Overview

The function of this page is to identify the frequency and proportion of children screened as having potential developmental disabilities (DD), intellectual disabilities (ID), epilepsy and more. Additionally, we are interested in identifying the proportion of non-cognitive difficulties (seeing, hearing) in these children. This will help us understand what the expected number of recruited children will be by the end of the study. For right now, we will be using data from two screeners (CFM5-17 and Birbeck Epilepsy Screener) to determine dry-EEG eligibility. The CFM responses will inform us of who is eligible for rsEEG (eyes-open and eyes-closed), MMN, and CPT. The Birbeck Epilepsy responses will inform us of who is eligible for rsEEG only, since we will have an epilepsy sample as well.

• The following descriptives were analyzed on 9/13/2024. Since data collection is still on-going, the numbers may slightly change in script re-runs on a later date.

CFM5-17 data cleaning

• We have collected data from 2590 children so far with the CFM5_17 (we are using this screener for children between 5-18 years of age)
• 14 of these screeners were removed for containing inaccurate information (2576 remaining)
• 141 children were excluded (blind, deaf, or immobile for DD; hearing, seeing or moving problems for noDD; 2435 remaining)
• 1584 children were removed for being under the age of 13 years (851 remaining)

# Load in packages
library(tidyverse)
library(readxl)

# Set the working directory
setwd("~/KBB/Data/FINAL_DS/Screener")

# Load in the data
All_Children <- read_excel("All Children.xlsx")

# Keep only the CFM5-17 data
CFM5_17 <- filter(All_Children, Screener.Type == "CFM5_17")

# Number of children screened
nrow(CFM5_17)

# Number of children with incorrect screener data
incorrect.screener <- filter(CFM5_17, Screener.Test != "Correct: CFM5-17")
nrow(incorrect.screener)

# Remove incorrect screeners from the dataset
CFM5_17.correct <- filter(CFM5_17, !(Child_ID %in% incorrect.screener$Child_ID))
nrow(CFM5_17.correct)

# Number of children excluded
CFM5_17.excluded <- filter(CFM5_17.correct, Excluded == "Yes")
nrow(CFM5_17.excluded)

# Remove excluded children
CFM5_17.not.excluded <- filter(CFM5_17.correct, !(Child_ID %in% CFM5_17.excluded$Child_ID))
nrow(CFM5_17.not.excluded)  

# Number of children younger than the age of 13
CFM5_17.under13 <- filter(CFM5_17.not.excluded, Child_age < 13)
nrow(CFM5_17.under13)

# Remove the number of children under the age of 13
CFM5_17.final <-  filter(CFM5_17.not.excluded, !(Child_ID %in% CFM5_17.under13$Child_ID))
nrow(CFM5_17.final)

CFM5-17 Scoring Intellectual Disability (ID)

• From the remaining 851 children, 154 have CFM-DD only, 12 have epilepsy only, 41 have CFM-DD and epilepsy, and 644 have no-DD.
• We will need to create the following five groups: ID, epilepsy, ID&epilepsy,other-DD, and no-DD.
• First we will start by creating an intellectual disability ID status variable. This will include anyone with 'at least some difficulty' for two of the following four questions from the CFM5_17 (CF15_Understood_Inside, CF16_Understood_Outside, CF17_Learning, CF18_Remembering). Children who have problems with communication (CF15 or CF16) must also have difficulties with learning (CF17) or remembering (CF18).

# Number of DD children from 5_17 screener
addmargins(table(CFM5_17.final$CFM_DD, CFM5_17.final$Epilepsy))

# All possible problems
sort(unique(do.call(c,str_split(CFM5_17$KBB_CFM_DD_type[!is.na(CFM5_17$KBB_CFM_DD_type)],pattern = "; "))))

# Creating the ID Group 
CFM5_17.final <- CFM5_17.final %>%
  mutate(ID.Status = case_when(
    grepl("CF15|CF16", KBB_CFM_DD_type) & grepl("CF17", KBB_CFM_DD_type) ~ "ID",
    grepl("CF15|CF16", KBB_CFM_DD_type) & grepl("CF18", KBB_CFM_DD_type) ~ "ID",
    grepl("CF17", KBB_CFM_DD_type) & grepl("CF18", KBB_CFM_DD_type) ~ "ID",
    TRUE~ "No-ID"
  ))

# Count the numbler of ID children
table(CFM5_17.final$ID.Status)

Creating Other Groupings

• Now that we have a variable for ID status, we can use that information to create the five groups of interest.
• The code below does just that and was visually inspected to determine it works as intended.

# Create a variable for the EEG groups
CFM5_17.final <- CFM5_17.final %>%
  mutate(EEG.Group = case_when(
    ID.Status == "ID" & Epilepsy == "No" ~ "ID",
    ID.Status == "ID" & Epilepsy == "Yes" ~ "ID&Epilepsy",
    ID.Status == "No-ID" & Epilepsy == "Yes" ~ "Epilepsy",
    ID.Status == "No-ID" & Epilepsy == "No" & KBB_DD_status == "Yes" ~ "CFM-DD only",
    KBB_DD_status == "No" ~ "No-DD",
    TRUE ~ NA
  ))

# Check for NA's
sum(is.na(CFM5_17.final$EEG.Group))

# Check the sample size of each subgroup
data.frame(Frequency= cbind(sort(table(CFM5_17.final$EEG.Group))))

Group Type	Total
ID & Epilepsy	21
Epilepsy	32
ID	44
Other DD (No Epilepsy)	123
No-DD	631

Prevalence of physical difficulties comorbid with DD group

• Since we are collecting EEG data during an MMN (auditory) and CPT (visual) task, we need to know how prevalent seeing and hearing problems are in these subgroups. If they are not too prevalent, then we can exclude individuals with hearing or seeing problems from undergoing dry-EEG testing. However, if there is a large percentage of difficulty in these children, which indicate these problems tend to come with the cognitive difficulties, then we should test them as is.

• We created a new variable that indicates whether someone has no difficulty, some difficulty, or a lot of difficulty in either seeing or hearing.

# Comorbidity of Seeing or Hearing Problems
CFM5_17.final <- CFM5_17.final %>%
  mutate(Seeing.Hearing = case_when(
    Seeing == "A lot of Difficulty" | Hearing == "A lot of Difficulty" ~ "A lot of Difficulty",
    Seeing == "Some Difficulty" | Hearing == "Some Difficulty" ~ "Some Difficulty",
    Seeing == "No difficulty" & Hearing == "No difficulty" ~ "No Difficulty"
    
  ))

# Create a table for EEG Groups and Seeing/Hearing Difficulties
table(CFM5_17.final$EEG.Group, CFM5_17.final$Seeing.Hearing)

Frequency of seeing/hearing difficulties by severity for each group

Group Type	No Difficulty	Some Difficulty	A lot of Difficulty	Total
ID & Epilepsy	14	4	3	21
Epilepsy	28	2	2	32
ID	31	7	6	44
Other DD (No Epilepsy)	99	17	7	123
No-DD	631	0	0	631

Decision for expected sample size

• It makes sense that those who have epilepsy or ID with epilepsy will do resting-state EEG (rsEEG) only. Since that records brain activity during lack of stimulus presentation, it may not matter much if they have seeing or hearing problems, with the exception of of extremes ('A lot of Difficulty').

• For the groups that do not have epilepsy, which are ID, Other DD, and No-DD, then it would be best if participants did not have any seeing or hearing impairments.

• Thus these tables propose the expected numbers from of different sub-groups eligible for dry-EEG data collection from 2576 screened children.

rsEEG (Only)

Group Type	Sensory Difficulties?	n
ID & Epilepsy	No/Some Difficulty	18
Epilepsy	No/Some Difficulty	30

All EEG Tasks

Group Type	Sensory Difficulties?	n
ID	No	31
Other CFM DD (No Epilepsy)	No	99
No-DD	No	631

• We are aiming for 50 people for each of these groups!

Takeaway from these preliminary results

We will be collecting dry-EEG data from children that are 13 years and older that may have difficulties with seeing or hearing depending on which group they are in. We will be creating 5 mutually exclusive groups that must have at least 50 children in them. The groups are as follows:

1. Epilepsy: The child has epilepsy only or epilepsy and "at least some difficulties" in one or more domains that do not meet criteria for our definition of ID.

2. ID: The child does not have epilepsy and has "at least some difficulties" in two of the following domains: communication(CF15_Understood_Inside; CF16_Understood_Outside), learning (CF17_Learning), and remembering (CF18_Remembering).

3. ID & Epilepsy; The child does have epilepsy and meets criteria for our definition of ID

4. Other-CFM DD: The child does not have epilepsy and has "at least some difficulties" in one or more domains that do not meet criteria for our definition of ID

5. No-DD: The child does not have epilepsy and has no reported difficulties for any of the CFM domains.

Final Comprehensive Table of Group Requirements and Tasks

Group	Group Requirements	Age	Seeing/Hearing Difficulty	Task	Hair Style
Epilepsy	must have epilepsy and may have at least some difficulties in domains that do not meet ID	13+	No/Some Difficulty	rsEEG only	Must provide good EEG recording
ID & Epilepsy	ID criteria and epilepsy	13+	No/Some Difficulty	rsEEG only	Must provide good EEG recording
ID	no epilepsy + at least some difficulties in at least two domains: communication, learning, remembering	13+	No Difficulties	rsEEG, MMN, CPT	Must provide good EEG recording
Other-CFM DD	no epilepsy + at least some difficulties in domains that do not meet ID	13+	No Difficulties	rsEEG, MMN, CPT	Must provide good EEG recording
No-DD	no epilepsy + no difficulties in the CFM domains	13+	No Difficulties	rsEEG, MMN, CPT	Must provide good EEG recording