Prevalence and associated factors of skin diseases in aged nursing home residents: a multicentre prevalence study.

OBJECTIVES: The aim of this study was to measure the prevalence of skin diseases in aged nursing home residents and to explore possible associations with demographic and medical characteristics.
DESIGN: Descriptive multicentre prevalence study. SETTING AND PARTICIPANTS: The study was conducted in a random sample of ten institutional long-term care facilities in the federal state of Berlin, Germany. In total, n=223 residents were included.
RESULTS: In total, 60 dermatological diseases were diagnosed. The most frequently diagnosed skin disease was xerosis cutis (99.1%, 95% CI 97.7% to 100.0%) followed by tinea ungium (62.3%, 95% CI 56.0% to 69.1%) and seborrheic keratosis (56.5%, 95% CI 50.2% to 63.0%). Only few bivariate associations have been detected between skin diseases and demographic and medical characteristics.
CONCLUSION: Study results indicate that almost every resident living in residential care has at least one dermatological diagnosis. Dermatological findings range from highly prevalent xerosis and cutaneous infection up to skin cancer. Not all conditions require immediate dermatological treatment and can be managed by targeted skin care interventions. Caregivers need knowledge and diagnostic skills to make appropriate clinical decisions. It is unlikely that specialised dermatological care will be delivered widely in the growing long-term care sector. TRIAL REGISTRATION NUMBER: This study is registered at https://clinicaltrials.gov/ct2/show/NCT02216526. © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted.

This was the largest randomly selected sample of long-term care residents aged ≥65 years undergoing a head-to-toe skin examination by board certified dermatologists.

Skin diseases, medications and concomitant diseases were classified according to international definitions and functional assessments were conducted according to established methods to support the generalisability of results.

Although three additional institutional long-term care facilities were included, the anticipated sample size of n=280 was not achieved.

There were differences between participating and non-participating long-term care institutions.

Systemic diseases were not specified and laboratory and histology data were not available.

Introduction

Background

Skin ageing, functional limitations, chronic diseases, polypharmacy, personal skin care and hygiene habits in populations aged ≥65 years cause an increased vulnerability to skin diseases and cutaneous problems.1–3 Epidemiological studies suggest that skin diseases are highly prevalent in the elderly population.4–6 For example, the prevalence of xerosis cutis range up to 85.5%, benign skin tumours up to 74.5%, fungal infections up to 77% and pressure ulcer (PU) up to 46%.7 8 However, most published epidemiological figures were obtained in hospital settings. The epidemiology of cutaneous diseases in institutional long-term care settings is largely unknown,7 although the number of multimorbid residents living in institutional long-term care is increasing.9

In addition to the high prevalence, the burden of skin diseases also increases with age.6 They are associated with reduced quality of life.10 It was shown that geriatric patients with dermatological diseases have an increased risk for mental and behavioural disorders, primarily depression.11 The medical treatment of the mulitmorbidities in nursing home residents may also result in polypharmacy.12 Associated adverse drug reactions, non-adherence or drug–drug interactions are common13 14 and linked to dermatological disorders. Immobility, cognitive impairment and organisational or reimbursement factors may also limit the opportunity for these population to receive specialised dermatological care. Traditionally, nurses and other healthcare professionals focus on PUs and incontinence-associated dermatitis (IAD) but may ignore other skin problems which may also require attention. On the other hand, not all dermatological conditions require specialised pharmacological treatment.

According to the latest statistics, there are 800.000 residents living in 13.600 long-term care institutions in Germany15 and these figures are expected to increase. At the same time, the prevalence of skin diseases in this care setting is largely unknown. In order to gain a detailed picture about the epidemiology of skin diseases in institutional long-term care this study was conducted.

Objectives

The aim of this study was to measure the prevalence of skin conditions and diseases in aged residents living in institutional long-term care facilities and to explore possible associations with demographic and medical characteristics.

Materials and methods

Study design

This was an observational, cross-sectional prevalence study and it was approved by the ethics committee of the Charité-Universitätsmedizin Berlin (EA1/190/14). The study protocol was published previously.16

Setting

The study was conducted from September 2014 to May 2015 in 10 institutional long-term care facilities in Berlin, Germany. In Germany, institutional long-term care facilities or residential care facilities are full-time accommodations with professional care. The staff is a mix between registered nurses and nursing assistants. Using computer-generated random numbers, institutional long-term care facilities from a list of all existing facilities (n=291) in the federal state of Berlin, Germany were contacted. In case of non-response, the next randomly selected nursing home was invited.

Participants

The inclusion criteria were (1) being resident of the respective residential care facility, (2) aged ≥65 years, and (3) written informed consent given personally or by legal representative. Only residents being able to give informed consent by themselves or having a legal representative who decided on behalf of the resident took part in this study. The exclusion criterion was residents at the end of life to avoid unnecessary burden due to the examinations. All residents (or their legal representatives) living in the residential care facility at time of data collection were invited to participate.

Variables

Skin diseases were classified according to the International Coding of Diseases (ICD-10) classification, with the exception of IAD and skin tears. IAD was diagnosed according to the IAD-IT classification of Junkin 2008.17 According to an international consensus, skin tears are caused by shear, friction and/or blunt force causing the separation of the layers of the skin (partial or full thickness wound) most commonly on the extremities.18 Skin tears were recorded as present/absent. Xerosis cutis was measured using the Overall Dry Skin score (ODS) with a 5-point scale ranging from ‘0’ (no skin dryness) to ‘4’ (advanced skin roughness, large scales, inflammation and cracks).19 20 Concomitant diseases (ICD-10 classification level 1) and medications were extracted from the medical records. These contain documentation of anamnesis, diagnoses, examination results, therapies and results, interventions and medical letters. Demographic variables of the nursing home residents (eg, age and sex) were collected. The physical function related to the daily activities was assessed using the Barthel Index. The scores range from 0 (very care dependent) to 100 (not care dependent).21 The Braden scale was used to measure PU risk. Scores range from 6 (high PU risk) to 23 (no PU risk).22 The educational qualification was classified into the following six categories: ‘no school qualification’, ‘primary school’, ‘secondary school’, ‘grammar school/A-level’, ‘vocational training’ and ‘university qualification’.

Data sources and measurement

All participating nursing home residents underwent a head-to-toe skin examination conducted by a board certified dermatologist (UBP, NGB, IJ). Examinations were done by clinical evaluation and using dermatoscopes (Dermogenius basic, DermoScan GmbH, Germany). Demographic characteristics (eg, age and sex) and information regarding school qualification were extracted from the medical records by trained study assistants or the residents were interviewed, if possible. PU risk and care dependency (Braden scale and Barthel Index) were extracted from the medical records or assessed by a registered nurse. All study data were continuously documented in data collection forms by the investigator and authorised staff.

Bias

Institutional long-term care facilities in the state of Berlin differ in terms of ownership, size, and specialisation. In order to reduce selection bias, institutions were randomly selected from all facilities of the state of Berlin. All study-related procedures and measurements were conducted by trained dermatologists and study assistants according to standard operating procedures. The board certified dermatologists had no access to medical history data of the residents prior and during examinations to reduce the risk of detection bias.

Study size

Assuming a prevalence of 0.5 of skin diseases, approximately 280 residents would have been needed to measure this proportion with a desired width of a 95% CI of ±0.06. According to the latest Nursing Care Statistics (2013), the size of the long-term care population in Berlin was approximately 30.000.23 Assuming 80 residents per institution and a participation rate of 50% (n=40), it was planned to include seven institutions which results in n=280 (7 x n=40) cases.

Quantitative variables

The duration of residency was measured in months. The Barthel Index and Braden scale scores were used as metric variables. In order to investigate possible associations with skin diseases, the variable ‘educational qualification’ was dichotomized into ‘university qualification’ (yes/no). Residents taking four or more medications were regarded as having ‘polypharmacy’.12

Statistical methods

Depending on the level of measurement (nominal, ordinal and continuous), demographic characteristics, functional assessment scores and dermatological diseases were described using means, medians, proportions, frequencies and associated spread estimates, standard deviations, ranges and interquartile ranges. The 95% CIs were calculated around point estimates of dermatological diseases. Exploratory data analysis to investigate possible bivariate associations were conducted using logistic regression analysis for all skin diseases with a prevalence of at least 8%. 95% CIs of the ORs excluding 1 were considered to be statistically significant. ORs being statistically significant or with values lower than 0.5 or higher than2.0 were considered to be likely associated. In case of multiple bivariate associations, multivariable logistic regression analyses were conducted. Models were built iteratively to increase model fit indicated by Nagelkerke’s R2.

Results

Fifty-five long-term care facilities were contacted. Finally, 10 long-term care facilities agreed to participate. Compared with participating institutions, non-participating institutions were larger in terms of number of beds (mean beds per institution: 104.5 vs 73.7) privately owned (76% vs 60%) and non-profit (30% vs 22%).

All residents of the eligible long-term care facilities were invited, but participation rate was <50%. In order to achieve the planned number, three additional long-term care facilities were recruited (in total 10). In total, n=811 long-term care residents were assessed for eligibility, n=58 residents (23%) provided written informed consent by themselves and for n=194 residents (77%), the legal representative gave consent for participation. In total, n=29 residents declined participation prior examination resulting in n=223 included long-term care residents (figure 1).

Figure 1

Flow chart of participants.

Descriptive data

Sample characteristics are shown in table 1. Most residents were women (67.7%) and the mean age was 83.6 (SD 8.0) years. Mean Barthel Index score was 45.1 (SD 23.8) and mean Braden scale score was 17.3 (SD 3.7). The median time of long-term care residence until data collection was 27 months. A vocational training was the highest educational level for the majority (48.9%). The most common concomitant diseases (ICD-10 system level 1) were diseases of the circulatory system (82.5%) and mental and behavioural disorders (70.4%). In total, 84.6% of the residents received four or more medications (polypharmacy). The mean number of medications used was 6.8 (SD 3.4) per resident.

Table 1

Demographic characteristics of participants (n=223)

Female, n (%)	151 (67.7)
Age (years)
Mean (SD)	83.6 (8.0)
Median (IQR)	84 (78–89)
Barthel Index total score*
Mean (SD)	45.1 (23.8)
Median (IQR)	45.0 (25.0–65.0)
Braden score*
Mean (SD)	17.3 (3.7)
Median (IQR)	18.0 (14.0–21.0)
BMI (kg/m²)†
Mean (SD)	25.3 (5.1)
Median (IQR)	24.6 (21.9–28.3)
Duration of residency (months)
Mean (SD)	42.6 (49.1)
Median (IQR)	27.0 (14.0–52.0)
Highest educational qualification, n (%)
No school qualification	3/184 (1.6)
Primary school	34/184 (18.5)
Secondary school	24/184 (10.8)
Grammar school/A-level	7/184 (3.8)
Vocational training	90/184 (48.9)
University	26/184 (14.1)
Number of medications per resident‡
Mean (SD)	6.84 (3.41)
Median (IQR)	6.0 (4.0–9.0)
Polypharmacy (≥4 medications), n (%)	186/221 (84.2)
Common concomitant diseases, ICD-10 system level 1, n (%)
Diseases of the circulatory system (I.00–I.99)	184/223 (82.5)
Mental and behavioural disorders (F.00–F.99)	157/223 (70.4)
Endocrine, nutritional and metabolic diseases (E.00–E.99)	122/223 (54.7)
Diseases of the genitourinary system (N.00–N.99)	106/223 (47.5)
Diseases of the nervous system (G.00–G.99)	99/223 (44.4)
Diseases of the musculoskeletal system and connective tissue (M.00–M.99)	96/223 (43.0)

*n, 222; †n, 216; ‡n, 221.

BMI, body mass index; ICD-10, International Coding of Diseases classification.

Main results

In total, 60 dermatological diseases were diagnosed. The complete list of dermatological findings is shown in the online supplementary table S1. Xerosis cutis was most frequent (99.1%, 95% CI 97.7% to 100.0%) followed by tinea ungium (62.3%, 95% CI 56.0% to 69.1%) and seborrheic keratosis (56.5%, 95% CI 50.2% to 63.0%). Thirty-two dermatological diseases were diagnosed for five residents or fewer (eg, Bowen’s disease, 5/223, allergic contact dermatitis, 2/223, atopic dermatitis 1/223).

The results of the bivariate associations are shown in table 2. Higher age was associated with the increased prevalence of seborrheic keratosis (OR=1.041, 95% CI 1.007 to 1.077) and intertrigo (OR=1.052, 95% 1.004 to 1.102). On the other hand, the occurrence of seborrheic dermatitis decreased with increasing age (OR=0.951, 95% CI 0.909 to 0.996). Female sex showed a decreased occurrence of androgenetic alopecia (OR 0.187, 95% CI 0.099 to 0.354), tinea pedis (OR=0.435, 95% CI 0.241 to 0.786) and actinic keratosis (OR=0.321, 95% CI 0.165 to 0.622). There were statistically significant associations between the Barthel Index and tinea pedis (OR=1.013, 95% CI 1.001 to 1.025) as well as venous insufficiency (OR=1.019, 95% CI 1.005 to 1.034); and between the duration of residency and tinea ungium (OR=0.992, 95% CI 0.987 to 0.998) as well as tinea pedis (OR=0.987, 95% CI 0.978 to 0.996), but the strength of association were small. Having a university qualification was associated with less occurrence of xerosis cutis (OR=0.462, 95% CI 0.175 to 1.223). The number of medications used was associated with the occurrence of venous insufficiency (OR=1.108, 95% CI 1.011 to 1.214) and scar and fibrosis (OR=1.103, 95% CI 1.000 to 1.217).

Table 2

Associations between skin diseases and demographic characteristics (bivariate)

Skin diseases (ICD-10)	Age (years) (OR, 95% CI)	Gender (OR, 95% CI) (0=man, 1=woman)	Barthel Index (OR, 95% CI)	University qualification (OR, 95% CI) (0=no, 1=yes)	Duration of residency (OR, 95% CI)	Number of medications used (OR, 95% CI)
Xerosis cutis (L85.3) ODS >0	–	–	–	–	–	–
Xerosis cutis (L85.3) ODS >1	1.037 (0.991 to 1.084)	1.166 (0.555 to 2.449)	1.001 (0.986 to 1.016)	0.462 (0.175 to 1.223)	0.998 (0.991 to 1.005)	1.012 (0.911 to 1.123)
Xerosis cutis (L85.3) ODS >2	1.027 (0.993 to 1.063)	1.039 (0.583 to 1.855)	1.008 (0.996 to 1.020)	0.636 (0.261 to 1.550)	0.998 (0.992 to 1.004)	0.934 (0.860 to 1.015)
Xerosis cutis (L85.3) ODS >3	1.022 (0.969 to 1.078)	1.178 (0.465 to 2.982)	1.005 (0.987 to 1.023)	0.610 (0.133 to 2.788)	0.988 (0.974 to 1.003)	0.971 (0.854 to 1.103)
Tinea unguium (B35.1)	1.022 (0.988 to 1.057)	0.829 (0.462 to 1.489)	0.998 (0.987 to 1.010)	1.253 (0.526 to 2.985)	0.992 (0.987 to 0.998)	0.969 (0.895 to 1.049)
Seborrheic keratosis (L82)	1.041 (1.007 to 1.077)	0.896 (0.507 to 1.580)	1.002 (0.991 to 1.014)	1.542 (0.648 to 3.667)	0.997 (0.991 to 1.002)	1.076 (0.992 to 1.166)
Androgenetic alopaecia (L64.9)	0.984 (0.952 to 1.017)	0.187 (0.099 to 0.354)	0.999 (0.988 to 1.010)	1.028 (0.447 to 2.362)	0.998 (0.993 to 1.004)	1.026 (0.949 to 1.109)
Incontinence-associated dermatitis	1.003 (0.969 to 1.037)	0.801 (0.448 to 1.434)	0.998 (0.986 to 1.009)	1.873 (0.812 to 4.319)	0.996 (0.989 to 1.002)	0.991 (0.914 to 1.074)
Tinea pedis (B35.3)	0.989 (0.955 to 1.024)	0.435 (0.241 to 0.786)	1.013 (1.001 to 1.025)	1.569 (0.663 to 3.717)	0.987 (0.978 to 0.996)	0.997 (0.917 to 1.083)
Other pigmentation disorders (L81.-)	1.028 (0.989 to 1.068)	1.305 (0.673 to 2.527)	1.006 (0.993 to 1.019)	0.856 (0.322 to 2.279)	0.996 (0.989 to 1.003)	0.998 (0.913 to 1.091)
Venous insufficiency (I87.2)	1.007 (0.968 to 1.047)	0.807 (0.417 to 1.562)	1.019 (1.005 to 1.034)	0.837 (0.294 to 2.379)	0.998 (0.992 to 1.005)	1.108 (1.011 to 1.214)
Actinic keratosis (L57.0)	1.029 (0.988 to 1.071)	0.321 (0.165 to 0.622)	1.004 (0.991 to 1.018)	1.017 (0.380 to 2.723)	0.993 (0.984 to 1.001)	1.034 (0.942 to 1.134)
Scar and fibrosis (L90.5)	1.022 (0.978 to 1.067)	1.262 (0.589 to 2.705)	0.993 (0.979 to 1.008)	0.882 (0.281 to 2.767)	0.997 (0.989 to 1.005)	1.103 (1.00 to 1.217)
Seborrheic dermatitis (L21.-)	0.951 (0.909 to 0.996)	0.992 (0.467 to 2.108)	0.993 (0.978 to 1.008)	0.809 (0.259 to 2.526)	0.996 (0.987 to 1.004)	0.991 (0.893 to 1.100)
Intertrigo (L30.4)	1.052 (1.004 to 1.102)	1.290 (0.585 to 2.842)	1.002 (0.987 to 1.017)	0.766 (0.212 to 2.758)	1.001 (0.994 to 1.008)	1.099 (0.994 to 1.216)
Haemangioma (D18.0)	0.997 (0.932 to 1.024)	1.057 (0.472 to 2.369)	1.008 (0.992 to 1.024)	1.067 (0.337 to 3.382)	0.998 (0.990 to 1.006)	1.091 (0.982 to 1.212)
Melanocytic naevi (D22.9)	0.955 (0.908 to 1.004)	0.634 (0.285 to 1.411)	1.008 (0.992 to 1.025)	1.255 (0.392 to 4.018)	0.994 (0.984 to 1.005)	1.080 (0.968 to 1.205)
Haemorrhage (R23.3)	1.031 (0.981 to 1.083)	0.538 (0.243 to 1.189)	1.003 (0.987 to 1.020)	1.553 (0.529 to 4.565)	1.001 (0.993 to 1.009)	0.985 (0.877 to 1.106)
Rosacea (L71.9)	0.974 (0.927 to 1.024)	0.704 (0.311 to 1.592)	1.004 (0.987 to 1.021)	1.255 (0.392 to 4.018)	0.996 (0.986 to 1.006)	0.987 (0.878 to 1.111)
Pressure ulcer (L89.-, all categories)	1.034 (0.974 to 1.097)	1.369 (0.473 to 3.961)	0.986 (0.966 to 1.007)	0.857 (0.183 to 4.012)	1.006 (0.998 to 1.013)	1.049 (0.918 to 1.198)
Neoplasm (C44.9)	1.065 (0.999 to 1.134)	1.737 (0.551 to 5.479)	1.001 (0.981 to 1.022)	1.587 (0.416 to 6.057)	1.002 (0.993 to 1.011)	0.957 (0.825 to 1.109)

Bold values indicate statistical significance, underlined values indicate OR ≥2.0; OR ≤ 0.5.

ICD 10: International Coding of Diseases classification; ODS, overall dry skin score.

Results of the multivariable logistic regression model with tinea pedis as dependent variable is displayed in the online supplementary table S2. Adjusted to the Barthel Index and the duration of residency, the occurrence of tinea pedis was lower in female residents (OR=0.454, 95% CI 0.245 to 0.893). Results of the multivariable logistic regression model with venous insufficiency as dependent variable is displayed in the online supplementary table S3. The occurrence of venous insufficiency was more likely in residents with higher Barthel Index scores (OR=1.019, 95% CI 1.004 to 1.033) and higher numbers of drugs (OR=1.110, 95% CI 1.010 to 1.220). None of the other skin diseases showed multiple associations in the bivariate regression.

Discussion

Key results

This prevalence study showed that nearly every resident in institutional long-term care is affected by at least one dermatological disease. In total, 60 dermatological diseases were diagnosed, which was unexpectedly high. The highest prevalence was observed for xerosis cutis followed by tinea unguium, seborrheic keratosis, androgenetic alopecia, IAD and tinea pedis. Only few bivariate associations have been detected between skin diseases and demographic and other characteristics. In the majority, the strengths of associations were small. Male sex was strongly associated with androgenetic alopecia, tinea pedis and actinic keratosis. A university qualification may be protective against xerosis cutis. Increasing age leads to increased risks of seborrheic keratosis and intertrigo and to decreased risks of having seborrheic dermatitis. Overall, the Barthel Index and the duration of residency seem to be unrelated to the occurrence of skin diseases in this population.

Limitations

Although three additional long-term care facilities were included, the anticipated sample size of n=280 was not achieved. In total, n=559/811 residents living in the institutional long-term care at the time of data collection did not responded, which may had led to a possible selection bias. Even though we performed a randomised selection of all long-term care facilities, there were differences between participating and non-participating institutions. Whether this has an effect on the results is unclear. We also excluded residents at the end of life which may have led to a selection bias. Although we collected numerous data, the systemic diseases were not further specified. This restricts detailed analyses of possible associations. Furthermore, we did not perform laboratory or histology. We also had no control over the documentation quality of the medical records.

Interpretation

Research in this setting is challenging due to difficulties of gathering written informed consent (eg, due to dementia and associated cognitive impairments).24 Irrespectively from that, besides a study published in Turkey in 2007 by Kilic et al,25 this was the largest randomly selected sample of residents aged 65 years or older undergoing a head-to-toe skin examination by board certified dermatologists in institutional long-term care facilities, compared with previous studies.5 26 27 In our study, prevalence estimates are higher compared with previous studies in this setting, for instance, the prevalence of xerosis cutis, IAD and actinic keratosis.5 25 28–30 Otherwise the study of Kilic et al reported a lower prevalence for actinic keratosis,25 which may be explained by the geographical region and the assumed darker skin types of examined nursing home residents. Prevalences for tinea pedis, pruritus and candidiasis were similar to previous reports.25 27

The PU prevalence of 9% was substantially higher compared with previous studies31 32 of the German long-term care setting. The main reason for this finding is unclear. Underreporting is a well-known phenomenon in epidemiological PU research.33 34 The full head-to-toe skin examination supports the internal validity and the accuracy of this point estimate. This indicates that PUs are a substantial problem in German long-term care settings.

We diagnosed a broad spectrum of dermatological conditions in our study population with a total number of 60 diagnoses, which is unexpectedly high. A study by Makrantonaki et al reported 72 dermatological disorders in a sample of 110 hospitalised elderly patients.35 These findings underscore the importance of dermatological examinations in geriatric patients and long-term care residents. However, the prevalence of >50% of the reported skin diseases was 2% or lower. Looking at the clinical spectrum of the diagnosed conditions, a large number are benign, easy to manage or seem to be of minor pathological relevance. Empirical evidence suggests the significant improvement of xerosis cutis in the elderly when using structured skin care regimens.36–40 Therefore our data may suggest a possible undersupply. Untreated dry skin is most often related to enhanced pruritus,38 and may lead to superficial injuries or wounds with superinfection.41 IAD or intertrigo may also be addressed by basic skin care interventions and/or antimycotic therapies.39 42 Other diseases like androgenetic alopecia, seborrheic keratosis or pigmentary disorders may be aesthetically disturbing but they do not require imperative medical treatment. However, also psychosocial well-being may be affected possibly leading to restrictions in mental health.41 43 Thus, in the elderly and especially in aged long-term care residents, we do have different challenges: realisation of regular dermatological examinations, detecting clinically relevant dermatoses obligatory to be treated, benign skin conditions for facultative treatment and aesthetically disturbing skin conditions with direct implications for physical and psychological well-being.

Some of the conditions identified in our study, like PUs, neoplasm, stasis dermatitis, venous insufficiency or superficial wounds require immediate medical attention. These diseases are frequently observed in this elderly population and may lead to several complications (eg, basal cell carcinoma, ulcus cruris and osteomyelitis) if not treated appropriately. It is important that healthcare practitioners are trained to screen for the most important and significant dermatological conditions in order to path the way for correct and adequate management.

In our study, we also identified conditions which may be considered borderline and may have lower or no importance but others may be simple and frequent conditions with severe consequences if not treated adequately. For instance, tinea pedis is frequent, with frequent relapses and often takes a chronic course. If tinea pedis is not treated properly, it bears the risk to spread to tinea corporis or to lead to onychomycosis and subsequent complications.44 The dermatophytes disturb the natural defence of the skin barrier, whereby bacteria and viruses can penetrate into deeper skin layers more easily. The risk of developing lower extremities cellulitis45 is increased. Another example of borderline conditions is actinic keratosis, which is a carcinoma in situ with the risk of progressing to squamous cell carcinoma.46 The distinction between actinic keratosis and squamous cell carcinoma can be challenging,46 but actinic keratosis may progress to a malignant disease.47

Interestingly, only few associations between skin diseases and demographic characteristics have been detected. Overall, the presence of skin diseases seems not to be associated with care dependency (Barthel Index) and the duration of residency. This indicates that residents are already affected by the skin disease when being admitted. Apparently they do not develop these conditions de novo in the institutions, but may develop them as a consequence to lifetime exposure to well-known risk factors such as ultraviolet (UV) exposure increasing the risk of skin cancer.48 The reason why higher education is associated with less dry skin is unclear. The educational level may be associated with skin self-care behaviour like the regular application of leave-on products.

The association of male sex and androgenetic alopecia was expected, because in the Caucasian population, the prevalence increases with age in men up to 80% and in women up to 42%.49 This may be also associated with actinic keratosis. Because men have a higher prevalence of pattern baldness, there is a reduced natural UV protection on the scalp skin which caused a higher occurrence of actinic keratosis. Also, increased manifestation of tinea pedis in the male gender may possibly be explained to increased hyperhidrosis, lower awareness for skin care (eg, regularly drying between toes, regularly checking feet and inappropriate hygiene habits).50

During the last decades, many studies were published reporting the high occurrence of dermatological disorders and the necessity to pay increasing attention to specialised dermatological care in the elderly population. However, is more specialised medical (dermatological) care feasible in this setting and is it cost-effective? A discussion of prioritisation in this vulnerable population is missing so far. Although there is an obvious need of dermatological care in institutional long-term care, it is unlikely that board certified dermatologists will solve this problem.51 Telemedicine applications and better medical training of healthcare providers in the institutional long-term care facilities were assumed as being suitable strategies.51 52 Frequent examinations by a dermatologists, as proposed by others,35 53 are unlikely to be affordable and manageable in this setting. Caregivers might be the key because they may have a gatekeeper function. They need to have skills to decide whether residents need medical or basic care and they need to decide when to refer to a specialist. They need to have an evidence-based algorithm for skin care and diagnostic skills to distinguish whether the skin condition is a cosmetic issue, whether it is crucial for skin care, whether it is a borderline disease needing observation or special attention and if it needs urgent medical attention. Therefore, we strongly recommend an algorithm which clarifies the ‘who?’, ‘what?’ and ‘when’ regarding skin care interventions and treatment for nursing and clinical decision making.

Generalisability

Using a population-based approach, n=223 residents living in institutional long-term care facilities were included. In comparison to the German care statistics, the participating institutional long-term care facilities were more private owned (60% vs 40.8% in the German care statistic) and there were less non-profit institutions (30% vs 55.8% in the German care statistic) which may limit the generalisability of results.54 Despite a response rate of 27.5% of residents living in the residential care facilities at time of data collection, demographic data like age, sex and care dependency are well comparable with the general German long-term care population statistics (eg, women 67.7% vs 72.7%; care-level I: 38.6% vs 39%; care-level II: 40.8% vs 40.5%; care-level III: 18.4% vs 21%)23 which supports the generalisability of the study results. However, a systematic exclusion of for instance highly care depended residents who might also been at higher PU risk may have introduced non-response bias. A response bias due to the informed consent procedure cannot be excluded as well.