Fitting a Discrete-time survival analysis

Question

I am trying to fit a discrete-time survival analysis using R. My overall goal is to check if the variable pain impacts smoking. Every time I include the variable as.factor(time_to_smoking) the model does not converge, even if I remove all other covariates. If I don't include it, the model converges.

I need help finding the model that would most accurately represent the plot.

merged_data_surv <- merged_data_any %>%
  mutate(
    time_to_smoking = case_when(
      smoking_any == 1 & smoking_w2 == 1 ~ 1,
      smoking_any == 1 & smoking_w3 == 1 ~ 2,
      smoking_any == 1 & smoking_w4 == 1 ~ 3,
      smoking_any == 1 & smoking_w5 == 1 ~ 4,
      smoking_any == 1 & smoking_w6 == 1 ~ 5,
      TRUE ~ 6
    )
  ) %>%
  filter(
    !is.na(R01R_A_SEX_IMP),
    !is.na(R01R_A_AGECAT7_IMP),
    !is.na(R01R_A_AM0018),
    !is.na(R01R_A_RACECAT3_IMP),
    !is.na(R01R_A_HISP_IMP)
  ) %>%
  left_join(da36498.6111 %>% select(PERSONID, R06_A_A01WGT), by = "PERSONID")


PATH_design_surv <- svrepdesign(
  data = merged_data_surv,
  weights = ~R06_A_A01WGT.x,
  repweights = "R06_A_A01WGT[1-9]+",
  type = "Fay",
  rho = 0.3,
  scale = 1,
  rscales = 1
)

cox_model <- svyglm(smoking_any ~ as.factor(time_to_smoking) + pain + R01R_A_SEX_IMP + 
                      R01R_A_AGECAT7_IMP + R01R_A_AM0018 + R01R_A_RACECAT3_IMP + 
                      R01R_A_HISP_IMP, design = PATH_design_surv, 
                    family = quasibinomial(link = "logit"))

cox_summary <- summary(cox_model)


merged_data_surv$pain_cat <- factor(merged_data_surv$pain, levels = c(0, 1), labels = c("No or Low Pain", "Moderate or Severe Pain"))
# Plot the 1-KM curve using ggsurvplot with custom modifications
ggsurvplot(
    survfit(Surv(time_to_smoking, smoking_any) ~ pain_cat, data = merged_data_surv),
    data = merged_data_surv,
    conf.int = TRUE,
    legend.labs = c("No or Low Pain", "Moderate or Severe Pain"),
    risk.table = "nrisk_cumevents",
    legend = "none",
    xlab = "Waves",
    ylab = "Cumulative Incidence of Daily Smoking (%)",
    xlim = c(1, 6),
    ylim = c(0,8),
    break.time.by = 1,
    font.family = "Aptos",
    palette = c("#705fec", "#f87d5c"),
    ggtheme = theme_bw(base_family = "Aptos") +
      theme(plot.title = element_text(family = "Aptos", size = 14, face = "bold"),
            axis.title = element_text(family = "Aptos", size = 12, face = "bold"),
            axis.title.y = element_text(margin = margin(t = 0, r = 20, b = 0, l = 0)),
            axis.text = element_text(family = "Aptos", size = 10.3, face = "bold"),
            legend.text = element_text(family = "Aptos", size = 12, face = "bold"),
            panel.grid.major = element_blank(),  # Remove all major grid lines
            panel.grid.minor = element_blank(),  # Remove all minor grid lines
            panel.border = element_blank(),  # Remove all borders
            axis.line = element_line(colour = "black"),  # Add axis lines on the left and bottom
            legend.title = element_blank()),
    tables.theme = theme(axis.title.y = element_blank(),
                         plot.title = element_text(family = "Aptos", size = 12, face = "bold"),
                         axis.title.x = element_blank(),
                         axis.text.x = element_blank(),
                         panel.border = element_blank(),
                         axis.line = element_blank(),
                         axis.text.y = element_text(family = "Aptos", face = "bold", size = 12)),
    surv.plot.height = 0.75,  # Adjust the height of the survival plot
    risk.table.height = 0.25,  # Adjust the height of the risk table
    risk.table.title ="No. at risk (no. of events)",
    fun = function(x) (1 - x) * 100  # Transform the survival probability to cumulative incidence
    )