Skip to main content

Table 4 Sensitivity, Specificity and Positive Predictive Value of Name Recognition Software

From: Use of name recognition software, census data and multiple imputation to predict missing data on ethnicity: application to cancer registry records

Name recognition software Ethnic group Sensitivity Specificity Positive predictive value
Onomap White 99.8% 51.5% 98.0%
  South Asian 82.1% 99.9% 92.9%
  Black 4.4% 99.9% 70.8%
  Chinese/Other 0.0% 100.0%  
Nam Pehchan* South Asian 71.1% 99.9% 94.5%
Onomap and Nam Pehchan combined South Asian 90.5% 99.9% 93.3%
  1. Includes 85 352 cases where a single ethnic group was recorded in the HES dataset. Individual logistic models (positive outcome threshold: p > = 0.5).
  2. * Matched on forename and surname separately.