Incidence and mortality of bullous pemphigoid and pemphigus vulgaris in the UK

1
Incidence and mortality of bullous pemphigoid and
pemphigus vulgaris in the UK

• Sinéad Langan
• Joe West

2
(No Transcript)
3
Bullous pemphigoid

• Incidence rates between 0.2 and 3 per
• 100,000 person-years
• Estimated incidence of 1.4 per
• 100,000 person-years in Scotland
• Reported 1-year mortality rates
• between 6 (USA) and 41 (France)

4
Pemphigus vulgaris

• Few studies, incidence between 0.08 and
• 1.6 per 100,000 person-years
• No data from the UK
• 1 year mortality rates between 4.8 and
• 54 no clear estimate of overall disease-
• specific mortality rates

5
Objectives

• To determine incidence and
• mortality of bullous pemphigoid and
• pemphigus vulgaris in the United
• Kingdom (UK)

6
Data source

• Computerised medical records from
• the Health Improvement Network
• (THIN), a large population-based UK
• general practice database were
• analysed

7
Confounders

• Age group in 10 year bands
• Gender
• Region
• Townsend scores

8
Study populations

• All patients gt20 years with bullous pemphigoid
  and pemphigus vulgaris diagnosed 1996-2006
• First 3 months following registration excluded
• 4 matched controls selected per case

9
Incidence

• Incidence rates by age category, gender and
  calendar period
• Incidence rate ratios (IRR) adjusting for changes
  in age, gender structure and calendar period over
  time

10
Mortality

• Deaths in cases and controls
• Kaplan-Meier techniques used to calculate 1 year
  mortality and 5 year survival rates
• Cox regression to compare mortality experience of
  cases and controls

11
ONS data

• Incidence and mortality rates from this
• study applied to ONS population data
• to calculate expected numbers of new
• cases and number of deaths

12
Results
13
Age at diagnosis
14
Bullous pemphigoid

• 868 people with bullous pemphigoid and 3453
  matched controls
• Median (IQR) age at first presentation for
  bullous pemphigoid 80 (15) years
• 534 (61) were female

15
Pemphigus vulgaris

• 139 people with pemphigus vulgaris and 551
  matched controls
• Median (IQR) age at first presentation for
  pemphigus vulgaris 71 (30) years
• 91 (66) were female

16
Incidence rates
17
Bullous pemphigoid incidence

• Crude incidence 4.3 (95 CI 4.0-4.6) per 100,000
  py
• 17 increase per year (IRR 1.2, 95 CI 1.1-1.2),
  adjusted for gender and age
• No association with region or Townsend scores

18
Survival rates
19
Bullous pemphigoid mortality

• 1-year mortality rate19
• (95 CI 16-22)
• Absolute excess mortality for
• cases70 deaths per 1000py
• Hazard ratios
• Crude 2.1 (95 CI 1.8-2.4)
• Adjusted 2.3 (95 CI 2.0-2.7)

20
Incidence rates
21
Pemphigus vulgarisincidence

• Crude incidence 0.7 (95 CI 0.6-0.8) per 100,000
  py
• 11 increase per year (IRR 1.1, 95 CI 1.0-1.2),
  adjusted for gender and age
• No association with region or Townsend scores

22
Survival rates
23
Pemphigus vulgarismortality

• 1-year mortality rate12
• (95 CI 8-19)
• Absolute excess mortality for
• cases62 deaths per 1000py
• Hazard ratios
• Crude 2.8 (95 CI 1.9-4.3)
• Adjusted 3.4 (95 CI 2.2-5.2)

24
Mortality rates
25
Incidence over time
26
Deaths over time
27
How many new cases per year?

• Bullous pemphigoid
• Applied to ONS data, 2370 new cases
• per year
• Pemphigus vulgaris
• Applied to ONS data, 467 new
• Cases per year

28
How many deaths between 2001 and 2005?

• Bullous pemphigoid
• Estimated 1977 deaths over 5
• years (vs. 190, ONS)
• Pemphigus vulgaris
• Estimated 221 deaths over
• 5 years (vs. 36, ONS)

29
Key findings

• Increased incidence both diseases between 1996
  and 2006
• No reduction in mortality rates
• Therefore unlikely to be due to ascertainment

30
Strengths

• Large population-based study
• Less susceptible to selection bias
• Power
• Diagnoses usually made with skin biopsy in
  secondary care
• Previous studies confirm validity of diagnoses in
  THIN
• Limited to up to standard data

31
Limitations

• Routine data
• Cannot comment on disease severity
• Possible misclassification of prevalent cases
  possible underestimation of mortality rates

32
Comparison to other research
33
Conclusion

• Increasing incidence rates over time
• No variation by region or socioeconomic status
• Mortality rates twice (pemphigoid) and three
  times (pemphigus) higher than matched controls

34
Implications

• Increasing population age, likely increase in
  morbidity and mortality from these diseases
• Other measures of disease occurrence
  underestimate associated mortality

35
(No Transcript)
36
Acknowledgements

• Dr Chris Smith, Professor Liam
• Smeeth, Dr Kate Fleming and
• Professor Richard Hubbard
• Professor Hywel Williams for guidance
• and support