Skip to main content

Table 4 Estimated adjusted odds ratios for attrition in wave 2, stratified by racef, g

From: Modelling attrition and nonparticipation in a longitudinal study of prostate cancer

   Deceased or Frailb, c Lost to Follow-Upb, d Refusale
   AAa EAa AA EA AA EA
Age at Diagnosis 70–79 2.03 (0.89–4.61) 2.30 (0.98–5.44) 0.36 (0.14–0.93) 0.27 (0.06–1.23) 1.03 (0.5–2.11) 1.38 (0.73–2.60)
60–69 0.81 (0.37–1.73) 1.06 (0.45–2.48) 0.37 (0.18–0.73) 0.47 (0.15–1.53) 0.99 (0.57–1.73) 0.84 (0.46–1.53)
Study Phase Post-Ka 0.41 (0.20–0.88) 0.46 (0.22–0.94) 1.12 (0.48–2.62) 0.96 (0.25–3.72) 0.76 (0.40–1.45) 0.73 (0.38–1.40)
Income $0–30,000 13.42 (0.84–214.98) 1.49 (0.58–3.78) 2.48 (0.86–7.13) 8.65 (1.88–39.77) 1.01 (0.44–2.30) 2.79 (1.34–5.81)
$30,001–70,000 14.13 (0.88–225.87) 1.21 (0.56–2.63) 1.48 (0.49–4.45) 1.99 (0.44–9.05) 1.34 (0.61–2.95) 1.96 (1.07–3.58)
Education ≤ High School 1.37 (0.63–2.99) 1.15 (0.55–2.42) 1.92 (0.90–4.09) 2.18 (0.68–7.05) 1.34 (0.75–2.41) 1.29 (0.73–2.30)
REALMa Score ≤ 6th 1.36 (0.58–3.21) 0.95 (0.26–3.44) 1.18 (0.51–2.74) 0.46 (0.07–2.91) 0.94 (0.47–1.86) 0.81 (0.33–2.00)
7th - 8th 0.47 (0.18–1.20) 2.20 (0.98–4.92) 0.74 (0.30–1.80) 1.25 (0.35–4.52) 0.62 (0.31–1.21) 0.64 (0.30–1.39)
Cancer Aggressiveness Intermediate 1.12 (0.55–2.30) 1.16 (0.55–2.45) 1.01 (0.49–2.06) 0.66 (0.20–2.23) 1.03 (0.59–1.80) 0.57 (0.32–1.01)
High 2.98 (1.34–6.62) 4.09 (1.94–8.63) 2.17 (0.99–4.76) 1.17 (0.28–4.90) 2.51 (1.34–4.72) 1.07 (0.54–2.10)
PPCa Score   0.77 (0.54–1.09) 0.91 (0.60–1.38) 1.23 (0.83–1.81) 1.05 (0.53–2.10) 0.76 (0.58–1.01) 0.70 (0.51–0.96)
CCIa   1.26 (1.03–1.54) 1.17 (0.98–1.39) 0.86 (0.65–1.12) 0.91 (0.61–1.37) 1.00 (0.83–1.21) 0.96 (0.80–1.15)
  1. aAA african american, EA european american, Post-K post-katrina, REALM rapid estimate of adult literacy in medicine, PPC patient provider communication, CCI charlson comorbidity index
  2. bFirth’s logistic regression models were used to reduce potential effects of low event rates
  3. cThe outcome variable was dichotomized as deceased or frail vs participant
  4. dThe outcome variable was dichotomized as lost to follow-up vs participant
  5. eThe outcome variable was dichotomized as refusal vs participant
  6. fSignificant p-values at Type I error 0.05 are bolded
  7. gReferent categories: Age: 40–59 age group, Race: EA, Study phase: Pre-K, Income: ≥$70,001, Education: high school, REALM score: high school, and Cancer aggressiveness: low