Undiagnosed and Untreated Peripheral Complications of Diabetes: Findings from a Pilot Study on Diabetes-Related Foot Diseases (DFD) in Patients with Glycemic Disorders

Abstract

Background

Diabetes-related foot disease (DFD) is a serious complication of diabetes, increasing the risk of amputation. Coimplications are preventable, but most diabetics do not receive proper screening and treatment, despite indications. This study was a pilot screening of diabetes-related foot disease in a group of people with glycemic disorders.

Material/Methods

We recruited 143 volunteers over 40 years of age. In the final analysis, we included 85 people diagnosed with glycemic disorders (diabetes or prediabetes), for whom we performed a total of 170 foot measurements. We screened for peripheral artery disease using: foot pulse, ankle-brachial index (manual and automatic), toe-brachial index, and transcutaneous oxygen pressure (TcPO2). To screen for diabetic peripheral neuropathy, we used indicators of loss of protective sensation: pressure perception and temperature perception, and plantar pressure distribution.

Results

A history of diabetes was reported by 26 (30.6%) of the subjects. Disorders of at least 1 foot occurred in 20 (66.7%) subjects with diagnosed diabetes and in 10 (17%) subjects declaring no diabetes. Higher risk and DFD category were correlated with duration of diabetes (r=0.68, p=0.007), glycemic levels (r=0.56, p=0.001), age (r=0.57, p=0.007), and the presence of other diabetes complications. The best predictor of risk in DFD was manual ABI, p=0.001; followed by automatic ABI, p=0.006.

Conclusions

Our results showed that peripheral complications of diabetes, such as DFD, often remain undiagnosed and untreated despite the high risk of developing ulcers. There is a need for multi-center screening studies.

Introduction

Diabetes-related foot disease (DFD) is one of the most serious complications of diabetes. According to the updated definition, it includes at least 1 of the following: diabetic peripheral neuropathy (DPN), peripheral artery disease (PAD), infection, ulcer(s), neuro-osteoarthropathy, gangrene, or amputation [1]. The global prevalence of these disorders, estimated in the largest systematic review to date, is 6.3%. The highest rate was 13.0% in North America, and the lowest was 3.0% in Oceania. The prevalence of DFD on the European continent was estimated at 5.1% [2]. In Poland, the prevalence was 1.2% in specialized diabetes care centers and 2.2% in primary health care. It should be emphasized that this value was estimated on the basis of a single study and only for people with diabetes newly diagnosed within the last 2 years [3]. In addition, these are not the latest data, so the actual prevalence of all stages of DFD in Poland may be much higher, but remains underestimated due to insufficient screening diagnostics. Our experience shows that in patients with foot disorders, etiological diagnosis is rarely performed before the first ulcer occurs, and if it does, the diagnosis is limited to examining the pulse and visual assessment of the feet [4–6].

International [7,8] and national [9,10] expert groups recommend screening and systematic examination of the feet in diabetic patients at least once a year, and even more frequently in the case of arterial and/or sensory disorders. In addition to visual assessment of the patient’s feet and footwear, they recommend examining the 2 most important pathophysiological factors of DFD – peripheral arterial disease and diabetic peripheral neuropathy. The first examination should be immediately after the diagnosis of type 2 diabetes (T2D), and not, as is common practice, at the onset of DFD symptoms. Primary health care (PHC), which cares for the majority of diabetic patients, lacks the necessary resources and tools, including screening equipment and financial procedures. Thus, most patients in Poland are not diagnosed early enough and do not have the chance to receive appropriate preventive interventions as recommended [7–10]. An additional problem is that diabetes can cover up typical symptoms of ischemia, such as intermittent claudication and resting pain. Some patients with neuropathy do not feel pain despite ulcers, deformities, and bone fractures, and therefore wait a long time before consulting a specialist. Many of them, despite a high awareness of the risk of DFD, do not consider systematic foot self-assessment as part of diabetes self-management [11]. All this results in the early and/or asymptomatic stages of the disease often remaining undiagnosed until the development of non-healing ulcers, necrosis, and infections, consequently lead to limb amputation [7,8,12]. The prognosis after diabetic amputation is poor. The 30-day postoperative mortality rate is 3.5–34% [13], and the 5-year mortality rate is 50% [7]. Arterial comorbidities also increase cardiovascular risk [14]. In Poland, major non-traumatic lower limb amputations were performed 19 times more often for diabetes than for other reasons [14]. Worldwide, diabetic amputations are performed as often as once every 30 s [15].

The aim of this study was pilot screening of diabetes-related foot disease in a group of people with glycemic disorders.

Material and Methods

Ethics

The study was conducted in accordance with the principles of the Declaration of Helsinki. The protocol of the study was approved by the local Bioethics Committee at the Collegium Medicum of the Nicolaus Copernicus University in Bydgoszcz (No. KB 515/2018). All participants were informed about the course and purpose of the study and gave written, informed consent to participate in it. We maintained data confidentiality by removing identifying information about participants from the database.

Study Design

For this screening study, we recruited volunteers aged >40 years with glycemic disorders: T2D or prediabetes diagnosed by a primary care physician or diabetologist. In accordance with the guidelines, we adopted 1 of the following criteria as diabetes: symptoms of hyperglycemia and random blood glucose ≥200 mg/dl; or fasting blood glucose ≥126 mg/dl in 2 separate tests; or blood glucose level at 120 minutes of an oral glucose tolerance test (OGTT) ≥200 mg/dl; or HbA1c level ≥6.5%. We adopted the following as the criteria for prediabetes: impaired fasting glucose (IFG): 100–125 mg/dl and/or impaired glucose tolerance (IGT): blood glucose level at 120 minutes of an OGTT between 140 and 199 mg/dl [9]. We excluded those with active leg or foot ulcers, after a lower-extremity amputation (minor or major), mobility impairments, limb edema, and other conditions that prevent blood pressure measurement at the extremities upper and lower limbs (eg, arteriovenous fistula, cast for limb fracture), and non-diabetic peripheral neuropathy. We conducted the research in a specialist chronic wound management clinic in Poland. We planned the recruitment from March 2020 to March 2021, but due to the global COVID-19 pandemic and periods of restrictions, we extended the research period to 2022. A total of 143 subjects were enrolled, but we included 85 eligible subjects in the final statistical analysis, with a total of 170 leg measurements (Figure 1).

Figure 1.

Recruitment and inclusion scheme of participants in the study (PowerPoint, Microsoft Office 2019).

In order to assess PAD, we measured: foot pulse, ankle-brachial index (ABI) (manual and automatic), toe-brachial index (TBI), transcutaneous oxygen pressure (TcPO₂). In order to assess the symptoms of DPN, we assessed indicators of loss of protective sensation (LOPS) – pressure perception and temperature perception, and plantar pressure distribution. We also collected a general medical history (including the reason for seeking screening), performed simple capillary blood tests and used an Intermittent Claudication Questionnaire (ICQ).

We asked participants to fast and avoid alcohol, caffeine, exercise, and smoking 2 hours before the visit. The immediate preparation for the examination included a 15-minute rest in the supine position. We took measurements in a room with a relatively constant ambient temperature of 21–24°C. The conditions were in accordance with the American Heart Association (AHA) [16,17], TransAtlantic Inter-Society Consensus (TASC) [18], and International Working Group on the Diabetic Foot (IWGDF) [7,8] guidelines. All measurements were performed by the same nurse, who was a specialist in surgical nursing with 15 years of experience.

When choosing tools for detecting disorders, we took into account their availability in Poland and the possibility of using them in screening diagnostics on a national scale.

Manual Measurement of ABI and TBI

For manual ABI and TBI measurements, we used the Ankle & Toe Pressure Kit with DMX Digital Doppler (Huntleigh Healthcare, Arjo, Inc., Addison, USA). For ABI measurement, we used the EZ8XS Widebeam 8MHz Doppler Probe and standard adult cuffs in appropriate sizes. For TBI measurement, we used an arterial plethysmography probe and toe cuffs. We kept the same order of measurements for all subjects: right arm, right foot, left foot (on each foot: dorsalis pedis artery, posterior tibial artery). At the end, we repeated the first measurement (on the right brachial artery) to avoid falsely elevated blood pressure associated with emotions and white-coat hypertension. Next, we measured the toe systolic pressure. We calculated the ABI as the ratio of the higher of the dorsalis pedis and posterior tibial systolic blood pressures to the higher of the left or right brachial systolic blood pressures. Similarly, we calculated the TBI: the ratio of the toe systolic pressure to the higher of the left or right brachial systolic blood pressures. We considered ABI 0.9–1.4 [16–21] and TBI ≥0.75 as normal [7,19,20].

Automatic Measurement of ABI

For the automatic measurement, we used an oscillometric device (BOSO ABI-system 100; BOSCH & SOHN, Jungingen, Germany). We measured the blood pressure of the upper and lower extremities simultaneously using 4 cuffs. Then, we calculated the index value as the ratio of ankle pressure to the higher of the left or right brachial systolic blood pressures.

TcPO₂ Measurements

We measured the transcutaneous oxygen pressure with the PRÉCISE 800 device (Medicap Homecare GmbH; Ulrichstein, Germany) using the fluorescence method. We placed the sensor on the dorsum of the foot, 2 cm proximal to the base of the third toe or as close to this location as possible. We avoided areas directly overlying bone or superficial veins. We carefully degreased the measurement site and then glued the sensor to the skin using adhesive discs and contact fluid provided by the manufacturer [22]. After starting the measurement, the skin was heated to 44°C. We read the measurement value according to the manufacturer’s instructions on the 10.1 LCD display. We considered a TcPO₂ value >60 mmHg to be normal [7,8,18].

Loss of Protective Sensation (LOPS)

We assessed pressure perception using a 5.07/10 g Semmes-Weinstein monofilament [7,8,23–25]. We applied the monofilament perpendicular to the test site until it bent for about 2 seconds. Patients were instructed to say “yes” every time they felt a monofilament on their foot. If patients failed to sense the monofilament after it bent, the test site was considered to be insensate [23,24]. The monofilament was applied to 4 plantar points (the big toe, and the heads of the first, third, and fifth metatarsal bones) and in 1 point on the dorsal surface of the foot (between the base of the first and second toes). We pressed each test point 3 times in total, but not directly one after another. We considered a single test point insensitive if at least 2 in 3 pressures were not felt. Subjects who failed to detect 1 or more points were classified as at-risk [23].

We also assessed temperature perception on the skin of the foot using the Thermo Feel instrument (GIME S.p.A.; Gessate; Italy). The tool has 2 flat tips – plastic (which feels “warmer”) and metal (which feels “cooler”). Due to the lack of method standardization, we followed the instructions on the provided leaflet. We applied a metal tip and a plastic tip to the skin surface, asking the subject to distinguish between “warmer” and “cooler” touches. We repeated the assessment 3 times. In total, we assessed 3 points on the foot (2 dorsal and 1 plantar). We assumed that 2 or 3 wrong answers at a given point indicates a distorted perception of temperature. Both tests of perception were performed outside the subject’s field of vision.

Capillary Blood Test

We performed a simple capillary blood glucose test using the reliable Accutrend Plus handheld device and test strips (Accutrend^® Plus system; Roche Diagnostics International AG; Rotkreuz, Switzerland) [26,27]. Measurements were made with a single puncture of the fingertip, with the free-hanging blood drop applied directly to the strip. We considered normal values for glucose to be 70–99 mg/dl [9].

Health-Related Quality of Life and Functional Status

We also used a specific tool to assess the quality of life of patients with PAD – the Polish version of the Intermittent Claudication Questionnaire (ICQ) [28]. The questionnaire consists of 16 questions assessing the impact of claudication on the functional capability of patients and health-related quality of life. The subjects completed the questionnaire on their own. The maximum number of ICQ points was 80. We calculated a coefficient (points/50*100) whose value was in the range of 0–100 points (higher value means worse quality of life and capability).

Plantar Pressure Measurement

Finally, we measured the plantar pressure distribution (static and dynamic) using the Footscan^® device (RSscan International N.V.; Paal, Belgium).

Risk Stratification Criteria DFD

We used the DFD risk stratification criteria based on IWGDF [7,8] and Polish [10] guidelines, according to which, presence of diabetes was sufficient to qualify the subject for a very low risk of disorders (category 0). The presence of any changes in the feet (PAD, LOPS, foot deformity, others) qualify for higher categories (categories 1–3) of DFD risk. Their characteristics are presented in Table 1.

Table 1.

DFD risk categories according to IWGDF guidelines in the study group.

Category ulcer risk		Characteristics	Feet number
			T2D	Pre-D
0.	Very low	No LOPS and No PAD (diabetes only)	20	–
1.	Low	LOPS or PAD	8	13
2.	Moderate	LOPS + PAD, or LOPS + foot deformity or PAD + foot deformity (by visual assessment or by plantar pressure distribution test)	6	0
3.	High	LOPS or PAD, and one or more of the following: history of a foot ulcer end-stage renal disease	18	1
Total*		Feet (patients)	52 (26)	14 (10)
		All patients at risk 1–3/patients at risk 1–3 of one foot/patients at risk 1–3 of both feet	20/7/13	10/5/5

Pre-D – prediabetes; LOPS – loss of protective sensation; PAD – peripheral artery disease; T2D – type 2 diabetes.

^*The IWGDF guidelines also include a lower-extremity amputation, but we excluded amputees from the study.

We defined LOPS as perceptual disorders diagnosed with one of the methods used. We defined PAD as abnormal ABI (either method) or TBI or TcPO₂ or ankle systolic pressure <100 mmHg [7,8,16–18]. If only 1 value was borderline (eg, ABI manual/automatic=0.9), but the others were normal, we did not diagnose PAD.

Statistical Analysis

We presented the descriptive analysis of quantitative and qualitative variables in the form of mean, median, standard deviation, skewness, and kurtosis, as well as minimum and maximum values. We assessed the normality of distribution using the Shapiro-Wilk test. We performed correlation analysis using the Pearson’s r test, and for dichotomous variables, we used point-biserial correlation. We assessed the predictive power of individual screening tools using multiple linear regression. Explanatory variables were included in the model using the entry method. The results of the F test and the adjusted value of the coefficient of determination R2 (R-square) were taken into account as measures of model fit. In the interpretation of indicators, standardized and unstandardized regression coefficients were taken into account and their significance was estimated based on the p-value for the t-test. We considered the value of p≤0.05 to be significant. We performed all calculations in Statistica 10 (StatSoft, Tulsa, OK, USA).

Results

Screening Results

In the study group, 26 (30.6%) subjects were diagnosed with diabetes and 59 (69.4%) with prediabetes; 36 (42.6%) of them had blood glucose levels above normal. Nearly 3/4 of the subjects had at least 1 comorbid condition, most commonly hypertension and cardiovascular disorders. The characteristics of the study group are shown in Table 2. Disorders of at least 1 foot occurred in 20 (66.7%) subjects with diagnosed diabetes and in 10 (17%) subjects with prediabetes, in total in the study group in 30 (35.3%) people on 46 (27.1%) feet.

Table 2.

Characteristics of the study group.

Variable	X (SD)	Me	Min–Max	SK	K	S-W test	p-value
Age, years	58.59 (11.88)	56.00	40.00–88.00	0.48	−0.58	0.95	<0.001
Gender*	Male, n=31(36.5%); Female, n=54 (63,5%)
Comorbidities (diagnosed)*	Yes, n=63 (74.1%); No, n=22 (25.9%)
Hypertension	Yes, n=49 (57.6%); No, n=36 (42.4%)
Heart attack in the past	Yes, n=14 (16.5%); No, n=71 (83.5%)
Stroke in the past	Yes, n=7 (8.2%); No, n=78 (91.8%)
Peripheral arterial disease	Yes, n=4 (4.7%); No, n=81 (95.3%)
Retinopathy	Yes, n=6 (7.1%); No, n=79 (92.9%)
Nephropathy	Yes, n=8 (9.4%); No, n=77 (90.6%)
Type 2 diabetes vs prediabetes	T2D, n=26 (30.6%); pre-D, n=59 (69.4%)
Diabetes duration, months	40.20 (42.2)	20.00	1.00–120.00	7.16	57.24	0.26	<0.001
Diabetes duration, years	3.35 (3.52)	2.00	0.08–10.00	7.16	57.24	0.26	<0.001
Smoking duration, years	16.19 (15.35)	15.00	0.00–55.00	0.56	−0.69	0.89	<0.001
Smoker/past smoker/passive smoker/non-smoker*	n=25 (29,4%)/n=26 (30,6%)/n=8 (9,4%)/n=26 (30,6%)
BMI, kg/m2	26.95 (4.29)	26.77	19.04–39.95	0.38	−0.14	0.97	<0.001
Below normal/normal/above normal*	n=0(0.0%)/n=27 (31.8%)/n=58 (68.2%)
Glucose level, mg/dl	99.94 (25.65)	97.00	45.00–197.00	1.07	2.88	0.92	<0.001
Below normal/normal/above normal*	n=5 (5.8%)/n=44 (51.8%)/n=36 (42.4%)
SBP, right arm	140.88 (22.19)	140.00	100.00–190.00	0.11	−0.78	0.98	0.004
SBP, left arm	136.05 (21.80)	135.00	100.00–185.00	0.33	−0.59	0.97	0.025
DBP, right arm	87.57 (11.03)	90.00	60.00–110.00	−0.29	−0.47	0.97	<0.001
DBP, left arm	86.55 (10.80)	90.00	60.00–110.00	−0.28	−0.45	0.97	<0.001
SBP, right DPA	127.84 (35.71)	130.00	0.00–200.00	−1.28	2.17	0.90	<0.001
SBP, left DPA	128.49 (39.42)	130.00	0.00–200.00	−1.27	2.01	0.91	<0.001
SBP, right PTA	131.10 (32.21)	135.00	30.00–195.00	−0.78	1.03	0.95	0.004
SBP, left PTA	130.06 (32.72)	140.00	20.00–205.00	−0.85	1.17	0.95	<0.001
SBP on the right toe	132.26 (46.72)	133.00	30.00–227.00	−0.51	1.07	0.96	0.006
SBP on the left toe	134.40 (46.67)	130.00	35.00–253.00	0.11	0.75	0.96	0.005
ICQ sum	8.45 (15.84)	0.00	0.00–55.25	1.87	2.05	0.59	<0.001
ICQ index	10.56 (19.81)	0.00	0.00–69.07	1.87	2.05	0.59	<0.001

T2D – type 2 diabetes; pre-D – prediabetes; DBP – diastolic blood pressure; DPA – dorsalis pedis artery; ICQ – intermittent claudication questionnaire; K – kurtosis; Me – median; Min–Max – minimum and maximum distribution value; PTA – posterior tibial artery; p – value of significance; SBP – systolic blood pressure; SD – standard deviation; SK – skewness; S-W – value of the Shapiro-Wilk test; X – mean.

^*Data presented as numbers and percentages.

Only 6 (23.1%) diabetics had both feet without disease, 7 (26.9%) had DFD in 1 foot, and 13 (50%) had DFD in both.

All PAD/LOPS diagnoses occurred in subjects with above-normal fasting glucose. Eighteen (34.6%) feet in patients with diabetes and 1 foot in patients with prediabetes were at high risk of ulceration (Table 1).

We identified PAD in 39 (23%) feet by manual ABI, in 32 (19%) by automatic ABI, and in 25 (14.7%) by TBI and TcPO2. We identified arterial stiffness in only 1 patient by manual ABI measurement. We diagnosed loss of protective sensation in 14 (16.5%) subjects with loss of pressure perception, and 18 (10.6%) subjects had loss of temperature perception (Table 3).

Table 3.

Values of assessed indicators of peripheral arterial disease and loss of protective sensation.

Variable	X±SD (Me)	Min–Max	SK	K	S-W test	p-value
Manual ABI	1.00±0.23 (1.01)	0.25–1.45	−1.29	2.28	0.90	<0.001
Below normal/normal/above normal*	R: n=19 (22.3%)/n=65 (76.5%)/n=1 (1.2%) L: n=20 (23.5%)/n=64 (75.3%)/n=1 (1.2%)
≤0.50/0.90–0.51/0.99–0.91/1.4–1.0/≥1.4	R: n=4 (4.7%)/n=15 (17.6%)/n=8 (9.5%)/n=57 (67.0%)/n=1 (1.2%) L: n=7 (8.2%)/n=13 (15.3%)/n=8 (9.5%)/n=56 (65.9%)/n=1 (1.2%)
Automatic ABI	0.99±0.25 (1.07)	0.26–1.34	−2.27	5.58	0.74	<0.001
Below normal/normal/above normal*	R: n=16 (18.8%)/n=69 (81.2%)/n=0 (0.0%) L: n=16 (18.8%)/n=69 (81.2%)/n=0 (0.0%)
≤0.50/0.90–0.51/0.99–0.91/1.4–1.0/≥1.4	R: n=6 (7.1%)/n=10 (11.7%)/n=9 (10.6%)/n=60 (70.6%)/n=0 (0.0%) L: n=6 (7.1%)/n=10 (11.7%)/n=7 (8.20%)/n=62 (73.0%)/n=0 (0.0%)
Manual TBI	0.97±0.32 (0.95)	0.21–1.86	−0.40	1.81	0.95	<0.001
Below normal/normal i.e. ≥7.5*	R: n=11 (12.9%)/n=74 (87.1%) L: n=14 (16.5%)/n=71 (83.5%)
TcP02	73.54±13.60 (76.00)	28.00–96.00	−1.16	1.16	0.90	<0.001
Below normal/normal i.e. ≥60 mmHg	R: n=13 (15.4%)/n=72 (84.6%) L: n=12 (14.1%)/n=73 (85.9%)
LOPS	R: 8 (9.5%)/9 (10.6%)	L: 6 (7.1%)/9 (10.6%)				<0.001
SWME/TFE

ABI – ankle-brachial index; K – kurtosis; LOPS – loss of protective sensation; Me – median; Min–Max – minimum and maximum distribution value; p – value of significance; SD – standard deviation; SK – skewness; S-W – value of the Shapiro-Wilk test; SWME – Semmes-Weinstein monofilament examination; TBI – toe-brachial index; TFE – Thermo Feel examination; X – mean.

^*Data presented as numbers and percentages.

Only 5 (16.7%) of 36 participants with DFD had been previously examined for foot disorders, including 4 due to limb pain (severe limb ischemia). Despite this, they all claimed that they did not have DFD because they did not have an ulcer. The remaining diabetics had not had even a single visual assessment since their diabetes diagnosis.

DFD Risk Categories

Higher DFD risk categories were significantly correlated with ankle systolic blood pressure (r=0.70, p=0.001), diagnosis of other diabetes complications such as PAD (r=0.69, p=0.01), nephropathy (r=0.69, p=0.035), coronary artery disease and/or previous myocardial infarction (r=0.58, p=0.02), duration of diabetes (r=0.68, p=0.007), and glycemic levels (r=0.56, p=0.001), as well as age (r=0.57, p=0.007) and male sex (r=0.46, p=0.035).

Predictive Values of Detection Tools

Linear regression analysis showed that the best predictor of DFD risk was manual and automatic ABI and ICQ (p<0.01). Other PAD screening tools were statistically significant predictors of only high-risk DFD (categories 2 and 3). The LOPS assessment tools we used were not significant risk predictors in any of the analyzes performed, as the sample size with a positive diagnosis was too small (Table 4).

Table 4.

Linear regression model describing the relationship between ulcer risk and its predictors.

Dependent variable		F	df1	df2	p	R 2
Ulcer risk	L	29.11	7	79	<0.001	0.63
	R	45.35	7	79	<0.001	0.73
	All	70.24	7	164	<0.001	0.68
		B	SE	Beta	t	p
I. constant		3.02	0.25	–	12.10	<0.001
ICQ		0.00	0.00	0.18	2.09	0.040
TCP02		−0.02	0.01	−0.17	−1.91	0.058
Automatic ABI		−0.59	0.21	−0.28	−2.81	0.006
Manual ABI		−0.80	0.23	−0.35	−3.49	0.001
TBI		−0.11	0.11	−0.07	−0.95	0.347
SWME		0.17	0.10	0.14	1.76	0.081
TFE		<0.01	0.02	−0.02	−0.16	0.875
II. constant		1.03	0.17	–	6.14	<0.001
ICQ		0.01	<0.01	0.28	3.69	<0.001
TCP02		−0.01	<0.01	−0.33	−2.70	0.008
Automatic ABI		−0.59	0.21	−0.28	−2.81	0.006
Manual ABI		−0.80	0.23	−0.35	−3.49	0.001
TBI		−0.02	0.01	−0.20	−2.36	0.019
SWME		0.48	0.25	0.18	1.94	0.054
TFE		0.44	0.45	0.13	0.99	0.324

ABI – ankle-brachial index; Beta – predictor parameters; df – degrees of freedom; F – the test statistic value; ICQ – intermittent claudication questionnaire; L – left foot analysis; p – value of significance; R – right foot analysis; R2 – coefficient of determination; SE – Standard Error; SWME – Semmes-Weinstein monofilament examination; t – student’s t-test statistic value; TBI – toe-brachial index; TFE – Thermo Feel examination; I – analysis of all risk levels (0–3); II – analysis of only moderate and high risk (2 and 3).

Discussion

In this pilot study, we showed that the incidence of foot disorders was high, both in the group with diagnosed diabetes and in the prediabetic group. Ulcer risk categories ranged from low to high, but it is worrying that such a high percentage of high and moderate risk had not been diagnosed and treated before. More than half of the subjects had diabetes for 2–10 years, and only a few were examined, and only after symptoms of significant limb ischemia occurred. Many of them also had other long-term complications of diabetes, including nephropathy, listed as an additional high-risk factor in the IWGDF risk stratification system [7]. These complications were significantly correlated with the level of DFD risk. Their diagnosis indicates long-term and advanced diabetes [3,29] and should be an alarm signal for the primary physician to examine the feet, even if the patient does not report problems. Poor screening can delay diagnosis and treatment of the disorder, and underestimates the actual number of patients with DFD. Underreporting the number of patients leads to underestimation of the problem. Zhang et al [2] noted that the low incidence of DFD in some regions of the world, such as Australia and Oceania, was due to poor DFD screening, and not due to the actual lower incidence. Our study shows that the prevalence of DFD in Poland is also underestimated.

The ARETAEUS1 analysis [3] cited in Zhang’s literature review [2] relied on questionnaire data from physicians rather than patient diagnosis. The authors did not report what diagnostic criteria for DFD were used by reporting physicians, especially in non-diabetes clinics where access to screening diagnostic tools is limited. Therefore, we assume that mainly symptomatic and advanced stages of disorders that could be diagnosed without specialized diagnostics were reported. However, it is difficult to speculate whether they included only deformations, necrosis, and foot ulcers visible to the naked eye, or also symptomatic forms of PAD and DPN. In addition, the ARETAEUS1 analysis included only patients with newly diagnosed diabetes and a short average disease duration of about 10 months [3]. Therefore, we believe that the actual incidence of all stages of DFD in Poland may be close to the value reported by other European countries with a similar degree of industrialization [2] and may be even 3 times higher than the suggested 1.7% [3]. Both diabetes and its complications can be asymptomatic for a long time [30,31]. Therefore, DFD symptoms occurred in patients who were probably unaware of their foot conditions. We would like to emphasize that foot disorders such as PAD, LOPS and deformities also occurred in the prediabetic group. Therefore, further screening in at-risk groups is necessary, as well as assessment of the actual prevalence of all stages of DFD, not only the most advanced ones in the form of ulcers.

In this study, we used several of the recommended methods for diagnosing PAD and LOPS to increase detection, especially of subclinical cases. The results confirmed that the best predictor of PAD-related DFD risk was the ankle-brachial index. Previous concerns about its usefulness in patients with T2D [19,21,32] have not been confirmed [7,19,33]. ABI was diagnostically consistent with duplex ultrasound and angiography in patients with T2D [33,34]. Some authors report that automated oscillometric methods of measuring ABI are more reliable in diabetic patients than manual Doppler measurement [19,34,35]. In our analysis, the predictive value of both methods was similar. Others, including TBI and TcPO2 were significantly predictive, but only in patients with more advanced disorders and higher DFD risk category. According to the authors, ABI is more specific for PAD than TBI (despite similar sensitivity) [21,35–37]. In contrast, TcPO2 measurement is recommended more for predicting the healing process [12,37]. The ICQ, despite its good predictive value, is recommended as a complement to objective methods, as this tool is used to subjectively assess the impact of disorders on patients’ functional ability and quality of life [30]. None of the LOPS screening methods we evaluated (due to the small number of diagnosed disorders) showed sufficient predictive power. A review of the literature suggests that the monofilament test is sensitive and comparable to a nerve conduction test [23–25]. Due to the wide variation in DPN, some authors have suggested combining several simple screening methods to assess detectability, such as assessing temperature perception with pressure, vibration, and tendon reflex perception [38]. This area certainly needs further research.

Limitation

The limitation of this study was the small sample size. However, this was only a pilot study, and its purpose was not to assess the actual prevalence of DFD, but to assess the need for population-based screening in the at-risk group. We used only a simple capillary blood test to assess fasting blood glucose. Future population studies should include both DFD screening and full diagnosis of diabetes and prediabetes.

Conclusions

To our knowledge, this is the first risk screening study for diabetic foot disease (DFD) in Poland using so many diagnostic tools. Despite the small sample size and pilot nature of the study, we showed that peripheral complications of diabetes are extremely common with many patients remaining undiagnosed and untreated despite the high risk of ulceration. This study also confirms that systematic DFD screening should begin immediately after T2D diagnosis, as disorders can progress even in the subclinical phase of the disease. Multi-center screening studies are necessary. A sufficient test for rapid screening is manual or automatic ABI. The assessment of DPN requires further research into the use of combined screening methods.