This study used a mixed methods approach combining key informant interviews, document analysis, focus group discussions, secondary data analyses and interpretation by an international dengue expert meeting. In order to get representative information on dengue endemic countries, five countries in Latin America (Brazil, Colombia, Dominican Republic, Mexico, Peru) and five countries in Asia (Indonesia, Malaysia, Maldives, Sri Lanka, Vietnam) were selected, based on the following criteria: a) representation of large (e.g. Brazil, Indonesia), intermediate (e.g. Peru, Vietnam) and small countries (e.g. Dominican Republic, Maldives) in each region; b) recent dengue outbreaks; c) existing relations with Ministries of Health and academic institutions to ensure openness and confidence between respondents and interviewers. Three countries were excluded as they had been analysed previously in a WHO-TDR supported study: Thailand, Cambodia and Bolivia (Runge-Ranzinger et al. unpublished data). Within the study countries, academic institutions and/or Ministries of Health were approached to select country interviewers who had public health knowledge, understanding of both disease surveillance and dengue, and skills in undertaking interviews. The interviewers received detailed instructions and data collection forms that roughly followed the Centers for Disease and Prevention “framework for evaluating public health surveillance systems for early detection of outbreaks” [3
] and had been pre-tested and revised following use in 3 countries (Nepal, Bangladesh, and Colombia). The set of data collection forms comprised:
Questionnaire for interviewers about the country context (data to be extracted from published and unpublished documents in the country).
Questionnaire/topic guide “Evaluating dengue surveillance and response” that included 83 items to be covered in relation to dengue disease surveillance, case notification and dengue classification, virological surveillance, routine vector surveillance and control, community participation during non-epidemic periods, outbreak preparedness, outbreak detection, recent experiences with a dengue outbreak and opinions about the success of the outbreak response.
Hospital questionnaire with 38 items about outbreak preparedness in hospitals, available resources in non-epidemic periods, characteristics of the last dengue outbreak, availability of resources during the outbreak, opinions about successes and failures of outbreak management.
Topic guide for focus group discussions to be applied in hospitals with a recent dengue outbreak
Matrix for the summarisation of collected information.
The work within countries included the following:
General appraisal of the epidemiology of dengue in the study country (questionnaire for interviewers mainly to be filled at national and State/Provincial level).
Semi-structured interviews using a list of 83 topics to be discussed with the interviewees (questionnaire/topic guide “evaluating dengue surveillance and response”). These data collection forms, developed with the above-mentioned framework for evaluating surveillance systems in mind (Buehler et al. 2004), were applied to key informants in each country at different levels: a) Government officials (central level): decision or policy maker, epidemiologist/ surveillance expert, laboratory expert (microbiologist or technician), entomologist, other. b) Government officials (state/province/district, sub-district level): epidemiologist/surveillance expert, laboratory expert (microbiologist and/or laboratory technician), entomologist, other. The number of interviews to be conducted followed the “saturation principle” (when no further information could be collected from informants the interview series was terminated) and included in general 20 to 30 respondents per country.
The hospital questionnaire with 38 items was applied in each country in 3 to 5 hospitals of different levels (teaching hospital, district hospital, sub-district hospital) using a purposive sampling strategy. At most sites there were none or only a few more similar hospitals in the study area.
Focus group discussions  with hospital staff involved in the last dengue outbreak were conducted for complementing the information obtained from individual interviews.
Combining complementary methodologies and information from varied sources, as well as several rounds of data verification before, during and after the expert meeting ensured a high level of internal validity providing more reliable results.
Ethical approval was received from WHO Regional Offices (Ethical Review Boards at the Pan American Health Organisation (PAHO), South-East Asian Regional Office (SEARO) and the Western Pacific Regional Office (WPRO) which was accepted by the study countries except for Peru where an additional ethical approval was obtained from the Institutional Review Board at Cayetano Heredia University. Verbal, and in some cases, written consent was obtained from respondents. The interviewees were assured of the anonymisation of their responses and interviews took place in closed rooms with no other persons being present. The completed forms were kept in separate files and no individual names of the respondents were recorded.
The country interviewers completed the data collection in the following areas: Brazil: : Brasilia Ministry of Health (National level). Pernambuco, Amazonas, Rio de Janeiro and Goiás (State level). Interviews at health units and hospitals in Amazonas and Rio de Janeiro (local level).; Colombia: : Bogota (National Institute of Health National level), Departamento Valle del Cauca (State level).; Dominican Republic: : Santo Domingo (National level) Santiago Valverde, San Juan, Azua, and Hato Mayor (Provincial level), Laguna Salada, Esperanza y Cienfuegos (Municipality level).; Mexico: : Mexico City Ministry of Health (National level), Yukatan and Guerrero (State level).; Peru: : Lima Ministry of Health (National level), Loreto and San Martin (State or Regional level), interviews in hospitals of Iquitos (Loreto), Tarapoto and Moyobamba (San Martin local level).; Indonesia: : Jakarta (National Provincial and District level) Yogyakarta (Provincial level) Bantul (District level).; Malaysia: : Kuala Lumpur Ministry of Health, Federal Territory of Putrajaya (National level), Selangor, Penang (State level) Klang Hulu Langat, Gombak , Petaling, Banting, Kuala Selangor(District level).; Maldives: : Male Ministry of Health (National. level), Hulumale, Thinadoo, Addu (Regional and District level).; Sri Lanka: : Colombo Ministry of Health, Medical Research Institute (National level), Colombo and Gampaha district (District level), hospital interviews in three hospitals in Colombo and Gampaha District (local level).; Vietnam: : Ha Noi Ministry of Health and National Hospital of Infectious Diseases (National level) Ho Chi Minh City (Hospital for Tropical Diseases District 8 hospital, Pasteur Institute, University of Medicine and Pharmacy, Preventive Medicine Centre of Ho Chi Minh City) Dong Thap province (Provincial Hospital Dong Thap Preventive Medicine Centre and Volunteer Group for Dengue).
The field work was done within a 5 month period (October 2011 to March 2012). The interviewers produced a comprehensive report which included a detailed analysis of the dengue epidemiology in their country, the completed data collection forms and the completed matrix with the summary of findings of each item in the questionnaire/topic guide “Evaluating dengue surveillance and response”. The information package was sent to the central team at WHO-TDR for compilation and preliminary comparative analysis. A 24 page synopsis of findings for both Latin American and Asian countries was produced. It was circulated among interviewers, Ministry of Health staff in the participating countries and WHO focal points for verification and complementation. Thereafter in June 2012 a three-day international expert workshop was organised by WHO-TDR involving all interviewers and two representatives from Ministries of Health of each country. The 45 participants were tasked to a) further validate the collected information; b) interpret and regionalise country findings and c) discuss recommendations. Expert consensus was gained through mediation by the Chairpersons. The following section provides the results of a synoptic analysis of the 10 study countries.