AHRQ Vaginal birth after Cesarean

By Christian McDonald, Assistant Professor of Practice
School of Journalism and Media, Moody College of Communication
University of Texas at Austin

This is an analysis of the Vaginal Birth After Cesareans, as defined by IQI 22 Vaginal Birth After Cesarean (VBAC) Delivery Rate, Uncomplicated in 2020.

Definition

Numerator

Number of vaginal deliveries among cases meeting the inclusion and exclusion rules for the denominator. Vaginal deliveries are identified by any-listed ICD-10-PCS procedure code for vaginal delivery (VAGDELP* ).

Denominator

Discharges with an ICD-10-CM diagnosis code for birth delivery outcome (DELOCMD) with any listed ICD-10-CM diagnosis codes for previous Cesarean delivery (PRVBACD).

Exclude cases:

• with any-listed ICD-10-CM diagnosis codes for abnormal presentation, fetal death, or multiple gestation (Appendix A: PRCSECD )
• with an ungroupable DRG (DRG=999)
• with missing gender (SEX=missing), age (AGE=missing), quarter (DQTR=missing), year (YEAR=missing) or principal diagnosis (DX1=missing)

library(tidyverse)
library(janitor)
library(DT)
library(tigris)

# suppresses grouping warning
options(dplyr.summarise.inform = FALSE)

Import of deliveries

I start here with “All deliveries, identified by any-listed ICD-10-CM diagnosis code for outcome of delivery (DELOCMD)”. This was processed in 01-process-loop.

# set test flag to FALSE to run production data
test_flag <- F

### test data
path_test <- "data-test/ahrq_del_all_loop_test.rds"

### production data
path_prod <- "data-processed/ahrq_del_cleaned.rds"

### import based on flag
if (test_flag == T) del <- read_rds(path_test) else del <- read_rds(path_prod)

del %>% nrow()

## [1] 1457413

Set up various processing lists

These are lists from the IQI 22 reference used for filtering and such. See 01-process-lists for definitions.

# diagnostic columns
diag_cols <- read_rds("procedures-lists/cols_diag.rds") %>% .$diag

# surgical procedure columns
surg_cols <- read_rds("procedures-lists/cols_surg.rds") %>% .$surg

# appendix a complications list
prcsecd_list <- read_rds("procedures-lists/ahrq_prcsecd.rds") %>% .$prcsecd

# Previous Cesarean delivery diagnosis codes: (PRVBACD)
prvbacd_list <- read_rds("procedures-lists/ahrq_prvbacd.rds") %>% .$prvbacd

# Vaginal deliveries (VAGDELP)
vagdelp_list <- read_rds("procedures-lists/ahrq_vagdelp.rds") %>% .$vagdelp

Filter deliveries to set denominator

Filter out complicated deliveries

Start by filtering out rows “with any-listed ICD-10-CM diagnosis codes for abnormal presentation, preterm, fetal death, or multiple gestation (Appendix A: PRCSECD)”.

del_ucmp <- del %>% 
  filter_at(
    vars(all_of(diag_cols)),
    all_vars(
      !(. %in% prcsecd_list)
    )
  )

del_ucmp %>% nrow()

## [1] 1315670

Filter out ungroupable DRG (DRG=999)

There are two Diagnosis Related Groups in the data, MS_DRG: Centers for Medicare and Medicaid Services (CMS) Diagnosis Related Group (DRG), as assigned for hospital payment for Medicare beneficiaries; and APR_DRG: All Patient Refined (APR) Diagnosis Related Group (DRG) as assigned by 3M APR-DRG Grouper). I filter out records with “999” in either of them.

del_ucmp <- del_ucmp %>% 
  filter(
    (APR_DRG != "999"),
    (MS_DRG != "999")
  )

del_ucmp %>% nrow()

## [1] 1315631

Filter to keep only previous Cesarean delivery (PRVBACD).

del_denom <- del_ucmp %>% 
  filter_at(
    vars(all_of(diag_cols)),
    any_vars(
      . %in% prvbacd_list
    )
  )

del_denom %>% nrow()

## [1] 263469

This completes our “denominator” set of data. This is saved in del_ucmp.

Numerator updates

Number of vaginal deliveries among cases meeting the inclusion and exclusion rules for the denominator. Vaginal deliveries are identified by any-listed ICD-10-PCS procedure code for vaginal delivery (VAGDELP).

We’ll add a new column VAGDELP based on the codes. If the code is present, the value is TRUE.

ahrq_vbac <- del_denom %>% 
  mutate(
    VBAC = case_when(
      PRINC_SURG_PROC_CODE %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_1 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_2 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_3 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_4 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_5 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_6 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_7 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_8 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_9 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_10 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_11 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_12 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_13 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_14 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_15 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_16 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_17 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_18 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_19 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_20 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_21 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_22 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_23 %in% vagdelp_list ~ TRUE,
      OTH_SURG_PROC_CODE_24 %in% vagdelp_list ~ TRUE,
      TRUE                     ~  FALSE
    )
  )

ahrq_vbac %>% nrow()

## [1] 263469

This completes the processing of the data. ahrq_vbac is our working dataframe.

---

VBAC analysis

• Do hospitals perform them?
• How many performed in 2019?

VBAC by hospital by year

This data is filtered for hospitals that have 30+ deliveries of patients with previous ceseareans in a given year.

ahrq_vbac_rate_hosp_yr <- ahrq_vbac %>% 
  group_by(YR, THCIC_ID, PROVIDER_NAME) %>% 
  count(VBAC) %>% 
  # pivot to get rate for vbac
  pivot_wider(names_from = VBAC, values_from = n) %>% 
  rename(
    NVBAC = `FALSE`,
    YVBAC = `TRUE`
  ) %>% 
  mutate(
    PCTOT = NVBAC + YVBAC,
    VBACRATE = ((YVBAC / PCTOT) * 100) %>% round_half_up(1)
  ) %>% 
  filter(PCTOT >= 30 ) %>% 
  arrange(YR, PROVIDER_NAME)
  
ahrq_vbac_rate_hosp_yr %>% datatable()

Rates in a table by year

vbac_hosp_yr_table <- ahrq_vbac_rate_hosp_yr %>%
  select(YR, THCIC_ID, PROVIDER_NAME, VBACRATE) %>% 
  pivot_wider(names_from = YR, values_from = c(VBACRATE)) %>% 
  arrange(PROVIDER_NAME)

vbac_hosp_yr_table %>% datatable()

VBACs by Laredo hospitals

vbac_hosp_yr_table %>% 
  filter(str_detect(PROVIDER_NAME, "Laredo"))

Summaries

Statewide VBAC percentage across data

ahrq_vbac %>% 
  tabyl(VBAC) %>% 
  rename(count = n) %>% 
  adorn_pct_formatting()

ahrq_vbac_rate_tx_summary <- ahrq_vbac %>% 
  group_by(VBAC) %>% 
  summarize(CNT = n()) %>% 
  pivot_wider(names_from = VBAC, values_from = CNT) %>% 
  rename(
    NVBAC_CNT = "FALSE",
    VBAC_CNT = "TRUE"
  ) %>% 
  mutate(
    TOTAL = NVBAC_CNT + VBAC_CNT
  ) %>% 
  mutate(
   SUMMARY = "TX",
   CATEGORY = "VAGINAL_BIRTH_AFTER_CESAREAN",
   MEASUREMENT = "RATE",
   VALUE = round_half_up((VBAC_CNT / TOTAL) * 100,1) # RATE
  ) %>%
  select(SUMMARY, CATEGORY, MEASUREMENT, VALUE)

ahrq_vbac_rate_tx_summary

VBAC rate statewide by year

Excludes hospitals with fewer than 30 deliveries a year.

ahrq_vbac_rate_tx_yr_summary <- ahrq_vbac %>% 
  group_by(VBAC, YR) %>% 
  summarize(CNT = n()) %>% 
  pivot_wider(names_from = VBAC, values_from = CNT) %>% 
  rename(
    NVBAC_CNT = "FALSE",
    VBAC_CNT = "TRUE"
  ) %>% 
  mutate(
    TOTAL = NVBAC_CNT + VBAC_CNT
  ) %>% 
  filter(TOTAL >= 30) %>% 
  mutate(
   SUMMARY = "TX",
   CATEGORY = "VAGINAL_BIRTH_AFTER_CESAREAN",
   MEASUREMENT = "RATE",
   VALUE = round_half_up((VBAC_CNT / TOTAL) * 100,1) # RATE
  ) %>%
  select(YR, SUMMARY, CATEGORY, MEASUREMENT, VALUE)

ahrq_vbac_rate_tx_yr_summary

VBAC rate by hospital: Averaged, by year

ahrq_vbac_rate_hosp_yr_summary <- ahrq_vbac_rate_hosp_yr %>%
  ungroup() %>% 
  group_by(YR) %>% 
  summarize(
    SUMMARY = "HOSPITAL",
    CATEGORY = "VAGINAL_BIRTH_AFTER_CESAREAN",
    MEASUREMENT = "MEAN_OF_RATE",
    # MEDIAN = median(PCRATE),
    VALUE = round_half_up(mean(VBACRATE),1) #MEAN
  )

ahrq_vbac_rate_hosp_yr_summary

Exports

Writing out some summary data for other notebooks, including:

• ahrq_vbac_rate_hosp_yr
• ahrq_vbac_rate_hosp_yr_summary
• ahrq_vbac_rate_tx_yr_summary

if (test_flag == F) ahrq_vbac_rate_tx_yr_summary %>%
  write_rds("data-processed/ahrq_vbac_rate_tx_yr_summary.rds")

if (test_flag == F) ahrq_vbac_rate_hosp_yr_summary %>%
  write_rds("data-processed/ahrq_vbac_rate_hosp_yr_summary.rds")

if (test_flag == F) ahrq_vbac_rate_hosp_yr %>%
  write_rds("data-processed/ahrq_vbac_rate_hosp_yr.rds")

Closing

beepr::beep(4)