Mortality in dermatology: A retrospective study of cause and course

INTRODUCTION

Dermatological disorders contribute to the global health burden, ranking as the fourth leading cause of non-fatal diseases worldwide.¹ Although often perceived as not being life-threatening, certain dermatological conditions such as severe cutaneous adverse drug reactions, toxic epidermal necrolysis (TEN), drug hypersensitivity syndrome (DHS) and autoimmune bullous disorders have the potential to cause mortality. However, the available literature focuses on disease-specific mortality, resulting in a knowledge gap regarding the overall mortality in dermatology and the aetio-pathological profile of sepsis and complications in such patients.

We investigated the specific dermatological disorders that lead to inpatient mortality.

METHODS

After obtaining approval from our institution’s ethics committee, we conducted a record review of all inpatients admitted over the past 5 years (2018–22) for dermatological disorders with systemic complications leading to death.

Demographic details, history of illness, disease progression, investigations, treatment details, and patient response to the same during the hospital stay were noted. The data regarding the cause of death were obtained from the death certificates as recorded by the physician or dermatologist at the time of death.

RESULTS

Of the 2874 patients admitted for primarily dermatological diseases during the study period, 13 died, resulting in a mortality rate of 0.45%. All patients, except two, were adults, with a mean (SD) age of 47 (24.8) years. The youngest patient was 15 years old and the oldest was 90 years old. Most of the patients belonged to the age group of 41–60 years (n=5), followed by 21–40 (n=3), 41–60 (n=2), 61–90 (n=2) and 0–20 years (n=1). Of the 13 patients, 8 were females, with a male-to-female ratio of 1:1.6.

The most common dermatological disease responsible for mortality was severe cutaneous adverse drug reactions (6; 46.2%), followed by autoimmune bullous disorders (3, 23.1%). Other conditions, including psoriasis, eczema, cytomegalovirus infections with toxic shock syndrome (CMV+TSS), and haemophago- cytic lymphohistiocytosis with neutrophilic dermatosis and toxic shock syndrome (HLH+TSS+neutrophilic dermatosis), accounted for one each (Table 1). The mean (SD) duration of all chronic dermatological conditions, excluding adverse drug reactions (TEN and DHS), before hospitalization was 17.42 (7.99) days.

DHS accounted for most of the severe cutaneous adverse drug reactions (4, 30.8%), followed by TEN (2, 15.4%). The most common implicated drugs were dapsone and antituberculous drugs (2 each, 15.4%), followed by leflunomide (1, 7.7%). The causative drug was unknown in 1 patient with TEN. The mean (SD) duration of the drug rash before hospitalization was 7.16 (4.66) days.

Among the immunobullous disorders, pemphigus vulgaris (2, 15.4%) was the most common cause of mortality, followed by bullous pemphigoid (1, 7.7%).

A mean body surface area of 65.7 (27.6)% was involved in all the patients. Six patients (46.2%) were in erythroderma at the time of presentation. Amongst these, 4 patients had drug rash and 1 patient each had psoriasis and eczema.

Seven patients had concomitant infections at the time of hospitalization––2 patients each had Herpes simplex virus, urinary tract and dengue infections while 1 had pulmonary tuberculosis and fungal pneumonia. Comorbid conditions such as diabetes, hypertension, hypothyroidism, rheumatoid arthritis, ischaemic heart disease, chronic kidney disease, non-Hodgkin lymphoma and Parkinson disease were seen in 9 patients.

The mean (SD) duration of hospitalization for all patients was 10.5 (6.6) days. Ten patients received treatment in the intensive care unit (ICU), of which 4 patients were admitted directly to the ICU. Of the six ICU transfers, 4 occurred within 72 hours of admission and 2 beyond 72 hours.

DISCUSSION

Our study revealed a mortality rate of 0.45%. This is much lower than the mortality reported in India by Gupta et al.³ (1.7%), Bhandare et al.⁴ (1.4%), Nair et al.⁵ (3.58%) and Chowdhury et al.⁶ (8.4%) and internationally by Saber et al.⁷ in Iran (2.24%), de Paula Samorano-Lima et al.⁸ in Brazil (2.5%), JournetTollhupp et al.⁹ in France (3.7%–5.3%), Keita et al.¹⁰ in Guinea (7.9%) and Ecra et al.¹¹ in sub-Saharan Africa (10.26%). The highest mortality rate was reported in the UK study (13.9%), which contrasts starkly with our study.¹² The favourable outcome in our study could be attributed to strict barrier nursing and prompt transfer to the ICU in the event of any impending signs of sepsis, as well as multidisciplinary management of these patients by dermatologists, physicians and intensivists.

Typically, advanced age is a risk factor contributing to mortality in diseases such as pemphigus, TEN and malignancies. Most Indian and international studies have reported a higher mortality rate in the 6th–7th decade.^4–7 Surprisingly, the highest number of deaths in our study occurred in the 4th–6th decade, which is similar to the results by Ecra et al.,¹¹ where the mean age was 43.1 years. The mean duration of hospitalization in our study was comparable to the findings of Saber et al.⁷ (12.39 days) and Keita et al. (13 days).¹⁰

Extensive skin involvement and acute skin failure as seen in adverse drug reactions and autoimmune bullous disorders, respectively, led to rapid clinical deterioration in our study, accounting for the top two causes of mortality. This aligns with other studies in the literature, which reported immunobullous disorders and cutaneous adverse drug reactions as the most common cause of mortality (Table 3).

Among the cutaneous adverse drug reactions, the majority of deaths in our study were attributed to DHS. The mortality rate amongst these patients has been reported to be around 10% in the literature.¹³ Most common drugs known to cause DHS are anticonvulsants and sulphonamides. Although rare, leflunomide has been known to cause severe cutaneous hypersensitivity reactions such as drug reaction with eosinophilia and systemic symptoms and Stevens–Johnson syndrome/TEN.¹⁴

Both the TEN patients in our study had extensive skin denudation (>90%), mucosal involvement, and a high baseline SCORTEN (SCORe of Toxic Epidermal Necrosis) score, which is consistent with previous studies.^5,6 Despite early initiation of IVIG and cyclosporine, these patients experienced rapid progression. Lower mortality rates of 9.52% and 10% were reported by Nizamoglu et al.¹⁵ and McCullough et al.,¹⁶ respectively, where TEN patients were treated in specialized burn wards. Early transportation to a burns unit can improve outcomes in SJS and TEN, as demonstrated in the multicentric study by Palmieri et al.¹⁷

Mortality rates in pemphigus, reported in the literature, range from 5% to 20%.^18,19 The lower mortality rate amongst patients with bullous pemphigoid in our study (7.7%) as compared to the study by Gupta¹² (38.4%) could be attributed to early initiation of systemic steroids and supportive therapy.

Cutaneous malignancies, particularly squamous cell carcinoma, have been identified as a leading cause of mortality in Iran,⁷ UK,¹² France⁹ and Guinea.¹⁰ Kaposi sarcoma was the most fatal malignancy in a sub-Saharan African study.¹¹ We did not record any deaths due to cutaneous malignancies. In India, the incidence of skin cancer is lower (0.04%–6.2% among males and 0.5%–4.8% in females) compared to other regions globally,²⁰ potentially explaining the differences in mortality rates between Indian and international studies.

Septic shock was the most common immediate cause of death in our study, consistent with the findings in the literature (Table 3).

This underscores the critical importance of infection control protocols and measures, such as barrier nursing, as acute skin failure and immunosuppression in these patients predispose them to infections and sepsis. Furthermore, antimicrobial stewardship and emphasis on preventing nosocomial infections are equally important. While not necessarily life-threatening, factors that contribute to morbidity and mortality in dermatology patients include extensive skin loss leading to loss of barrier function and concomitant immunosuppression.

Our study is limited by its small smaple size. In addition, the absence of a control population resulted in an inability to compute disease-specific mortality rates and draw statistically significant comparisons. Identifying factors that influence mortality, early detection, prompt initiation of treatment and a multidisciplinary approach can improve outcomes in these patients.