Akerke Baibergenova, md, phd, mph University of Toronto Division of Dermatology Sunnybrook Health Sciences Centre 2075 Bayview Ave, M1-700 Toronto ON, M4N 3M5 Canada E-mail: email@example.com
Background Bullous skin diseases are known to be associated with significant morbidity and mortality. There have been no studies on mortality from severe bullous skin diseases in Canada.
Methods We used mortality data from the Statistics Canada website from 2000 to 2007 for three major bullous skin diseases: bullous pemphigoid; pemphigus; and toxic epidermal necrolysis (TEN). Crude and age-standardized mortality rates were calculated and compared with the corresponding US mortality rates. Linear regression was used to assess time trend and effect of gender and age on mortality rates.
Results During the period of eight years, there were 115 deaths attributed to pemphigoid, 84 to pemphigus, and 44 to TEN. The crude annual mortality rate was the highest for pemphigoid (0.045 per 100,000), followed by pemphigus (0.033), and TEN (0.017). None of these conditions demonstrated significant time trends in mortality rates over the eight-year period, although a trend towards decreasing pemphigus mortality was observed (P = 0.07). No gender difference in mortality was observed, but advanced age was associated with mortality in all three conditions.
Conclusion Among bullous skin diseases, pemphigoid is the leading cause of mortality in Canada. This is in contrast to the USA, where TEN is the leading cause of mortality from bullous skin diseases. It is not clear whether differences in healthcare systems explain these findings.
Bullous skin diseases are associated with significant morbidity and mortality and present a challenging condition to treat for internists, dermatologists, and ophthalmologists. The incidence of acquired bullous skin diseases in North America is unknown. In spite of being relatively uncommon, bullous diseases are a major cause of mortality from skin diseases following skin cancers, ulcers, and bacterial skin infections.1
A recently published study from the USA looked at time trends in mortality rates from four severe bullous diseases: pemphigus, pemphigoid, toxic epidermal necrolysis (TEN), and epidermolysis bullosa from 1979 through 2002.2 This study demonstrated decreasing mortality rates from pemphigus but rising mortality rates from pemphigoid – a finding that might be attributable to changes in incidence and management of these diseases over time.
There have been no epidemiologic studies looking at bullous skin disease mortality in Canada. In the present study, we report mortality data on three major acquired bullous diseases of adults: pemphigus; pemphigoid; and TEN. We also compare our findings with findings of the US study.
Materials and methods
We obtained mortality data from the Statistics Canada website (http://www.statcan.gc.ca). The mortality data is a part of CANSIM data – a comprehensive socioeconomic database maintained by Statistics Canada.3 Disease-specific mortality is publicly available for years 2000 through 2007. The data were accessed in November 2010.
Mortality was analyzed in three general bullous disease categories: pemphigoid disorders (hereinafter referred to as pemphigoid); pemphigus disorders (hereinafter referred to as pemphigus); and TEN (including erythema multiforme and Stevens–Johnson syndrome, which share the same codes in ICD-10).
Since 2000, the cause of death in Canada is recorded using the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10). The ICD-10 codes used in our study included L12, L10, and L51 and their subgroups for pemphigus, pemphigoid, and TEN, respectively.
We used Canadian population estimates for July 1 of each year provided by Statistics Canada as a denominator for calculating crude mortality rates per 100,000 population. These population estimates were based on the most recent census data with post-censal retrospective adjustments. In addition to annual rates, mortality rates using two consecutive years were also calculated to have a better stability of rates for visualization of time trends. Mortality rates for each disease were also analyzed by gender.
To compare our findings with the published US mortality rates,2 we also calculated age-standardized mortality rates using the US 2000 standard population.
We used linear regression to model annual rate as a function of calendar year, gender, and age. The goals of linear regression were (i) to assess for presence of any significant time trends, and (ii) to estimate effect of gender and age on mortality. We used SAS® (Cary, NC, USA) software, version 9.2, for linear regression calculations. Values with P < 0.05 were considered to be statistically significant. Research ethics board approval was not applicable to this study because it used publicly available data.
Over a period of eight years (2000–2007), there were 115 deaths attributed to pemphigoid, 84 to pemphigus, and 44 to TEN.
The crude annual mortality rate was the highest for pemphigoid (0.045 per 100,000), followed by pemphigus (0.033 per 100,000), and TEN (0.017 per 100,000). Table 1 displays both crude and age-standardized rates as well as direct comparison of age-standardized Canadian and US rates for years 2000–2002 (overlap period between the US and present study). In comparison with the USA, pemphigus and pemphigoid mortality rates were higher in Canada. In contrast, the TEN mortality rate in Canada was only a half of that in the USA.
Table 1. Acquired bullous disease mortality in Canada 2000–2007. Crude and age-standardized mortality rates per 100 000 population
Crude rates (n), (2000–2007)
Age-standardized rates, Canada (2000–2007)*
Comparison of age-standardized mortality rates (n) between Canada and USA, 2000–2002*
Rate ratio (95% CI)
*Age-standardized to the US 2000 population.
1.51 (1.03; 2.20)
1.62 (1.18; 2.22)
Toxic epidermal necrolysis
0.44 (0.27; 0.72)
Mortality rates per 100,000 by age and gender are shown in Figure 1. During the study period there was a slight decrease in pemphigoid mortality rates but mortality rates from pemphigus and TEN appeared to be stable (Fig. 2). This was confirmed by results of linear regressions (Table 2): there was little change in mortality from these causes over the eight years studied, although there was a trend towards decreased pemphigus mortality (from 0.036 to 0.027) that was of borderline statistical significance (P = 0.07). After adjusting for age, gender was not a predictor of mortality. However, advanced age was associated with significantly higher mortality rates for all three bullous conditions (see Table 2).
Table 2. Mortality rate as a function of year, gender, and age, Canada 2000–2007
Toxic epidermal necrolysis
Year (95% CI)
−0.053 (−0.110; 0.004)
−0.006 (−0.050; 0.038)
−0.004 (−0.040; 0.032)
Female gender (95% CI)
−0.060 (−0.320; 0.200)
0.124 (−0.076; 0.324)
−0.135 (−0.299; 0.029)
Age 65–74 years old (95% CI)
0.043 (−0.325; 0.410)
0.048 (−0.235; 0.331)
0.095 (−0.136; 0.327)
Age 75–84 years old (95% CI)
0.204 (−0.163; 0.571)
0.275 (−0.008; 0.557)
0.110 (−0.122; 0.341)
Age 85 years and older (95% CI)
1.265 (0.897; 1.632)
1.730 (1.446; 2.012)
0.387 (0.155; 0.619)
We provide a descriptive analysis of bullous diseases mortality in Canada. We looked at a relatively short period and did not detect significant time trends in bullous disease mortality, in contrast to the US study.
The average mortality rates in Canada are generally similar to mortality rates from bullous skin diseases reported from the USA.2 However, the differences may be important. When comparison was restricted to 2000–2002, we noted that Canadian mortality from pemphigus and pemphigoid was nearly two-thirds higher than in the USA, but the TEN mortality was less than half the US rate. It should be mentioned that the mortality from TEN in the USA also demonstrated significant racial disparity, with deaths among African-American people 7.5 times higher than among Caucasian people. TEN mortality rates among Caucasian people in the USA (0.025) were closer to (but still higher than) Canadian rates (0.018).
Without further data, we can only speculate about the causes behind these differences. Lower mortality from pemphigus and pemphigoid in the USA might be due to the more aggressive therapeutic approaches used there or due to lower incidence. In contrast to pemphigus and pemphigoid, TEN is a dermatologic emergency with rapid onset of symptoms and progression. Therefore, differences in TEN mortality rates between countries may reflect differences in access to urgent care. One could also speculate that differences in ethnic composition between Canadian and US populations4 might cause higher incidence of pemphigus in Canada leading to higher pemphigus mortality rates. Variation in pemphigus incidence by ethnicity has been observed.5 The ethnic composition of Canada may also contribute to the lower TEN mortality rates. The US study cited above showed that TEN mortality rates are significantly higher in African-American people: Canada has about five times lower percentage of African-American people compared with the USA.4
There are no data on the incidence of each of these three diseases in North America. The higher mortality rates due to pemphigoid found in our study may be a reflection of higher incidence of this disease compared with pemphigus and TEN, as well as its occurrence in older people with possibly multiple comorbidities. The UK study based on a general practice database (1996–2006) showed that incidence of pemphigoid is about six times higher compared with that of pemphigus.6 The gap between incidence of pemphigoid and pemphigus was even higher in the recent Swiss study: the incidence of pemphigoid was shown to be about 20 times as high as that of pemphigus.5 Ethnic homogeneity and higher proportion of elderly people in Switzerland may contribute to this difference.
Similar to other studies, we found that advanced age is associated with significantly higher mortality rates from pemphigoid.7–9 Gender, on the other hand, was not a predictor of pemphigoid mortality in our study. There is no consensus on the effect of gender on pemphigoid mortality. While some reports showed no effect,7,9 others suggested that female gender was associated with higher mortality rates.8
The key strengths of our study include ascertainment of deaths from the entire population of Canada. To put our findings in perspective, the average annual (years 2000–2007) mortality rate per 100,000 from all skin diseases combined, excluding skin cancers (L0-L99, ICD-10), was 9.3 for females and 8.2 for males.3 Therefore, acquired bullous skin diseases comprise about 8–9% of mortality from all skin diseases combined, excluding skin cancers. This makes acquired bullous skin diseases an important cause of mortality from skin conditions.
There are several limitations of our investigation. We have to acknowledge that it is difficult to ascertain pemphigoid or pemphigus as a direct cause of death as patients with either of these diseases frequently die from secondary complications of the disease itself or its treatment. Therefore, we cannot exclude a measurement error in pemphigoid and pemphigus-related mortality. However, similarities in mortality rates between Canada and the USA reported by two different national agencies are reassuring. In addition, the lack of details about the persons who died of these diseases, beyond simple demographic data, limits interpretation of our findings.
We have provided a snapshot of trends in mortality from three major bullous skin diseases in Canada. Our study found that pemphigoid is the leading cause of mortality among these bullous skin diseases. Further investigation into potential causes behind our findings, particularly reasons for differences in mortality rates between Canada and the USA, will help us find better ways to reduce mortality from this group of skin diseases.