From 54f61fcfe6e52a6fac122aca863a054658115018 Mon Sep 17 00:00:00 2001
From: Yamina Boukari <87201452+YaminaB@users.noreply.github.com>
Date: Thu, 2 Apr 2026 12:55:50 +0000
Subject: [PATCH 1/2] amended denominator criteria for annual counts

---
 analysis/utilities.py | 31 ++++++++++++++++++++++++++-----
 1 file changed, 26 insertions(+), 5 deletions(-)

diff --git a/analysis/utilities.py b/analysis/utilities.py
index 772effa..496f5cd 100644
--- a/analysis/utilities.py
+++ b/analysis/utilities.py
@@ -14,10 +14,31 @@ def build_common_vars(INTERVAL):
     # -------------------
     was_alive_on_1Jan = patients.is_alive_on(INTERVAL.start_date)
 
-    was_registered_on1Jan = (
-        practice_registrations.for_patient_on(INTERVAL.start_date)
-        .exists_for_patient()
-    )
+    was_registered_at_any_point_during_interval = practice_registrations.where(
+        # registered for the entire interval
+        ((practice_registrations.start_date.is_on_or_before(INTERVAL.start_date)) 
+        & (practice_registrations.end_date.is_on_or_after(INTERVAL.end_date))) |
+        
+        # registered during the interval and end date is after the interval end date
+        ((practice_registrations.start_date.is_after(INTERVAL.start_date))
+         & (practice_registrations.end_date.is_on_or_after(INTERVAL.end_date))) |
+         
+         # registered before the interval and registration is ongoing
+         ((practice_registrations.start_date.is_on_or_before(INTERVAL.start_date)) &
+          (practice_registrations.end_date.is_null())) |
+          
+          # registered after interval start date and registration is ongoing
+          ((practice_registrations.start_date.is_after(INTERVAL.start_date)) &
+           (practice_registrations.end_date.is_null())) |
+           
+           # registered before the interval start date and end date is before the end date, but after the start date 
+           ((practice_registrations.start_date.is_before(INTERVAL.start_date)) &
+           (practice_registrations.end_date.is_between_but_not_on(INTERVAL.start_date, INTERVAL.end_date))) |
+           
+           # registered for part of the interval only
+           ((practice_registrations.start_date.is_between_but_not_on(INTERVAL.start_date, INTERVAL.end_date)) &
+            (practice_registrations.end_date.is_between_but_not_on(INTERVAL.start_date, INTERVAL.end_date)))
+    ).exists_for_patient()
 
     has_recorded_sex = patients.sex.is_in(["male", "female"])
 
@@ -28,7 +49,7 @@ def build_common_vars(INTERVAL):
 
     denominator = (
         was_alive_on_1Jan
-        & was_registered_on1Jan
+        & was_registered_at_any_point_during_interval
         & has_recorded_sex
         & has_possible_age
     )

From 8eae6ef5c838e9b0f994f8bc71842702f10757ed Mon Sep 17 00:00:00 2001
From: Yamina Boukari <87201452+YaminaB@users.noreply.github.com>
Date: Thu, 2 Apr 2026 14:03:16 +0000
Subject: [PATCH 2/2] fixed overall row in date of uk entry cohort file

---
 analysis/process_date_of_uk_entry_cohort.R | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/analysis/process_date_of_uk_entry_cohort.R b/analysis/process_date_of_uk_entry_cohort.R
index 33f7726..cfa8c60 100644
--- a/analysis/process_date_of_uk_entry_cohort.R
+++ b/analysis/process_date_of_uk_entry_cohort.R
@@ -25,7 +25,6 @@ output_file <- "output/tables/demographics_date_of_uk_entry_cohort.csv"
 args <- commandArgs(trailingOnly=TRUE)
 print(commandArgs(trailingOnly=TRUE))
 
-
 # Import data ----
 cohort <- read_feather(cohort_file) %>%
   mutate(
@@ -45,7 +44,6 @@ vars_to_summarise <- c(
   "imd_quintile"
 )
 
-
 rounding <- function(vars) {
   case_when(vars == 0 ~ 0,
             vars > 7 ~ round(vars / 5) * 5)
@@ -54,7 +52,7 @@ rounding <- function(vars) {
 table_freq_overall <- cohort %>%
   group_by(any_migrant) %>%
   summarise(
-    n = rounding(nrow(cohort)),
+    n = rounding(n()),
     percentage = 100) %>%
   mutate(
     subgroup = "All",