ABSTRACT
Contemporary population‐based data on incidence trends and in‐hospital outcomes of pemphigus vulgaris remain limited, particularly in Southeast Asia. This study aimed to characterize incidence patterns and identify factors associated with severe in‐hospital outcomes. We conducted a nationwide retrospective cohort study using Thailand's national administrative health database from 2015 to 2024. Incident cases were defined using a two‐year washout period. Crude and age‐standardized incidence rates were estimated, and temporal trends were evaluated using Poisson regression. Hospitalization analyses were restricted to index hospitalizations. Severe in‐hospital outcomes were defined as a composite of in‐hospital mortality, invasive mechanical ventilation, renal replacement therapy, or shock. Associations with severe in‐hospital outcomes were evaluated using multivariable logistic regression, with sensitivity analyses accounting for repeated admissions using generalized estimating equations. A total of 2 542 incident adult cases were identified, corresponding to a mean annual incidence of 6.03 per 1 000 000 population. Incidence increased over time (adjusted incidence rate ratio 1.20; 95% CI 1.18–1.22), was higher with advancing age, and among females (female‐to‐male ratio 1.41:1). Increasing age‐standardized incidence rates suggested that the temporal increase was not fully explained by population aging. This trend was accompanied by increasing annual hospitalizations. Among 1 537 index hospitalizations, 8.1% experienced severe in‐hospital outcomes. Acute kidney injury, malnutrition, and infection were associated with severe in‐hospital outcomes. In sensitivity analyses accounting for repeated admissions, advancing age and cardiovascular comorbidity were also associated with increased risk. Pemphigus vulgaris demonstrates a rising incidence beyond the effects of population aging, accompanied by increasing hospitalizations and a growing burden of severe in‐hospital outcomes. These findings support early identification of high‐risk patients and appropriate clinical management.
Keywords: hospitalization, incidence, pemphigus vulgaris, Thailand
1. Introduction
Pemphigus vulgaris is a rare but potentially life‐threatening autoimmune blistering disease characterized by autoantibody‐mediated loss of keratinocyte adhesion [1]. Despite therapeutic advances, the disease remains associated with significant morbidity, particularly among patients who develop severe complications requiring hospitalization [2, 3]. A comprehensive understanding of its epidemiology and clinical burden is therefore essential for anticipating healthcare needs and improving patient outcomes.
Population‐based studies have reported substantial geographic variation in incidence, with a pooled global estimate of approximately 2–3 cases per 1 000 000 person‐years [4]. However, data from Southeast Asia remains limited. Temporal trends have also been inconsistent across regions, underscoring the need for contemporary nationwide data from underrepresented populations [5, 6, 7].
Severe in‐hospital outcomes remain a major source of morbidity, mortality, and healthcare utilization in pemphigus vulgaris. Although prior studies have examined predictors of in‐hospital outcomes, most were derived from single‐center or selected cohorts. Consequently, the population‐level determinants of severe outcomes, including demographic characteristics, comorbidities, acute complications, and healthcare system factors, remain incompletely characterized.
We conducted a nationwide population‐based study using administrative health data from Thailand to estimate the incidence and temporal trends of pemphigus vulgaris in adults, and to identify factors associated with severe in‐hospital outcomes. By integrating population‐level incidence with hospitalization outcomes, this study provides a comprehensive assessment of the epidemiologic and clinical burden of pemphigus vulgaris in a Southeast Asian setting.
2. Methods
This retrospective study used administrative health data from the National Health Security Office (NHSO) of Thailand. Patients with a diagnosis of pemphigus vulgaris (ICD‐10‐TM code L10.0) between January 1, 2015, and December 31, 2024, were identified across outpatient and inpatient settings. Patients aged < 18 years and those with pemphigus vulgaris recorded only as a secondary diagnosis were excluded to restrict the analysis to adults and improve diagnostic specificity. Cohort construction and exclusion criteria are summarized in Figure 1.
FIGURE 1.
Flow diagram illustrating patient selection, identification of adult patients with pemphigus vulgaris (ICD‐10‐TM code L10.0) from the nationwide National Health Security Office database in Thailand between 2015 and 2024, application of exclusion criteria, and derivation of the incidence and hospitalization cohorts.
Demographic variables included age and sex. Age was analyzed both continuously and in predefined categories (18–29, 30–39, 40–49, 50–59, 60–69, 70–79, and ≥ 80 years). Clinical variables included length of hospital stay, discharge status, and inpatient healthcare expenditures. Geographic region and hospital type were categorized according to the Thai Ministry of Public Health classification system (university‐affiliated hospitals, regional hospitals, general hospitals, and community hospitals).
Baseline comorbidities were identified using ICD‐10‐TM codes according to the Charlson Comorbidity Index framework and included diabetes mellitus, hypertension, cardiovascular disease, neurologic disease, chronic kidney disease, chronic pulmonary disease, connective tissue disease, peptic ulcer disease, chronic liver disease, malignancy, tuberculosis, and HIV infection. Conditions of specific clinical relevance to pemphigus vulgaris, including myasthenia gravis, Cushing syndrome, adrenal insufficiency, and protein–energy malnutrition, were also identified. Comorbidities were defined using ICD‐10‐TM codes corresponding to clinically relevant disease categories and analyzed as grouped comorbidity categories rather than stratified by individual diagnostic subtypes or severity levels. Detailed ICD codes are provided in Supporting Information.
In‐hospital complications were classified as infectious or non‐infectious. Infectious complications included sepsis, pneumonia, urinary tract infection, skin and soft tissue infection, and selected opportunistic infections, including nocardiosis, eczema herpeticum, cytomegalovirus infection, invasive candidiasis, and Pneumocystis jirovecii pneumonia. Non‐infectious complications included acute kidney injury and shock. Organ support and procedure‐related interventions were identified using ICD‐9‐CM procedure codes and included renal replacement therapy, invasive mechanical ventilation, intravenous immunoglobulin infusion, and systemic corticosteroid administration. Rituximab use could not be specifically ascertained due to the absence of a dedicated procedure code. Complications were not mutually exclusive, and individual patients could experience multiple complications during a single hospitalization.
In‐hospital diagnoses were identified based on codes recorded within the same admission. As the temporal sequence of diagnoses cannot be determined from administrative data, these conditions were analyzed as coexisting diagnoses associated with severe in‐hospital outcomes rather than time‐ordered events.
For analyses of in‐hospital outcomes, the primary analytic cohort was based on the index hospitalization, defined as the first recorded admission for pemphigus vulgaris for each patient during the study period. Severe in‐hospital outcomes were defined as a composite outcome of mortality, invasive mechanical ventilation, renal replacement therapy, or shock, representing life‐threatening disease requiring major organ support or hemodynamic compromise.
3. Statistical Analysis
3.1. Incidence Analyses
Incident cases were defined as individuals with a first recorded diagnosis during the study period. A 2‐year washout period was applied to reduce misclassification of prevalent cases. Crude incidence rates were calculated per 1 000 000 population using mid‐year population denominators, with 95% confidence intervals (CIs) estimated assuming a Poisson distribution. Age‐specific incidence rates were estimated using 10‐year age bands. Age‐standardized incidence rates (ASIRs) were calculated using direct standardization with the aggregated study population as the standard, and 95% CIs were derived using the Dobson method.
As a sensitivity analysis to further reduce potential misclassification from single‐visit coding, a stricter case definition was applied, requiring at least two separate encounters with a primary diagnosis of pemphigus vulgaris. Incident cases were redefined using the first qualifying encounter, and temporal trends and age‐specific incidence patterns were re‐evaluated under this definition.
Temporal trends in incidence were assessed using Poisson regression models with the log population as an offset. Calendar year was modeled as a continuous variable, and multivariable models were adjusted for age group and sex. Results were reported as incidence rate ratios (IRRs) with 95% CIs.
3.2. Hospitalization Cohort and Baseline Characteristics
For analyses of in‐hospital outcomes, we included patients admitted with a primary diagnosis of pemphigus vulgaris. Factors associated with severe in‐hospital outcomes were evaluated using the index hospitalization to reflect baseline severity at first admission and minimize within‐patient correlation.
Baseline demographic characteristics, hospital type, geographic region, comorbidities, and complications at index hospitalization were summarized by severe in‐hospital outcome status. Continuous variables were reported as means with standard deviations or medians with interquartile ranges, as appropriate, and categorical variables as counts and percentages. Between‐group comparisons were performed using parametric or nonparametric tests for continuous variables and chi‐square or Fisher's exact tests for categorical variables. Clinical course and in‐hospital outcomes, including individual components of the composite outcome, infectious complications, length of stay, and hospital costs, were also summarized by outcome status.
3.3. Multivariable Analysis of Severe In‐Hospital Outcomes
Multivariable logistic regression models were used to identify factors associated with severe in‐hospital outcomes at the index hospitalization. Candidate factors were selected based on clinical relevance and/or evidence of association in descriptive analyses, and variables with p < 0.20 in univariable comparisons were considered for inclusion in the multivariable model. Adjusted odds ratios (aORs) with 95% CIs were reported.
Sensitivity analyses, including all hospitalizations, were performed using logistic regression models fitted with generalized estimating equations and robust standard errors to account for within‐patient correlation. All analyses were performed using R statistical software (version 4.5.2, R Foundation for Statistical Computing, Vienna, Austria), and a two‐sided p < 0.05 was considered statistically significant.
4. Results
4.1. Incidence of Pemphigus Vulgaris
During 2017–2024, a total of 2 542 incident cases of pemphigus vulgaris were identified, corresponding to a mean annual incidence of 6.03 per 1 000 000 population (95% CI, 5.80–6.27) (Table S1). Incidence increased markedly with advancing age, ranging from 1.77 per million in individuals aged 18–29 years to 17.33 per million among those aged ≥ 80 years, demonstrating a clear stepwise gradient (Figure 2). Incidence rates were consistently higher in females than males (7.21 and 4.77 per million, respectively).
FIGURE 2.
Age‐specific incidence of pemphigus vulgaris per 1 000 000 population, pooled across the study period (2017–2024). Incidence increased progressively with advancing age, with the highest rates observed among adults aged ≥ 80 years. Error bars represent 95% confidence intervals.
Temporal trends in crude and ASIRs are shown in Figure 3. Both measures increased steadily over time, with crude incidence rates remaining consistently higher than age‐standardized rates. The increasing separation between crude and age‐standardized rates suggests a growing contribution of population aging to the overall incidence. However, age‐standardized incidence also increased, indicating that demographic aging alone does not fully explain the observed temporal rise. The annual number of hospital admissions showed a parallel increase over time (Figure 4).
FIGURE 3.
Temporal trends in crude and age‐standardized incidence of pemphigus vulgaris, 2017–2024. Incidence increased over time (p for trend < 0.001, Poisson regression). Shaded areas represent 95% confidence intervals; blue and red lines denote crude and age‐standardized incidence, respectively.
FIGURE 4.
Annual number of hospital admissions for pemphigus vulgaris in Thailand from 2017 to 2024. Each data point represents the total number of hospitalizations recorded per calendar year. A progressive increase in annual admissions was observed over the study period.
In Poisson regression models adjusted for age group and sex, the incidence of pemphigus vulgaris increased by approximately 20% per year (adjusted IRR 1.20; 95% CI, 1.18–1.22; p < 0.001; Table 1). Age remained a strong independent factor, with progressively increasing IRRs across older age groups compared with adults aged 18–29 years. Females had a higher incidence than males (adjusted IRR 1.41; 95% CI, 1.30–1.54). In age‐stratified analyses, female predominance persisted across most age groups, with the greatest differences observed in younger and older adults (Table S2).
TABLE 1.
Poisson regression analysis of incidence of pemphigus vulgaris by age, sex, and calendar year.
| Variable | Incidence rate ratio (95% CI)c | p |
|---|---|---|
| Year (per 1‐year increase) | 1.20 (1.18–1.22) | < 0.001 |
| Age group (reference: Age 18–29) | ||
| Age 30–39 | 1.31 (1.05–1.62) | 0.015 |
| Age 40–49 | 2.97 (2.49–3.58) | < 0.001 |
| Age 50–59 | 4.17 (3.51–4.99) | < 0.001 |
| Age 60–69 | 5.45 (4.57–6.53) | < 0.001 |
| Age 70–79 | 7.03 (5.85–8.51) | < 0.001 |
| Age 80+ | 9.10 (7.44–11.15) | < 0.001 |
| Female sex (reference: Male sex) | 1.41 (1.30–1.54) | < 0.001 |
Abbreviation: CI, confidence interval.
In sensitivity analyses using a stricter case definition requiring at least two pemphigus vulgaris‐coded encounters, 1 613 incident cases were identified. The increasing temporal trend remained robust after adjustment for age group and sex (adjusted IRR 1.24 per year; 95% CI, 1.21–1.27). The age‐specific pattern was also preserved, with incidence increasing progressively across age groups and peaking in older adults, with a slight attenuation in the ≥ 80‐year group (Figure S1).
4.2. Severe In‐Hospital Outcomes
During the study period, a total of 3 042 hospitalizations for pemphigus vulgaris were identified. Of these, 1 537 represented index hospitalizations and were included in the primary analysis. Among index hospitalizations, 124 (8.1%) experienced severe in‐hospital outcomes.
Baseline characteristics stratified by severe in‐hospital outcome status are presented in Table 2. Patients with severe in‐hospital outcomes were slightly older, although the difference was not statistically significant. Sex distribution was comparable between groups. Geographic region differed significantly (p = 0.011), with a higher proportion of severe cases in Bangkok. Hospital type was strongly associated with severity (p < 0.001); severe cases were more frequently managed in university‐affiliated hospitals, whereas non‐severe cases were predominantly treated in regional, general, and community hospitals.
TABLE 2.
Baseline characteristics of patients hospitalized with pemphigus at index admission, stratified by severe in‐hospital outcome status.
| Characteristic | Overall (N = 1 537) | Non‐severe (N = 1 413) | Severe a (N = 124) | p |
|---|---|---|---|---|
| Age (mean ± SD) | 57.59 ± 17.48 | 57.37 ± 17.52 | 60.19 ± 16.96 | 0.084 |
| Male, n (%) | 626 (40.7) | 571 (40.4) | 55 (44.4) | 0.446 |
| Region, n (%) | 0.011 | |||
| Bangkok | 62 (4.0) | 49 (3.5) | 13 (10.5) | |
| Central | 174 (11.3) | 159 (11.3) | 15 (12.1) | |
| Northeastern | 690 (45.0) | 635 (45.0) | 55 (44.4) | |
| Northern | 314 (20.5) | 296 (21.0) | 18 (14.5) | |
| Eastern | 89 (5.8) | 82 (5.8) | 7 (5.6) | |
| Western | 52 (3.4) | 47 (3.3) | 5 (4.0) | |
| Southern | 154 (10.0) | 143 (10.1) | 11 (8.9) | |
| Hospital type, n (%) | < 0.001 | |||
| University‐affiliated hospital | 84 (5.5) | 51 (3.6) | 33 (26.6) | |
| Regional hospital | 564 (36.7) | 530 (37.5) | 34 (27.4) | |
| General hospital | 415 (27.0) | 389 (27.5) | 26 (21.0) | |
| Community hospital | 356 (23.2) | 338 (23.9) | 18 (14.5) | |
| Non‐ministry of public health hospital | 93 (6.1) | 84 (5.9) | 9 (7.3) | |
| Private hospital | 25 (1.6) | 21 (1.5) | 4 (3.2) | |
| Baseline comorbidities | ||||
| Diabetes mellitus, n (%) | 276 (18.0) | 255 (18.0) | 21 (16.9) | 0.852 |
| Hypertension, n (%) | 407 (26.5) | 368 (26.0) | 39 (31.5) | 0.229 |
| Cardiovascular disease, n (%) | 10 (0.7) | 7 (0.5) | 3 (2.4) | 0.049 |
| Neurologic disease, n (%) | 94 (6.1) | 77 (5.4) | 17 (13.7) | < 0.001 |
| Chronic kidney disease, n (%) | 85 (5.5) | 76 (5.4) | 9 (7.3) | 0.501 |
| Chronic liver disease, n (%) | 35 (2.3) | 33 (2.3) | 2 (1.6) | 0.839 |
| Malignancy, n (%) | 16 (1.0) | 15 (1.1) | 1 (0.8) | 1.00 |
| Malnutrition, n (%) | 34 (2.2) | 23 (1.6) | 11 (8.9) | < 0.001 |
| Complications | ||||
| Infection during admission, n (%) | 345 (22.4) | 285 (20.2) | 60 (48.4) | < 0.001 |
| Acute kidney injury, n (%) | 79 (5.1) | 52 (3.7) | 27 (21.8) | < 0.001 |
Abbreviation: SD, standard deviation.
Severe in‐hospital outcomes were defined as a composite of mortality, shock, invasive mechanical ventilation, and renal replacement therapy.
Several baseline comorbidities were more prevalent among patients with severe in‐hospital outcomes, including cardiovascular disease, neurologic disease, and malnutrition. In contrast, diabetes mellitus, hypertension, chronic kidney disease, chronic liver disease, and malignancy did not differ significantly. Infection and acute kidney injury were more frequently observed among patients with severe in‐hospital outcomes. Rare comorbidities (myasthenia gravis, n = 1; Cushing syndrome, n = 0; adrenal disorders, n = 4) and infrequently recorded treatments (intravenous immunoglobulin, n = 14; systemic corticosteroid infusion, n = 1) were excluded from regression models because of sparse data.
Multivariable logistic regression results are presented in Table 3. After multivariable adjustment, hospital type remained independently associated with severe in‐hospital outcomes. Malnutrition was independently associated with severe in‐hospital outcomes (adjusted odds ratio [aOR] 2.86; 95% CI, 1.11–6.84). Infection (aOR, 2.82; 95% CI, 1.81–4.39) and acute kidney injury (aOR, 4.84; 95% CI, 2.60–8.83) showed the strongest associations with severe in‐hospital outcomes. Age and sex were not independently associated with severe in‐hospital outcomes at index hospitalization.
TABLE 3.
Multivariable logistic regression analysis of factors associated with severe in‐hospital outcomes among patients with pemphigus vulgaris at index admission.
| Characteristic | Adjusted OR (95% CI) | p |
|---|---|---|
| Age (per 10‐year increase) | 1.13 (0.99–1.29) | 0.073 |
| Male sex | 1.23 (0.81–1.87) | 0.328 |
| Region (reference: Bangkok) | ||
| Central | 0.85 (0.31–2.30) | 0.739 |
| Northeastern | 1.76 (0.71–4.58) | 0.235 |
| Northern | 1.01 (0.38–2.72) | 0.986 |
| Eastern | 1.13 (0.32–3.77) | 0.850 |
| Western | 1.59 (0.41–5.65) | 0.484 |
| Southern | 1.45 (0.47–4.46) | 0.520 |
| Hospital type (reference: University‐affiliated hospital) | ||
| Regional hospital | 0.06 (0.03–0.13) | < 0.001 |
| General hospital | 0.07 (0.04–0.15) | < 0.001 |
| Community hospital | 0.07 (0.03–0.14) | < 0.001 |
| Non‐ministry of public health hospital | 0.13 (0.05–0.30) | < 0.001 |
| Private hospital | 0.40 (0.10–1.26) | 0.142 |
| Cardiovascular disease | 3.97 (0.68–17.66) | 0.087 |
| Neurologic disease | 1.82 (0.91–3.46) | 0.078 |
| Malnutrition | 2.86 (1.11–6.84) | 0.023 |
| Infection during admission | 2.82 (1.81–4.39) | < 0.001 |
| Acute kidney injury | 4.84 (2.60–8.83) | < 0.001 |
Abbreviations: CI, confidence interval; OR, odds ratio.
Clinical course and in‐hospital outcomes stratified by severe in‐hospital outcome status are presented in Table 4. Patients who experienced severe in‐hospital outcomes had high in‐hospital mortality (46.0%) and frequently required organ support, including mechanical ventilation (29.0%) and renal replacement therapy (6.5%). Shock occurred in 37.1% of severe cases. Infectious complications were more frequent in the severe group, including sepsis (26.6%), pneumonia (17.7%), urinary tract infection (12.9%), and skin and soft tissue infections (12.1%). Opportunistic infections were uncommon but occurred predominantly in the severe group, with cases restricted to eczema herpeticum, cytomegalovirus infection, and invasive candidiasis. In addition, patients with severe outcomes had longer median hospital stays and higher hospital costs.
TABLE 4.
Clinical course and in‐hospital outcomes at index admission according to severe in‐hospital outcome status.
| Variable | Non‐severe (N = 1 413) | Severe a (N = 124) |
|---|---|---|
| Components of severe composite | ||
| In‐hospital mortality, n (%) | — | 57 (46.0) |
| Shock, n (%) | — | 46 (37.1) |
| Mechanical ventilation, n (%) | — | 36 (29.0) |
| Renal replacement therapy, n (%) | — | 8 (6.5) |
| Infection complications | ||
| Sepsis, n (%) | 93 (6.6) | 33 (26.6) |
| Pneumonia, n (%) | 33 (2.3) | 22 (17.7) |
| Urinary tract infection, n (%) | 48 (3.4) | 16 (12.9) |
| Skin and soft tissue infection, n (%) | 138 (9.8) | 15 (12.1) |
| Eczema herpeticum, n (%) | 19 (1.3) | 1 (0.8) |
| Cytomegalovirus infection, n (%) | 0 (0.0) | 2 (1.6) |
| Invasive candidiasis, n (%) | 1 (0.1) | 3 (2.4) |
| Length of stay (days), median (IQR) | 7.0 (4.0–11.0) | 13.0 (6.0–26.2) |
| Hospital cost (USD), median (IQR) | 388.5 (189.8–739.1) | 1436.1 (363.6–4136.2) |
Abbreviations: IQR, interquartile range; USD, United States dollars.
Severe in‐hospital outcomes were defined as a composite of mortality, shock, invasive mechanical ventilation, and renal replacement therapy. Patients could meet more than one component; therefore, percentages may sum to more than 100%.
Sensitivity analyses, including all 3 042 hospitalizations and accounting for within‐patient correlation using generalized estimating equations, yielded results consistent with the primary index‐hospitalization analysis (Table S3). Acute kidney injury and infection during admission remained associated with severe in‐hospital outcomes, while cardiovascular disease and older age were additionally identified as significant factors.
Consistent with this finding, age‐specific adjusted odds ratios demonstrated a progressive increase in the probability of severe in‐hospital outcomes with advancing age (Table S4), highlighting the cumulative impact of aging on disease severity across repeated hospitalizations.
5. Discussion
5.1. Incidence of Pemphigus Vulgaris
In this nationwide study, the mean annual incidence was 6.03 per 1 000 000 population, exceeding the pooled global incidence of 2.83 per 1 000 000 person‐years [4, 8]. Although within the reported global range, the higher incidence likely reflects temporal and regional differences. Most studies contributing to pooled estimates were conducted before 2019, whereas our analysis covers a more recent period during which incidence increased steadily, suggesting that earlier reports may underestimate current disease burden.
Geographic variation in incidence has been widely documented. Genetic susceptibility, including HLA class II alleles, may contribute to regional heterogeneity [8, 9]. However, population‐level data directly linking genetic susceptibility to regional incidence remain limited, and further region‐specific studies integrating epidemiologic and genetic data are needed.
We observed a consistent female predominance, consistent with previous epidemiologic studies [8]. Notably, the pattern persisted across age groups, suggesting that sex differences are unlikely to be explained solely by reproductive hormonal factors. Although sex‐related differences in immune regulation have been proposed in autoimmune diseases, the biological basis of female predominance in pemphigus vulgaris remains incompletely understood [10].
Advancing age emerged as a major determinant of pemphigus vulgaris incidence, with rates increasing markedly across older age groups. Similar age‐associated increases have been reported in several population‐based cohorts [5, 6], supporting age‐related susceptibility as an important epidemiologic feature. The observed pattern is biologically plausible, as aging is associated with immune dysregulation, including immunosenescence and chronic low‐grade inflammation, which may increase autoreactivity and susceptibility to autoantibody‐mediated diseases such as pemphigus vulgaris [11].
Nevertheless, classical descriptions of pemphigus vulgaris often emphasize a peak incidence in middle‐aged individuals [1, 12]. The higher incidence observed in older age groups in this study may partly reflect age at diagnosis rather than true disease onset, as delayed recognition and diagnostic overlap with other autoimmune blistering diseases, particularly bullous pemphigoid, may be more common in elderly populations. However, in sensitivity analyses using a stricter case definition requiring at least two PV‐coded encounters, the age‐specific pattern remained consistent, suggesting that the observed age gradient is unlikely to be explained solely by single‐visit misclassification or diagnostic uncertainty. These findings suggest that the observed age gradient may reflect both diagnostic factors and a potential shift in disease occurrence in aging populations.
Comparison of crude incidence and ASIRs provides important insight into temporal trends. While population aging is expected to contribute to increases in crude incidence, the continued rise in age‐standardized rates suggests that demographic change alone does not fully explain the observed trend, indicating the presence of additional age‐independent influences. Improved case ascertainment and heightened diagnostic awareness, potentially supported by broader access to immunopathologic and serologic testing, may partially explain the observed increase. However, concurrent environmental changes, including climatic factors and pesticide exposure, have been implicated in pemphigus vulgaris and may have intensified over time, suggesting that a true rise in disease occurrence cannot be excluded [13, 14, 15, 16].
Collectively, our findings suggest that the increasing incidence of pemphigus vulgaris reflects a combination of demographic shifts, improved case ascertainment, and potential environmental influences. The persistence of rising age‐standardized incidence rates indicates that factors beyond population aging contribute to disease occurrence, supporting a potential epidemiologic transition. This evolving pattern is accompanied by increasing hospitalizations, underscoring a growing disease burden. In this context, identifying factors associated with severe in‐hospital outcomes is essential for risk stratification and clinical management.
5.2. Hospitalization and Outcomes
Although only a minority of hospitalized patients experienced severe in‐hospital outcomes at index hospitalization (8.1%), these events were associated with high mortality and frequent need for organ support. Hemodynamic instability and serious infections were common, leading to prolonged hospital stays and increased healthcare costs. These findings indicate that severe in‐hospital outcomes carry substantial prognostic and resource implications.
Hospital type emerged as a strong factor associated with severe in‐hospital outcomes. Patients admitted to university‐affiliated hospitals were more likely to experience severe in‐hospital outcomes than those treated in regional, general, or community hospitals. This likely reflects referral and case‐mix differences, whereby patients with greater clinical instability are preferentially managed at tertiary centers, rather than differences in quality of care. Healthcare system characteristics have been shown to influence outcomes in pemphigus, underscoring the importance of contextual factors in interpreting severity patterns [17]. Therefore, hospital type should be interpreted as a proxy for disease severity and referral patterns, rather than a direct determinant of clinical outcomes.
Protein–energy malnutrition was independently associated with severe in‐hospital outcomes. This finding is biologically plausible, as pemphigus vulgaris may impair nutritional status through hypercatabolism, corticosteroid‐associated muscle loss, and reduced oral intake due to mucosal involvement [18]. Compromised nutritional reserve may predispose patients to clinical deterioration. Prior studies have also linked hypoalbuminemia with adverse outcomes in pemphigus vulgaris [19].
Cardiovascular and neurologic diseases were more prevalent among patients with severe in‐hospital outcomes in univariable analyses. Although these associations did not retain statistical significance after multivariable adjustment, effect estimates remained elevated. Prior studies have reported increased cardiovascular morbidity and neurologic comorbidity in pemphigus vulgaris [19, 20, 21, 22], and reduced physiologic reserve related to chronic systemic disease may increase vulnerability to acute deterioration during hospitalization. The relatively low prevalence of these comorbidities in our cohort likely limited statistical power, resulting in wide confidence intervals. Nonetheless, these findings suggest that chronic cardiovascular and neurologic conditions may contribute to susceptibility to severe in‐hospital outcomes.
Infectious complications identified during hospitalization were strongly and consistently associated with severe in‐hospital outcomes across primary and sensitivity analyses. These findings likely reflect the close interplay between infection, disease severity, and immunosuppression in hospitalized patients with pemphigus vulgaris, rather than a strictly causal relationship. Disruption of epithelial barriers, combined with immunosuppressive therapy and advanced age, likely increases susceptibility to infection in patients with pemphigus vulgaris. Prior studies have similarly identified infection as a major contributor to morbidity and mortality [2, 23, 24]. Although opportunistic infections were uncommon, their near‐exclusive occurrence in severe cases supports the role of immune vulnerability in adverse outcomes.
Acute kidney injury (AKI) was strongly associated with severe in‐hospital outcomes, whereas chronic kidney disease showed no clear association with severity. AKI was observed in approximately 20% of severe cases at index hospitalization, underscoring its clinical relevance. Volume depletion, systemic inflammation, and critical illness may contribute to renal vulnerability in this setting. Given the lack of temporal information, AKI likely reflects acute clinical deterioration during hospitalization rather than a baseline risk factor. Collectively, these findings suggest that AKI is a key marker of acute deterioration in hospitalized patients with pemphigus vulgaris.
Sensitivity analyses incorporating all hospitalizations and accounting for within‐patient correlation provided further insight into factors associated with severe in‐hospital outcomes. In these analyses, advancing age was associated with a graded increase in the likelihood of severe in‐hospital outcomes across older age groups. While most previous studies have focused on age at diagnosis in relation to mortality [19, 25, 26], our findings suggest that age‐related vulnerability becomes more apparent when cumulative risk across repeated admissions is considered.
Cardiovascular disease also became significantly associated with severe in‐hospital outcomes in sensitivity analyses. This pattern suggests that chronic cardiovascular comorbidity may contribute to vulnerability over time, especially in the context of recurrent hospitalizations. Prior longitudinal studies have demonstrated persistently elevated cardiovascular risk in pemphigus vulgaris [21], which may contribute to cumulative susceptibility to severe in‐hospital outcomes. These findings suggest that vulnerability to severe in‐hospital outcomes accumulates with advancing age and cardiovascular comorbidity across repeated hospitalizations.
5.3. Strengths and Limitations
A key strength of this study is its integrated assessment of pemphigus vulgaris across the disease spectrum, linking population‐level incidence with severe in‐hospital outcomes. Using nationwide administrative data from Thailand, we generated population‐based estimates over an extended study period, allowing assessment across age, sex, and calendar year. These findings provide contemporary regional evidence from Southeast Asia. Incident cases were identified using a washout period to minimize misclassification of prevalent disease, and both crude and ASIRs were examined to distinguish demographic aging from temporal changes in disease risk. Hospitalization analyses focused on index hospitalizations to avoid repeated counting of individuals. Sensitivity analyses, including all admissions, yielded consistent results, supporting the robustness of the findings.
Several limitations should be acknowledged. The use of administrative claims data may introduce misclassification, as diagnoses, comorbidities, and complications were identified using ICD codes without clinical validation or access to detailed disease severity measures, laboratory data, or immunologic and genetic information. In addition, comorbidities were analyzed as grouped categories, and further stratification by disease severity (e.g., chronic kidney disease stage or liver disease severity) was not performed. Undercoding may have resulted in the underestimation of certain complications and interventions. Furthermore, several conditions identified during hospitalization (e.g., infection and acute kidney injury) may represent coexisting clinical events rather than baseline characteristics. As temporal information is unavailable in administrative data, their sequence relative to severe outcomes cannot be determined, and potential mediator effects cannot be fully distinguished. Outpatient disease activity and longitudinal treatment data, including detailed information on systemic corticosteroids, immunosuppressive therapies, and adjunctive treatments, were not available, limiting evaluation of treatment‐related effects; rituximab use could not be specifically identified due to the absence of a dedicated procedure code. Variability in treatment strategies may have influenced disease course and outcomes. Complications, including opportunistic infections, may be incompletely captured in administrative data. As this study was conducted within a single national healthcare system, generalizability to other settings may be limited. Residual confounding cannot be excluded despite multivariable adjustment.
5.4. Conclusions
In this nationwide population‐based study, pemphigus vulgaris in Thailand demonstrated a rising incidence not fully explained by population aging, with substantially higher rates observed in older age groups. Among hospitalized patients, infection, acute kidney injury, and malnutrition were associated with severe in‐hospital outcomes. Sensitivity analyses further indicated that advancing age and cardiovascular comorbidity were associated with cumulative vulnerability across repeated admissions. Together, these findings suggest an increasing clinical and healthcare burden and support the importance of early risk stratification and proactive management.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Ethics Statement
The study protocol was approved by the Institutional Review Board of Khon Kaen University (HE681587).
Consent
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting Information: S1 International Classification of Diseases (ICD‐10‐TM and ICD‐9‐CM) codes used to define comorbidities, complications, and in‐hospital procedures.
Figure S1: Age‐specific incidence of pemphigus vulgaris per 1 000 000 population using a stricter case definition (≥ 2 coded encounters), pooled across 2017–2024. Incidence increased with age, peaking in older adults with slight attenuation in those aged ≥ 80 years. Error bars indicate 95% confidence intervals.
Table S1: Incidence of pemphigus vulgaris in adults by sex and age group, 2017–2024.
Table S2: Age‐specific female‐to‐male incidence rate ratios of pemphigus vulgaris.
Table S3: Sensitivity analysis of predictors of severe in‐hospital outcomes among all hospitalizations for pemphigus using generalized estimating equations.
Table S4: Estimated adjusted odds ratios for severe in‐hospital outcomes by age.
Acknowledgments
We would like to thank the National Health Security Office (NHSO), Bangkok, Thailand, and the Research Program of Khon Kaen University (RP68‐5‐001) for their support.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Schmidt E., Kasperkiewicz M., and Joly P., “Pemphigus,” Lancet 394, no. 10201 (2019): 882–894. [DOI] [PubMed] [Google Scholar]
- 2. Hsu D. Y., Brieva J., Sinha A. A., Langan S. M., and Silverberg J. I., “Comorbidities and Inpatient Mortality for Pemphigus in the U.S.A,” British Journal of Dermatology 174, no. 6 (2016): 1290–1298. [DOI] [PubMed] [Google Scholar]
- 3. Murrell D. F., Peña S., Joly P., et al., “Diagnosis and Management of Pemphigus: Recommendations of an International Panel of Experts,” Journal of the American Academy of Dermatology 82, no. 3 (2020): 85.e1–85.e575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Zhao L., Chen Y., and Wang M., “The Global Incidence Rate of Pemphigus Vulgaris: A Systematic Review and Meta‐Analysis,” Dermatology 239, no. 4 (2023): 514–522. [DOI] [PubMed] [Google Scholar]
- 5. Langan S. M., Smeeth L., Hubbard R., Fleming K. M., Smith C. J., and West J., “Bullous Pemphigoid and Pemphigus Vulgaris – Incidence and Mortality in the UK: Population Based Cohort Study,” BMJ 337, no. 7662 (2008): a180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jelti L., Cordel N., Gillibert A., et al., “Incidence and Mortality of Pemphigus in France,” Journal of Investigative Dermatology 139, no. 2 (2019): 469–473. [DOI] [PubMed] [Google Scholar]
- 7. Chiu H. Y., Chang C. J., Lin Y. J., and Tsai T. F., “National Trends in Incidence, Mortality, Hospitalizations, and Expenditures for Pemphigus in Taiwan,” Journal of Dermatological Science 99, no. 3 (2020): 203–208. [DOI] [PubMed] [Google Scholar]
- 8. Rosi‐Schumacher M., Baker J., Waris J., Seiffert‐Sinha K., and Sinha A. A., “Worldwide Epidemiologic Factors in Pemphigus Vulgaris and Bullous Pemphigoid,” Frontiers in Immunology 14 (2023): 1159351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Egami S., Yamagami J., and Amagai M., “Autoimmune Bullous Skin Diseases, Pemphigus and Pemphigoid,” Journal of Allergy and Clinical Immunology 145, no. 4 (2020): 1031–1047. [DOI] [PubMed] [Google Scholar]
- 10. Zhao C. Y. and Murrell D. F., “Autoimmune Blistering Diseases in Females: A Review,” Int J Womens Dermatol 1 (2015): 4–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Vadasz Z., Haj T., Kessel A., and Toubi E., “Age‐Related Autoimmunity,” BMC Medicine 11 (2013): 94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kasperkiewicz M., Ellebrecht C. T., Takahashi H., et al., “Pemphigus,” Nature Reviews Disease Primers 3 (2017): 17026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ren Z., Hsu D., Brieva J., and Silverberg J. I., “Association Between Climate, Pollution and Hospitalization for Pemphigus in the USA,” Clinical and Experimental Dermatology 44, no. 2 (2019): 135–143. [DOI] [PubMed] [Google Scholar]
- 14. Kyriakis K. P., Vareltzidis A. G., and Tosca A. D., “Environmental Factors Influencing the Biologic Behavior of Patterns of Pemphigus Vulgaris: Epidemiologic Approach,” International Journal of Dermatology 34, no. 3 (1995): 181–185. [DOI] [PubMed] [Google Scholar]
- 15. Stone C., Bak G., Oh D., et al., “Environmental Triggers of Pemphigus Vulgaris and Bullous Pemphigoid: A Case Control Study,” Front Med (Lausanne) 11 (2024): 1441369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Valikhani M., Kavusi S., Chams‐Davatchi C., et al., “Pemphigus and Associated Environmental Factors: A Case‐Control Study,” Clinical and Experimental Dermatology 32, no. 3 (2007): 256–260. [DOI] [PubMed] [Google Scholar]
- 17. Kasperkiewicz M., Lai O., Recke A., et al., “Clinical Outcomes of Patients With Pemphigus Treated by the Same Physicians in a Public Safety Net Healthcare System vs. a Private Healthcare System,” British Journal of Dermatology 181, no. 4 (2019): 850–852. [DOI] [PubMed] [Google Scholar]
- 18. Kajdas A. A., Żebrowska A., Zalewska‐Janowska A., and Czerwonogrodzka‐Senczyna A., “The Role of Nutrition in the Pathogenesis and Treatment of Autoimmune Bullous Diseases‐A Narrative Review,” Nutrients 16, no. 22 (2024): 3961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kridin K., Zelber‐Sagi S., and Bergman R., “Risk Factors for Lethal Outcome in Patients With Pemphigus: A Retrospective Cohort Study,” European Journal of Dermatology 28, no. 1 (2018): 26–37. [DOI] [PubMed] [Google Scholar]
- 20. Rokni A. M., Ayasse M., Ahmed A., Guggina L., Kantor R. W., and Silverberg J. I., “Association of Autoimmune Blistering Disease, and Specifically, Pemphigus Vulgaris, With Cardiovascular Disease and Its Risk Factors: A Systematic Review and Meta‐Analysis,” Archives of Dermatological Research 315, no. 2 (2023): 207–213. [DOI] [PubMed] [Google Scholar]
- 21. Bonnesen K., Poulsen C. F. B., Schmidt S. A. J., Sørensen H. T., and Schmidt M., “Autoimmune Blistering Disorders and Cardiovascular Risks: A Population‐Based Cohort Study,” Journal of the American Academy of Dermatology 91, no. 1 (2024): 82–90. [DOI] [PubMed] [Google Scholar]
- 22. Kridin K., Hübner F., Recke A., Linder R., and Schmidt E., “The Burden of Neurological Comorbidities in Six Autoimmune Bullous Diseases: A Population‐Based Study,” Journal of the European Academy of Dermatology and Venereology 35, no. 10 (2021): 2074–2078. [DOI] [PubMed] [Google Scholar]
- 23. Ren Z., Narla S., Hsu D. Y., and Silverberg J. I., “Association of Serious Infections With Pemphigus and Pemphigoid: Analysis of the Nationwide Inpatient Sample,” Journal of the European Academy of Dermatology and Venereology 32, no. 10 (2018): 1768–1776. [DOI] [PubMed] [Google Scholar]
- 24. Leshem Y. A., Gdalevich M., Ziv M., David M., Hodak E., and Mimouni D., “Opportunistic Infections in Patients With Pemphigus,” Journal of the American Academy of Dermatology 71, no. 2 (2014): 284–292. [DOI] [PubMed] [Google Scholar]
- 25. Bai Y. X., Zhang L. M., Xiao T., and Chen H. D., “A 6‐Year Treatment Experience for Pemphigus: Retrospective Study of 69 Chinese Patients,” Dermatologic Therapy 29, no. 2 (2016): 84–87. [DOI] [PubMed] [Google Scholar]
- 26. Shamsadini S., Fekri A. R., Esfandiarpoor I., et al., “Determination of Survival and Hazard Functions for Pemphigus Patients in Kerman, a Southern Province of Iran,” International Journal of Dermatology 45, no. 6 (2006): 668–671. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information: S1 International Classification of Diseases (ICD‐10‐TM and ICD‐9‐CM) codes used to define comorbidities, complications, and in‐hospital procedures.
Figure S1: Age‐specific incidence of pemphigus vulgaris per 1 000 000 population using a stricter case definition (≥ 2 coded encounters), pooled across 2017–2024. Incidence increased with age, peaking in older adults with slight attenuation in those aged ≥ 80 years. Error bars indicate 95% confidence intervals.
Table S1: Incidence of pemphigus vulgaris in adults by sex and age group, 2017–2024.
Table S2: Age‐specific female‐to‐male incidence rate ratios of pemphigus vulgaris.
Table S3: Sensitivity analysis of predictors of severe in‐hospital outcomes among all hospitalizations for pemphigus using generalized estimating equations.
Table S4: Estimated adjusted odds ratios for severe in‐hospital outcomes by age.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.