and Vaccines (2016) 2:8 Page 10 of 10

questionnaire, and data collection. MJBC was involved in study design, 11. Kularatne SA. Survey on the management of dengue infection in Sri Lanka:
creation of the questionnaire, and interpretation of results. AGF participated opinions of physicians and pediatricians. Southeast Asian J Trop Med Public
in statistical analysis and interpretation of results. JLF participated in Health. 2005;36(5):1198–200.
interpretation of results and review of the manuscript draft. RXRE was 12. Lee LK, Thein TL, Kurukularatne C, Gan VC, Lye DC, Leo YS. Dengue
involved in study design, creation of the questionnaire, and data collection. knowledge, attitudes, and practices among primary care physicians in
SJR was involved in study design, creation of the questionnaire, interpretation Singapore. Ann Acad Med Singapore. 2011;40(12):533–8.
of results, and review of the manuscript. AMSI conceived of the study, and was 13. Thaver AM, Sobani ZA, Qazi F, Khan M, Zafar A, Beg MA. Assessing the need
involved in study design, creation of the questionnaire, data collection, for training: general practitioners’ knowledge, attitude and practice
statistical analysis, interpretation of results, and coordinating and drafting the concerning dengue and malaria in Karachi, Pakistan. Int Health. 2011;3(2):
manuscript. 126–30.
14. Huang HL, Chiu TY, Huang KC, Cheng SY, Yao CA, Lee LT. Travel-related
Competing interests mosquito-transmitted disease questionnaire survey among health
The authors declare that they have no competing interests with the research. professionals in Taiwan. J Travel Med. 2011;18(1):34–8.
15. Ho TS, Huang MC, Wang SM, Hsu HC, Liu CC. Knowledge, attitude, and
practice of dengue disease among healthcare professionals in Southern
Consent for publication Taiwan. J Formos Med Assoc. 2013;112(1):18–23.
N/A 16. Tomashek KM, Biggerstaff BJ, Ramos MM, Pérez-Guerra CL, Garcia Rivera EJ,
Sun W. Physician survey to determine how dengue is diagnosed, treated
Ethics approval and consent to participate and reported in Puerto Rico. PLoS Negl Trop Dis. 2014;8(10):e3192.
The investigation protocol was reviewed and approved by the Institutional 17. PAHO/WHO. Manejo Clínico Del Dengue - Dengue Diagnóstico Y Tratamiento
Review Boards (IRB) of the State University of New York (SUNY) Upstate De Casos. 2010. http://www.paho.org/ecu/index.php?option=com_
Medical University in Syracuse, New York, and the Ecuadorian Ministry of content&view=article&id=544:materiales-enfrentar-dengue&catid=692:ecu.13-
Health. The study was certified as exempt by the IRBs (as all participants comunicacin-social&Itemid=289. Accessed 27 Sep 2015.
were over the age of 18 and no personal identifying information was 18. Stewart-Ibarra AM, Ryan SJ, Belt Beltrán E, Mejía R, Silva M, Muñoz A.
collected), and no informed consent was required during the administration Dengue vector dynamics (Aedes Aegypti) influenced by climate and social
of the questionnaire. factors in Ecuador: implications for targeted control. PLoS One. 2013;8(11):
e78263.
Author details 19. Stewart-Ibarra AM, Lowe R. Climate and non-climate drivers of dengue
1
Department of Pediatrics, Stony Brook Children’s Hospital, Stony Brook, NY, epidemics in Southern Coastal Ecuador. Am J Trop Med Hyg. 2013;88(5):
USA. 2Universidad Técnica de Machala, Machala, Ecuador. 3Ministerio de 971–81.
Salud Publica, Machala, El Oro, Ecuador. 4Facultad de Ingenieria Maritima, 20. Anuario Epidemiologico. 1994–2014. Direccion Nacional de Vigilancia
Ciencias Oceanicas y Recursos Naturales, Escuela Superior Politécnica del Epidemiologica, Ministerio de Salud Publica de Ecuador. http://www.salud.gob.
Litoral (ESPOL), Guayaquil, Guayas, Ecuador. 5Division of Nutritional Sciences, ec/direccion-nacional-de-vigilancia-epidemiologica/. Accessed 1 Apr 2016.
Cornell University, Ithaca, NY, USA. 6Department of Geography, University of 21. Stewart-Ibarra AM, Luzadis VA, Borbor-Cordova MJ, Silva M, Ordoñez T,
Florida, Gainesville, FL, USA. 7Emerging Pathogens Institute, University of Beltrán E, Ryan SJ. A social-ecological analysis of community perceptions of
Florida, Gainesvillee, fl, USA. 8Center for Global Health and Translational dengue fever and Aedes aegypti in Machala, Ecuador. BMC Public Health.
Science, State University of New York Upstate Medical University, Syracuse, 2014;14(1):1135.
NY, USA. 22. Tomashek KM, Sharp RM, Morgolis HS. Chapter 3: Infectious Diseases
Related to Travel: Dengue. In: CDC Health Information for International
Received: 2 February 2016 Accepted: 25 May 2016 Travel. New York: Oxford University Press; 2014.

References
1. Halstead S. Dengue. Lancet. 2007;370(9599):1644–52.
2. Hotez PJ, Bottazzi ME, Franco-Paredes C, Ault SK, Periago MR. The neglected
tropical diseases of Latin America and the Caribbean: a review of disease
burden and distribution and a roadmap for control and elimination. PLoS
Negl Trop Dis. 2008;2(9):e300.
3. PAHO (n.d.) Number of reported cases of dengue and severe dengue (DS)
in the Americas by country (2010–2014). Washington: Pan American Health
Organization. http://www.paho.org/hq/index.php?option=com_
topics&view=readall&cid=3273&Itemid=40734&lang=en. Accessed 27 Sep
2015.
4. Díaz-Quijano FA, Waldman EA. Factors associated with dengue mortality in
Latin America and the Caribbean, 1995–2009: an ecological study. Am J
Trop Med Hyg. 2012;86(2):328–34.
5. Shepard DS, Coudeville L, Halasa YA, Zambrano B, Dayan GH. Economic
impact of dengue illness in the Americas. Am J Trop Med Hyg. 2011;84(2): Submit your next manuscript to BioMed Central
200–7. and we will help you at every step:
6. World Health Organization. Dengue: guidelines for diagnosis, treatment,
prevention and control. Geneva: World Health Organization; 2009. • We accept pre-submission inquiries
7. Farrar JJ, Hien TT, Hortisck O, Hung NT, Jaenisch T, Junghanns T, et al. • Our selector tool helps you to find the most relevant journal
Dogma in classifying dengue disease. Am J Trop Med Hyg. 2013;89(2):198–201.
• We provide round the clock customer support
8. Halstead SB. Dengue: the syndromic basis to pathogenesis research. Inutility
of the 2009 WHO case definition. Am J Trop Med Hyg. 2013;88(2):212–5. • Convenient online submission
9. Prasad D, Kumar C, Jain A, Kumar R. Accuracy and applicability of the • Thorough peer review
revised WHO classification (2009) of dengue in children seen at a tertiary
• Inclusion in PubMed and all major indexing services
healthcare facility in Northern India. Infection. 2013;41(4):775–82.
10. Barniol J, Gaczkowski R, Barbato EV, da Cunha RV, Salgado D, Martínez E, et • Maximum visibility for your research
al. Usefulness and applicability of the revised dengue case classification by
disease: multi-centre study in 18 countries. BMC Infect Dis. 2011;11:106. Submit your manuscript at
www.biomedcentral.com/submit