Codelist development for and anxiety disorders and depression diagnoses

Overview

This report assesses the suitability of NHS England Primary Care Domain Reference Sets for identifying diagnoses of common mental health conditions, specifically anxiety diagnoses and depression, and compares them with codelists constructed to match the definitions used in the Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4.

The analysis proceeds in two stages. First, the coverage and usage of the NHS Refsets are assessed using data from the NHS England SNOMED Code Usage in Primary Care dataset. Second, condition-specific codelists are constructed to mirror APMS diagnostic definitions, and usage patterns of these lists are evaluated over time.

NHS Primary Care Refsets

The Primary Care Domain Refset contain codelist for anxiety and depression diagnostic (anx_cod,depr_cod) and resolved (anx_res_cod,dep_res_cod) codes.

Diagnosis

Analysis of codelists

The below figure shows the coding of anxiety disorders and depression diagnoses, as specified in the NHS Primary Care Domain Refsets (anx_cod, depr_cod).

Show the code

anx_cod_depr_cod_usage <- snomed_usage |> 
  mutate(
    depr_cod = snomed_code %in% depr_cod$code,
    anx_cod = snomed_code %in% anx_cod$code
    ) |> 
  pivot_longer(
    cols = c(depr_cod, anx_cod),
    names_to = "codelist") |> 
  filter(value == TRUE) |> 
  group_by(end_date, codelist) |> 
  summarise(count = sum(usage, na.rm = TRUE))

fig_depr_anx_cod_trends <- anx_cod_depr_cod_usage |> 
  mutate(codelist = factor(
    codelist,
    levels = c("depr_cod", "anx_cod"), 
    labels = c("Depression diagnostic codes", "Anxiety diagnostic codes"))) |> 
  ggplot(
      aes(
        y = count,
        x = end_date,
        colour = codelist,
        shape = codelist
      )
    ) +
    geom_line(alpha = 0.2) +
    geom_point() +
    labs(
      title = "Yearly counts of anxiety disorder and depression diagnoistic codes",
      x = NULL,
      y = NULL,
      colour = NULL,
      shape = NULL,
      caption = "Data from NHS England SNOMED Code Usage in Primary Care."
    ) +
    scale_x_date(
      date_breaks = "1 year",
      labels = label_date_short()
    ) +
    scale_y_continuous(
      labels = label_number(accuracy = 1),
      limits = c(0, NA)
    ) +
    theme_bw() +
    theme(legend.position = "bottom") +
    scale_colour_viridis_d(end = .8)

ggsave(
  filename = here("output", "figures", "cod_devel", "dx", "fig_depr_anx_cod_trends.png"),
  fig_depr_anx_cod_trends,
  height = 4,
  width = 8
)

fig_depr_anx_cod_trends

Figure 1: Yearly counts of anxiety disorder and depression diagnoistic codes as specified in the NHS Primary Care Domain Refsets.

Analysis of individual codes

Show the code

overlap <- intersect(depr_cod$code, anx_cod$code)

snomed_usage |> 
  filter(snomed_code %in% overlap) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  create_table_sem_tag(
    title = md("Codes appearing in both the NHS Refsets for depression and anxiety diagnoses"),
    subtitle = "Timeframe: August 2011 to July 2024"
  ) |> 
  cols_hide(ratio_usage)

SNOMED code	Description	Usage
Codes appearing in both the NHS Refsets for depression and anxiety diagnoses
Timeframe: August 2011 to July 2024
Disorder
231504006	Mixed anxiety and depressive disorder	20,659,770
16264901000119109	Recurrent moderate major depressive disorder co-occurrent with anxiety	810
16264621000119109	Recurrent mild major depressive disorder co-occurrent with anxiety	695
16264821000119108	Recurrent severe major depressive disorder co-occurrent with anxiety	475
16265301000119106	Recurrent major depressive disorder in partial remission co-occurrent with anxiety	40
Data scource: NHS England SNOMED Usage in Primary Care.

This table shows the 20 most used codes covering 99.35% of the total code usage of the codelist. The full list of included codes is avilable on OpenCodelists, see depr_cod.

Show the code

# Display formatted table
tab_depr_cod_counts <- depr_cod_counts |>
  create_table_sem_tag(
    title = md("*Depression* diagnosis codes (`depr_cod`)"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_depr_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_depr_cod_usage.png")
  )

tab_depr_cod_counts

SNOMED code	Description	Usage	%
Depression diagnosis codes (`depr_cod`)
Timeframe: August 2011 to July 2024
Disorder
231504006	Mixed anxiety and depressive disorder	20,659,770	43.751%
35489007	Depressive disorder	20,623,000	43.673%
310496002	Moderate depression	1,098,060	2.325%
87414006	Reactive depression (situational)	1,042,110	2.207%
191616006	Recurrent depression	990,730	2.098%
310495003	Mild depression	434,190	0.919%
192080009	Chronic depression	357,170	0.756%
310497006	Severe depression	287,310	0.608%
300706003	Endogenous depression	251,490	0.533%
36923009	Major depression, single episode	217,360	0.460%
78667006	Dysthymia	171,700	0.364%
83458005	Agitated depression	158,640	0.336%
274948002	Endogenous depression - recurrent	134,630	0.285%
191604000	Single major depressive episode, severe, with psychosis	122,470	0.259%
231499006	Endogenous depression first episode	96,870	0.205%
268621008	Recurrent major depressive episodes	68,420	0.145%
73867007	Severe major depression with psychotic features	54,090	0.115%
247803002	Seasonal affective disorder	51,760	0.110%
15639000	Moderate major depression, single episode	47,470	0.101%
191676002	Reactive depressive psychosis	46,570	0.099%
Data scource: NHS England SNOMED Usage in Primary Care.

This table shows the 20 most used codes covering 99.58% of the total code usage of the codelist. The full list of included codes is avilable on OpenCodelists, see anx_cod.

Show the code

# Display formatted table
tab_anx_cod_counts <- anx_cod_counts |>
  create_table_sem_tag(
    title = md("*Anxiety* diagnosis codes (`anx_cod`)"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_anx_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_anx_cod_usage.png")
  )

tab_anx_cod_counts

SNOMED code	Description	Usage	%
Anxiety diagnosis codes (`anx_cod`)
Timeframe: August 2011 to July 2024
Disorder
231504006	Mixed anxiety and depressive disorder	20,659,770	44.792%
197480006	Anxiety disorder	13,705,140	29.714%
21897009	Generalized anxiety disorder	1,493,400	3.238%
371631005	Panic disorder	580,480	1.259%
191722009	Agoraphobia with panic attacks	121,340	0.263%
386810004	Phobic disorder	114,790	0.249%
70691001	Agoraphobia	104,740	0.227%
25501002	Social phobia	82,920	0.180%
Finding
198288003	Anxiety state	4,085,680	8.858%
48694002	Anxiety	3,572,040	7.745%
191708009	Chronic anxiety	544,060	1.180%
191709001	Recurrent anxiety	234,810	0.509%
231501003	Needle phobia	100,780	0.219%
247854002	Flying phobia	53,610	0.116%
19887002	Claustrophobia	53,010	0.115%
247805009	Anxiety and fear	41,720	0.090%
386808001	Phobia	39,290	0.085%
247808006	Anxiety about body function or health	36,480	0.079%
38617005	Dental phobia	28,550	0.062%
Situation
161470009	History of anxiety state	275,320	0.597%
Data scource: NHS England SNOMED Usage in Primary Care.

Resolved

Analysis of codelists

Show the code

nhsrefset_res_cod_usage <- snomed_usage |> 
  mutate(
    anx_res = snomed_code %in% anx_res_cod$code,
    dep_res = snomed_code %in% dep_res_cod$code
    ) |> 
  pivot_longer(
    cols = c(
      anx_res,
      dep_res
    ),
    names_to = "codelist") |> 
  filter(value == TRUE) |> 
  group_by(end_date, codelist) |> 
  summarise(count = sum(usage, na.rm = TRUE))

fig_nhsrefset_res_cod_trends <- nhsrefset_res_cod_usage |> 
  mutate(codelist = factor(
    codelist,
    levels = c(
      "anx_res",
      "dep_res"
    ),
    labels = c(
      "Anxiety disorder resolved",
      "Depression resolved"
    ))) |> 
  ggplot(
      aes(
        y = count,
        x = end_date,
        colour = codelist,
        shape = codelist
      )
    ) +
    geom_line(alpha = 0.4) +
    geom_point(size = 2) +
    labs(
      title = "Yearly counts of resolved codes as specified by NHS Refsets",
      x = NULL,
      y = NULL,
      colour = NULL,
      shape = NULL,
      caption = "Data from NHS England SNOMED Code Usage in Primary Care."
    ) +
    scale_x_date(
      date_breaks = "1 year",
      labels = label_date_short()
    ) +
    scale_y_continuous(
      labels = label_number(accuracy = 1),
      limits = c(0, NA)
    ) +
    theme_bw() +
    theme(legend.position = "bottom") +
    scale_color_okabe_ito()

ggsave(
  filename = here("output", "figures", "cod_devel", "dx", "fig_nhsrefset_res_cod_trends.png"),
  fig_nhsrefset_res_cod_trends,
  height = 4,
  width = 8
)

fig_nhsrefset_res_cod_trends

Figure 2: Yearly counts of resolved codes as specified by NHS Refsets.

Analysis of individual codes

OpenCodelists: Anxiety disorder resolved codes (anx_res_cod)

Show the code

anx_res_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% anx_res_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_anx_res_cod_counts <- anx_res_cod_counts |>
  create_table_sem_tag(
    title = md("*Anxiety* disorder resolved codes (`anx_res_cod`)"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_anx_res_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_anx_res_nhsrefset_cod_usage.png")
  )

tab_anx_res_cod_counts

SNOMED code	Description	Usage	%
Anxiety disorder resolved codes (`anx_res_cod`)
Timeframe: August 2011 to July 2024
Finding
1037391000000102	Anxiety resolved	3,755	98.556%
Disorder
16265061000119105	Recurrent major depressive disorder co-occurrent with anxiety in full remission	45	1.181%
11941006	Panic disorder with agoraphobia, agoraphobic avoidance in full remission AND panic attacks in full remission	5	0.131%
82494000	Panic disorder without agoraphobia with panic attacks in full remission	5	0.131%
Data scource: NHS England SNOMED Usage in Primary Care.

OpenCodelists: Depression dignosis resolved codes (depres_cod)

Show the code

dep_res_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% dep_res_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_dep_res_cod_counts <- dep_res_cod_counts |>
  create_table_sem_tag(
    title = md("*Depression* dignosis resolved (`dep_res_cod`)"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_dep_res_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_dep_res_nhsrefset_cod_usage.png")
  )

tab_dep_res_cod_counts

SNOMED code	Description	Usage	%
Depression dignosis resolved (`dep_res_cod`)
Timeframe: August 2011 to July 2024
Finding
196381000000100	Depression resolved	440,010	100.000%
Data scource: NHS England SNOMED Usage in Primary Care.

APM Survey

In the Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4 the following mental health conditions are assessed in Chapter 1: Common mental health conditions and Chapter 3: Posttraumatic stress disorder of the survey, see Appendix A: Assessment of disorders for more details.

Show the code

condition_names <- list(
  gad = "Generalised anxiety disorder (GAD)",
  cmhc_nos = "Common mental health conditions not otherwise specified (CMHC-NOS)",
  ocd = "Obsessive and compulsive disorder (OCD)",
  dep = "Depressive episode",
  pd = "Panic disorder (PD)",
  phobia = "Phobia",
  ptsd = "Posttraumatic stress disorder (PTSD)"
)

measures_mental_health_conditions <- list(
  gad = list(
    condition = "Generalised anxiety disorder (GAD)",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  cmhc_nos = list(
    condition = "Common mental health conditions not otherwise specified (CMHC-NOS)",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  ocd = list(
    condition = "Obsessive and compulsive disorder (OCD)",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  dep = list(
    condition = "Depressive episode",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  pd = list(
    condition = "Panic disorder",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  phobia = list(
    condition = "Phobia",
    diagnostic_status = "Present to diagnostic criteria",
    classification_system = "ICD-10",
    assessment_tool = "CIS-R",
    survey_phase = "1",
    reference_period = "Past week"
  ),
  ptsd = list(
    condition = "Posttraumatic stress disorder (PTSD)",
    diagnostic_status = "Screen positive",
    classification_system = "DSM-IV",
    assessment_tool = "PTSD Check List",
    survey_phase = "1",
    reference_period = "Past week"
  )
)

df_measures_mental_health_conditions <- map_dfr(measures_mental_health_conditions, ~.x, .id = "condition_tag")

gt(df_measures_mental_health_conditions) |> 
  cols_label(
    condition = "Condition",
    diagnostic_status = "Diagnostic status",
    classification_system = "Classification system",
    assessment_tool = "Assessment tool",
    survey_phase = "Survey phase",
    reference_period = "Reference period"
  ) |> 
  cols_hide(
    c(
      condition_tag,
      diagnostic_status,
      survey_phase,
      reference_period
      )
    )

Condition	Classification system	Assessment tool
Generalised anxiety disorder (GAD)	ICD-10	CIS-R
Common mental health conditions not otherwise specified (CMHC-NOS)	ICD-10	CIS-R
Obsessive and compulsive disorder (OCD)	ICD-10	CIS-R
Depressive episode	ICD-10	CIS-R
Panic disorder	ICD-10	CIS-R
Phobia	ICD-10	CIS-R
Posttraumatic stress disorder (PTSD)	DSM-IV	PTSD Check List

Diagnosis

Analysis of codelists

Show the code

amps_cod_usage <- snomed_usage |> 
  mutate(
    gad_cod = snomed_code %in% gad_cod$code,
    cmhcnos_cod = snomed_code %in% cmhcnos_cod$code,
    ocd_cod = snomed_code %in% ocd_cod$code,
    depression_cod = snomed_code %in% depression_cod$code,
    pd_cod = snomed_code %in% pd_cod$code,
    phobia_cod = snomed_code %in% phobia_cod$code,
    ptsd_cod = snomed_code %in% ptsd_cod$code
    ) |> 
  pivot_longer(
    cols = c(
      gad_cod,
      cmhcnos_cod,
      ocd_cod,
      depression_cod,
      pd_cod,
      phobia_cod,
      ptsd_cod
    ),
    names_to = "codelist") |> 
  filter(value == TRUE) |> 
  group_by(end_date, codelist) |> 
  summarise(count = sum(usage, na.rm = TRUE))

fig_amps_cod_trends <- amps_cod_usage |> 
  mutate(codelist = factor(
    codelist,
    levels = c(
      "gad_cod",
      "cmhcnos_cod",
      "ocd_cod",
      "depression_cod",
      "pd_cod",
      "phobia_cod",
      "ptsd_cod"
    ),
    labels = c(
      "GAD",
      "CMHC-NOS",
      "OCD",
      "Depression",
      "PD",
      "Phobia",
      "PTSD"
    ))) |> 
  ggplot(
      aes(
        y = count,
        x = end_date,
        colour = codelist,
        shape = codelist
      )
    ) +
    geom_line(alpha = 0.4) +
    geom_point(size = 2) +
    labs(
      title = "Yearly counts of diagnoistic codes following AMPS methods",
      x = NULL,
      y = NULL,
      colour = NULL,
      shape = NULL,
      caption = "Data from NHS England SNOMED Code Usage in Primary Care."
    ) +
    scale_x_date(
      date_breaks = "1 year",
      labels = label_date_short()
    ) +
    scale_y_continuous(
      labels = label_number(accuracy = 1),
      limits = c(0, NA)
    ) +
    theme_bw() +
    theme(legend.position = "bottom") +
    scale_color_okabe_ito()

ggsave(
  filename = here("output", "figures", "cod_devel", "dx", "fig_amps_cod_trends.png"),
  fig_amps_cod_trends,
  height = 4,
  width = 8
)

fig_amps_cod_trends

Figure 3: Yearly counts of diagnoistic codes following AMPS methods.

Show the code

fig_amps_cod_trends_zoom <- fig_amps_cod_trends +
    facet_zoom(ylim = c(30000, 200000), zoom.size = 3) +
    scale_x_date(
      breaks = c(
        as.Date("2012-07-31"),
        as.Date("2016-01-01"),
        as.Date("2020-01-01"),
        as.Date("2024-07-31")
      ),
      labels = label_date("%Y")
    )

ggsave(
  filename = here("output", "figures", "cod_devel", "dx", "fig_amps_cod_trends_zoom.png"),
  fig_amps_cod_trends_zoom,
  height = 4,
  width = 8
)

fig_amps_cod_trends_zoom

Figure 4: Yearly counts of diagnoistic codes following AMPS methods. The y axis of the left panel is zoomed into codelists with low counts.

Analysis of individual codes

ICD-10 codes: F41.1 (Generalised anxiety disorder).
OpenCodelists: Generalised anxiety disorder diagnostic codes

Show the code

gad_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% gad_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_gad_cod_counts <- gad_cod_counts |>
  create_table_sem_tag(
    title = md("Generalised Anxiety Disorder"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_gad_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_gad_usage.png")
  )

tab_gad_cod_counts

SNOMED code	Description	Usage	%
Generalised Anxiety Disorder
Timeframe: August 2011 to July 2024
Disorder
21897009	Generalized anxiety disorder	1,493,400	100.000%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F41.2 (Mixed anxiety and depressive disorder).
OpenCodelists: Common mental health conditions not otherwise specified

Show the code

cmhcnos_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% cmhcnos_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_cmhcnos_cod_counts <- cmhcnos_cod_counts |>
  create_table_sem_tag(
    title = md("CMHC-NOS diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_cmhcnos_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_cmhcnos_cod_usage.png")
  )

tab_cmhcnos_cod_counts

SNOMED code	Description	Usage	%
CMHC-NOS diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
231504006	Mixed anxiety and depressive disorder	20,659,770	100.000%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F42 (Obsessive-compulsive disorder).
OpenCodelists: Obsessive-compulsive disorder diagnostic codes

Show the code

ocd_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% ocd_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_ocd_cod_counts <- ocd_cod_counts |>
  create_table_sem_tag(
    title = md("Obsessive-compulsive disorder diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_ocd_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_ocd_cod_usage.png")
  )

tab_ocd_cod_counts

SNOMED code	Description	Usage	%
Obsessive-compulsive disorder diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
191736004	Obsessive-compulsive disorder	889,580	96.476%
83482000	Body dysmorphic disorder	32,490	3.524%
723913009	Olfactory reference disorder	5	0.001%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F32.00 (Mild depressive episode without somatic symptoms) and F32.01 (Mild depressive episode with somatic symptoms) were grouped to produce mild depressive episode (i.e. with or without somatic symptoms). F32.10 (Moderate depressive episode without somatic symptoms) and F32.11 (Moderate depressive episode with somatic symptoms) were similarly grouped to produce moderate depressive episode. Mild depressive episode, moderate depressive episode and severe depressive episode were then combined to produce the final category of depressive episode.
OpenCodelists: Depression diagnostic codes

Show the code

depression_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% depression_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_depression_cod_counts <- depression_cod_counts |>
  create_table_sem_tag(
    title = md("Depression diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_depression_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_depression_cod_usage.png")
  )

tab_depression_cod_counts

SNOMED code	Description	Usage	%
Depression diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
35489007	Depressive disorder	20,623,000	74.271%
310496002	Moderate depression	1,098,060	3.955%
87414006	Reactive depression (situational)	1,042,110	3.753%
191616006	Recurrent depression	990,730	3.568%
58703003	Postpartum depression	764,080	2.752%
310495003	Mild depression	434,190	1.564%
192080009	Chronic depression	357,170	1.286%
310497006	Severe depression	287,310	1.035%
300706003	Endogenous depression	251,490	0.906%
36923009	Major depression, single episode	217,360	0.783%
78667006	Dysthymia	171,700	0.618%
83458005	Agitated depression	158,640	0.571%
274948002	Endogenous depression - recurrent	134,630	0.485%
191604000	Single major depressive episode, severe, with psychosis	122,470	0.441%
231499006	Endogenous depression first episode	96,870	0.349%
268621008	Recurrent major depressive episodes	68,420	0.246%
73867007	Severe major depression with psychotic features	54,090	0.195%
247803002	Seasonal affective disorder	51,760	0.186%
Finding
871840004	Episode of depression	269,510	0.971%
Situation
473126001	Suspected depressive disorder	162,930	0.587%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F41.0 Panic disorder
OpenCodelists: Panic disorder diagnostic codes

Show the code

pd_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% pd_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_pd_cod_counts <- pd_cod_counts |>
  create_table_sem_tag(
    title = md("Panic Disorder (PD) diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_pd_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_pd_cod_usage.png")
  )

tab_pd_cod_counts

SNOMED code	Description	Usage	%
Panic Disorder (PD) diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
371631005	Panic disorder	580,480	99.212%
192037000	Acute panic state due to acute stress reaction	4,190	0.716%
56576003	Panic disorder without agoraphobia	230	0.039%
89948007	Panic disorder with agoraphobia AND mild panic attacks	35	0.006%
35607004	Panic disorder with agoraphobia	25	0.004%
43150009	Panic disorder without agoraphobia with severe panic attacks	25	0.004%
65064003	Panic disorder without agoraphobia with moderate panic attacks	20	0.003%
72861004	Panic disorder without agoraphobia with mild panic attacks	20	0.003%
5509004	Panic disorder with agoraphobia AND severe panic attacks	15	0.003%
53956006	Panic disorder without agoraphobia with panic attacks in partial remission	10	0.002%
61212007	Panic disorder with agoraphobia, severe agoraphobic avoidance AND severe panic attacks	10	0.002%
8185002	Panic disorder with agoraphobia AND moderate panic attacks	10	0.002%
11941006	Panic disorder with agoraphobia, agoraphobic avoidance in full remission AND panic attacks in full remission	5	0.001%
82494000	Panic disorder without agoraphobia with panic attacks in full remission	5	0.001%
82738004	Panic disorder with agoraphobia, moderate agoraphobic avoidance AND moderate panic attacks	5	0.001%
83631006	Panic disorder with agoraphobia, moderate agoraphobic avoidance AND severe panic attacks	5	0.001%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F40.00 (Agoraphobia without panic disorder), F40.01 (Agoraphobia with panic disorder), F40.1 (Social phobias), and F40.2 (Specific (isolated) phobias), were combined into one category of phobia.
OpenCodelists: Phobia diagnostic codes

Show the code

phobia_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% phobia_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_phobia_cod_counts <- phobia_cod_counts |>
  create_table_sem_tag(
    title = md("Phobia diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_phobia_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_phobia_cod_usage.png")
  )

tab_phobia_cod_counts

SNOMED code	Description	Usage	%
Phobia diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
191722009	Agoraphobia with panic attacks	121,340	23.533%
386810004	Phobic disorder	114,790	22.262%
70691001	Agoraphobia	104,740	20.313%
25501002	Social phobia	82,920	16.081%
54587008	Simple phobia	20,330	3.943%
61569007	Agoraphobia without history of panic disorder	5,110	0.991%
191725006	Social phobia, fear of public speaking	3,250	0.630%
191724005	Social phobia, fear of eating in public	1,230	0.239%
191726007	Social phobia, fear of public washing	130	0.025%
89948007	Panic disorder with agoraphobia AND mild panic attacks	35	0.007%
35607004	Panic disorder with agoraphobia	25	0.005%
5509004	Panic disorder with agoraphobia AND severe panic attacks	15	0.003%
61212007	Panic disorder with agoraphobia, severe agoraphobic avoidance AND severe panic attacks	10	0.002%
8185002	Panic disorder with agoraphobia AND moderate panic attacks	10	0.002%
11941006	Panic disorder with agoraphobia, agoraphobic avoidance in full remission AND panic attacks in full remission	5	0.001%
238965007	Venereophobia	5	0.001%
82738004	Panic disorder with agoraphobia, moderate agoraphobic avoidance AND moderate panic attacks	5	0.001%
83631006	Panic disorder with agoraphobia, moderate agoraphobic avoidance AND severe panic attacks	5	0.001%
Finding
19887002	Claustrophobia	53,010	10.281%
54307006	Zoophobia	7,460	1.447%
58963008	Acrophobia	850	0.165%
280947008	Examination phobia	340	0.066%
102916005	Arachnophobia	10	0.002%
Data scource: NHS England SNOMED Usage in Primary Care.

ICD-10 codes: F43.1 (Post-traumatic stress disorder), assessed using DSM-IV symptoms of PTSD.
OpenCodelists: PTSD diagnostic codes

Show the code

ptsd_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% ptsd_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_ptsd_cod_counts <- ptsd_cod_counts |>
  create_table_sem_tag(
    title = md("Posttraumatic Stress Disorder (PTSD) diagnosis codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_ptsd_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_ptsd_cod_usage.png")
  )

tab_ptsd_cod_counts

SNOMED code	Description	Usage	%
Posttraumatic Stress Disorder (PTSD) diagnosis codes
Timeframe: August 2011 to July 2024
Disorder
47505003	Posttraumatic stress disorder	1,197,030	97.856%
192042008	Acute post-trauma stress state	9,290	0.759%
443919007	Complex posttraumatic stress disorder	8,810	0.720%
313182004	Chronic post-traumatic stress disorder	4,080	0.334%
699241002	Chronic post-traumatic stress disorder following military combat	2,500	0.204%
446180007	Delayed posttraumatic stress disorder following military combat	760	0.062%
446175003	Acute posttraumatic stress disorder following military combat	710	0.058%
620181000000106	Other post-traumatic stress disorder	50	0.004%
318784009	Posttraumatic stress disorder, delayed onset	25	0.002%
Data scource: NHS England SNOMED Usage in Primary Care.

Resolved

Analysis of codelists

Show the code

amps_res_cod_usage <- snomed_usage |> 
  mutate(
    anxiety_resolved = snomed_code %in% anxiety_resolved_cod$code,
    pd_resolved = snomed_code %in% pd_resolved_cod$code,
    depression_resolved = snomed_code %in% depression_resolved_cod$code,
    ) |> 
  pivot_longer(
    cols = c(
      anxiety_resolved,
      pd_resolved,
      depression_resolved
    ),
    names_to = "codelist") |> 
  filter(value == TRUE) |> 
  group_by(end_date, codelist) |> 
  summarise(count = sum(usage, na.rm = TRUE))

fig_amps_res_cod_trends <- amps_res_cod_usage |> 
  mutate(codelist = factor(
    codelist,
    levels = c(
      "anxiety_resolved",
      "pd_resolved",
      "depression_resolved"
    ),
    labels = c(
      "Anxiety disorder resolved",
      "Panic disorder resolved",
      "Depression resolved"
    ))) |> 
  ggplot(
      aes(
        y = count,
        x = end_date,
        colour = codelist,
        shape = codelist
      )
    ) +
    geom_line(alpha = 0.4) +
    geom_point(size = 2) +
    labs(
      title = "Yearly counts of resolved codes following AMPS methods",
      x = NULL,
      y = NULL,
      colour = NULL,
      shape = NULL,
      caption = "Data from NHS England SNOMED Code Usage in Primary Care."
    ) +
    scale_x_date(
      date_breaks = "1 year",
      labels = label_date_short()
    ) +
    scale_y_continuous(
      labels = label_number(accuracy = 1),
      limits = c(0, NA)
    ) +
    theme_bw() +
    theme(legend.position = "bottom") +
    scale_color_okabe_ito()

ggsave(
  filename = here("output", "figures", "cod_devel", "dx", "fig_amps_res_cod_trends.png"),
  fig_amps_res_cod_trends,
  height = 4,
  width = 8
)

fig_amps_res_cod_trends

Figure 5: Yearly counts of resolved codes following AMPS methods.

Analysis of individual codes

OpenCodelists: Anxiety disorder resolved codes

Show the code

anxiety_res_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% anxiety_resolved_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_anxiety_res_cod_counts <- anxiety_res_cod_counts |>
  create_table_sem_tag(
    title = md("Anxiety disorder resolved codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_anxiety_res_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_anxiety_res_usage.png")
  )

tab_anxiety_res_cod_counts

SNOMED code	Description	Usage	%
Anxiety disorder resolved codes
Timeframe: August 2011 to July 2024
Finding
1037391000000102	Anxiety resolved	3,755	100.000%
Data scource: NHS England SNOMED Usage in Primary Care.

OpenCodelists: Depression dignosis resolved or full remission codes

Show the code

depression_res_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% depression_resolved_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_depression_res_cod_counts <- depression_res_cod_counts |>
  create_table_sem_tag(
    title = md("Depression dignosis resolved or full remission codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_depression_res_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_depression_res_cod_usage.png")
  )

tab_depression_res_cod_counts

SNOMED code	Description	Usage	%
Depression dignosis resolved or full remission codes
Timeframe: August 2011 to July 2024
Finding
196381000000100	Depression resolved	440,010	99.568%
Disorder
46244001	Recurrent major depression in full remission	800	0.181%
19527009	Single episode of major depression in full remission	760	0.172%
68019004	Recurrent major depression in remission	260	0.059%
16265061000119105	Recurrent major depressive disorder co-occurrent with anxiety in full remission	45	0.010%
63412003	Major depression in full remission	45	0.010%
Data scource: NHS England SNOMED Usage in Primary Care.

OpenCodelists: Panic disorder resolved or full remission codes

Show the code

pd_res_cod_counts <- snomed_usage |> 
  filter(snomed_code %in% pd_resolved_cod$code) |> 
  get_semantic_tag() |> 
  summarise_code_counts() |> 
  slice_max(n = 20, order_by = usage)

# Display formatted table
tab_pd_res_cod_counts <- pd_res_cod_counts |>
  create_table_sem_tag(
    title = md("Panic disorder resolved or full remission codes"),
    subtitle = "Timeframe: August 2011 to July 2024"
    )

gtsave(
  tab_pd_res_cod_counts,
  here("output", "tables", "cod_devel", "dx", "tab_pd_res_cod_usage.png")
  )

tab_pd_res_cod_counts

SNOMED code	Description	Usage	%
Panic disorder resolved or full remission codes
Timeframe: August 2011 to July 2024
Disorder
11941006	Panic disorder with agoraphobia, agoraphobic avoidance in full remission AND panic attacks in full remission	5	50.000%
82494000	Panic disorder without agoraphobia with panic attacks in full remission	5	50.000%
Data scource: NHS England SNOMED Usage in Primary Care.