Skip to main content

Table 4 Association between any gap in care coordination and preventable adverse events, among participants with intermediate-frailty or frailty, and separately by intermediate-frailty and frailty

From: Frailty, gaps in care coordination, and preventable adverse events

 

No gap in care coordination

Gap in care coordination

Participants with intermediate-frailty or frailty

 N

1535

1091

 Participants with ≥ 1 preventable adverse event, n (%)

154 (10.0)

160 (14.7)

Prevalence ratios (95% confidence intervals) for ≥ 1 preventable adverse event

 Model 1 (n = 2,626)

1 (Ref)

1.46 (1.19 – 1.80)

 Model 2 (n = 2,234)

1 (Ref)

1.34 (1.07 – 1.68)

 Model 3 (n = 1,892)

1 (Ref)

1.38 (1.08 – 1.77)

 Final modela (n = 2,626)

1 (Ref)

1.45 (1.18 – 1.78)

Participants with intermediate-frailty

 N

1442

994

 Participants with ≥ 1 preventable adverse event, n (%)

139 (9.6)

137 (13.8)

 Prevalence ratios (95% confidence intervals) for any preventable adverse event

 Model 1 (n = 2,436)

1 (Ref)

1.43 (1.15 – 1.78)

 Model 2 (n = 2,070)

1 (Ref)

1.34 (1.05 – 1.70)

 Model 3 (n = 1,756)

1 (Ref)

1.33 (1.02 – 1.74)

 Final modela (n = 2,436)

1 (Ref)

1.44 (1.15 – 1.79)

Participants with frailty

 N

93

97

 Participants with ≥ 1 preventable adverse event, n (%)

15 (16.1)

23 (23.7)

 Prevalence ratios (95% confidence intervals) for any preventable adverse event

 Model 1 (n = 190)

1 (Ref)

1.47 (0.82 – 2.64)

 Model 2 (n = 164)

1 (Ref)

1.11 (0.57 – 2.18)

 Model 3 (n = 136)

1 (Ref)

1.51 (0.68 – 3.38)

 Final modela (n = 190)

1 (Ref)

1.29 (0.69 – 2.42)

  1. Model 1 is unadjusted
  2. Model 2 includes adjustment for age, gender, race, education, annual household income, marital status, geographic region of residence and rural area
  3. Model 3 includes adjustment for the variables in model 2 and hypertension, hyperlipidemia, diabetes, history of myocardial infarction, prior coronary revascularization, history of stroke, chronic kidney disease, atrial fibrillation, and self-rated health
  4. a In the final model, multiple imputation was used to retain participants with missing data in covariates