Skip to main content

Table 2 Agreement between self-report and administrative records of public service utilization among ‘At Home’ study participants (n = 433a)

From: Accuracy of reported service use in a cohort of people who are chronically homeless and seriously mentally ill

Variables Recall period Agreement Disagreement Observed kappa (95 % CI) Prevalence index, Bias index PABAKb (95 % CI) McNemar test, p value
   Overall Both yes Both no Overall Over reporting Under reporting     
Hospitalized over six months for mental illness 5 years 88 % 2 % 86 % 12 % 8 % 4 % 0.21 (0.12, 0.30) 0.84, 0.05 0.77 (0.71, 0.83) 0.007
Hospitalized at least two times for mental illness 5 years 75 % 30 % 45 % 25 % 19 % 6 % 0.50 (0.41, 0.59) 0.14, 0.13 0.50 (0.42, 0.58) <0.001
Any Hospitalization 6 months 73 % 28 % 45 % 27 % 16 % 11 % 0.44 (0.34, 0.53) 0.18, 0.04 0.46 (0.37, 0.54) 0.117
Any emergency room visitc 6 months 73 % 46 % 27 % 27 % 9 % 18 % 0.44 (0.34, 0.53) 0.19, 0.08 0.45 (0.36, 0.54) 0.002
Jaild 6 months 87 % 11 % 76 % 13 % 4 % 9 % 0.55 (0.46, 0.64) 0.66, 0.06 0.74 (0.68, 0.81) 0.001
Income source: disability or income assistance Past month 91 % 86 % 5 % 9 % 8 % 1 % 0.47 (0.38, 0.56) 0.81, 0.07 0.82 (0.76, 0.87) <0.001
  1. aOut of 497 participants, 433 provided consent to access to administrative data and were linkable to health records
  2. bPABAK means prevalence adjusted and bias adjusted kappa
  3. cRestricted to 387 participants
  4. dAnalysis was restricted to 428 participants since consent status varied across public service domains