Available datasets in opencodecounts
Source:vignettes/available-datasets.Rmd
available-datasets.RmdClinical codes are unique numeric or alphanumeric identifiers used in healthcare settings for administrative, billing and clinical care purposes. They are grouped into different classification systems to allow standardised recording of diagnoses, procedures or medications, as well as other clinical, demographic and administrative data. In England, three clinical coding classification systems are used: SNOMED Clinical Terms (SNOMED CT) in primary care (GP practices); and International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) and OPCS Classification of Interventions and Procedures, version 4 (OPCS-4) in secondary care.
NHS England makes annual summaries of clinical coding activity in general practice (SNOMED CT UK) and NHS hospitals (ICD-10 and OPCS-4, with the most complete data available for inpatient admissions) openly available. For each of the three classification systems, the summaries provide annual usage of each code used at least once each year. The opencodecounts package aggregates these summaries to allow data exploration. The key consideration is that instances of clinical code recording should not be misinterpreted as estimates of incidence, prevalence or healthcare utilisation.
R setup
To work with the datasets in R, ensure that the opencodecounts package is loaded.
# Load opencodecounts package
library(opencodecounts)SNOMED CT code usage in GP practices in England
SNOMED CT is the world’s most comprehensive clinical terminology system, encoding clinical findings, observations, anatomical terms, disease causes, products and procedures. In England, the SNOMED CT UK edition has been used in general practices since 2019, and contains 357,000 globally common codes, alongside UK-specific extension for local screening procedures and products. In most cases, the information coded prior to 2019 has been translated to SNOMED CT.
The annual code usage is published every October by NHS England, covering the preceding August to July. The number of individuals included has been increasing, reaching over 62 million patient records across 6,600 providers in 2023/24. The annual usage count reflects how many times each listed SNOMED code was added to a GP patient record in England in a given year. The codes are included if they have been used at least ones across the selected years. The counts are rounded to the nearest 10, while counts between 1 and 4 are withheld. The codes with no usage are excluded.
This is only a selection of the full dataset published by NHS
Digital, for data pre-processing see
/data-raw/snomed_code_usage.R.
# Return SNOMED code usage data
snomed_usage
#> # A tibble: 1,523,967 × 7
#> start_date end_date snomed_code description usage active_at_start
#> <date> <date> <chr> <chr> <int> <lgl>
#> 1 2023-08-01 2024-07-31 279991000000102 Short message … 4.41e8 TRUE
#> 2 2023-08-01 2024-07-31 184103008 Patient teleph… 1.91e8 TRUE
#> 3 2023-08-01 2024-07-31 428481002 Patient mobile… 1.16e8 TRUE
#> 4 2023-08-01 2024-07-31 423876004 Clinical docum… 7.81e7 TRUE
#> 5 2023-08-01 2024-07-31 72313002 Systolic arter… 6.87e7 TRUE
#> 6 2023-08-01 2024-07-31 1091811000000102 Diastolic arte… 6.87e7 TRUE
#> 7 2023-08-01 2024-07-31 1000731000000107 Serum creatini… 4.82e7 TRUE
#> 8 2023-08-01 2024-07-31 60621009 Body mass inde… 4.65e7 TRUE
#> 9 2023-08-01 2024-07-31 1000661000000107 Serum sodium l… 4.63e7 TRUE
#> 10 2023-08-01 2024-07-31 1000651000000109 Serum potassiu… 4.62e7 TRUE
#> # ℹ 1,523,957 more rows
#> # ℹ 1 more variable: active_at_end <lgl>ICD-10 code usage in inpatient admissions in England
ICD-10 is a global classification system, containing 18,000 diagnoses coded as four or five-character alphanumeric codes. It is used to determine financial reimbursement for admitted patient care in all NHS-commissioned acute hospitals in England. The coding data submitted by the hospitals is aggregated and published openly by NHS England as the Admitted Patient Care Activity of the Hospital Episode Statistics (HES-APC). It is released in annual intervals, covering April to March. Hospital Episode Statistics also include information on the emergency, outpatient and critical care, which is seperate from the Admitted Patient Care activity presented in opencodecounts.
Activity in HES-APC is captured in episodes of care under one consultant, known as Finished Consultant Episodes (FCEs). Each FCE can be assigned up to 20 clinical diagnoses (ICD-10 codes). The usage count for ICD-10 codes equals the number of FCEs with a specific four-character ICD-10 code recorded in any diagnostic position. This means that each diagnoses can be counted only once per FCE. To be included in the dataset, each diagnosis code must have been used at least once across the selected years. The counts in HES-APC are not rounded and codes with no usage are excluded.
This is only a selection of the full dataset published by NHS
Digital, for data pre-processing see
/data-raw/icd10_usage.R.
# Return ICD-10 code usage data
icd10_usage
#> # A tibble: 136,136 × 5
#> start_date end_date icd10_code description usage
#> <date> <date> <chr> <chr> <int>
#> 1 2023-04-01 2024-03-31 A000 Cholera due to Vibrio cholerae 01, bi… 2
#> 2 2023-04-01 2024-03-31 A009 Cholera, unspecified 40
#> 3 2023-04-01 2024-03-31 A010 Typhoid fever 884
#> 4 2023-04-01 2024-03-31 A011 Paratyphoid fever A 139
#> 5 2023-04-01 2024-03-31 A012 Paratyphoid fever B 13
#> 6 2023-04-01 2024-03-31 A013 Paratyphoid fever C 2
#> 7 2023-04-01 2024-03-31 A014 Paratyphoid fever, unspecified 68
#> 8 2023-04-01 2024-03-31 A020 Salmonella enteritis 2165
#> 9 2023-04-01 2024-03-31 A021 Salmonella sepsis 319
#> 10 2023-04-01 2024-03-31 A022 Localized salmonella infections 82
#> # ℹ 136,126 more rowsOPCS-4 code usage in inpatient admissions in England
OPCS-4 consists of 11,500 four-letter alphanumeric codes and was developed by NHS Digital for recording procedures. Similarly to ICD-10, it determines financial reimbursement for admitted patient care in all NHS-commissioned acute hospitals in England and is published in annual April - March intervals by the NHS England as the Admitted Patient Care Activity of the Hospital Episode Statistics (HES-APC).
Activity in HES-APC is captured in episodes of care under one consultant, known as Finished Consultant Episodes (FCEs). Each FCE can be assigned up to 24 procedures (OPCS-4 codes). The usage count for the OPCS-4 codes equals the number of times each 4-character OPCS-4 code is listed across all FCEs in a given year. This means that unlike diagnoses, a procedure can be counted multiple times within one FCE. To be included in the dataset, each procedure code must have been used at least once across the selected years. The counts in HES-APC are not rounded and codes with no usage are excluded.
This is only a selection of the full dataset published by NHS
Digital, for the data pre-processing see
/data-raw/opcs4_usage.R.
# Return OPCS-4 code usage data
opcs4_usage
#> # A tibble: 107,379 × 5
#> start_date end_date opcs4_code description usage
#> <date> <date> <chr> <chr> <int>
#> 1 2023-04-01 2024-03-31 A011 Hemispherectomy 7
#> 2 2023-04-01 2024-03-31 A012 Total lobectomy of brain 36
#> 3 2023-04-01 2024-03-31 A013 Partial lobectomy of brain 134
#> 4 2023-04-01 2024-03-31 A018 Other specified major excision of tis… 27
#> 5 2023-04-01 2024-03-31 A019 Unspecified major excision of tissue … 3
#> 6 2023-04-01 2024-03-31 A021 Excision of lesion of tissue of front… 1399
#> 7 2023-04-01 2024-03-31 A022 Excision of lesion of tissue of tempo… 971
#> 8 2023-04-01 2024-03-31 A023 Excision of lesion of tissue of parie… 704
#> 9 2023-04-01 2024-03-31 A024 Excision of lesion of tissue of occip… 260
#> 10 2023-04-01 2024-03-31 A025 Excision of lesion of tissue of cereb… 604
#> # ℹ 107,369 more rows