Skin disease case mix in the early 1920s: diagnoses prevalent at the time of the founding of the British Association of Dermatologists
Dear Editor, Dermatology is a ceaselessly changing specialty, no more so than currently, so it is unsurprising that the case mix of skin diseases presenting to a dermatologist a century ago, when our association was founded, should be different from today’s. For one thing, the overall healthcare context has changed. Life expectancy in 1920 at birth for a man was 55·6 years, and for a woman 59·6 years.1 Causes of death had transitioned from the nineteenth-century pattern of being overwhelmingly due to infectious disease, to a predominance of respiratory and cardiovascular disease, with contribution from cancer.2 Access to secondary care was variable and often patchy in the interwar years, prior to the foundation of the National Health Service in 1948. Secondary care was provided by voluntary hospitals, often supported by charities and endowments, and by some municipal hospitals, but was also provided by private specialists.3, 4 For people of restricted means, options were often limited.
Details of new cases primarily referred by general practitioners to the outpatient clinics of the Royal Infirmary, Edinburgh (a voluntary hospital with substantial endowments) and at the Edinburgh Skin Dispensary (a voluntary and charitable organization) were collated from the disease index record cards for the years 1921 and 1922. Out of 3915 cases recorded, eczema, tinea, impetigo, scabies and ‘seborrhoea’ accounted for two-thirds of all those seen (Table 1). Syphilis and tuberculosis of the skin made up 2% each. There were only 37 cases of ‘rodent ulcer’ (basal cell carcinoma) and nine of ‘carcinoma cutis’ (presumed squamous cell carcinoma), and no cases of malignant melanoma.
| Study | McCall Anderson5 | Radcliffe Crocker6 | Royal Infirmary & Dispensary | Hall7 | Harris8 | Svensson9 | |
|---|---|---|---|---|---|---|---|
| Location | 1873, Glasgow | 1903, London | 1921–22, Edinburgh | 1954–6, Northern Ireland | 1988, Edinburgh | 2017, Europe | |
| Type of study | Hospital practice | Hospital clinic | Private clinic | Public clinics | Hospital clinics | Hospital clinics | Point prevalence |
| Cases | 10 000 | 10 000 | 5000 | 3915 | 29 553 | 1592 | 12 377 |
| Eczema | 25 | 27 | 20 | 17 | 15 | 13 | 16 |
| Tinea | 6 | 13 | 3 | 15 | 3·5 | ||
| Impetigo | 0·1 | 10 | 1 | 13 | 4 | ||
| Scabies | 25 | 8 | 1·1 | 11 | 1 | ||
| Seborrhoea | 0·1 | 1 | 6 | 11 | 6 | ||
| Psoriasis | 7 | 7 | 6 | 5 | 5 | 4 | 3 |
| Alopecia | 0·6 | 3 | 16 | 4 | 2·5 | ||
| Acne/rosacea | 3 | 5 | 13 | 3 | 5 | 3 | 5 |
| Syphilis | 5 | 6 | 3 | 2 | |||
| Tuberculosis | 5 | 2 | 1·1 | 2 | |||
| Pediculosis | 3 | 4 | 0·4 | 2 | 0·5 | ||
| Urticaria | 1·5 | 4 | 4 | 1 | 1·5 | 1·7 | |
| Skin cancer | 0·4 | 0·2 | 1·2 | 1 | 3 | 8 | 0·5 |
| Leg ulcer | 4 | 0·3 | 2 | 0·2 | |||
| Benign tumours | 0·3 | 0·3 | 1·3 | 1·5 | 40 | ||
| Viral warts | 0·3 | 0 | 0·9 | 1·3 | 18 | 14 | 7 |
Comparisons can be made with case mix studies both before and after this period, accepting that there will be inaccuracies from differences in nosological preference and channelled populations. Eczema was the predominant diagnosis made by both McCall Anderson in Glasgow in 1873, and by Radcliffe Crocker in London in 1903, accounting for about a quarter of all cases in both instances.5, 6 If ‘seborrhoea’, from 1920s descriptions clearly now regarded as seborrhoeic dermatitis, is included, then eczemas in 1920s Edinburgh had a similar overall prevalence to that found by McCall Anderson and Radcliffe Crocker.5, 6 The 11% frequency of seborrhoeic dermatitis in 1921–22 Edinburgh figures is about double that seen by Hall and Burrows from 1950s Northern Ireland, and higher than clinical experience suggests for the present day.7 The reasons are unclear but it might be explained by the greater frequency of washing and bathing as a consequence of more baths being installed in housing after the 1920s.
In the Edinburgh survey of 1988, 62% of referrals were of benign and malignant skin tumours or viral warts, diagnoses that accounted for only 3·8% of cases in 1921–22.8 With this in mind, it is evident that psoriasis and the eczemas accounted for a higher proportion of the remaining referrals in the later survey, namely 10% and 34%, respectively in 1988, compared with 5% and 28% (if seborrhoea is included) or 18% (for eczema excluding seborrhoea).8 The differences might indicate a lower threshold for referral or an actual increase in disease prevalence.
The apparent surge in impetigo and tinea in the first quarter of the twentieth century, as judged by Radcliffe Crocker’s hospital data and the 1921–22 Edinburgh figures, in comparison with McCall Anderson’s 1873 observations, is possibly from close contact of school children.5, 6 It is notable that Radcliffe Crocker’s 1903 private patients had lower prevalences of infections and infestations, suggesting a social class difference.6 By the mid-twentieth century, effective treatments had largely conquered such contagions irrespectively of social class.
The biggest surprise for the present-day dermatologist must be the very low rate of skin cancers of any description in the 1920s, something that was still apparent in 1950s Northern Ireland.7 Accepting the evidence that skin cancer incidence increased in the last quarter of the twentieth century, an additional explanation for the low prevalence of skin cancers may be that in the 1920s presentations were delayed and, because life expectancies, as mentioned, were lower, those affected may have died from other causes before their cutaneous malignancies precipitated referral. An additional possibility, impossible to quantify, is that some or indeed most skin cancers, such as basal cell carcinoma, were referred directly to surgeons.
A major criticism of this and other clinic-based case mix studies is one of selection referral, something that can only be overcome by true population studies, of which there are very few for skin disease. A recent such study by Svensson et al. of 12 377 participants aged 18–74 years, drawn from Germany, Italy, the Netherlands, Portugal and Sweden, showed that, for ‘active manifestation’ (rather than lifetime prevalence), eczema was the most common diagnosis, at a point prevalence of 16% (most frequently, contact dermatitis), with psoriasis at 3%, acne 5%, vitiligo 1·4% and skin cancer 0·5%.9
The study by Svensson et al. suffers from the drawback that it excludes children and those older than 74 years, thus skewing the findings away from atopic dermatitis and skin cancers. No similar population study exists for the early part of the twentieth century to indicate whether the 1921–22 data are applicable to the populace at large. However, some inference might be drawn by looking at the ranking of diseases. In 1921–22 data, tinea, impetigo and scabies were the second, third and fourth most common diagnoses (accounting for 39% of referrals), implying significant prevalence in the public at large, whereas for Svensson et al. these conditions did not merit their own diagnostic categories.9
The dermatological case mix of a century ago at the founding of the British Association of Dermatologists shows both similarities and differences from that of the present day. A wide range of factors, including environmental, social, medical referral practice and demographic, seem to be involved in the observed shifting pattern of case mix.
Author Contribution
David Gawkrodger: Writing-original draft (equal).
