Skip to main content

Table 4 Proportion and predictors of locating women who required tracing, n = 372

From: Predictors of locating women six to eight years after contact: internet resources at recruitment may help to improve response rates in longitudinal research

  Located Crude odds ratio Adjusted* odds ratio
  (n = 181) OR (95% CI) OR (95% CI)
  n (%)     
Age group (years)       
   65–69 37 (55.2) 1.95 (1.09–3.50) 1.73 (0.93–3.22)
   70–79 87 (55.1) 1.94 (1.23–3.05) 2.13 (1.31–3.46)
   80–89 57 (38.8) 1.00 (referent) 1.00 (referent)
Region (population density)       
   Low 91 (54.8) 1.00 (referent) 1.00 (referent)
   High 90 (43.7) 0.64 (0.42–0.97) 0.55 (0.35–0.86)
2001 web-search       
   Yes 70 (56.5) 1.60 (1.04–2.47) 1.82 (1.14–2.90)
   No 111 (44.8) 1.00 (referent) 1.00 (referent)
Study address status§       
   No change 41 (60.3) 1.00 (referent) 1.00 (referent)
   New address 10 (58.8) 0.64 (0.25–1.69) 1.00 (0.33–3.02)
   Possible address 66 (61.1) 0.93 (0.49–1.77) 1.09 (0.57–2.07)
   Unable to determine 64 (35.8) 0.33 (0.18–0.60) 0.34 (0.19–0.61)
Reason required tracing       
   Returned mail 128 (50.0) 1.00 (referent)   
   Incorrect number 53 (45.7) 1.19 (0.77–1.85)   
  1. CI = confidence interval.
  2. OR = odds ratio.
  3. *Adjusted for all variables indicated in the table (all variable except reason in tracing which was not significant).
  4. Low population density = Oxford County, population density = 49/km2; high population density = East York within Toronto, population density = 5,418/km2 [9].
  5. Included in pilot study having web-search completed in June 2001; results only used if tracing was required after the study web-search prior to recruitment (March–May 2003).
  6. §Address status as determined by March–May 2003 web-search [10-12]; completed blind of 2001 web-search results.