Info

From Henry, J., Magruder, S., Snyder, M. (2004a). Using Kaiser Permanente Nurse Advice Hotline Data for syndromic surveillance in the national capital area. MMWR Morb Mortal Wkly Rep Under review.

From Henry, J., Magruder, S., Snyder, M. (2004a). Using Kaiser Permanente Nurse Advice Hotline Data for syndromic surveillance in the national capital area. MMWR Morb Mortal Wkly Rep Under review.

and (3) the time of the hotline call was earlier than or simultaneous with the time of the call for the appointment.

They assigned zero, one, or more than one syndrome to the office visit based on the ICD-9 codes assigned for the visit, using the existing ESSENCE II ICD-9-to-syndrome mappings (Chapter 23, Table 23.3). For each syndrome, they calculated the sensitivity, specificity, and positive predictive value, using office visits as a diagnostic standard.

Table 28.4 shows the number of hotline calls and outpatient office visits by syndrome. Calls categorized as respiratory were most likely to result in a (linkable) subsequent office visit, and calls for neurologic least likely.

Of the syndrome groups studied, the sensitivity of hotline calls for respiratory was highest (74.7%), followed by hotline calls for gastrointestinal (72%). Sensitivity for hemorrhagic was lowest (13.3%). The specificity ranged from 88.9% to 99.9% (Table 28.5). The mean time lag between the nurse advice call and the corresponding physician encounter by syndrome ranged from 8.2 to 25 hours. Note that the actual time difference experienced by a biosurveillance system will likely be greater as the physician diagnosis is recorded after the encounter, whereas the nurse selected protocol is available very early in the telephone call.

Doroshenko et al. (2004) used correlation analysis to study calls to the United Kingdom's Nurse Direct helpline and one outbreak of influenza. NHS Direct is a nurse-led telephone helpline that provides health information and advice to callers with symptoms. Nurses at the 22 NHS Direct sites use a computerized clinical decision support system containing approximately 200 clinical algorithms, each with a series of questions relating to symptoms. The NHS Direct syndromic surveillance system automatically assigns a patient into one of 10 syndromes (listed above).

The evaluation measured the Spearman rank correlation coefficient between weekly counts of cold/influenza syndrome, as determined by the NHS Direct syndromic surveillance system with weekly counts the Royal College of General Practitioner's Weekly Return Service (WRS) (Figure 28.2). WRS is an established sentinel clinician surveillance system for ILI.The Spearman correlation of 0.85 is a measure of the

figure 28.2 Time series of weekly counts of cold/influenza from the NHS (National Health Service) Direct syndromic surveillance system (upper line) and the Royal College of General Practitioner's Weekly Return Service (lower line). These data are from England and Wales for the period August 2001-August 2004. (From Doroshenko, A., Cooper, D., Smith, G., et al. (2004). Evaluation of syndromic surveillance based on NHS Direct derived data in England and Wales. MMWR Morb Mortal Wkly Rep 54(Suppl):117-22, with permission.)

1 15 30 43 57 7t 85 SO 113 127 141 155

Weeks5

figure 28.2 Time series of weekly counts of cold/influenza from the NHS (National Health Service) Direct syndromic surveillance system (upper line) and the Royal College of General Practitioner's Weekly Return Service (lower line). These data are from England and Wales for the period August 2001-August 2004. (From Doroshenko, A., Cooper, D., Smith, G., et al. (2004). Evaluation of syndromic surveillance based on NHS Direct derived data in England and Wales. MMWR Morb Mortal Wkly Rep 54(Suppl):117-22, with permission.)

correlation without a time offset, and the result indicates that the two time series are strongly correlated.3

Doroshenko et al. (2004) also reported that, in structured interviews with stakeholders, the majority of respondents indicated that the NHS Direct syndromic surveillance system registered an increase in calls about diarrhea and vomiting at the times when traditional public health surveillance systems indicated a national increase in norovirus.

Platt et al. (2003) operate the National Syndromic Surveillance Demonstration Project. Included among its data sources are telephone triage data from Optum, a company that operates triage call centers that serve healthcare facilities in 50 states (Yih et al., 2004). There have been no published studies as of the time of this writing of the value of the call center data collected.4

0 0

Post a comment