Online Respondent-Driven Detection (RDD)
1 like475 views
The document discusses online respondent-driven detection (RDD) for identifying and analyzing contact networks relevant to disease transmission, emphasizing its application in surveillance and outbreak detection. It highlights the significance of network factors, disease metrics, and the use of online surveys to gather data on infectious disease cases. The pilot studies and methods presented illustrate the approach to detect clustering behaviors and improve disease control strategies.
![Hepatitis B screening intention
Moroccan immigrants
The Netherlands
“Social ties influence having
the same screening intention”
OR 1.70 [1.15 – 2.51]
Hamdiui et al. (in preparation)
Peer-driven interventions
www.respondentdrivensampling.org
Ice Bucket Challenge
Hamdiui et al. BMC Medicine 2018](https://image.slidesharecdn.com/escaidepresentationmlsteinforprint-190130104753/85/Online-Respondent-Driven-Detection-RDD-43-320.jpg)
Online Respondent-Driven Detection (RDD)
- 1. Online Respondent-Driven Detection (RDD) Mart Stein, PhD mart.stein@rivm.nl National Coordination Centre for Communicable Disease Control (LCI CIb RIVM) ESCAIDE 2018 | 23-11-2018
- 2. Titel | 30-05-2018 2 Prof. Mirjam Kretzschmar Prof. dr. ir Vincent Buskens Dr. Jim van Steenbergen Prof. dr. Anna Thorson Dr. Linus BengtssonDrs. Nora Hamdiui Prof. dr. Aura Timen Prof. dr. Maria van den Muijsenbergh
- 3. Respondent-driven sampling for surveillance % ‘Typical’ RDS % % …
- 4. Types of networks Family Neighbours Coworkers Malta network
- 8. Respondent-Driven Detection to identify cases
- 9. Respondent-Driven Detection to identify cases Topics
- 10. Contact network factors relevant for pathogen transmission SARS Ebola Tuberculosis Rotavirus Varicella Rubella … Pertussis Mumps Measles Influenza
- 11. Contact network factors relevant for pathogen transmission Network density Number of contact persons (degree / network size) Mixing patterns Contact type Duration Frequency
- 12. Pathogen transmission Pathogen‐ Virulence ‐ Infectivity ‐ Modes of transmission ‐ … EnvironmentHost ‐ Age ‐ Prior exposure ‐ Susceptibility ‐ Co‐infection ‐ Immune response ‐ … Disease ‐ Contact network factors ‐ Climate ‐ Physical structures ‐ … Epidemiological triangle
- 13. Basic reproduction number R0 “The expected number of secondary cases produced by a typical infected individual during its entire infectious period in a completely susceptible population” When R0 < 1: outbreak will die out R0 > 1: pathogen is able to spread in population Anderson & May, 1991 O. Diekmann, et al. 1990 R0 = Contact rate x Probability of transmission upon contact x Duration infectious period
- 15. R0 Disease metrics Empirical data on disease and contact networks Statistics Mathematical disease models Predictions Interventions
- 16. Contact diary surveys (self‐reported) Pioneered by Edmunds et al 1997 Online survey Beutels et al 2006 Combine with serological data Wallinga et al 2006 Prospective vs retrospective design Mikolajczyk & Kretzschmar 2008 Contacts of school children Mikolajczyk et al 2008 Contact survey in 8 European countries POLYMOD, Mossong et al 2008 Contact surveys in Asian countries SMILI study Detrimental and beneficial effects social networks Brinkhues et al 2018 Human–livestock contacts Klous et al 2016; Webster et al 2017 HCWs’ contacts in and around health care settings Jiang et al 2018; English et al 2018 See review by Read et al 2012 Quantifying potential infectious contacts
- 17. D. Klinkenberg et al IB 2018 Simulation of a zombie epidemic…
- 20. Contact tracing
- 21. NEJM, The Wall Street Journal (2014) Example of contact tracing 2014–2016 West Africa Ebola virus disease epidemic Clustering of cases Heterogeneous networks ...
- 22. Traditional contact tracing ? ? ? ? ? ? ? Labour intensive Complex networks Too late?! Detrimental effects Resource consuming
- 24. Respondent Driven Detection (RDD) Objective: detection of cases or clusters of disease General population Invite persons with whom you had face-to-face contact in a certain time period Degree: contacts that relevant for (potential) disease transmission *Heckathorn, Douglas D. Social Problems 1997 Wejnert and Heckathorn et al Web-Based Network Sampling. 2008 Respondent Driven Sampling (RDS)* Objective: estimate disease prevalence using a statistical model hard-to-reach populations (injecting drug users) Invite individuals from same social network Degree: known persons from the same social network
- 27. Set of symptoms in preceding 2 weeks disease‐related questions Basic demographics age gender education … WELCOME Only use the buttons ‘Previous’ and ‘Next’ below the pages to navigate through this survey. Question 1. Is this the first time that you fill in the survey? Yes No Question 2. I am a: Male Female Methods - Survey Influenza‐like‐illness (ILI) • human‐to‐human transmission • common illness • seasonal
- 28. WELCOME Only use the buttons ‘Previous’ and ‘Next’ below the pages to navigate through this survey. Question 1. Is this the first time that you fill in the survey? Yes No Question 2. I am a: Male Female Defined a contact person persons touched or talked within a distance of about one arm’s length Contact persons number and age group during one day at four settings individual’s degree
- 29. Invite online max. 4 contact persons “Face‐to‐face contact in past two weeks” Incentives ‐ Join a raffle for gift card €25 ‐ Your recruitment tree ‐ Anonymous survey results Recruitment criteria
- 30. Marquet et al. BMC Public Health, 2006 InfluenzaNet • Weekly newsletters • Volunteers provide ILI data • 14,000 active in NL 2013‐2014 • Reaching those not visiting the GP Recruitment of seeds Web based participatory syndromic surveillance
- 31. Sample composition number of individuals in tree number of waves Freq. (log10) Freq. (log10) Panel members ±14000 active volunteers 1105 seeds 455 recruits
- 32. Stein ML, et al. BMC Infectious Diseases 2015 Age of recruiter Age recruited contact Contact network Who recruits whom? Degree of recruiter Degree recruited contact
- 33. Recruitment generation interval Probability density time in hours (log10) Spatial recruitment patterns median distance 3.0 km (SD 38.5) 180 (39.6% of 455) with same postal code 76.4% within province Stein ML, et al. BMC Infectious Diseases 2015
- 34. dseeds + Recruits + Recruits ‐ Recruits + Recruits ‐ 109 (69.0%) 49 (31.0%) 118 (52.0%) 109 (48.0%) 71 (43.0%) dseeds ‐ 94 (57.0%) 17% P = 0.001 X2 test with continuity correction. Fisher's exact test (<10). More symptomatic recruits by symptomatic seeds One or more general symptoms Stein ML, et al. American Journal of Public Health 2015
- 35. dseeds + Recruits + Recruits ‐ Recruits + Recruits ‐ 4 (16.7%) 20 (83.3%) 21 (5.8%) 340 (94.2%) 16 (9.7%) dseeds ‐ 149 (90.3%) 10.9% P = 0.061 Common cold symptoms Stein ML, et al. American Journal of Public Health 2015
- 36. dseeds + Recruits + Recruits ‐ Recruits + Recruits ‐ 4 (18.2%) 18 (81.8%) 14 (3.9%) 349 (96.1%) 12 (7.3%) dseeds ‐ 153 (92.7%) 14.3% P = 0.015 Symptom Fever Stein ML, et al. American Journal of Public Health 2015
- 37. 14 lab confirmed cases day of onset symptoms <4w Intensive contact <6w Online RDD in practice Pilot study pertussis Follow up after 2 weeks April-May 2017
- 38. Online RDD Pilot study pertussis unpublished data 9 seeds 19 contact persons
- 39. Online RDD In perspective Empirical data on contact networks Detect ‘hidden’ cases “Saves time”Diverse groups of individuals Short time frame Large geographic coverage
- 40. Online RDD Challenges Online recruitment “R0” > 1 Added value? When? Disease severity and situation PertussisMumps Measles ………… Ebola ? RubellaShigella Polio Stein ML, et al. PLOS ONE 2018
- 41. Next steps… Drivers online recruitment Use online RDD to detect clustering of behaviour CDC Public Health ID 11162
- 42. RDD for studying clustering of behaviour
- 43. Hepatitis B screening intention Moroccan immigrants The Netherlands “Social ties influence having the same screening intention” OR 1.70 [1.15 – 2.51] Hamdiui et al. (in preparation) Peer-driven interventions www.respondentdrivensampling.org Ice Bucket Challenge Hamdiui et al. BMC Medicine 2018
- 44. Julius Center Martin Bootsma University of Utrecht, Netherlands Vincent Buskens Peter van der Heijden Karolinska Institutet, Sweden Anna Thorson Linus Bengtsson University of Greenwich Luis E.C. Rocha Stockholm University, Sweden Fredrik Liljeros RIVM Nora Hamdiui Jim van Steenbergen Aura Timen Mirjam Kretzschmar Hans Lobach Danielle Timmermans Albert Wong Kata Ottovay Felix ter Schegget Lisette van Breen Mireille Bakker Hella Smit Mahidol University, Thailand Charnchudhi Chanyasanha Mathuros Tipayamongkholgul Wasamon Sabaiwan Science in Action Carl Koppeschaar Ronald Smallenburg Games for Health Acknowledgements Thank you! GGD Regio Utrecht Diederik Brandwagt Gerda van den Berg GGD Zuid Limburg Christian Hoebe Nicole Dukers Carlijn Soomers GGD Gelderland-Zuid Helma Ruijs Anne de Munter-Mulder GGD NOG Lutje Huijbregts Pharos / Radboud University Maria van den Muijsenbergh mart.stein@rivm.nl