Table 3 Multivariate model accounting for significant variables found in the single-covariate models showing estimated effects of patient characteristics on visual acuity mean and slope.

From: Clinical outcomes in neovascular age-related macular degeneration: a cohort study of patients with care delay due to the COVID-19 pandemic

(a) Patient characteristic

Estimate of effect on visual acuity slope

Significance

95% confidence interval

Age (years)

 − 0.00146

p = 0.582

 − 0.00657 to 0.00365

Sex

 Male

0.070445

p = 0.152

 − 0.02628 to 0.16717

 Female

0a

pa

–

Delay interval (months)

 − 0.03285

p = 0.044

 − 0.0657 to − 0.002343

Intraretinal fluid absent at visit prior to lockdown

0.048545

p = 0.370

 − 0.058765 to 0.155855

(b) Patient characteristic

Estimate of effect on mean visual acuity (intercept)

Significance

95% confidence interval

Age (years)

0.007720

p = 0.143

 − 0.002666 to 0.018107

Sex

 Male

 − 0.075951

p = 0.476

 − 0.286894 to 0.134991

 Female

0a

pa

–

Delay interval (days)

0.000349

p = 0.746

 − 0.001773 to 0.002471

Intraretinal fluid absent at visit prior to lockdown

 − 0.249428

p = 0.039

 − 0.486404 to − 0.012451

  1. When accounting for intraretinal fluid, age, sex, and delay interval, the delay interval was the single significant factor that affected visual acuity slope with a greater interval leading to a lower rate of vision loss. Patients with intraretinal fluid at the visit prior to lockdown had significantly worse visual acuity mean.
  2. (a) Patient characteristics that are categorical (e.g., smoking status, anti-VEGF agent, etc.) are interpreted as the change in visual acuity slope in comparison to the reference variable. A reference variable defined as the last categorical variable is set for each categorical variable. For instance, male sex has a 0.070445 greater visual acuity slope (logMAR/year) than female sex (not significant p > 0.05). Continuous variables (e.g., age) are interpreted as the change per variable unit to visual acuity slope. For instance, a patient with a 1-month delay interval would have a 1 × (0.03285) lower visual acuity slope (logMAR/year).
  3. (b) Patient characteristics that are categorical (e.g., smoking status, anti-VEGF agent, etc.) are interpreted as the change in visual acuity mean in comparison to the reference category. A reference category is defined as the last category within a given variable. For example, presence of intraretinal fluid is the reference category for the intraretinal fluid variable. Thus, a coefficient of − 0.249428 indicates that patients without intraretinal fluid at the visit prior to lockdown had 0.249428 lower logMAR visual acuity compared to those who had intraretinal fluid. Continuous variables (e.g., age) are interpreted as the change per relevant unit to visual acuity mean. For example, an 80-year-old patient would have an 80 × (0.007720) greater mean logMAR visual acuity (not significant p > 0.05).
  4. ap-values are redundant as these categories are set as the reference category; therefore, estimates are not shown.
Back to article page