|
|||||||||||||||||
|
|
EID Home | Ahead of Print | Past Issues | EID Search | Contact Us | Announcements | Suggested Citation | Submit Manuscript PDF Version | Comments | Email this article |
|
Computer-assisted Telephone Interview TechniquesTo the Editor: Fox et al. used computer-assisted telephone interview (CATI) techniques in an outbreak of cryptosporidiosis (1). Australian health agencies have used CATI for several years. A case-control study during an outbreak of Salmonella Mbandaka in 1996 employed CATI to interview 15 case-patients and 45 controls; contaminated peanut butter was implicated (2). Foodborne disease outbreaks are often geographically widespread and suited to using CATI. Australian health authorities investigate ≈100 outbreaks of foodborne disease each year, with 3–4 using CATI-based case-control studies. Some jurisdictions investigate outbreaks by using CATI interviews of controls sampled from a bank of potential study participants (3). Potential study participants are recruited at the conclusion of rolling risk factor survey interviews, similar to the Behavioral Risk Factor Surveillance System. A "control bank" allows investigators to rapidly obtain contact details for appropriately matched controls because age and sex of all household members are recorded in a database. Using control banks with CATI allows completion of studies quicker than CATI or traditional methods alone (4). South Australia has used CATI during 11 case-control studies of salmonellosis, legionellosis, Q fever, campylobacteriosis, Shiga toxin–producing Escherichia coli, and cryptosporidiosis (http://www.dh.sa.gov.au/pehs/notifiable-diseases-summary/current-outbreak-table.htm). During an Australian CATI survey about gastroenteritis, 5,123 (84%) of 6,087 households agreed to be in a control bank (5). This bank of 14,024 potential controls was used in 4 case-control studies of sporadic salmonellosis and campylobacteriosis. This system avoided randomly dialing thousands of households to enroll controls in young age groups. The control bank was used for 3 years after initial collection, although many jurisdictions update banks annually. Investigators may find CATI useful, although it can be costly and introduce biases (4). Programming questionnaires can delay investigations, which makes paper-based collection better in small outbreaks (4). CATI cannot be used in areas where a small proportion of the population has telephones. Despite limitations, CATI, when combined with control banks, may improve outbreak investigations. Martyn Kirk,*
Ingrid Tribe,† Rod Givney,† Jane Raupach,† and Russell Stafford‡ Suggested citation
for this article: References
In response: We appreciate the comments of Martyn Kirk and colleagues, who describe their experience using computer-assisted telephone interview (CATI) techniques in Australia with geographically widespread foodborne outbreaks (1). The intent of our article was to illustrate 1 example of the use of the CATI infrastructure in investigating a large communitywide cryptosporidiosis outbreak (2); yet we recognize the applicability of this infrastructure to multiple acute infectious disease outbreak scenarios. In our article, we comment that the use of existing CATI systems, like the Behavioral Risk Factor Surveillance System (BRFSS), can provide a practical means for obtaining controls in case-control studies, and the letter by Kirk and colleagues describes their use of the CATI infrastructure to create a "control bank" for acute infectious disease outbreak investigations. This control bank includes participants of longitudinal risk factor surveys, like BRFSS, who are subsequently recruited as controls for outbreak investigations. We acknowledge that a "bank" of these readily accessible controls could permit more rapid recruitment of participants in numerous age strata and obviate the need for extensive random digit dialing to recruit an adequate age-matched control population in many investigations. Nevertheless, in most epidemiologic investigations, controls need to be selected from the same geographic area as the case-patients, and even in large telephone surveys, the number of respondents in any given region can be small. This would make it difficult to recruit enough controls within the small areas affected by most outbreaks, particularly within specific age strata. A control bank may therefore be more practical for use in large communitywide outbreaks, outbreaks that occur over large regions (i.e., an entire state), or in densely populated areas. Additionally, the lengthy start-up time required for questionnaire programming with a CATI system also supports the view that the CATI method may be maximally applicable in large-scale investigations. LeAnne M. Fox,*
D. Charles Hunt,†‡ and Michael J. Beach* Suggested citation
for this article References
|
|
||||||||||||
|
||||||||||||
|
EID Home | Top of Page | Ahead-of-Print | Past Issues | Suggested Citation | EID Search | Contact Us | Accessibility | Privacy Policy Notice | CDC Home | CDC Search | Health Topics A-Z |
||
This page
posted March 21, 2006 |
||
Emerging
Infectious Diseases Journal |
||