Important prognostic factors include age, extent of epidermal detachment, serum glucose and creatinine levels, neutropaenia and lymphopaenia and serum urea level [2]. Pulmonary involvement, characterised by dyspnoea, haemoptysis and desaturation, is associated with the need for mechanical ventilation and a poor prognosis [3].

SCORTEN is a prognostic tool, developed in 2000, to predict the mortality of a patient with TEN [4]. Severity of illness is assessed by various parameters including age >40 years old, presence of a malignancy, tachycardia (heart rate >120 beats/min), epidermal detachment >10 % of total BSA as well as biochemical markers (serum urea, bicarbonate and glucose). A score, ranging from 0 to 7, is calculated within 24 h of admission and repeated on day 3 [5, 6]. The higher the scores, the worse the prognosis. A SCORTEN score of 5 or more is associated with 90 % mortality.

Early death due to sepsis is common, typically triggered by Pseudomonas aeruginosa and Staphylococcus aureus [2]. Death can also be due to dehydration and metabolic abnormalities, multiorgan failure, pulmonary embolus and gastrointestinal bleeding [5].

Cutaneous lesions take days to start healing and eventually heal within 2 weeks, while mucosal lesions persist longer and on the glans penis may last up to 2 months [2, 5]. Wounds in trauma or pressure-prone locations such as the back and buttock take longer to heal [2]. Scarring with hypopigmentation is more common than hyperpigmentation, though the latter can be treated with intense pulse light therapy [5, 7].

Avoiding long-term complications after TEN is very unlikely. Ocular involvement can include conjunctival injection, symblepharon, pseudomembrane formation, photophobia, keratitis and corneal erosions. Chronic photosensitivity and dry and irritative eyes are common complaints [8]. Genital complications include stenoses, strictures and adhesions occurring in both genders [2]. Psychiatric disorders (e.g. post-traumatic stress syndrome) and buccal, dental and genital lesions have also been reported.

Prevention and early detection of SJS and TEN is the key. Once the diagnosis is suspected, early withdrawal of the offending agent is imperative and has been shown to reduce mortality [9]. Patients should be transferred early to specialised centres such as intensive care units, burns units or tertiary hospitals which have the facilities and skills and resources to manage such complex patients [10]. A burns centre, however, is not the only facility that can provide for the needs of a patient with TEN. A burns patient is also different to a patient with TEN. The aetiology, histology and clinical requirements differ. TEN affects the epidermis requiring less fluid resuscitation than burns patients. Intravenous fluid administration should be titrated against clinical signs of dehydration, vital signs (especially blood pressure), urine output and blood parameters [11].

Specialised care with intense nursing and medical expertise comes at a cost to the health system and the individual. The financial impact on the individual and the hospital system should not be underestimated, given the utilisation of resources such as intensive care units, burns units and expensive therapies such as IVIg and dressings. Long-term complications, rehospitalisations and use of more expensive alternative drugs (e.g. a different class of antibiotics) all contribute to the significant health burden caused by SJS/ TEN.

Because SJS and TEN are uncommon, there is a paucity of RCTs and hence lack of evidence supporting the proposed treatment regimes. A Cochrane review in 2002 [12] found only one RCT which compared thalidomide to placebo in 22 patients. This study was discontinued as the mortality rate was higher with thalidomide compared to placebo (83 % vs. 30 %). It is thought that thalidomide may increase the TNF-a production and make the TEN worse [13]. This is the only RCT published to date. Given the low incidence and findings of only one RCT, it is clear that therapies directed at acute and long-term management of SJS and TEN are not based on level 1 evidence but rather on case series and expert opinion. When assessing the evidence, it is important to keep in mind that common therapies such as corticosteroids, immunosuppressive, antibiotics and biologics are not supported by RCTs.

Once the culprit drug has been removed and the patient transferred to a specialised centre, special attention should be made to eye care with an early ophthalmological review. Supportive management with adequate fluid hydration and nutritional support is essential. General measures include measuring the extent of epidermal detachment as a percentage of BSA on a daily basis [5], blood tests including blood cultures and urine cultures assessing for treatable infections and prescribing an alternative medication to the offending agent.

Antibiotics should be used judiciously when signs of sepsis are evident. Prophylactic use of antibiotics may confer a poorer prognosis and produce resistant organisms [14]. Sources of infection such as indwelling catheters and lines should be minimised and avoided if possible [15]. Dressings are a contentious issue. Some advocate necrotic tissue debridement followed by artificial or biologic (e.g. porcine xenograft or human skin allograft) dressings [5], though it is questionable whether this approach is necessary. The alternative is to leave the detached epidermis in situ as a biological dressing. A recent paper described two patients with TEN who were managed successfully in the high dependency unit of a tertiary hospital without transfer to a burns unit. The bullae were dehisced with a sterile needle and the epidermis used as a protective covering for the dermis underneath. Denuded skin was covered with nonstick dressings and the skin was kept clean and free from abrasives by using a thermal water spray. Both patients remained conscious throughout their admission and made a speedy recovery [16].

Despite the use of various treatments in the early stages of SJS/TEN to stop progression, no one therapy has been shown to be consistently effective. Systemic corticosteroids in SJS/TEN are also controversial. The proposed beneficial immunosuppressive and anti-inflammatory effects are counteracted by the risks of sepsis and poor wound healing. Early administration of corticosteroids should be avoided as it is associated with longer hospital stay and increased mortality [5, 17, 18]. Pulse therapy, however, seems promising in arresting disease progression and preventing ocular complications, though further studies are required to confirm this [19, 20].

Since 1988, IVIg has been used in increasing frequency in the acute management of TEN. It is thought IVIg contains anti-Fas ligand antibodies which may, in theory, stop the progression of disease by blocking the Fas receptor and hence preventing keratinocyte apoptosis [21, 22]. Improvements with IVIg in case reports have never been supported by RCTs. A retrospective analysis shows no clear benefit of using IVIg or corticosteroids [23]. Its beneficial effect which was shown in earlier studies may have been overestimated due to publication bias [24], and a meta-analysis found that its effects were not as favourable as initially thought, though there may be a role for high-dose IVIg [25]. High-dose IVIg has a nonstatistical beneficial effect in TEN compared to low-dose IVIg. Low-dose IVIg is no better than placebo.

Cyclosporine has also been used successfully in SJS/TEN. It inhibits CD8 T-cell activation and reduces the release of granulysin, granzyme and perforin [22]. Cyclosporine has been used with granulocyte colony-stimulating factor in a neutropenic patient [26] and also in TEN which did not respond to corticosteroids [27]. Nonrandomised trials and case series have reported lower than expected mortality rates and epidermal detachment [28, 29].

Case reports have indicated that plasmapheresis may have a beneficial effect possibly by eliminating the offending drug and removing cytokines such as Fas ligands [30–33]. However, plasmapheresis is invasive, time consuming and has not been shown to be superior to normal treatment [34].

Though cyclophosphamide has been reported to be effective in arresting the disease [35, 36], a comparison between cyclophosphamide and cyclosporine shows that cyclosporine results in a shorter time to stopping progression and re-epithelisation [37].

Biologic agents have also been reported as having some benefit. Promising as they seem, biologics need to be tested in a RCT setting before they are used as established therapeutic options for SJS and TEN. Lee et al. report a case of a 32-year-old man with HIV who presented with SJS initially which progressed into TEN after taking amoxicillin and mefenamic acid. One dose of intravenous methylprednisolone was administered but ceased due to concerns regarding immunosuppression and infections. Etanercept was administered at a dose of 50 and 25 mg on days 3 and 5, respectively. Healing started to occur at day 8 of admission and the patient had total mucosal resolution by day 14 of admission. This patient’s CD4 count also increased after the etanercept from 85 to 202 cells/microL [38].