First look at Fitbit Data

Let’s read in data from first 18 completers, only FCP, CRP, and outcome.

data <- read_excel('FitBIt_FCP_CRP_05-10-21.xlsx',
        col_types = c("text","numeric","numeric","numeric","numeric","numeric","text","numeric","numeric","numeric"))  %>% 
  mutate(success = case_when(status ==2 ~ 1, TRUE ~ 0))
## New names:
## * steroid_dose -> steroid_dose...4
## * steroid_dose -> steroid_dose...8

Plotting

Going for a spaghetti plot by patient, with summary lines by group.

First FCP

data %>% 
ggplot(aes(x = visit, y = fcp, 
             col = factor(success), group = factor(sbj_id))) +
  geom_line(size = 0.5, alpha = 0.3) +
  geom_smooth(aes(group = factor(success)), 
      se = FALSE, method = "lm", size = 2) +
  scale_color_manual(values = c("indianred1", "dodgerblue")) +
  labs(y = "Fecal Calprotectin",
       x = "Visit Number",
       col = "Taper Success") +
  theme_minimal(base_size = 14) +
  theme(legend.position = c(0.75, 0.8))
## `geom_smooth()` using formula 'y ~ x'

Tiny difference in intercept for FCP, but slopes look the same for failure vs success of taper.

Now CRP

data %>% 
ggplot(aes(x = visit, y = crp, 
             col = factor(success), group = factor(sbj_id))) +
  geom_line(size = 0.5, alpha = 0.3) +
  geom_smooth(aes(group = factor(success)), 
      se = FALSE, method = "lm", size = 2) +
  scale_color_manual(values = c("indianred1", "dodgerblue")) +
  labs(y = "C-Reactive Protein",
       x = "Visit Number",
       col = "Taper Success") +
  theme_minimal(base_size = 14) +
  theme(legend.position = c(0.75, 0.8))
## `geom_smooth()` using formula 'y ~ x'

Clearly different slopes for CRP in success vs failure of taper