Abstract
Patients with bullous pemphigoid are susceptible to serious infections, which are the leading cause of death in these patients. The aims of this population-based cohort study were to investigate the incidence and spectrum of serious infections in patients with bullous pemphigoid and to identify associated risk factors. The outcome measure was any infection requiring hospitalization. Hazard ratios with 95% confidence intervals were estimated using subdistribution hazard models. In total, 12,300 patients with bullous pemphigoid and 49,200 matched controls were identified through the National Health Insurance Research Database in Taiwan. Within 2 years of bullous pemphigoid diagnosis, 5,006 (40.7%) patients developed serious infections, with an incidence of 385.5/1,000 person-years. Patients with bullous pemphigoid were twice as likely to develop serious infections as controls (adjusted hazard ratio, 2.01; 95% confidence interval 1.92–2.10). Systemic corticosteroid use was the strongest risk factor, resulting in a 2-fold increase in the risk for serious infections. Other independent risk factors were advanced age, female sex, low income, and certain comorbidities. In conclusion, this study demonstrated an increased risk of serious infections following a diagnosis of bullous pemphigoid. Prophylaxis of serious infections through active intervention with the risk factors may be essential in reducing the morbidity and mortality associated with bullous pemphigoid.
SIGNIFICANCE
Infections were found to be common and severe following diagnosis of bullous pemphigoid. Patients with bullous pemphigoid were 3.3 times as likely to develop serious infections as those without bullous pemphigoid. Bullous pemphigoid per se doubled the risk of serious infections. Systemic corticosteroid use, the first-line therapy for bullous pemphigoid, was the strongest risk factor for serious infection in patients with bullous pemphigoid. Other independent risk factors associated with serious infections were advanced age, female sex, low income, and certain comorbidities. Prophylaxis of serious infections through active intervention with the risk factors might be essential in reducing the morbidity and mortality associated with bullous pemphigoid.
Key words: bullous pemphigoid, cohort studies, infections, pneumonia, sepsis, vesiculobullous skin disease
Bullous pemphigoid (BP), the most common autoimmune bullous disease (AIBD), typically manifests as generalized crops of itchy urticarial plaques and tense blisters in elderly patients. The annual incidence of BP has increased over the decades, especially in people aged over 60 years, which imposes a heavy burden on countries with an ageing population (1, 2). The overall 1-year mortality rate is 23.5% (95% confidence interval (95% CI) 20.2–26.8) and the standardized mortality ratio is estimated to be 3.61 (95% CI 2.59–5.03) in patients with BP (3, 4). Risk factors for mortality in patients with BP include advanced age, presence of circulating anti-BP180 autoantibodies, treatment with high-dose corticosteroids, and comorbidities such as dementia and stroke (5, 6).
Infections are common and severe following the onset of BP, accounting for 20–87% of deaths in patients with BP (7–11). Infection-specific standardized mortality ratio significantly increases in these patients, suggesting that infection is a mediator in the association between BP and death (11, 12). Investigating infectious complications is necessary for better preventive strategies, which may lead to a reduction in mortality associated with BP. However, only a few studies have evaluated the incidence and risk factors of serious infections in BP, most of which were single-centre studies limited by small sample sizes and a lack of matched controls (7, 8, 13, 14). Therefore, a nationwide retrospective cohort study was conducted, which aimed to evaluate: (i) the incidence of different categories of serious infections in patients with BP, (ii) the risk of serious infections in patients with BP compared with age- and sex-matched controls, and (iii) the risk factors for serious infections associated with BP by analysing sociodemographic characteristics, comorbidities, and treatment modalities.
MATERIALS AND METHODS
Data source
Data were retrieved from Taiwan’s National Health Insurance Research Database (NHIRD), which covers more than 99% of Taiwanese residents (approximately 23 million people) (15). These contain scrambled individual data (including sex, birth date, and insurance premium), inpatient and outpatient department diagnostic data, and data on drug dispensations from pharmacies and medical facilities. To protect the privacy of the participants, each was assigned a unique and anonymous identifier before the data was released to researchers. Participants were followed-up continuously using a unique identifier. The reliability of the NHIRD has been documented in several validation studies, and it has been widely used in epidemiological studies in Taiwan (16–20). This study was approved by the Institutional Review Board of the Taipei Veterans General Hospital (VGHIRB 2019-01-007AC); they granted a waiver of informed consent for retrospective studies utilizing data from the NHIRD.
Inclusion criteria
All individuals diagnosed with BP between 1 January 2007 and 31 December 2018 were identified. The diagnosis of BP in the NHIRD was based on the Ninth Edition of the International Classification of Diseases, Clinical Modification (ICD-9-CM) code 694.5 and ICD-10-CM codes L12.0, before and after 1 January 2016, respectively. To enhance diagnostic validity, only patients with at least 3 outpatient diagnoses or 1 inpatient diagnosis of BP from a board-certified dermatologist within any given 1-year period were enrolled into the BP cohort. The positive predictive value of BP diagnosis using this algorithm in the NHIRD is 98.0% (95% CI 96.3–99.7%) (18). Each patient with BP was randomly matched with 4 reference individuals (those without a diagnosis of BP during the study period) on age and sex, identified through the NHIRD. To reduce detection bias, more than 5 outpatient visits within 1 year prior to enrollment were required for matched individuals without BP.
Study outcome
The study outcome was serious infections, defined as any infection requiring hospitalization. The diagnosis of serious infections was identified through the major hospital discharge ICD-9-CM and/or ICD-10-CM codes (Table SI). Serious infections were categorized based on the specific organ system affected; infections leading to intensive care unit (ICU) admission were also identified.
Follow-up
Patients with BP were followed-up from the date of first BP diagnosis (index date). Reference individuals were followed-up from the date of BP diagnosis for the matched patients with BP. Since only 28.4% (3497/12300) of patients with BP remained in the cohort 2 years after the index date, the follow-up period was restricted to 2 years. Follow-up ended at the time of diagnosis of serious infections, death, 2 years after the index date, or on 31 December 2018, whichever occurred first. In the analysis of serious infections in specific organ systems, follow-up continued when serious infections occurred in other organ systems.
Covariates
Directed acyclic graphs were performed to identify confounders and mediators in each analysis, to prevent collider bias. The income-related monthly premium at cohort entry was representative of socioeconomic status. Chronic comorbidities that might have contributed to serious infections were identified, including diabetes mellitus (DM), heart failure, cerebrovascular disease (CVD), chronic obstructive pulmonary disease (COPD), chronic liver disease (CLD), chronic kidney disease (CKD), nephrotic syndrome, connective tissue disease (CTD), psoriasis, dementia, Parkinson’s disease, depression, and malignancies. Comorbidities were defined as at least 3 outpatient diagnoses or 1 inpatient diagnosis based on their corresponding ICD-9-CM and/or ICD-10-CM codes during the follow-up period (Table SII). Systemic immunosuppressant use for BP was assessed during the follow-up period. These immunosuppressants included corticosteroids, azathioprine, mycophenolate mofetil, methotrexate, cyclosporine, cyclophosphamide, and chlorambucil. The dose of systemic corticosteroids was calculated using the defined daily dose (DDD) assigned by the World Health Organization Collaborating Centre for Drug Statistics Methodology. The DDD of systemic corticosteroids is prednisolone 10 mg/day, or equivalent doses of other corticosteroid compounds.
Statistical analysis
Sociodemographic characteristics and comorbidities are expressed as mean with standard deviation or median with interquartile range (IQR) for continuous variables and count with percentage for categorical variables. The independent t-test and Wilcoxon rank-sum test were used to compare continuous variables between cohorts, while Pearson’s χ2 test was used for categorical variables. Cumulative incidences of serious infections were compared between cohorts using the Kaplan–Meier method and the modified log-rank test proposed by Gray (21). Incidence rates (IRs) of serious infections were calculated by dividing the number of study outcomes by total person-time at risk, presented as case number per 1,000 person-years. IR ratios (IRRs) with 95% CIs were computed as the ratio of IR between the cohorts.
Subdistribution hazard models were used to estimate hazard ratios (HRs) with 95% CIs of the association between BP and each category of serious infection, adjusting for the competing risk of death (crude model) (22). Covariates adjusted in the multivariable models were age, sex, monthly premium, annual number of outpatient visits, annual number of hospitalizations, and chronic comorbidities. A sensitivity analysis was performed by excluding study outcomes within the first 4 weeks of the index date to avoid the possible scenario that the serious infection preceded BP onset.
We also applied subdistribution hazard models to determine risk factors for serious infections in patients with BP. The risk factors analysed included age, sex, monthly premium, chronic comorbidities, and systemic immunosuppressant use within 12 weeks before the study endpoint. Two separate multivariable models were built for the 2 potential risk factors regarding systemic immunosuppressant use (i.e. systemic immunosuppressant regimens and systemic corticosteroid doses) due to presence of collinearity when they were included in the same regression model.
All data processing and statistical analyses were performed using Statistical Analysis Software, version 9.4 (SAS Institute, Cary, NC, USA). A 2-tailed p-value < 0.05 was considered statistically significant. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
RESULTS
Baseline characteristics
A total of 12,300 patients with BP and 49,200 controls were identified (Table I). Age (median 81 years) and sex (men 51.6%) were well matched between cohorts. Most comorbidities were significantly more common in patients with BP than in controls. The median follow-up duration was 0.9 (IQR 0.3–2.0) and 2.0 (IQR 0.9–2.0) years for patients with BP and controls (p < 0.001), respectively.
Table I.
Descriptive characteristics of the bullous pemphigoid (BP) and reference (patients without BP) cohorts
| BP cohort (n = 12,300) | Non-BP cohort (n = 49,200) | p-value | |
|---|---|---|---|
| Age, years, mean (SD)a | 78.2 (12.9) | 78.2 (12.9) | 0.92 |
| Median (IQR) | 81.4 (74.0–86.5) | 81.3 (73.7–86.8) | 0.93 |
| Sex, n (%) | |||
| Female | 5,956 (48.4) | 23,824 (48.4) | > 0.99 |
| Male | 6,344 (51.6) | 25,376 (51.6) | > 0.99 |
| Monthly premium, US dollars, mean (SD)a | 672.1 (820.7) | 754.0 (809.5) | < 0.001 |
| Median (IQR) | 730.0 (41.6–800.0) | 800.0 (41.6–800.0) | < 0.001 |
| Annual number of hospitalizations, mean (SD)b | 2.4 (3.6) | 0.5 (1.4) | < 0.001 |
| Median (IQR) | 0.9 (0.0–3.2) | 0.0 (0.0–0.4) | < 0.001 |
| Annual number of outpatient visits, mean (SD)b | 43.4 (34.9) | 25.2 (20.4) | < 0.001 |
| Median (IQR) | 37.8 (24.5–54.6) | 20.6 (12.3–32.7) | < 0.001 |
| Comorbidities, n (%) | |||
| Diabetes mellitus | 6,213 (50.5) | 18,170 (36.9) | < 0.001 |
| Heart failure | 3,518 (28.6) | 9,896 (20.1) | < 0.001 |
| CVD | 5,424 (44.1) | 8,439 (17.2) | < 0.001 |
| COPD | 3,552 (28.9) | 8,052 (16.4) | < 0.001 |
| CLD or cirrhosis | 2,026 (16.5) | 8,073 (16.4) | 0.88 |
| CKD or renal failure | 2,526 (20.5) | 8,867 (18.0) | < 0.001 |
| Nephrotic syndrome | 180 (1.5) | 477 (1.0) | < 0.001 |
| CTD or systemic vasculitis | 296 (2.4) | 1,450 (3.0) | < 0.001 |
| Psoriasis | 351 (2.9) | 549 (1.1) | < 0.001 |
| Dementia | 5,535 (45.0) | 10,354 (21.0) | < 0.001 |
| Parkinson’s disease | 1,895 (15.4) | 2,871 (5.8) | < 0.001 |
| Depression | 623 (5.1) | 1,998 (4.1) | < 0.001 |
| Malignancy | 1158 (9.4) | 6545 (13.3) | < 0.001 |
| Systemic immunosuppressants for BPc, n (%) | |||
| Corticosteroids +other immunosuppressants | 1,382 (11.2) | N/A | |
| Corticosteroids alone | 10,128 (82.3) | N/A | |
| Immunosuppressants other than corticosteroids | 15 (0.1) | N/A | |
| No systemic immunosuppressants | 775 (6.3) | N/A | |
| Follow-up duration, years, mean (SD)d | 1.1 (0.8) | 1.5 (0.7) | < 0.001 |
| Median (IQR) | 0.9 (0.3–2.0) | 2.0 (0.9–2.0) | < 0.001 |
| Serious infection, n (%) | 5,006 (40.7) | 7,418 (15.1) | < 0.001 |
| Mortality after index date, n (%) | 2,612 (21.2) | 2,984 (6.1) | < 0.001 |
At cohort entry.
During the follow-up period.
Systemic immunosuppressant use for bullous pemphigoid during the follow-up period; these immunosuppressants included corticosteroids, azathioprine, mycophenolate mofetil, methotrexate, cyclosporine, cyclophosphamide, and chlorambucil.
Follow-up for 2 years at most, ended at the time of serious infection or death or study end, whichever occurred first.
BP: bullous pemphigoid; CKD: chronic kidney disease; CLD: chronic liver disease; COPD: chronic obstructive pulmonary disease; CTD: connective tissue disease; CVD: cerebrovascular disease; N/A: not applicable; IQR: interquartile range; SD: standard deviation.
Incidence of serious infections
The cumulative incidence is shown in Fig. 1. The cumulative incidence of serious infections in patients with BP was significantly higher than that in controls (p < 0.001). A total of 5,006 (40.7%) patients with BP and 7,418 (15.1%) controls had serious infections during the follow-up period (p < 0.001) (Table II). The IR (per 1,000 person-years) of any serious infection was 385.5 (95% CI 375.0–396.4) for patients with BP and 102.7 (95% CI 100.3–105.0) for controls (IRR, 3.76; 95% CI 3.62–3.89). Serious infections occurred within the first year of the index date in 3,795 (75.8 %) patients with BP and 4,516 (60.9 %) controls. Common serious infections in patients with BP were lower respiratory tract infections (LRTIs), sepsis/septicaemia, urinary tract infections (UTIs), and skin and soft tissue infections (SSTIs), similar to the pattern in the reference cohort. The IRs of serious infections in specific organ systems were all higher in patients with BP, with the highest IRR of 4.77 (95% CI 4.35–5.24) for SSTIs.
Fig. 1.
Kaplan–Meier curves. Cumulative incidence of serious infections in bullous pemphigoid (BP) and reference cohorts (patients without BP). The difference between the cohorts was calculated using the modified log-rank test.
Table II.
Incidence rate of serious infections in the bullous pemphigoid (BP) and reference cohorts (patients without BP)
| Serious infections | BP cohort (n = 12,300) | Non-BP cohort (n = 49,200) | IRR (95% CI) | ||||
|---|---|---|---|---|---|---|---|
| Events, n | PYs | IRa (95% CI) | Events, n | PYs | IRa (95% CI) | ||
| Any | 5,006 | 12,984.7 | 385.5 (375.0–396.4) | 7,418 | 72,251.9 | 102.7 (100.3–105.0) | 3.76 (3.62–3.89) |
| Skin and soft tissue infections | 898 | 15,855.5 | 56.6 (53.1–60.5) | 910 | 76,721.5 | 11.9 (11.1–12.7) | 4.77 (4.35–5.24) |
| Musculoskeletal infections | 119 | 16,495.8 | 7.2 (6.0–8.6) | 141 | 77,322.6 | 1.8 (1.5–2.2) | 3.96 (3.09–5.05) |
| Upper respiratory tract infections | 64 | 16,541.5 | 3.9 (3.0–4.9) | 126 | 77,313.0 | 1.6 (1.4–1.9) | 2.37 (1.76–3.21) |
| Lower respiratory tract infections | 2,316 | 15,139.3 | 153.0 (146.9–159.3) | 3,655 | 75,108.4 | 48.7 (47.1–50.3) | 3.14 (2.98–3.31) |
| Heart infections | 46 | 16,564.5 | 2.8 (2.1–3.7) | 113 | 77,349.3 | 1.5 (1.2–1.8) | 1.90 (1.35–2.69) |
| Sepsis/septicaemia | 2,256 | 15,433.7 | 146.2 (140.3–152.3) | 2,409 | 76,288.8 | 31.6 (30.3–32.9) | 4.63 (4.37–4.90) |
| Central nervous system infections | 27 | 16,577.8 | 1.6 (1.1–2.4) | 49 | 77,397.5 | 0.6 (0.5–0.8) | 2.57 (1.61–4.12) |
| Gastrointestinal infections | 306 | 16,371.5 | 18.7 (16.7–20.9) | 660 | 76,956.1 | 8.6 (7.9–9.3) | 2.18 (1.90–2.49) |
| Urinary tract infections | 1,725 | 15,222.9 | 113.3 (108.1–118.8) | 2,609 | 75,520.7 | 34.5 (33.2–35.9) | 3.28 (3.09–3.49) |
| Infections leading to ICU admission | 979 | 16,053.4 | 61.0 (57.3–64.9) | 1,386 | 76,717.0 | 18.1 (17.1–19.0) | 3.38 (3.11–3.66) |
Incidence rate (IR) per 1000 person-years.
CI: confidence interval; ICU: intensive care unit; IRR: incidence rate ratio; PYs: person-years.
Association between bullous pemphigoid and risk of serious infections
The subdistribution hazard models showed significant associations between BP and the risk of any serious infection before (crude HR 3.27; 95% CI 3.15–3.38) and after (adjusted HR [aHR] 2.01; 95% CI 1.92–2.10) adjusting for potential confounders (Table III). A significant association was observed between BP and the risk of SSTIs (aHR 2.63; 95% CI 2.37–2.92), sepsis/septicaemia (aHR 2.49; 95% CI 2.32–2.67), musculoskeletal infections (aHR 1.86; 95% CI 1.40–2.47), LRTIs (aHR 1.75; 95% CI 1.65–1.87), UTIs (aHR 1.69; 95% CI 1.57–1.82), and gastrointestinal infections (aHR 1.25; 95% CI 1.06–1.47); no association was observed for upper respiratory tract, heart, and central nervous system infections. A significant association was also observed between BP and serious infections leading to ICU admission (aHR 2.06; 95% CI 1.86–2.29). Table SIII and SIV show the results of sensitivity analysis. The association between BP and any serious infection (aHR 2.06; 95% CI 1.97–2.15) on sensitivity analysis was similar to that observed on primary analysis.
Table III.
Risk of serious infection in patients with bullous pemphigoid vs those without bullous pemphigoid
| Serious infections | cHR (95% CI) | p-value | aHRa (95% CI) | p-value |
|---|---|---|---|---|
| Any | 3.27 (3.15–3.38) | < 0.001 | 2.01 (1.92–2.10) | < 0.001 |
| Skin and soft tissue infections | 3.86 (3.51–4.23) | < 0.001 | 2.63 (2.37–2.92) | < 0.001 |
| Musculoskeletal infections | 3.21 (2.51–4.11) | < 0.001 | 1.86 (1.40–2.47) | < 0.001 |
| Upper respiratory tract infections | 1.90 (1.41–2.57) | < 0.001 | 1.02 (0.72–1.46) | 0.90 |
| Lower respiratory tract infections | 2.57 (2.44–2.71) | < 0.001 | 1.75 (1.65–1.87) | < 0.001 |
| Heart infections | 1.53 (1.09–2.16) | 0.01 | 1.16 (0.79–1.69) | 0.44 |
| Sepsis/septicaemia | 3.85 (3.63–4.08) | < 0.001 | 2.49 (2.32–2.67) | < 0.001 |
| Central nervous system infections | 2.08 (1.30–3.32) | 0.002 | 1.43 (0.81–2.54) | 0.22 |
| Gastrointestinal infections | 1.75 (1.53–2.01) | < 0.001 | 1.25 (1.06–1.47) | 0.009 |
| Urinary tract infections | 2.64 (2.48–2.81) | < 0.001 | 1.69 (1.57–1.82) | < 0.001 |
| Infections leading to ICU admission | 2.76 (2.55–3.00) | < 0.001 | 2.06 (1.86–2.29) | < 0.001 |
Adjusted for age, sex, monthly premium, annual number of outpatient visits, annual number of hospitalizations, and comorbidities including diabetes mellitus, heart failure, cerebrovascular disease, chronic obstructive pulmonary disease, chronic hepatic disease or cirrhosis, chronic kidney disease or renal failure, nephrotic syndrome, connective tissue disease or systemic vasculitis, psoriasis, dementia, Parkinson’s disease, depression, and malignancies.
aHR: adjusted hazard ratio; BP: bullous pemphigoid; cHR: crude hazard ratio; 95% CI: 95% confidence interval; ICU: intensive care unit.
Risk factors for serious infections
In the analysis of risk factors for any serious infection in patients with BP, a modest association was observed for age, female sex, lower monthly premium, and comorbidities including DM, heart failure, CVD, COPD, CKD, dementia, and Parkinson’s disease (Table IV). Sex, nephrotic syndrome, CLD, CTD, psoriasis, depression, and malignancy were not independently associated with the risk of serious infections. The strongest association was observed with systemic immunosuppressant use within 12 weeks before the study endpoint. Compared with those who did not receive systemic immunosuppressants, patients who received systemic corticosteroids alone (aHR 1.96; 95% CI 1.83–2.10) or in combination with other immunosuppressants (aHR 1.80; 95% CI 1.60–2.03) had higher risks of developing serious infections; those who received systemic immunosuppressants other than corticosteroids did not have an increased risk of serious infections. The increased risk of serious infections among systemic corticosteroid users was dose-dependent with an aHR of 1.53 (95% CI 1.41–1.66) for DDD < 0.5 up to 2.33 (95% CI 2.16–2.52) for DDD ≥ 1.
Table IV.
Risk factors of any serious infection in patients with bullous pemphigoid (N = 14,048)
| Risk factors | Serious infections (n = 5,006) n (%) | No serious infections (n = 7,294) n (%) | cHR (95% CI) | p - value | aHRa (95% CI) | p - value | aHRb (95% CI) | p - value |
|---|---|---|---|---|---|---|---|---|
| Age, years | ||||||||
| ≥81 | 2,848 (56.9) | 3,568 (48.9) | 1.30 (1.22–1.37) | < 0.001 | 1.13 (1.07–1.19) | < 0.001 | 1.15 (1.09–1.22) | < 0.001 |
| < 81 | 2,158 (43.1) | 3,726 (51.1) | Ref | Ref | Ref | |||
| Sex | ||||||||
| Female | 2,451 (49.0) | 3,505 (48.1) | 1.03 (0.97–1.09) | 0.32 | 1.05 (1.00–1.11) | 0.04 | 1.06 (1.01–1.12) | 0.02 |
| Male | 2,555 (51.0) | 3,789 (51.9) | Ref | Ref | Ref | |||
| Monthly premium, USDs | ||||||||
| ≥730 | 2,364 (47.2) | 4,073 (55.8) | 0.78 (0.740.83) | < 0.001 | 0.83 (0.78–0.87) | < 0.001 | 0.82 (0.78–0.87) | < 0.001 |
| < 730 | 2,642 (52.8) | 3,221 (44.2) | ref | ref | ref | |||
| Diabetes mellitus | ||||||||
| Present | 2,700 (53.9) | 3,513 (48.2) | 1.22 (1.16–1.29) | < 0.001 | 1.16 (1.10–1.22) | < 0.001 | 1.16 (1.10–1.23) | < 0.001 |
| Absent | 2,306 (46.1) | 3,781 (51.8) | Ref | Ref | Ref | |||
| Heart failure | ||||||||
| Present | 1,619 (32.3) | 1,899 (26.0) | 1.24 (1.17–1.31) | < 0.001 | 1.07 (1.07–1.13) | 0.03 | 1.07 (1.01–1.13) | 0.02 |
| Absent | 3,387 (67.7) | 5,395 (74.0) | Ref | Ref | Ref | |||
| CVD | ||||||||
| Present | 2,460 (49.1) | 2,964 (40.6) | 1.29 (1.22–1.36) | < 0.001 | 1.18 (1.12–1.25) | < 0.001 | 1.19 (1.13–1.25) | < 0.001 |
| Absent | 2,546 (50.9) | 4,330 (59.4) | Ref | Ref | Ref | |||
| COPD | ||||||||
| Present | 1,655 (33.1) | 1,897 (26.0) | 1.26 (1.19–1.34) | < 0.001 | 1.13 (1.06–1.19) | < 0.001 | 1.11 (1.05–1.18) | < 0.001 |
| Absent | 3,351 (66.9) | 5,397 (74.0) | Ref | Ref | Ref | |||
| CLD or cirrhosis | ||||||||
| Present | 797 (15.9) | 1,229 (16.8) | 0.94 (0.87–1.02) | 0.12 | 0.96 (0.90–1.04) | 0.32 | 0.97 (0.90–1.04) | 0.36 |
| Absent | 4,209 (84.1) | 6,065 (83.2) | Ref | Ref | Ref | |||
| CKD or renal failure | ||||||||
| Present | 1,114 (22.3) | 1,412 (19.4) | 1.15 (1.07–1.22) | < 0.001 | 1.10 (1.03–1.17) | 0.003 | 1.12 (1.05–1.19) | < 0.001 |
| Absent | 3,892 (77.7) | 65,882 (80.6) | Ref | Ref | Ref | |||
| Nephrotic syndrome | ||||||||
| Present | 80 (1.6) | 100 (1.4) | 1.18 (0.95–1.47) | 0.14 | 1.15 (0.94–1.41) | 0.18 | 1.16 (0.94–1.42) | 0.16 |
| Absent | 4,926 (98.4) | 7,194 (98.6) | Ref | Ref | Ref | |||
| CTD or systemic vasculitis | ||||||||
| Present | 105 (2.1) | 191 (2.6) | 0.84 (0.69–1.02) | 0.08 | 1.00 (0.83–1.21) | 0.97 | 0.98 (0.82–1.18) | 0.84 |
| Absent | 4,901 (97.9) | 7,103 (97.4) | Ref | Ref | Ref | |||
| Psoriasis | ||||||||
| Present | 121 (2.4) | 230 (3.2) | 0.80 (0.67–0.96) | 0.02 | 0.88 (0.74–1.04) | 0.13 | 0.86 (0.73–1.02) | 0.08 |
| Absent | 4,885 (97.6) | 7,064 (96.8) | Ref | Ref | Ref | |||
| Dementia | ||||||||
| Present | 2,523 (50.4) | 3,012 (41.3) | 1.31 (1.24–1.38) | < 0.001 | 1.17 (1.10–1.24) | < 0.001 | 1.17 (1.11–1.24) | < 0.001 |
| Absent | 2,483 (49.6) | 4,282 (58.7) | Ref | Ref | Ref | |||
| Parkinson’s disease | ||||||||
| Present | 849 (17.0) | 1,046 (14.3) | 1.15 (1.07–1.24) | < 0.001 | 1.07 (1.00–1.15) | 0.05 | 1.07 (1.00–1.15) | 0.05 |
| Absent | 4,157 (83.0) | 6,248 (85.7) | Ref | Ref | Ref | |||
| Depression | ||||||||
| Present | 246 (4.9) | 377 (5.2) | 0.97 (0.85–1.10) | 0.60 | 0.95 (0.85–1.08) | 0.45 | 0.95 (0.85–1.08) | 0.45 |
| Absent | 4,760 (95.1) | 6,917 (94.8) | Ref | Ref | Ref | |||
| Malignancy | ||||||||
| Present | 442 (8.8) | 716 (9.8) | 0.92 (0.83–1.01) | 0.07 | 0.94 (0.86–1.03) | 0.20 | 0.95 (0.86–1.04) | 0.25 |
| Absent | 4,564 (91.2) | 6,578 (90.2) | Ref | Ref | Ref | |||
| Regimens of systemic SISsc | ||||||||
| SCSs +other SISs | 299 (6.0) | 4,30 (5.9) | 1.88 (1.66–2.12) | < 0.001 | 1.80 (1.60–2.03) | < 0.001 | ||
| SCSs alone | 3,666 (73.2) | 4,070 (55.8) | 2.21 (2.07–2.36) | < 0.001 | 1.96 (1.83–2.10) | < 0.001 | ||
| SISs other than SCSs | 17 (0.3) | 48 (0.7) | 0.97 (0.61–1.54) | 0.89 | 1.10 (0.70–1.73) | 0.67 | ||
| No SISsd | 1,024 (20.5) | 2,746 (37.6) | Ref | Ref | ||||
| Mean dose of SCSse | ||||||||
| ≥ 1 DDD/day | 1,990 (39.8) | 1,901 (26.1) | 2.63 (2.45–2.82) | < 0.001 | 2.33 (2.16–2.52) | < 0.001 | ||
| ≥0.5, < 1 DDD/day | 1,017 (20.3) | 1,126 (15.4) | 2.16 (1.99–2.35) | < 0.001 | 1.97 (1.81–2.14) | < 0.001 | ||
| < 0.5 DDD/day | 958 (19.1) | 1,473 (20.2) | 1.62 (1.49–1.76) | < 0.001 | 1.53 (1.41–1.66) | < 0.001 | ||
| No SCSs | 1,041 (20.8) | 2,794 (38.3) | Ref | Ref |
Adjusted for age, sex, monthly premium, comorbidities listed in Table I, and therapeutic regimens of systemic immunosuppressants, excluding the factor used for categorization in individual analyses.
Adjusted for age, sex, monthly premium, comorbidities listed in Table I, and mean dose of systemic corticosteroids, excluding the factor used for categorization in individual analyses.
Systemic immunosuppressant use for bullous pemphigoid within 12 weeks before the study endpoint; these immunosuppressants included corticosteroids, azathioprine, mycophenolate mofetil, methotrexate, cyclosporine, cyclophosphamide, and chlorambucil.
Among these patients, 75.4% (2,841/3,770) had prescription records of topical corticosteroids.
Mean dose of systemic corticosteroids within 12 weeks before the study endpoint; the defined daily dose is prednisolone 10 mg/day or equivalent doses of other corticosteroid compounds.
aHR: adjusted hazard ratio; BP: bullous pemphigoid; cHR: crude hazard ratio; 95% CI: 95% confidence interval; CKD: chronic kidney disease; CLD: chronic liver disease; COPD: chronic obstructive pulmonary disease; CTD: connective tissue disease; CVD: cerebrovascular disease; DDD: defined daily dose; SCSs: systemic corticosteroids; SISs: systemic immunosuppressants; USDs: US dollars.
DISCUSSION
This study showed that patients with BP were 2.0 times as likely to develop serious infections as those without BP after adjusting for sociodemographic factors and comorbidities. Systemic corticosteroid use, the first-line therapy for BP, was the strongest risk factor for serious infection in patients with BP, doubling their risk of serious infections compared with those who did not receive systemic immunosuppressant therapy.
These findings are consistent with previous retrospective studies in which patients with BP were commonly affected by serious infections (7, 8, 13, 14, 23). Single-centre longitudinal studies showed that infections requiring hospitalization or leading to death occurred in 44.4–55.7% of patients with BP; 77.7% of these infections appeared in the first year following BP diagnosis (7, 13). Identical to the current findings, the most common infections were pneumonia, UTIs, bacteraemia, and SSTIs (7,8). However, these studies did not enroll a control group for comparison; hence, the additional risk of serious infections contributed by BP could not be ascertained. A cross-sectional study using representative inpatient samples from the USA revealed significant association between BP and serious infections (adjusted odds ratio (aOR), 2.38; 95% CI 2.28–2.49) (23). Types of infections found to be significantly associated with BP in this previous study including SSTIs, musculoskeletal infections, pneumonia, UTIs, gastrointestinal infections, and septicaemia were similar to those observed in the current study. However, this previous study only enrolled inpatients for analysis and the directionality of the association could not be determined due to the cross-sectional design. Although infections are common and severe following the diagnosis of BP, routine antimicrobial prophylaxis is not recommended by updated guidelines (24, 25). Instead, clinicians should aim for early recognition of signs of infection and timely management with appropriate antimicrobials.
Susceptibility to infections in patients with BP is probably multifactorial. Recent exposure to systemic corticosteroids was the strongest risk factor identified in the current study, a finding supported by previous findings that treatment with systemic corticosteroids is associated with an increased risk of infectious complications in patients with BP (8, 13, 23). The current study found a dose-dependent relationship between the daily dose of corticosteroids and the risk of serious infections, suggesting that treatment with the lowest effective dose and early tapering of systemic corticosteroids are important in the management of BP. Notably, the risk even increased in those who received prednisolone equivalent doses of 5 mg or less daily. A previous study reported an additional risk of infection in patients with BP complicated by Cushing syndrome, implying that the duration of corticosteroid exposure might also play a role (23). However, the duration-dependent effect of systemic corticosteroids on the risk of serious infections was not easily explored in the current study because of survivor treatment selection bias that potentially occurs in survival analysis (26). In addition to systemic corticosteroids, whole-body application of high-potency topical corticosteroids is also considered first-line treatment for BP, since it is equally effective and could lead to lower morbidity and mortality (24, 27). The current findings add evidence to this perspective, since those who did not use systemic immunosuppressants in our BP cohort had a significantly lower risk of serious infections, and most were treated with topical corticosteroids.
The use of systemic immunosuppressants other than corticosteroids was not an independent predictor of serious infections in the current study. This finding is in agreement with our previous analysis in which BP patients treated with corticosteroid-sparing immunosuppressants for more than 3 consecutive months had a significantly lower mortality rate than those treated with systemic corticosteroids alone (17), since serious infection is among the leading causes of mortality in these patients (7–11). Furthermore, corticosteroid use is associated with an increased risk of severe bacterial infections in patients with rheumatoid arthritis vs those solely treated with corticosteroid-sparing immunosuppressants, many of which are also commonly used as adjuvant therapy for BP (28, 29). Besides, several non-immunosuppressive agents with anti-inflammatory properties (e.g. doxycycline, omalizumab, and dupilumab) have been reported to be as effective as systemic corticosteroids for treating BP (30–32). Given these findings, adding corticosteroid-sparing agents to the therapeutic regimen with early withdrawal of corticosteroids may be crucial in preventing serious infections in patients with BP.
Strength and limitations
The strengths of this study include analysis of population-based cohorts with detailed treatment information, which allowed for adequate statistical power and less susceptibility to both selection and attrition bias compared with previous single-centre studies. The study also has certain limitations. First, disease severity was unavailable in the NHIRD, and hence was not evaluated in the current study. However, since patients with severe BP tend to receive more intensive immunosuppressive therapy, it would have been difficult to separate the effects of disease severity and immunosuppressant use on the risk of serious infections. Second, our analyses focused on infections that were severe enough for hospitalization. Minor infections that could be managed in outpatient departments were not analysed. Lastly, although we controlled for the most important factors, potential confounders associated with infections, such as smoking, alcohol, obesity, and nutritional status, remained uncontrolled because of lack of relevant data in the NHIRD.
Conclusion
Patients with BP had an increased risk of developing serious infections requiring hospitalization. Systemic corticosteroid use was the most significant factor in predicting serious infections, while advanced age, female sex, low income, and comorbidities also played a role. These findings highlight the importance of adopting topical corticosteroids or corticosteroid-sparing therapies in the treatment of BP, especially for patients with multiple risk factors for serious infection.
Supplementary Material
ACKNOWLEDGEMENTS
This study is funded by grants from the Ministry of Science and Technology, Taiwan, R.O.C. (MOST 111-2314-B-075-012-MY3) and Taipei Veterans General Hospital (V111C-022).
Institutional Review Board approval status: This study was reviewed and approved by the Institutional Review Board of the Taipei Veterans General Hospital (IRB number: VGHIRB 2019-01-007AC). Patient consent was waived by the review board because of the retrospective nature of data collection from the National Health Insurance Research Database.
Footnotes
The authors have no conflicts of interest to declare.
REFERENCES
- 1.Persson MSM, Harman KE, Vinogradova Y, Langan SM, Hippisley-Cox J, Thomas KS, et al. Incidence, prevalence and mortality of bullous pemphigoid in England 1998–2017: a population-based cohort study. Br J Dermatol 2021; 184: 68–77. [DOI] [PubMed] [Google Scholar]
- 2.Kridin K, Ludwig RJ. The growing incidence of bullous pemphigoid: overview and potential explanations. Front Med (Lausanne) 2018; 5: 220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kridin K, Schwartz N, Cohen AD, Zelber-Sagi S. Mortality in bullous pemphigoid: a systematic review and meta-analysis of standardized mortality ratios. J Dermatol 2018; 45: 1094–1100. [DOI] [PubMed] [Google Scholar]
- 4.Kridin K, Shihade W, Bergman R. Mortality in patients with bullous pemphigoid: a retrospective cohort study, systematic review and meta-analysis. Acta Derm Venereol 2019; 99: 72–77. [DOI] [PubMed] [Google Scholar]
- 5.Liu YD, Wang YH, Ye YC, Zhao WL, Li L. Prognostic factors for mortality in patients with bullous pemphigoid: a meta-analysis. Arch Dermatol Res 2017; 309: 335–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rzany B, Partscht K, Jung M, Kippes W, Mecking D, Baima B, et al. Risk factors for lethal outcome in patients with bullous pemphigoid: low serum albumin level, high dosage of glucocorticosteroids, and old age. Arch Dermatol 2002; 138: 903–908. [DOI] [PubMed] [Google Scholar]
- 7.Phoon YW, Fook-Chong SM, Koh HY, Thirumoorthy T, Pang SM, Lee HY. Infectious complications in bullous pemphigoid: an analysis of risk factors. J Am Acad Dermatol 2015; 72: 834–839. [DOI] [PubMed] [Google Scholar]
- 8.Chen J, Mao X, Zhao W, Zhang B, Chen X, Yu C, et al. Assessment of the characteristics and associated factors of infectious complications in bullous pemphigoid. Front Immunol 2020; 11: 1607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cai SC, Allen JC, Lim YL, Chua SH, Tan SH, Tang MB. Mortality of bullous pemphigoid in Singapore: risk factors and causes of death in 359 patients seen at the National Skin Centre. Br J Dermatol 2014; 170: 1319–1326. [DOI] [PubMed] [Google Scholar]
- 10.Joly P, Benichou J, Lok C, Hellot MF, Saiag P, Tancrede-Bohin E, et al. Prediction of survival for patients with bullous pemphigoid: a prospective study. Arch Dermatol 2005; 141: 691–698. [DOI] [PubMed] [Google Scholar]
- 11.Roujeau J-C, Lok C, Bastuji-Garin S, Mhalla S, Enginger V, Bernard P. High risk of death in elderly patients with extensive bullous pemphigoid. Arch Dermatol 1998; 134: 465–469. [DOI] [PubMed] [Google Scholar]
- 12.Cortés B, Marazza G, Naldi L, Combescure C, Borradori L. Mortality of bullous pemphigoid in Switzerland: a prospective study. Br J Dermatol 2011; 165: 368–374. [DOI] [PubMed] [Google Scholar]
- 13.Lehman JS, Khunger M, Lohse CM. Infection in autoimmune bullous diseases: a retrospective comparative study. J Dermatol 2013; 40: 613–619. [DOI] [PubMed] [Google Scholar]
- 14.Langan SM, Hubbard R, Fleming K, West J. A population-based study of acute medical conditions associated with bullous pemphigoid. Br J Dermatol 2009; 161: 1149–1152. [DOI] [PubMed] [Google Scholar]
- 15.Hsieh CY, Su CC, Shao SC, Sung SF, Lin SJ, Kao Yang Y-H, et al. Taiwan’s National Health Insurance Research Database: past and future. Clin Epidemiol 2019; 11: 349–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wu CY, Wu CY, Lin YH, Chang YT. Statins did not reduce the mortality risk in patients with bullous pemphigoid: a population-based cohort study. Dermatol Sin 2021; 39: 153–154. [Google Scholar]
- 17.Wu CY, Wu CY, Li CP, Lin YH, Chang YT. Association of immunosuppressants with mortality of patients with bullous pemphigoid: a nationwide population-based cohort study. Dermatology 2022; 238: 378–385. [DOI] [PubMed] [Google Scholar]
- 18.Wu CY, Wu CY, Li CP, Chou YJ, Lin YH, Chang YT. Association between dipeptidyl peptidase-4 inhibitors and risk of bullous pemphigoid in patients with type 2 diabetes: a population-based cohort study. Diabetes Res Clin Pract 2021; 171: 108546. [DOI] [PubMed] [Google Scholar]
- 19.Dai YX, Tai YH, Chen CC, Chang YT, Chen TJ, Chen MH. Bidirectional association between alopecia areata and major depressive disorder among probands and unaffected siblings: a nationwide population-based study. J Am Acad Dermatol 2020; 82: 1131–1137. [DOI] [PubMed] [Google Scholar]
- 20.Dai YX, Tai YH, Chang YT, Chen TJ, Chen MH. Increased risk of alopecia areata among patients with endometriosis: a longitudinal study in Taiwan. Dermatol Sin 2021; 39: 41–44. [Google Scholar]
- 21.Robert JG. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 1988; 16: 1141–1154. [Google Scholar]
- 22.Fine JP, Gray RJ. A Proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509. [Google Scholar]
- 23.Ren Z, Narla S, Hsu DY, Silverberg JI. Association of serious infections with pemphigus and pemphigoid: analysis of the Nationwide Inpatient Sample. J Eur Acad Dermatol Venereol 2018; 32: 1768–1776. [DOI] [PubMed] [Google Scholar]
- 24.Borradori L, Van Beek N, Feliciani C, Tedbirt B, Antiga E, Bergman R, et al. Updated S2 K guidelines for the management of bullous pemphigoid initiated by the European Academy of Dermatology and Venereology (EADV). J Eur Acad Dermatol Venereol 2022; 36: 1689–1704. [DOI] [PubMed] [Google Scholar]
- 25.Venning VA, Taghipour K, Mohd Mustapa MF, Highet AS, Kirtschig G. British Association of Dermatologists’ guidelines for the management of bullous pemphigoid 2012. Br J Dermatol 2012; 167: 1200–1214. [DOI] [PubMed] [Google Scholar]
- 26.Glesby MJ, Hoover DR. Survivor treatment selection bias in observational studies: examples from the AIDS literature. Ann Intern Med 1996; 124: 999–1005. [DOI] [PubMed] [Google Scholar]
- 27.Joly P, Roujeau JC, Benichou J, Picard C, Dreno B, Delaporte E, et al. A comparison of oral and topical corticosteroids in patients with bullous pemphigoid. N Engl J Med 2002; 346: 321–327. [DOI] [PubMed] [Google Scholar]
- 28.Widdifield J, Bernatsky S, Paterson JM, Gunraj N, Thorne JC, Pope J, et al. Serious infections in a population-based cohort of 86,039 seniors with rheumatoid arthritis. Arthritis Care Res (Hoboken) 2013; 65: 353–361. [DOI] [PubMed] [Google Scholar]
- 29.Schneeweiss S, Setoguchi S, Weinblatt ME, Katz JN, Avorn J, Sax PE, et al. Anti-tumor necrosis factor alpha therapy and the risk of serious bacterial infections in elderly patients with rheumatoid arthritis. Arthritis Rheum 2007; 56: 1754–1764. [DOI] [PubMed] [Google Scholar]
- 30.Williams HC, Wojnarowska F, Kirtschig G, Mason J, Godec TR, Schmidt E, et al. Doxycycline versus prednisolone as an initial treatment strategy for bullous pemphigoid: a pragmatic, non-inferiority, randomised controlled trial. Lancet 2017; 389: 1630–1638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Abdat R, Waldman RA, de Bedout V, Czernik A, McLeod M, King B, et al. Dupilumab as a novel therapy for bullous pemphigoid: a multicenter case series. J Am Acad Dermatol 2020; 83: 46–52. [DOI] [PubMed] [Google Scholar]
- 32.Balakirski G, Alkhateeb A, Merk HF, Leverkus M, Megahed M. Successful treatment of bullous pemphigoid with omalizumab as corticosteroid-sparing agent: report of two cases and review of literature. J Eur Acad Dermatol Venereol 2016; 30: 1778–1782. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.