Clin Rheumatol DOI 10.1007/s10067-013-2428-7

ORIGINAL ARTICLE

The predictors of foot ulceration in patients with rheumatoid arthritis Jill Firth & Robin Waxman & Graham Law & E. Andrea Nelson & Philip Helliwell & Heidi Siddle & Simon Otter & Violet Butters & Lesley Baker & Rosemary Hryniw & Sarah Bradley & Lorraine Loughrey & Begonya Alcacer-Pitarch & Samantha Davies & Jennifer Tranter

Received: 30 August 2013 / Revised: 4 October 2013 / Accepted: 24 October 2013 # Clinical Rheumatology 2013

Abstract This study was conducted to determine the predictors of foot ulceration occurring in patients with rheumatoid arthritis (RA) without diabetes. A multi-centre case control study was undertaken; participants were recruited from eight sites (UK). Cases were adults diagnosed with RA (without

diabetes) and the presence of a validated foot ulcer, defined as a full thickness skin defect occurring in isolation on / below the midline of the malleoli and requiring > 14 days to heal. Controls met the same criteria but were ulcer naive. Clinical examination included loss of sensation (10g monofilament);

Significance and Innovations • ABPI, forefoot deformity and loss of sensation predict risk of foot ulceration in patients with RA • In contrast with diabetes, raised plantar pressures did not predict risk of ulceration in this study This work was funded by the award of the Ann Wilks Grant by the Arthritis Research Trust (Yorkshire). J. Firth (*) Pennine Musculoskeletal Partnership Ltd, Oldham Integrated Care Centre, New Radcliffe Street, Oldham, Lancashire, OL1 1NL, UK e-mail: [email protected]

R. Hryniw North Lancashire PCT and University Hospitals of Morecambe Bay, Lancashire, UK

R. Waxman : P. Helliwell : H. Siddle Division of Rheumatic and Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK

S. Bradley Calderdale and Huddersfield NHS Foundation Trust, West Yorkshire, UK

G. Law Division of Biostatistics, Leeds Institute of Genetics, Health and Therapeutics, University of Leeds, Leeds, UK J. Firth : E. A. Nelson School of Healthcare, University of Leeds, Leeds, UK S. Otter School of Health Professions, University of Brighton, Brighton, UK V. Butters Greater Glasgow and Clyde NHS Trust, Glasgow, Scotland, UK L. Baker East Sussex Hospitals NHS Trust, East Sussex, UK

L. Loughrey : B. Alcacer-Pitarch Leeds Teaching Hospitals NHS Trust, Leeds, UK B. Alcacer-Pitarch Bradford Teaching Hospitals NHS Foundation Trust, West Yorkshire, UK S. Davies Pennine Acute Hospitals NHS Trust, Lancashire, UK J. Tranter School of Community Health Sciences, University of Nottingham, Nottingham, UK

Clin Rheumatol

ankle-brachial pressure index (ABPI); forefoot deformity (Platto); plantar pressures (PressureStat); RA disease activity (36 swollen/tender joint counts) and the presence of vasculitis. History taking included past ulceration/foot surgery; current medication and smoking status. Participants completed the Health Assessment Questionnaire (HAQ) and Foot Impact Scale. A total of 83 cases with 112 current ulcers and 190 ulcer naïve controls participated. Cases were significantly older (mean age 71 years; 95 % confidence interval [CI], 69–73 vs. 62 years, 60– 64) and had longer RA disease duration (mean 22 years; 19–25 vs. 15, 13–17). Univariate analysis showed that risk of ulceration increases with loss of sensation; abnormality of ABPI and foot deformity. Plantar pressures and joint counts were not significant predictors. HAQ score and history of foot surgery were strongly associated with ulceration (odds ratio [OR]=1.704, 95 % CI 1.274–2.280 and OR=2.256, 95 % CI 1.294–3.932). Three cases and two controls presented with suspected cutaneous vasculitis. In logistic regression modelling, ABPI (OR=0.04; 95 % CI, 0.01–0.28) forefoot deformity (OR=1.14; 95 % CI, 1.08–1.21) and loss of sensation (OR=1.22; 95 % CI, 1.10–1.36) predicted risk of ulceration. In patients with RA, ABPI, forefoot deformity and loss of sensation predict risk of ulceration but, in contrast with diabetes, raised plantar pressures do not predict risk. Keywords Foot ulceration . Predictors . Rheumatoid arthritis

Introduction Foot ulceration is estimated to affect 10–13 % of patients with rheumatoid arthritis (RA) during the course of their disease and 47 % of those affected report experiencing recurrent episodes of ulceration involving multiple sites on the foot [1, 2]. Qualitative research has demonstrated that this problem has an additional impact on health-related quality of life over and above the impact of RA in every domain. Pain arising from ulceration creates new walking disability and restrictions in daily life and social participation with psychological consequences [3]. However, whilst the risk factors for foot ulceration in patients with diabetes have been extensively studied and care pathways are well established, much less is known about the aetiology of foot ulceration in RA. Furthermore, considerable variation in rheumatology foot health provision exists, certainly within the UK [4]. A lack of awareness amongst affected patients regarding prompt recognition of tissue damage, how to access care and preventative measures are among the information needs expressed by participants in a recent study [5]. This is a barrier to effective healthcare and a concern in the era of biologic therapies with their associated risks of serious soft tissue and skin infections that may rapidly progress if untreated [6]. Diabetes services have been able to develop specialised foot clinic services because the risk factors for diabetic foot

ulceration are understood. There are some apparent similarities in the aetiology of foot ulcers in diabetes and RA, such as the structural changes in the foot, but there are sufficient differences to require us to determine the predictors of foot ulceration in RA empirically. We cannot assume that the predictors in RA are the same as diabetic foot ulcer risk. Established risk factors for foot ulceration in patients with diabetes were identified in a literature review and include peripheral neuropathy, peripheral vascular disease, raised plantar pressures, foot deformity and ill-fitting footwear [7, 8]. These clinical manifestations of disease also occur to a greater or lesser extent in patients with RA, but the contribution of these factors to the risk of foot ulceration in RA is unknown. People with RA may also have local disease activity in the joints and soft tissues of the foot, cutaneous vasculitis and nodules that may contribute to risk. Furthermore, disease-modifying medication such as corticosteroids may cause skin fragility and hinder tissue repair [9]. The individual's ability to maintain foot health independently may also be compromised by reduced self-care capacity and manual dexterity. A pilot study indicated that the predictors of ulceration in RA included active disease and current steroid therapy but not loss of sensation, foot deformity and raised plantar pressures, which are important in diabetic foot ulceration [10]. The contribution of peripheral vascular disease to risk was unclear due to sparse data. The sample size in this nested single centre study was too small to represent a generalizable population, hence the need for a larger multi-centre study. The aim of this study was therefore to determine the predictors of foot ulceration in patients with RA unaffected by concomitant diabetes. This will help to underpin strategies to reduce both incidence and recurrence rates and improve quality of life.

Patients and methods This was a multi-centre case controlled study conducted in eight centres in the UK. Ethical approval was obtained from the Local Research Ethics Committee for Leeds (East) Research Ethics Committee (UK). The study was deemed exempt from site-specific assessment and Research and Development approval was obtained at each site. Sample size and power calculations were conducted using the powercal function in Stata (version 11) [11].The detectable odds ratio (OR) was estimated for an unmatched case control study, with a p value threshold of 0.05 and power of 80 %. A sample of 100 cases, with 200 controls, would enable the detection of a statistically significant risk factor (p =0.05) with an OR of 2.0 and power of 80 %. Therefore, we aimed to recruit 100 cases (patients reporting open foot ulceration) and 200 unmatched controls (patients with no reported history of foot ulceration). Cases were recruited by podiatrists who were

Clin Rheumatol

working in hospital, university clinic and community settings. The inclusion criteria were adult patients with a diagnosis of RA with the ability to give informed consent and the presence of a validated foot ulcer, defined as full thickness skin defect occurring in isolation on or below the midline of the malleoli and requiring >14 days to heal. The exclusion criteria were a concomitant diagnosis of diabetes as this in itself is a known risk factor for foot ulceration and heel ulceration because this was deemed more likely to be a pressure ulcer associated with immobility. All patients meeting these criteria were invited to take part. Controls met the same inclusion/ exclusion criteria but reported no history of ulceration. Podiatrists at each centre initially used random number tables to identify potential controls from rheumatology clinic lists to reduce bias. A random number table was selected to match the size of the clinic list and the podiatrist selected the next number shown in the random number table until two eligible controls per case were recruited. However, centres were instructed to approach all eligible controls from each clinic list when it became apparent this was necessary to achieve the recruitment target. Clinical examination was undertaken by the podiatrist at each participating centre following training to standardize data collection procedures. Data collected included the following: &

&

& &

Insensitivity to 10 g monofilament applied to the plantar surface of the forefoot over the metatarsal phalangeal joints (MTPJs) 1–5 and the plantar aspect of the great toe [12]. Ankle-brachial pressure index (ABPI) using handheld Doppler ultrasound, measured the maximum of the dorsalis pedis and posterior tibial systolic pressures in each limb. The maximum pressure from the ankle was then divided by the maximum of right and left brachial pressures to give the ABPI (right and left readings). A reading of less than 0.9 was considered abnormal [indicative of arterial disease [13]. Forefoot deformity measured by the Platto index [14] and Manchester grading scale for hallux valgus [15] Plantar pressures were measured using PressureStat™ footprint mat [16]. Patients were asked to walk over the mat barefoot at normal cadence and measurements were taken using a 'first step' approach. Peak plantar pressure and background greyness were classified as well as the site of maximum plantar pressure. Scores were recorded by two assessors; where there was disagreement, footprints were reviewed by both assessors and a final score agreed. Background noise subtracted to improve validity and reliability [17]. Pressures of less than 6 kg/cm2 are considered to be normal, and equal to or more than 6 kg/ cm2 indicative of high plantar pressure, based on the manufacturer's guidelines.

& & &

A 36 swollen and tender joint count. Clinician reported grading of callus if present at the ulcer site (1 = minimal; 2 = light; 3 = moderate; 4 = heavy). The presence of pre-ulcerative lesions, defined as a clear area of tissue damage associated either with subcutaneous haemorrhage or discolouration of the skin.

History taking included a record of any previous foot ulceration or surgery; duration of current ulceration; current disease modifying and steroid therapy. Participants also completed the Foot Impact Scale [18], a two sub-scale, 51-item questionnaire in two domains to compare impairments and footwear and activities and participation between cases and controls, and the Stanford Health Assessment Questionnaire (HAQ), modified for use with British patients, to assess self care capacity [19]. The research team comprising the principal investigator, a specialist nurse (JF); rheumatologist (PSH); podiatrist (HS); wound care expert (EAN); project manager (RW) and statistician (GRL) met to develop and agree a Directed Acyclic Graph (DAG) mapping out the potential causal pathway of ulcer formation in patients with RA. The version agreed by expert consensus is shown in Fig. 1. The DAG makes our assumptions explicit and this approach was used to help choose which covariates should be included in traditional statistical approaches in order to minimize the magnitude of the bias in the estimate produced [20]. The data was entered onto a database management system [21] and exported for statistical analysis. Stat Transfer was used to convert the data into suitable formats for Stata. Next, logistic regression was applied to estimate potential predictors of foot ulceration in RA for the individual variables. Each variable was carefully checked for linearity. Our modelling strategy was to build a model including known risk factors for foot ulceration in diabetes (loss of sensation; abnormal ABPI; foot deformity and raised plantar pressure) plus potential

gender

RA

ulcer

impaired function

sensory loss

RA duration

age

ABPI

foot surgery

foot deformity

raised plantar pressures

Fig. 1 Directed acyclic graph representing hypothesized causal pathway of foot ulceration in patients with RA

Clin Rheumatol

predictors in RA identified from the univariate analysis using the DAG. Proxy variables were acknowledged and excluded from the final model.

Results Eighty three cases and 190 controls were recruited between April 2009 and April 2011.The voluntary participation rate of cases was 90 %, compared with 61 % of eligible controls. Cases were significantly older than controls and had significantly longer RA disease duration, but there was no significant difference between cases and controls in terms of sex, ethnicity or smoking status. Cases were significantly more likely to be taking steroids and non-steroidal anti-inflammatory drugs (NSAIDS) but less likely to be taking disease-modifying anti-rheumatic drugs (DMARDS); there was no difference in terms of numbers prescribed biologic therapy. Table 1 displays analysis of cases and controls. Eighty three cases presented with 112 current ulcers and 52 of these cases (63 %) reported a history of ulceration. Sixty three cases (76 %) reported that this was their first episode of the current ulceration. Forty-four percent reported one or more previous episodes of the current ulceration; the mean number of previous episodes was 1.41 (95 % CI, 0.95–1.87). Twenty five percent of cases had a current ulcer at more than one site: 18 had two ulcers; two had three ulcers and two had four or more, with a maximum count of six. Ten cases (12 %) were experiencing ulceration bilaterally. The most common sites of current ulceration were the plantar aspect of the MTPJs (n =31); dorsal aspect of the inter-phalangeal joints (IPJs) (n =16); and the medial aspect of the 1st MTPJ (n =13). Callus was present at the site of current ulceration in 66 ulcers (59 %); this was graded as minimal or light at 42 sites (64 %) and moderate or heavy in the remaining 24 (36 %). Three cases and two controls presented Table 1 Analysis of cases and controls

with suspected cutaneous vasculitis. The mean duration of current ulceration was 12.41 weeks (95 % CI, 8.69–16.13), ranging up to 182 weeks. The initial univariate analysis is displayed in Table 2. The results displayed in Table 2 show that in the univariate analysis, plantar pressures, swollen and tender joint counts were not significant predictors of ulceration. HAQ score and history of foot surgery were strongly associated with ulceration. However, HAQ score correlated strongly with advancing age and surgery was considered a proxy variable for deformity. Both domains of the Foot Impact Scale (impairments and footwear and activities and participation) were strongly associated with ulceration but in a case control study the temporal relationship between ulceration and impact is unclear and higher scores may represent the known impact of ulceration in these domains. Hence the scores were excluded from the final modelling dues to their low predictive value. Table 3 gives the result of the logistic regression modelling where loss of sensation, ABPI and higher Platto score of forefoot deformity were all significant predictors. The analysis showed that risk of ulceration increases with increasing loss of sensation; lower APBI and increasing foot deformity. Thirteen cases (six controls) had an ABPI <0.9. Using ABPI as a categorical variable in the logistic regression modelling, the variable just failed to achieve significance. Within patients there were no significant differences in the ABPI data between the affected and non-affected limbs for unilateral ulceration. The DAG and univariate analyses informed model construction. Table 3 shows the model of best fit after multivariate analysis. We undertook a sub-group analysis of cases affected by plantar ulceration to explore whether the location of ulceration affected the significance of predictors, hypothesizing that

Variable

Cases (n =83)

Controls (n =190)

p value

Age (mean, 95 % CI) Gender (n, % male/female) Ethnicity (% white British) RA disease duration (years, mean, 95 % CI) Smoking status (n, %) Current smoker Ex-smoker Never smoked Current medication (≥1) NSAIDs DMARDs Steroids Biologic therapy

71 (69–73) 18 (22 %) 82 (99 %) 22 (19–25)

62 (60–64) 40 (21 %) 188 (99 %) 15 (13–17)

<0.0001 0.906 0.554 0.000

14 (17 %) 25 (30 %) 44 (53 %)

35 (18 %) 69 (36 %) 86 (45 %)

0.801 0.502 ref

44 (53 59 (71 31 (37 16 (19

%) %) %) %)

65 (34 167 (88 19 (10 36 (19

%) %) %) %)

0.004 0.001 <0.0001 0.949

Clin Rheumatol Table 2 Univariate analysis Variable Person level data Mean swollen joint count (SD) Mean tender joint count (SD) History of foot surgery (n, %) Mean HAQ score of impaired function (SD) Foot Impact Scale Limb level data Loss of sensation (mean n of sites lost, SD) ABPI right foota Mean Platto score of forefoot deformity right foot (SD)a Raised plantar pressurea a

Cases (n =83)

Controls (n =190)

Unadjusted OR (95 % CI)

4.976 (6.319) 7.301 (8.581) 33 (40 %) 1.953 (0.918) 34.470 (10.956)

4.821 (5.711) 8.168 (9.197) 43 (23 %) 1.489 (0.950) 27.428 (13.819)

1.004 (0.962–1.049) 0.989 (0.960–1.019) 2.256 (1.294–3.933) 1.704 (1.274–2.280) 1.0685 (1.034–1.104)

3.193 (3.878) 1.087 (0.214) 8.049 (2.484) 22/74 (30 %)

0.942 (2.264) 1.175 (0.158) 5.305 (3.429) 31/182(17 %)

1.267 (1.155–1.389) 0.471 (0.008–0.264) 1.331 (1.206–1.470) 1.795 (0.874–3.688)

No statistical difference between right and left so right foot data displayed in univariate analysis and entered into multivariate analysis

plantar ulceration would be more likely to be associated with raised plantar pressures. Thirty-three patients presented with 34 plantar ulcers at the time of data collection. The significant predictors of ulceration were unchanged in the univariate analysis and raised plantar pressures did not achieve significance. In the final model, loss of protective sensation (OR= 1.28; 95 % CI, 1.16–1.46) abnormal ABPI (OR=0.01; 95 % CI, 0.001–0.30) and forefoot deformity (OR=1.16; 95 % CI, 1.06–1.27) remained as significant predictors of risk ulceration in this sub group.

Discussion As far as we know, this is the first study to identify predictors of foot ulceration in patients with RA without diabetes. The findings of this study indicate that foot ulceration occurring in patients with RA shares similarities with the causal pathway of diabetic foot ulceration. We were not only able to identify the factors that predicted increased risk, but also the relative weighting for risk associated with these factors. The analysis showed that risk of ulceration rose with increasing loss of sensation and foot deformity and with falling ABPI values. Our conclusions should be viewed with a degree of caution however, as missing data contributed to the wide confidence intervals (CI) for ABPI. Data was missing because nine cases (11 %) and six controls (3 %) were unable to tolerate measurement. The presence of lower limb oedema was noted to be a contributory factor affecting the ease of measurement,

where missing data exists, which in itself may be suggestive of venous disease. Peripheral neuropathy was thought to be uncommon in RA but studies to determine its prevalence through clinical examination and electrophysiology indicate that it is under–estimated. Prevalence of electro-physiologic evidence of neuropathy has been estimated to be between 57 % and 65 % but the majority of affected patients fail to report signs or symptoms [22, 23]. Similarly, the prevalence of reduced protective sensation has been found to be as high as 59 % compared with 12.5 % in the feet of controls [24]. Whilst peripheral neuropathy in diabetes is undoubtedly more profound, targeted screening programmes mean that there are mechanisms in place for detection and intervention, including patient education to raise awareness of the risks to tissue viability arising from loss of protective sensation. It is the combination of loss of sensation with foot deformity that increases the risks of pressure from footwear. In the era of biologic therapies, we would hope that more effective treatment strategies reduces the erosive impact of disease and subsequent deformity, but over reliance on the DAS-28 score (which does not include assessment of foot and ankle disease) to inform clinical decision-making may increase the chances of disease activity in the feet being under treated. Indeed, foot examination would appear to be infrequent (mean 15.7 months) based upon recall even amongst patients with severe disease warranting anti-TNFα medication and regular review (mean 4.1 months) [25]. Research in diabetes indicated that 78 % of patients had rigid foot deformity directly

Table 3 Model of best fit Variable

Adjusted OR (95 % CI)

SE

z

P > |z|

Loss of protective sensation Abnormal ABPI Platto score of forefoot deformity

1.22 (1.10–1.36) 0.04 (0.01–0.28) 1.14 (1.08–1.21)

0.07 0.04 0.03

3.66 −3.27 4.63

0.000 0.001 0.000

Clin Rheumatol

associated with the site of ulceration [26] but forefoot deformity is also an indirect risk factor for ulceration. Forefoot deformity has been found to correlate significantly with peak pressures and the duration of loading in RA [27, 28], triggering a series of interacting component causes of ulceration. The interface between foot deformity and footwear is a potential source of pressure and shearing forces that poses a risk to tissue viability. Qualitative research involving patients with RA and open foot ulceration linked pressure and friction from footwear to the onset of ulceration and the challenges of finding footwear to accommodate deformity were a source of frustration to participants [3]. It was not possible to objectively measure footwear fit and robust data is difficult to collect: inshoe pressure and force analysis equipment was not available at participating centres and its use would in any event be limited by the cross sectional design of the study and the potential for footwear modifications to be made following the onset of ulceration. The potential importance of this as a risk factor is perhaps best illustrated by studies which show that without specialist footwear, 80 % of diabetic foot ulcers recur, compared with 20 % when specialist footwear is provided by an orthotist [29, 30]. Cutaneous manifestations of disease in RA, including ulceration, are often attributed to vasculitis but in this study, the incidence was low. Podiatrists participating in the study received training in the detection of possible vasculitis (nail fold infarcts; purpuric rash or inflamed wound margins) prior to data collection but vasculitis was only suspected in three cases and two controls. This is not to say that sub-clinical vasculitis may not exist but in a study where biopsy specimens were available for review in 12 of 16 patients with lower extremity ulceration, only three had biopsy evidence of vasculitis [31], suggesting that its role in the aetiology of ulceration may be have been over played historically. In contrast with the pilot study, disease activity measures (swollen and tender joint counts) were not significant predictors of ulceration and this was unchanged regardless of whether a 36 joint count or only foot and ankle disease was entered in the analysis. However, it should be noted that the mean duration of ulceration was greater than 3 months so this may not reflect disease activity at the time of onset of ulceration and a prospective study would be needed to confirm these results. At limb level, raised plantar pressures were significant predictors of foot ulceration in RA, but not in the multivariate analysis, which is surprising for two reasons. Firstly, the diabetic experience demonstrates a strong association between high plantar pressures and plantar ulceration, particularly in association with neuropathy [24, 32, 33]. Secondly, patients with RA consistently record high pressures over the metatarsal heads [27, 28] and peak pressures are accompanied by a higher prevalence of callosities [34]. Hence, expert consensus incorporated raised plantar pressures in the hypothesized causal pathway for foot ulceration in RA. Plantar ulceration was

the most common site for ulceration in our study, accounting for 34 current sites (30 %) of ulcer sites affecting 33 patients. However, even when we retested the multivariate analysis in this sub group the findings were unchanged. We acknowledge the limitations of using PressureStat™ as a measurement tool, which does not allow the measurement of loading times and pressure time integrals. These portable footprint mats were utilized in the absence of superior technology available at all sites; no two sites shared the same platform based equipment and most had no technology available. The accuracy of pressure measurement was improved by the instruction of the two readers (JF and PH) who dually coded the prints and the subtraction of background noise and which has been shown to improve the accuracy of pressure measurement [17]; any inter-observer differences were resolved by the PI. PressureStat™ readings are, however, likely to overestimate pressure, which is unlikely to have affected the findings in this instance. The lack of clear foot health pathways precluded a multicentre prospective cohort study and meant that it was difficult to identify and recruit participants at a number of centres, although fortunately the participation rate amongst cases was high at 90 %. Patients with concomitant diabetes were excluded to avoid skewing the results, which also affected recruitment. As a result, the study was under-powered but the final sample size was sufficient to generate preliminary findings from a UK population to inform clinicians regarding the factors that are associated with ulceration and may predict risk. Recent cluster analysis of the causal pathways in the development of diabetic foot ulceration points to diversity within patient populations [35]. Four dominant clusters were identified from 24 comprising (1) neuropathy, deformity, callus and elevated plantar pressures; (2) peripheral vascular disease; (3) penetrating trauma; and (4) ill-fitting shoes. The findings of our study indicate a need to screen patients with RA for neuropathy and peripheral vascular disease, implement care to accommodate or surgically correct foot deformity and educate patients regarding foot health to address risk factors and prevent ulceration. The multi-factorial nature of wounds and their healing cannot be under-estimated but should not delay the introduction of strategies to reduce risk and promote healing. Future work needs to be undertaken to develop and agree practical criteria for the prospective screening of patients and identification and management of those at risk. This includes the development and testing of patient education materials to address expressed information needs including the recognition and aetiology of foot ulceration; care pathways, prognosis and strategies to prevent recurrence [5]. Tackling incidence and recurrence rates has the potential to improve quality of life for affected patients and reduce the burden of healthcare demands arising from the health professional management of this problem. At service delivery level in the UK, it is necessary to continue work to implement national standards of care for foot health provision [36] to improve care

Clin Rheumatol

quality and to reduce regional variation. Clearer care pathways and the standardisation of clinical examination techniques are important steps in enabling the design of future multicentre research projects and will aid the implementation of evidence-based care.

Disclosure None.

References 1. Firth J, Hale C, Helliwell PS, Hill J, Nelson EA (2008) The prevalence of foot ulceration in patients with rheumatoid arthritis. Arthritis Care Res 59(2):200–205 2. Matricali GA, Boonen A, Verduyckt J, Taelman V, Verschueren P, Sileghem A, Corluy L, Westhovens R (2006) The presence of forefoot problems and the role of surgery in patients with rheumatoid arthritis. Ann Rheum Dis 65(9):1254–1255 3. Firth J, Nelson EA, Briggs M, Gorecki CA (2011) Qualitative study to explore the impact of foot ulceration on health-related quality of life in patients with rheumatoid arthritis. Int J Nurs Stud 48(11): 1401–1408 4. Redmond AC, Waxman R, Helliwell PS (2006) Provision of foot health services in rheumatology in the UK. Rheumatology (Oxford) 45(5):571–576 5. Firth J, Nelson EA, Briggs M, Gorecki C (2013) Experiences of healthcare provision for foot ulceration occurring in people with rheumatoid arthritis. Musculoskelet Care. 11930:159-67. doi:10.1002/msc.1036 6. Dixon WG, Watson K, Lunt M, Hyrich KL, Silman AJ, Symmons DPM (2006) Rates of serious infection, including site-specific and bacterial intracellular infection, in rheumatoid arthritis patients receiving anti-tumor necrosis factor therapy — results from the British Society for Rheumatology Biologics Register. Arthritis Rheum 54(8):2368–2376 7. Singh N, Armstrong DG, Lipsky BA (2005) Preventing foot ulcers in patients with diabetes. JAMA 293(2):217–228 8. Jeffcoate WJ, Harding KG (2003) Diabetic foot ulcers. Lancet 361(9368):1545–1551 9. Anstead GM (1998) Steroids, retinoids, and wound healing. Adv Wound Care 11:277–285 10. Firth J, Hale CA, Hill J, Helliwell PS, Nelson EA (2008) The predictors of foot ulceration in patients with rheumatoid arthritis: a preliminary investigation. Clin Rheumatol 27(11):1423–1428 11. Stata. StataCorp (2009) Stata statistical software: release 11. StataCorp LP, College Station 12. Bakker K, Apelqvist J, Schaper, NC (2012) On behalf of the International Working Group. Practical guidelines on the prevention and management of the diabetic foot 2011. Diabetes / metabolism research and reviews 28(1): 225–231 13. Vowden KR, Vowden P (2001) Doppler assessment and ABPI : interpretation in the management of leg ulceration. http://www. worldwidewounds.com/2001/march/Vowden/Doppler-assessmentand-ABPI.html. Accessed 03/10/13 14. Platto MJ, O'Connell PG, Hicks JE, Gerber LH (1991) The Relationship of pain and deformity of the rheumatoid foot to gait and an index of functional ambulation. J Rheumatol 18:1

15. Garrow AP, Papageorgiou A, Silman AJ, Thomas E, Jayson MIV, Macfarlane GJ (2001) The grading of hallux valgus: the Manchester scale. J Am Podiatr Med Assoc 91(2):74–78 16. Bailey Instruments Ltd: Manchester, UK http://www. baileyinstruments.co.uk/ 17. Firth J, Turner, Smith DE, Woodburn J, Helliwell PS (2007) The validity and reliability of PressureStat™ for measuring plantar foot pressures in patients with rheumatoid arthritis. Clin Biomech 22(5):603–606 18. Helliwell PS, Reay N, Gilworth G, Redmond A, Slade A, Tennant A, Woodburn J (2005) Development of a foot impact scale for rheumatoid arthritis. Arthritis Rheum 53(3):418–422 19. Kirwan JR, Reeback JS (1986) Stanford Health Assessment Questionnaire modified to assess disability in British patients with rheumatoid arthritis. Br J Rheumatol 25:206–209 20. Shrier I, Platt RW (2008) Reducing bias through directed acyclic graphs. BMC Med Res Methodol 8:70. doi:10.1186/1471-2288-8-70 21. IBM SPSS (2003) SPSS data entry: version 4. Chicago, IL, USA 22. Lanzanillo B, Pappone N, Crisci C, di Girolamo C, Massini R, Caruso G (1998) Subclinical peripheral nerve involvement in patients with rheumatoid arthritis. Arthritis Rheum 41(7):1196–1202 23. Agarwal VV, Singh RR, Wiclaf, Chauhan SS, Tahlan AA, Ahuja CK et al (2008) A clinical, electrophysiological, and pathological study of neuropathy in rheumatoid arthritis. Clin Rheumatol 27(7):841–844 24. Wilson O, Kirwan JR (2006) Measuring sensation in the feet of patients with rheumatoid arthritis. Musculoskelet Care 4(1):12–23 25. Otter SJ, Lucas K, Springett K, Moore A, Davies K, Young A, Walker-Bone K (2011) Comparison of foot pain and foot care among rheumatoid arthritis patients taking and not taking anti-TNFα therapy: an epidemiological study. Rheumatol Int 31(11):1515–1519 26. Lavery LA, Armstrong DG, Vela SA, Quebedeaux TL, Fleischli JG (1998) Practical criteria for screening patients at high risk for diabetic foot ulceration. Arch Intern Med 158(2):157–162 27. Tuna H, Birtane M, Tastekin N, Kokino S (2005) Pedobarography and its relation to radiologic erosion scores in rheumatoid arthritis. Rheumatol Int 26:42–47 28. Van der Leeden M, Steultjens M, Dekker JHM, Prins APA, Dekker J (2006) Forefoot joint damage, pain and disability in rheumatoid arthritis patients with foot complaints: the role of plantar pressure and gait characteristics. Rheumatology 45(4):465–469 29. Anderson RB, Hodge Davis W (1997) The pedorthic and orthotic care of the diabetic foot. Foot Ankle Clin 2(1):137–151 30. Edmonds ME, Blundell MP, Morris ME, Thomas EM, Cotton LT, Watkins PJ (1986) Improved survival of the diabetic foot: the role of a specialised foot clinic. Q J Med 60(232):763–771 31. Shanmugam KV, DeMaria DM, Attinger CE (2011) Lower extremity ulcers in rheumatoid arthritis: features and response to immunosuppression. Clin Rheumatol 30(6):849–853 32. Murray HJ, Young MJ, Hollis S, Boulton AJ (1996) The association between callus formation, high pressures and neuropathy in diabetic foot ulceration. Diabet Med 13(11):979–982 33. Veves A, Murray HJ, Young MJ, Boulton AJM (1992) The risk of foot ulceration in diabetic patients with high foot pressure: a prospective study. Diabetologia 35:660–663 34. Woodburn J, Helliwell PS (1996) Relationship between heel position and the distribution of forefoot plantar pressures and skin callosities in rheumatoid arthritis. Ann Rheum Dis 55(11):806–810 35. Lavery LA, Peters EJG, Armstrong DG (2008) What are the most effective interventions in preventing diabetic foot ulcers? Int Wound J 5:425–433 36. Podiatric Rheumatic Care Association (2008) Standards of care for people with musculoskeletal foot health problems. Arthritis and Musculoskelet Alliance. www.prcassoc.org.uk/standards-project