] Diab Comp, 62 67-174
Medical Risk Factors in Diabetic Patients With Foot Ulcers and Severe Peripheral Vascular Disease and Their Influence on Outcome fn tpelqvist an arsson Carl-David Agardh
ABSTRACT The association between medical risk factors and the outcome of foot ulcers was evaluated in 208 consecutive diabetic patients with severe peripheral vascular disease (systolic toe blood pressure 545 mm Hg). All patients were treated and followed by the same foot care team. Eighty patients healed primarily, 83 healed after a minor or major amputation, and 45 died. The systolic toe blood pressure was higher among primary healed (30 + 13 mm Hg) compared with amputated (22 & 15 mm Hg; p < 0.001) and deceased patients (20 + 14 mm Hg; p < 0.001). The patients were comparable regarding age, sex, and diabetes and wound duration. Only 41 (19%) patients had intermitten claudication, whereas 153 (77%) lacked palpable pedal pulses, 36% of whom healed primarily. Rest pain occurred in 72 (33%) patients,
F
INTRODUCTION oot ulcer is a common complication to diabetes mellitus.1-3 Accelerated atherosclerosis leading to impaired circulation in the legs is one of the most important causes of gangrene and
Departments of Internal Medicine (J.A., C.-D.A.), and Orthopedic Surgery (J.L.), University Hospital, Lund, Sweden Reprint requests to be sent to: Dr. Jan Apelqvist, Department of Internal Medicine, University Hospital, S-221 85 Lund, Sweden. 0 1992 Journal of Diabetes and Its Complications
38 (47%) of whom had an amputation and 18 (25%) who healed primarily (p < 0.01). Peripheral edema and proteinuria were more common among patients who healed after amputation compared with those who healed primarily (p < 0.001 and p --L0.01, respectively). Signs of sensory neuropathy were found in 158 (77%) patients. There were no differences concerning cardiovascular disease, smoking habits, or short-term metabolic control between patients who healed primarily or after an amputation. In conclusion, diabetic patients with foot ulcers and severe peripheral vascular disease with low systolic toe blood pressure were not excluded from the possibility of primary healing. The most important risk factors for amputation were a systolic toe pressure of less than 30 mm Hg, peripheral edema, rest pain, and proteinuria.
amputation. 4-6 In previous studies a critically low systolic toe blood pressure has been shown to indicate a high probability for development of gangrene and the need for amputation.7-‘2 The aim of the present prospective study was to describe the association between clinical risk factors and outcome in a large group of consecutively presenting diabetic patients with foot ulcers and a systolic toe blood pressure equal to or below 45 mm Hg. 0891-6632/92/$5,00
168
1 Diab Comp 1992; 6:3
AI’ELQVIST ET AL.
MATERIALS AND METHODS
within that period, as intact skin at the time of death. Presence of precipitating factors for the development of ulcers were defined as previously described.15
Patients. In this prospective study 208 consecutively presenting diabetic patients (104 males and 104 females) referred to the Department of Interna Medicine because of foot ulcer and with a systolic toe blood pressure of 45 mm Hg or less, were investigated. The patients were mainly treated as outpatients, but in cases of deep infection associated with septic conditions, amputation, or intercurrent disease, they were treated as inpatients. Patient characteristics at enrollment are given in Table 1. The median duration of foot ulceration prior to the first admission was 8 weeks (range, 0.5-157 weeks) and the wound healing time after admission was 4 months (range, l-44 months). The follow-up time for all patients was 7 months (range, 7-50 months).
Medical Treatment. A physical examination was performed at enrollment and regularly during the study. Saline dressings and local enzymatic treatment were used as cleansing agents. Exudative wounds were treated with absorption dressings. Clean ulcers were covered with occlusivelsemiocclusive membranes or ointment dressings to maintain moisture. Protective shoe wear were used in most patients. In a small proportion, total contact walking casts or special orthotic appliances were used. Bed rest was not recommended to patients except in cases of a septic or toxic condition. When resting in a sitting position they were instructed to keep the foot elevated to hip level. All patients with plantar ulcers had an individually fitted orthotic appliance (e.g., specially made shoes, insoles, walking cast) which maintained ambulation, but with a modification of shoe wear to avoid pressure on the ulcer area. There was no difference in outcome between different orthotic appliances (data not shown). Metabolic control was improved when possible. A majority of the patients were treated with insulin (Table 11, and there was no difference in treatment mode at admission and at final outcome (data not shown). Peripheral edema was usually treated by diuretics. P-blocking agents were usually changed to
Ulcer Definition. At enrollment and throughout the study all lesions were characterized by the same foot care team and documented by color photographs as previously described. 13-15 Each patient was represented by one lesion below the ankle. Lesions were classified according to Wagner,16 and the most superficial ulcer included in the study was a lesion through the full thickness of the dermis. The ulcers were classified as superficial (n = 73), deep (n = 23), osteomyelitis and/or abscess (n = 26), minor gangrene (n = 49), or major gangrene (n = 37). Wound healing was defined as intact skin for at least 6 months or, in cases when the patient died
TABLE 1. PATIENT CHARACTERISTICS Primary II Age (yr) Sex (M/F) Duration of diabetes (yr) Wound duration in weeks (range) Treatment (n) Diet Oral agents Insulin HbAl, (%) Smoking habits (n) Smokers/ex-smokers Never smoked Signs of neuropathy (VPT > 30) (n) Severe retinopathy (n) Diabetic nephropathy, Creatinine (FmoVL)
Healed
83 69 ? 11 41142 18 + 13 9 (0.5-105)
10 29 40 8.5 * 1.9
6 22 55 8.3 ? 1.8 34 48 66 (9) 27 28*” 128 f 108
62 (3) 26 11 111 * 71
VPT, vibration perception threshold. Note: Wound duration is given as median (range); other values are means f * p < 0.05, ** p < 0.01, *** p < 0.002 compared with prima y healed. Numbers within bars indicate patients with nonmeasurable
VPT.
Amputated
80 71 * 11 43137 16 c 12 8 (0.5-157)
38 42
Aibustix pos (n)
AT ENROLLMENT
SD
Deceased 45 77 * 11* 20125 13 f 11 6 (1.0?157) 5 21 19 8.1 2 2.1 15 29 30 (12) 7 187 -t :98*‘*
1 Diab Comp 2992; 6:3
calcium channel blockers, diuretics, or angiotensinconverting enzyme inhibitors. Antibiotics, usually flucloxacilline or cefalosporins continued with metronidazole, were used when clinical signs of infection were present, i.e., cellulitis, deep abscess, and/or osteomyelitis, or when progressive gangrene was present. Analgesic agents used were paracetamol and dextropropoxiphen, although in some patients narcotics had to be used, and in cases of painful polyneuropathy, carbamazepine or tricyclic antidepressive agents were given. Surgical Treatment. Surgical debridement of the lesions was performed when required. Incision and drainage were performed in cases of deep plantar abscess, and resection was performed in cases of severe osteomyelitis not responding to antibiotic treatment. Amputation was carried out in cases of progressive gangrene, intolerable pain despite analgesia, and toxic or septic conditions not responding to conservative treatment. The lowest level employed for amputation was at the metatarso-phalangeal level. Eighty-two patients had femoral angiography, and 29 of these patients had either reconstructive vascular surgery (n = 19) or percutaneous transluminal angioplasty (n = 10). Clinical Examinations and Definitions. Duration of diabetes was calculated from the year of diagnosis until enrollment in the study. Hypertension was considered present when the systolic and/or diastolic blood pressure exceeded 160 and 90 mm Hg, respectively, or when the patients were treated with antihypertensive medication. Angina pectoris was defined as a history of retrosternal chest pain with a duration of less than 30 min, aggravated by exercise or cold environment, and relieved within 5 min of administration of nitroglycerin. Definitive previous myocardial infarction (MI) was defined either by the presence of resting electrocardiogram (ECG) changes consistent with an MI according to the Minnesota code or by the presence of two of the following criteria: (1) a history of retrosternal pain at onset of the attack according to the WHO Myocardial Infarction Community Registers,17 (2) development of ECG changes indicative of MI according to the Minnesota Code,18 (3) peak values of S-ASAT (EC 2.6.1.1) twice the upper limit of normal within 72 h, or (4) autopsy evidence of previous MI. Congestive heart failure was defined according to the New York Heart Association” and was verified with clinical examination and chest x-ray. Nonischemic heart disease was considered when other causes of heart disease such as valvular heart disease and atria1 fibrillation were verified through clinical examination, x-ray, and ECG. Cerebrovascular disease was defined as a history of
DIABETIC FOOT ULCERS
169
neurological deficit according to the Veteran’s Administration Cooperative Study of antiplatelet agents.” Smoking habits were divided according to the following: (1) smokers, i.e., daily smokers including those who had stopped smoking within 1 year prior to the study, (2) previous smokers, and (3) nonsmokers. Nonsmokers were defined as persons who never had had any daily nicotine consumption. The amount of smoking was registered as number of years smoked. Retinal examination was performed through a dilated pupil by ophthalmoscopy or fundus photography. The patients were examined within 1 year of the time of enrollment and were classified into three groups depending on the degree of retinopathy. The most seriously affected eye was used for the evaluation. The groups were (1) no signs of retinopathy, (2) simplex retinopathy, and (3) severe retinopathy (preproliferative retinopathy, proliferative retinopathy, or previous photocoagulation). Four patients were not examined due to their medical condition. Clinical signs of nephropathy were considered to be of diabetic origin when persistent proteinuria (positive Albustix) was present for more than 6 months and as nondiabetic when other causes of kidney disease were verified. The vibration perception threshold (VPT) was measured with a biothesiometer (Bio Medical, Newbury, OH, USA; scale l-50) at the tip of digit I as previously described. l5 Values exceeding 30 arbitrary units were considered as an indication of sensory neuropathy. 21,22 In 24 patients, VPT was not possible to measure due to their mental status. Edema was defined as swelling of the foot so pronounced as to leave a clear imprint after pressure by a finger. Peripheral pulses of the lower extremity (dorsal pedal and posterior tibial) were examined by manual palpation and registered as present or absent. Only findings from the foot with the ulcer are presented. Rest pain was defined as severe and persistent when localized to the ulcerated foot and relieved by lowering of the foot. Intermittent claudication was defined as a history of recurrent cramping pain or tightness in the calf induced by exercise and relieved by rest. Deep venous thrombosis had been verified by previous venous phlebography. Ulcer duration was defined as the number of weeks from the development of the ulcer until entry in the study. Blood samples were taken at admission and every third month until healing had occurred. Glycosylated hemoglobin (HbA,,) levels were measured by ion-exchange chromatography using commercially available microcolumns (BIO RAD, Richmond, CA, USA). Norma1 value is less than 6,0%.
170
APELQVIST ET AL.
j Diab Comp 1992; 6:3
TABLE 2. TYPE OF ULCER IN RELATON TO FINAL OUTCOME
Primary Healed Type of Ulcer Superficial Deep Abscess/osteomyelitis Minor gangrene Major gangrene n
Deceased
n
n
n
73
51 (70)
1
21
22 26 49 37 208
12 (55) 13 (50) 4 (8) 0 (0) 80 -
2 10 37 32 83
8 3 8 5 45
The systolic ankle and toe blood pressure were measured with strain-gauge and doppler techniques as previously described.’ The ankle pressure was not measurable in 23 patients (mainly because of incompressible arteries). Statistics. Values are given as means -+-SD and, in the case of deviation from an estimated normal distribution, as median and range. Differences between or within groups were calculated using the Mann-Whitney U test (two-tailed) and the x2 test. The simultaneous influence of several possible risk factors for amputation was investigated by means of logistic regression. RESULTS Primary healing occurred in 80 (38%) patients, 83 (40%) healed after amputation, and 45 (21%) died before healing had occurred (9 of these patients had amputation prior to death) (Table 1). Of those 83 patients who healed after amputation, 27 had had an amputation below the ankle (15 toe/s; 12 TMT), 47 below the knee, and 8 at the thigh level. The cause of death was considered as cardio-cerebrovascular in 75% of the cases. There were no differences in age, sex, duration of diabetes, and wound duration between the groups with the exception of patients who died unhealed being older than patients with primary healing (p < 0.05; Table 1). There was a gradual decline in primary healing rate
TABLE 3. ULCER LOCALIZATION
(o/o)
with increasing tissue involvement of the ulcer (Table 2). Four patients with minor gangrene healed primarily after mummification and auto-amputation. There were no differences in healing rate at different ulcer localizations (Table 3). With multiple ulcers, however, only one of 25 patients healed primarily. The presence of external precipitating factors and their relation to outcome is described in Table 4. In 80% (148 of 208) of the patients such a factor could be verified, of which ill-fitting shoes/socks was the most common. Systolic Ankle and Toe Blood Pressure. There was no difference in mean systolic ankle blood pressure between patients with primary healing and those who healed after amputation or died (Table 2); however, a systolic ankle pressure below 40 mm Hg was not consistent with primary healing (data not shown). Patients who healed primarily had higher systolic toe blood pressure levels compared with those who had an amputation or died (p < 0.01 and p < 0.001, respectively). No difference was seen between patients who had amputation or died. A gradual increase in primary healing rate was seen with increasing toe pressure (Figure 1). Primary healing was more common among patients with a toe pressure greater than 30 mm Hg (41 of 73) compared with those below that level (39 of 135; p < 0.001). Still, only 56% healed primarily with a toe pressure between 31 and 45 mm Hg, and 19 patients developed gangrene within these pressure levels.
IN RELATION
TO FINAL OUTCOME
Ulcer Localization
n
Primary Healed n (o/o)
Dig I, dorsal and plantar surface Dig II-V, dorsal and plantar surface Metatarsal head I-V Mid-foot and heel Dorsum of the foot Multiple ulcers (23)
64 50 10 34 24 25 208
30 (47) 19 (38) 6 (60) 15 (44) 9 (38) 1 (4) 80 -
n
Amputated
n
Amputated n
20 23 1 14 6 18 83
Deceased
n
14 8 3 5 9 6 45
DIABETICFOOT ULCERS 171
1 Diab Comp 1992; 63
TABLE 4. EXTERNAL PRECIPITATING FACTORS AND THEIR RELATION TO OUTCOME Precipitating Factor Ill-fitting shoes/socks Stress ulcer Acute mechanical trauma Paronychia Decubital ulcer Miscellaneous Reason unknown n
n
Primary Healed n
Amputated n
Deceased n
58 17 26 12 13 22 60 208
29 9 9 7 2 4 20 80
26 6 12 0 4 8 27 83
3 2 5 5 7 10 13 45
Development of gangrene and presence of multiple ulcers was strongly related to low toe blood pressure levels (O-15 mm Hg, 60% gangrene; 16-30 mm Hg, 42%; 31-45 mm Hg, 26%). There was no correlation between distal blood pressure levels and depth, size, and localization of ulcer with the exception of patients with multiple ulcers or patients developing gangrene having lower ankle and toe pressure levels compared with all other sites or types of lesions (p < 0.001 and p < 0.001, respectively). There was no difference in systolic ankle pressure levels comparing patients with primary healing and toe pressure less than 30 mm Hg, and those healed after amputation and toe pressure less than 30 mm Hg (68 ? 36 mm Hg versus 74 + 33 mm Hg; n = 45 and n = 27, respectively. Eleven patients healed primarily after angioplasty or vascular surgery (O-15 mm Hg, n = 5; 16-30 mm Hg, n = 2; 31-45 mm Hg: n = 4) with a mean preoperative toe pressure of 28 mm Hg, (range, O-45 mm Hg) and a mean postoperative pressure of 63 mm Hg (range, 25-90 mm Hg).
74% of the patients had no palpable pedal pulses, but the presence of pedal pulses was not related to the outcome (Table 5). Only 41 (19%) had intermittent claudication, and 72 (33%) had rest pain. Rest pain was more common among patients who healed after amputation compared with primary healed (p < O.Ol), but 18 patients (25%) with rest pain healed primarily. Patients with rest pain and primary healing had a higher toe pressure compared with those with rest pain who healed after amputation 30 +- 14 mm Hg (n = 18) versus 20 rt 14 (n = 38) mm Hg (p < 0.001). There was no difference in ankle pressure levels in the corresponding groups (82 + 42 mm Hg versus 79 + 41 mm Hg). Four of 18 patients with rest pain and primary healing had vascular surgery or angioplasty and one healed after autoamputation. Previous deep venous thrombosis was more common among patients with amputation compared with primary healed (p < 0.05). Neuropathy, Retinopathy and Nephropathy. VPT values above 30 arbitrary units were seen in 158 patients (76%), which represent 85% of patients with a measurable WT. There was no difference in presence of sensory neuropathy between primary healed patients and those who healed after amputation or died (Table 1). Severe retinopathy was seen in 60 patients (29%) and was equally present in the three groups.
Edema, Pedal Pulses, Rest Pain, and Claudication. Peripheral edema was present in almost onehalf of the patients and was most common among patients who healed after amputation or died (p < 0.001 and p < 0.001, respectively; Table 5). A total of
Systolictoe Dressure
0
1
20 I
40 I
60 I
80 I
100% I
N
hmHg) 0 - 15
57
16 - 30
78
31 - 45
73
FIGURE 1 Systolic toe blood pressure levels (mm Hg) in patients with primary healing (open area), amputation (hatched area), or death before healing occurred (solid area). N indicates number of patients.
172
APELQVIST
J Diab Camp 1992; 6:3
ET AL.
TABLE 5. EDEMA,
PEDAL PULSES, REST PAIN, CLAUDICATION, AND PREVIOUS THROMBOSIS IN RELATION TO FINAL OUTCOME Primary Healed ?I (o/o)
n
80
Edema Absence of pedal pulses Rest pain Claudication Previous deep venous thrombosis
25 56 18 20 7
(31) (70) (22) (25) (9)
DEEP VENOUS
Amputated n (o/o)
Deceased n (%I
83
45
48** (58) 62 (75) 38** (46) 17 (20 17* (20)
28** (62) 35 (78) 16 (36) 4 (9) 9 (20)
* p < 0.05, ** p < 0.001 compared with primary heaIed.
Persistent albuminuria was most common among patients who healed after amputation (n = 28) compared to patients with primary healing (n = 11; p < 0.01) (Table 1). The majority of the patients had signs of cardiovascular disease; however, there were no differences concerning such diseases or smoking habits between patients who healed primarily or healed after amputation. Nevertheless, the number of years smoked was related to the outcome, because smokers with amputation had a longer duration of smoking compared to smokers with primary healing (17 < 0.05) (data not shown). Simultaneous Analysis of Risk Factors for Amputation. To investigate the simultaneous influence of different possible risk factors for amputation logistic regression analysis were performed. First, we removed those 45 patients who died before healing. Second, the conventional explanatory variables age and sex were investigated; we found that neither age nor sex had a significant influence on the risk for amputation. As a third step for each of a long list of possible explanatory variables, a logistic regression analysis of the amputation risk was performed using only those factors showing promise of influence, the latter being defined as p < 0.15. Finally, in a stepwise manner, logistic regression analysis was performed with remaining variables simultaneously, starting with a full model and removing nonsignificant variables one by one. The result was that only three variables remained, i.e., diabetic nephropathy, rest pain, and systolic toe blood pressure (Table 6). In the present
TABLE 6. MULTIPLE STEPWISE REGRESSION ANALYSES OF RISK FACTORS FOR AMPUTATION
Diabetic nephropathy Rest pain Systolic toe blood pressure
Odds Ratio
p Value
3.0 2.4 0.95
0.048 0.037 0.001
study a 1 mm Hg increase in the systolic toe pressure would decrease the risk for amputation by 5% (odds ratio, 0.95) and a corresponding increase by 10 mm Hg gives an odds ratio of 0.951° (0.61) and a risk reduction by 39%. DISCUSSION In the present study putative medical risk factors were studied in diabetic patients with foot ulcers and severe peripheral vascular disease. Primary healing occurred in 38% of the patients. To eliminate for the physician being influenced by knowledge of clinical data in his medical decision making, the indications accepted for amputation were progressive gangrene, septic condition, and rest pain not responding to conservative treatment. All patients were followed until final outcome and were treated by the same team of physicians in order to eliminate for differences in strategy of treatment. For the same reason strict amputation criteria were maintained. The selection of patients with foot ulcer and a systolic toe blood pressure equal to or below 45 mm Hg was based on a previous study where such pressure indicated a high probability for development of gangrenes7 Furthermore, amputations above that level were seldom preceeded by development of gangrene.7 In the present study the systolic toe pressure level was strongly related to the outcome. The finding indicates that patients with a systolic toe blood pressure level below 30 mm Hg are at high risk for development of gangrene, and it might be valuable to identify diabetic patients with foot lesions in need of further vascular intervention as suggested by the European Consensus Document on Critical Limb Ischaemia.23 However, one of four patients with a toe pressure between 31 and 45 mm Hg developed gangrene . There was no difference in systolic ankle pressure levels between patients with primary healing, those that healed after amputation, or deceased patients. But an ankle pressure below 40 mm Hg was inconsis-
DIABETIC
j Diab Comp 1992; 6:3
tent with primary healing as has previously been shown. 7 In the present study one of three patients (39%) had an angiography. Of these 35% had either angioplasty or vascular surgery, and 38% of these interventions resulted in primary healing. These results might have been different if the decision for angiography had been made earlier and in more patients whose condition would allow for mechanical angioplasty and with routine visualization of the pedal arch. This statement is further stressed by the finding that although 77% of these patients had absence of both pedal pulses, absence of pedal pulses were not more common among patients who healed primarily compared with those who required an amputation. Furthermore, only 19% had intermittent claudication, and 33% had rest pain. Of those patients with rest pain, 25% healed primarily. The fact that some patients with rest pain healed primarily indicates the importance of evaluating other causes of pain than ischemia. Despite the difficulties in evaluating pedal pulses, rest pain, and intermittent claudication in these patients, the findings indicate that a clinical evaluation of peripheral vascular disease can give valuable but imperfect information concerning the need for further vascular intervention. Peripheral edema was found in almost 50% of the patients and was most common among patients who had had amputation or died which is in agreement with previous observations where edema, often from multifactorial causes,l* has been associated with the development and of importance for the outcome of foot ulcers in diabetic subjects.‘4J4*25 The most common factors associated with edema have been congestive heart failure, proteinuria, and neuropaPeripheral edema impairs wound thy. 14,24-28,26~27~28 healing probably by decreasing the microcircuIation.27 Persistent proteinuria was seen in 22% of the patients and was most common among those who healed after amputation. These findings are in agreement with the study by Nelson et al.29 and in studies of atherosclerotic occlusive disease in the lower extremity in diabetic patients.30,3* Proteinuria has also been suggested to be a predictor of cardiovascular mortality32 and an indicator for widespread cardiovascular disease in diabetic patients.33 In summary, the present study shows that diabetic patients with foot.ulcers and low systolic toe blood pressure are not excluded from the possibility of primary healing, as two of five healed primarily. The most important factor related to the outcome was a systolic toe blood pressure of less than 30 mm Hg. However, rest pain, proteinuria, and peripheral edema were also important factors related to amputation and death. ClinicaI signs of peripheral vascular
FOOT ULCERS
173
disease give important but imperfect information concerning the need for vascular intervention in these patients. ACKNOWLEDGMENT This work was supported by the Swedish Research Council (Grant No. 02872), the Hoechst Diabetes Foundation, The Swedish Diabetes Association, and The Medical Faculty, University of Lund. REFERENCES 1.
Rosenqvist LJ: An epidemiological survey of diabetic foot problems in the Stockholm County 1982. Acfa Med Stand Suppl687:55-60,
1984.
2.
Levin ME, O’Neal LW (ed), The Diabetic Foot. St. Louis MO: Mosby, 1988.
3.
Edmonds ME: The diabetic foot: Pathophysiology and treatment, in Watkins PJ (ed), Long-term complications of diabetes. Baillieres Clin Endocrinol Mefab 15: 889-916,
1986.
4.
Klamer TW, Towne JB, Bandyk DF, Bonner MJ: The influence of sepsis and ischemia on the natural history of the diabetic foot. Am Surg 53:490-494, 1987.
5.
Gutman M, Kaplan 0, Skornick Y, Klausner JM, Leicuk S, Rozin RR: Gangrene of the lower limbs in diabetic patients: A malignant complication. Am J Surg 154:305-308, 1987.
6.
Bild DE, Selby JV, Sinnock P, Browner WS, Braveman P, Showstack JA: Lower-extremity amputation in people with diabetes. Epidemiology and prevention. Dia-
7.
Apelqvist J, Castenfors J, Larsson J, Stenstriim A, Agardh C-D: Prognostic value of systolic ankle and toe blood pressure levels in outcome of diabetic foot ulcer.
8.
Holstein P, Lassen NA: Healing of ulcers on the feet correlated with distal blood pressure measurements in occlusive arterial disease. Acfa Orthop Stand 51:
betes Care 12:24-31,
1989.
Diabetes Care 12:373-378,
995-1006,
9.
1989.
1980.
Ramsey DE, Manke DA, Sumner DS: Toe blood pressure a valuable adjunct to ankle pressure measurement for assessing peripheral arterial disease. 7 Cardiovasc Surg 24143-48,
1985.
10.
Faris I, Duncan H: Skin perfusion pressure in the prediction of healing in diabetic patients with ulcers or gangrene of the foot. J Vascular Surg 4:536-540, 1985.
11.
Carter SA: The relationship of distal systolic pressures to healing of skin lesions in limbs with arterial occlusive disease with special reference to diabetes mellitus.
12.
Bone GE, Pomajzl MJ: Toe blood pressure by photoplethysmography: An index of healing in forefoot amputation. Surgery 89:569-574, 1981.
13.
Apelqvist J, Castenfors J, Stenstrijm A, Agardh C-D: Wound classification is more important that site of ulceration in the outcome of diabetic foot ulcers. Diabetic
Stand J Clin Lab lnuest Suppl 128:239-243,
Med 6:526-530, 14.
1973.
1989.
Apelqvist J, Larsson J, Agardh C-D: The importance of peripheral pulses, peripheral oedema and local pain for the outcome of diabetic foot ulcers. Diabetic Med 7: 590-594,
1990.
174
APELQVIST ET AL.
1 Diab Camp 1992; 6:3
15.
Apelqvist J, Larsson J, Agardh C-D: The influence of external precipitating factors and peripheral neuropathy on the development and outcome of diabetic foot ulcers. J Diubet Complications 4:21-25, 1990.
24.
16.
Wagner FW: The dysvascular foot: A system for diagnosis and treatment. Foot Ankle 2:64-122, 1981.
25.
17.
Myocardial Infarction Community Register (WHO). Public Health in Europe no. 5, Copenhagen, WHO, 1976.
26.
Rayman G, Hassan A, Tooke JE: Blood flow in the skin of the foot related to posture in diabetes mellitus. EMI 292:87-90, 1986.
18.
Rose G, Blackburn H: Cardiovascular Geneva: World Health Organization, ries No. 56, 1968.
survey methods. Monograph Se-
27.
Tooke JE: The microcirculation Med 4189-196, 1987.
19.
Criteria Committee, New York Heart Association, Inc. Diseases of the heart and blood vessels. Nomenclature and Criteria JOY Diagnosis, 6th ed. Boston, Little Brown and Co, 1964, p. 114.
28.
Edmonds ME: The neuropathic foot in diabetes. 1: Blood flow. Diabetic Med 3:111-115, 1986.
29.
Nelson RG, Gohdes MD, Everhart JE, Hartner JA, Zwemer FL, Pettitt DJ, Knowler WC. Lower extremity amputation in NIDDM: 12-year follow up study in Pima Indians. Diabetes Cure 11:8-16, 1988.
30.
Bendick PJ, Glover JL, Kuebler TW, Dilley RS: Progression of atherosclerosis in diabetics. Surgery 93:834-838, 1983.
20.
Colwell JA, Bingham SF, Abraira C, Anderson JS, Kwaan HC. The cooperative study group: V.A. COoperative study on antiplatelet agents in diabetic patients after amputation for gangrene. III. Definition and review of design and baseline characteristics. Horm Metub Res Suppl 15:69-73, 1985.
Edmonds ME, Blundell Ml’, Morris ME, Maelor TE, Cotton LT, Watkins PH: Improved survival of the diabetic foot: The role of a specialised foot clinic. Q 1 Med 60:763-771, 1986. Lithner F, Tijrnblom N: Gangrene localized to the feet in diabetic patients. Acta Med Stand 215:75-79, 1984.
in diabetes.
Diabetic Part
21.
Boulton AJM, Kubrusly DB, Bowker JH, Gardia MT, Quintero L, Becker DM, Skyler JS, Sosenko JM: Impaired vibration perception and diabetic foot ulceration. Diabetic Med 3:335-337, 1986.
31.
Beach KW, Strandness DE: Arteriosclerosis obliterans and associated risk factors in insulin-dependent and non-insulin-dependent diabetes. Diabetes 29882-888, 1980.
22.
Boulton AJM, Hardisty CA, Betts RP, Franks CI, Worth RC, Ward JD, Duckworth T: Dynamic foot pressure and other studies as diagnostic and management aids in diabetic neuropathy. _ _ Diabetes Cure 6:26-33, 1983.
32.
Borch-Johnsen K, Kreiner S: Proteinuria-A predictor of cardiovascular mortality in insulin-dependent diabetes mellitus. BMJ 294:1651-16.54, 1987.
33.
23.
Dormandy J (ed), European Consensus Document on Critical Limb Ischaemia March 1989. Berlin, Springer Verlag, 1989.
Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T, Kofoed-Enevoldsen A: Albuminuria reflects widespread vascular damage. The Steno hypothesis. Diabetologia 32:219-229, 1989.
Medical Risk Factors in Diabetic Patients With Foot Ulcers and Severe Peripheral Vascular Disease and Their Influence on Outcome fn tpelqvist an arsson Carl-David Agardh
ABSTRACT The association between medical risk factors and the outcome of foot ulcers was evaluated in 208 consecutive diabetic patients with severe peripheral vascular disease (systolic toe blood pressure 545 mm Hg). All patients were treated and followed by the same foot care team. Eighty patients healed primarily, 83 healed after a minor or major amputation, and 45 died. The systolic toe blood pressure was higher among primary healed (30 + 13 mm Hg) compared with amputated (22 & 15 mm Hg; p < 0.001) and deceased patients (20 + 14 mm Hg; p < 0.001). The patients were comparable regarding age, sex, and diabetes and wound duration. Only 41 (19%) patients had intermitten claudication, whereas 153 (77%) lacked palpable pedal pulses, 36% of whom healed primarily. Rest pain occurred in 72 (33%) patients,
F
INTRODUCTION oot ulcer is a common complication to diabetes mellitus.1-3 Accelerated atherosclerosis leading to impaired circulation in the legs is one of the most important causes of gangrene and
Departments of Internal Medicine (J.A., C.-D.A.), and Orthopedic Surgery (J.L.), University Hospital, Lund, Sweden Reprint requests to be sent to: Dr. Jan Apelqvist, Department of Internal Medicine, University Hospital, S-221 85 Lund, Sweden. 0 1992 Journal of Diabetes and Its Complications
38 (47%) of whom had an amputation and 18 (25%) who healed primarily (p < 0.01). Peripheral edema and proteinuria were more common among patients who healed after amputation compared with those who healed primarily (p < 0.001 and p --L0.01, respectively). Signs of sensory neuropathy were found in 158 (77%) patients. There were no differences concerning cardiovascular disease, smoking habits, or short-term metabolic control between patients who healed primarily or after an amputation. In conclusion, diabetic patients with foot ulcers and severe peripheral vascular disease with low systolic toe blood pressure were not excluded from the possibility of primary healing. The most important risk factors for amputation were a systolic toe pressure of less than 30 mm Hg, peripheral edema, rest pain, and proteinuria.
amputation. 4-6 In previous studies a critically low systolic toe blood pressure has been shown to indicate a high probability for development of gangrene and the need for amputation.7-‘2 The aim of the present prospective study was to describe the association between clinical risk factors and outcome in a large group of consecutively presenting diabetic patients with foot ulcers and a systolic toe blood pressure equal to or below 45 mm Hg. 0891-6632/92/$5,00
168
1 Diab Comp 1992; 6:3
AI’ELQVIST ET AL.
MATERIALS AND METHODS
within that period, as intact skin at the time of death. Presence of precipitating factors for the development of ulcers were defined as previously described.15
Patients. In this prospective study 208 consecutively presenting diabetic patients (104 males and 104 females) referred to the Department of Interna Medicine because of foot ulcer and with a systolic toe blood pressure of 45 mm Hg or less, were investigated. The patients were mainly treated as outpatients, but in cases of deep infection associated with septic conditions, amputation, or intercurrent disease, they were treated as inpatients. Patient characteristics at enrollment are given in Table 1. The median duration of foot ulceration prior to the first admission was 8 weeks (range, 0.5-157 weeks) and the wound healing time after admission was 4 months (range, l-44 months). The follow-up time for all patients was 7 months (range, 7-50 months).
Medical Treatment. A physical examination was performed at enrollment and regularly during the study. Saline dressings and local enzymatic treatment were used as cleansing agents. Exudative wounds were treated with absorption dressings. Clean ulcers were covered with occlusivelsemiocclusive membranes or ointment dressings to maintain moisture. Protective shoe wear were used in most patients. In a small proportion, total contact walking casts or special orthotic appliances were used. Bed rest was not recommended to patients except in cases of a septic or toxic condition. When resting in a sitting position they were instructed to keep the foot elevated to hip level. All patients with plantar ulcers had an individually fitted orthotic appliance (e.g., specially made shoes, insoles, walking cast) which maintained ambulation, but with a modification of shoe wear to avoid pressure on the ulcer area. There was no difference in outcome between different orthotic appliances (data not shown). Metabolic control was improved when possible. A majority of the patients were treated with insulin (Table 11, and there was no difference in treatment mode at admission and at final outcome (data not shown). Peripheral edema was usually treated by diuretics. P-blocking agents were usually changed to
Ulcer Definition. At enrollment and throughout the study all lesions were characterized by the same foot care team and documented by color photographs as previously described. 13-15 Each patient was represented by one lesion below the ankle. Lesions were classified according to Wagner,16 and the most superficial ulcer included in the study was a lesion through the full thickness of the dermis. The ulcers were classified as superficial (n = 73), deep (n = 23), osteomyelitis and/or abscess (n = 26), minor gangrene (n = 49), or major gangrene (n = 37). Wound healing was defined as intact skin for at least 6 months or, in cases when the patient died
TABLE 1. PATIENT CHARACTERISTICS Primary II Age (yr) Sex (M/F) Duration of diabetes (yr) Wound duration in weeks (range) Treatment (n) Diet Oral agents Insulin HbAl, (%) Smoking habits (n) Smokers/ex-smokers Never smoked Signs of neuropathy (VPT > 30) (n) Severe retinopathy (n) Diabetic nephropathy, Creatinine (FmoVL)
Healed
83 69 ? 11 41142 18 + 13 9 (0.5-105)
10 29 40 8.5 * 1.9
6 22 55 8.3 ? 1.8 34 48 66 (9) 27 28*” 128 f 108
62 (3) 26 11 111 * 71
VPT, vibration perception threshold. Note: Wound duration is given as median (range); other values are means f * p < 0.05, ** p < 0.01, *** p < 0.002 compared with prima y healed. Numbers within bars indicate patients with nonmeasurable
VPT.
Amputated
80 71 * 11 43137 16 c 12 8 (0.5-157)
38 42
Aibustix pos (n)
AT ENROLLMENT
SD
Deceased 45 77 * 11* 20125 13 f 11 6 (1.0?157) 5 21 19 8.1 2 2.1 15 29 30 (12) 7 187 -t :98*‘*
1 Diab Comp 2992; 6:3
calcium channel blockers, diuretics, or angiotensinconverting enzyme inhibitors. Antibiotics, usually flucloxacilline or cefalosporins continued with metronidazole, were used when clinical signs of infection were present, i.e., cellulitis, deep abscess, and/or osteomyelitis, or when progressive gangrene was present. Analgesic agents used were paracetamol and dextropropoxiphen, although in some patients narcotics had to be used, and in cases of painful polyneuropathy, carbamazepine or tricyclic antidepressive agents were given. Surgical Treatment. Surgical debridement of the lesions was performed when required. Incision and drainage were performed in cases of deep plantar abscess, and resection was performed in cases of severe osteomyelitis not responding to antibiotic treatment. Amputation was carried out in cases of progressive gangrene, intolerable pain despite analgesia, and toxic or septic conditions not responding to conservative treatment. The lowest level employed for amputation was at the metatarso-phalangeal level. Eighty-two patients had femoral angiography, and 29 of these patients had either reconstructive vascular surgery (n = 19) or percutaneous transluminal angioplasty (n = 10). Clinical Examinations and Definitions. Duration of diabetes was calculated from the year of diagnosis until enrollment in the study. Hypertension was considered present when the systolic and/or diastolic blood pressure exceeded 160 and 90 mm Hg, respectively, or when the patients were treated with antihypertensive medication. Angina pectoris was defined as a history of retrosternal chest pain with a duration of less than 30 min, aggravated by exercise or cold environment, and relieved within 5 min of administration of nitroglycerin. Definitive previous myocardial infarction (MI) was defined either by the presence of resting electrocardiogram (ECG) changes consistent with an MI according to the Minnesota code or by the presence of two of the following criteria: (1) a history of retrosternal pain at onset of the attack according to the WHO Myocardial Infarction Community Registers,17 (2) development of ECG changes indicative of MI according to the Minnesota Code,18 (3) peak values of S-ASAT (EC 2.6.1.1) twice the upper limit of normal within 72 h, or (4) autopsy evidence of previous MI. Congestive heart failure was defined according to the New York Heart Association” and was verified with clinical examination and chest x-ray. Nonischemic heart disease was considered when other causes of heart disease such as valvular heart disease and atria1 fibrillation were verified through clinical examination, x-ray, and ECG. Cerebrovascular disease was defined as a history of
DIABETIC FOOT ULCERS
169
neurological deficit according to the Veteran’s Administration Cooperative Study of antiplatelet agents.” Smoking habits were divided according to the following: (1) smokers, i.e., daily smokers including those who had stopped smoking within 1 year prior to the study, (2) previous smokers, and (3) nonsmokers. Nonsmokers were defined as persons who never had had any daily nicotine consumption. The amount of smoking was registered as number of years smoked. Retinal examination was performed through a dilated pupil by ophthalmoscopy or fundus photography. The patients were examined within 1 year of the time of enrollment and were classified into three groups depending on the degree of retinopathy. The most seriously affected eye was used for the evaluation. The groups were (1) no signs of retinopathy, (2) simplex retinopathy, and (3) severe retinopathy (preproliferative retinopathy, proliferative retinopathy, or previous photocoagulation). Four patients were not examined due to their medical condition. Clinical signs of nephropathy were considered to be of diabetic origin when persistent proteinuria (positive Albustix) was present for more than 6 months and as nondiabetic when other causes of kidney disease were verified. The vibration perception threshold (VPT) was measured with a biothesiometer (Bio Medical, Newbury, OH, USA; scale l-50) at the tip of digit I as previously described. l5 Values exceeding 30 arbitrary units were considered as an indication of sensory neuropathy. 21,22 In 24 patients, VPT was not possible to measure due to their mental status. Edema was defined as swelling of the foot so pronounced as to leave a clear imprint after pressure by a finger. Peripheral pulses of the lower extremity (dorsal pedal and posterior tibial) were examined by manual palpation and registered as present or absent. Only findings from the foot with the ulcer are presented. Rest pain was defined as severe and persistent when localized to the ulcerated foot and relieved by lowering of the foot. Intermittent claudication was defined as a history of recurrent cramping pain or tightness in the calf induced by exercise and relieved by rest. Deep venous thrombosis had been verified by previous venous phlebography. Ulcer duration was defined as the number of weeks from the development of the ulcer until entry in the study. Blood samples were taken at admission and every third month until healing had occurred. Glycosylated hemoglobin (HbA,,) levels were measured by ion-exchange chromatography using commercially available microcolumns (BIO RAD, Richmond, CA, USA). Norma1 value is less than 6,0%.
170
APELQVIST ET AL.
j Diab Comp 1992; 6:3
TABLE 2. TYPE OF ULCER IN RELATON TO FINAL OUTCOME
Primary Healed Type of Ulcer Superficial Deep Abscess/osteomyelitis Minor gangrene Major gangrene n
Deceased
n
n
n
73
51 (70)
1
21
22 26 49 37 208
12 (55) 13 (50) 4 (8) 0 (0) 80 -
2 10 37 32 83
8 3 8 5 45
The systolic ankle and toe blood pressure were measured with strain-gauge and doppler techniques as previously described.’ The ankle pressure was not measurable in 23 patients (mainly because of incompressible arteries). Statistics. Values are given as means -+-SD and, in the case of deviation from an estimated normal distribution, as median and range. Differences between or within groups were calculated using the Mann-Whitney U test (two-tailed) and the x2 test. The simultaneous influence of several possible risk factors for amputation was investigated by means of logistic regression. RESULTS Primary healing occurred in 80 (38%) patients, 83 (40%) healed after amputation, and 45 (21%) died before healing had occurred (9 of these patients had amputation prior to death) (Table 1). Of those 83 patients who healed after amputation, 27 had had an amputation below the ankle (15 toe/s; 12 TMT), 47 below the knee, and 8 at the thigh level. The cause of death was considered as cardio-cerebrovascular in 75% of the cases. There were no differences in age, sex, duration of diabetes, and wound duration between the groups with the exception of patients who died unhealed being older than patients with primary healing (p < 0.05; Table 1). There was a gradual decline in primary healing rate
TABLE 3. ULCER LOCALIZATION
(o/o)
with increasing tissue involvement of the ulcer (Table 2). Four patients with minor gangrene healed primarily after mummification and auto-amputation. There were no differences in healing rate at different ulcer localizations (Table 3). With multiple ulcers, however, only one of 25 patients healed primarily. The presence of external precipitating factors and their relation to outcome is described in Table 4. In 80% (148 of 208) of the patients such a factor could be verified, of which ill-fitting shoes/socks was the most common. Systolic Ankle and Toe Blood Pressure. There was no difference in mean systolic ankle blood pressure between patients with primary healing and those who healed after amputation or died (Table 2); however, a systolic ankle pressure below 40 mm Hg was not consistent with primary healing (data not shown). Patients who healed primarily had higher systolic toe blood pressure levels compared with those who had an amputation or died (p < 0.01 and p < 0.001, respectively). No difference was seen between patients who had amputation or died. A gradual increase in primary healing rate was seen with increasing toe pressure (Figure 1). Primary healing was more common among patients with a toe pressure greater than 30 mm Hg (41 of 73) compared with those below that level (39 of 135; p < 0.001). Still, only 56% healed primarily with a toe pressure between 31 and 45 mm Hg, and 19 patients developed gangrene within these pressure levels.
IN RELATION
TO FINAL OUTCOME
Ulcer Localization
n
Primary Healed n (o/o)
Dig I, dorsal and plantar surface Dig II-V, dorsal and plantar surface Metatarsal head I-V Mid-foot and heel Dorsum of the foot Multiple ulcers (23)
64 50 10 34 24 25 208
30 (47) 19 (38) 6 (60) 15 (44) 9 (38) 1 (4) 80 -
n
Amputated
n
Amputated n
20 23 1 14 6 18 83
Deceased
n
14 8 3 5 9 6 45
DIABETICFOOT ULCERS 171
1 Diab Comp 1992; 63
TABLE 4. EXTERNAL PRECIPITATING FACTORS AND THEIR RELATION TO OUTCOME Precipitating Factor Ill-fitting shoes/socks Stress ulcer Acute mechanical trauma Paronychia Decubital ulcer Miscellaneous Reason unknown n
n
Primary Healed n
Amputated n
Deceased n
58 17 26 12 13 22 60 208
29 9 9 7 2 4 20 80
26 6 12 0 4 8 27 83
3 2 5 5 7 10 13 45
Development of gangrene and presence of multiple ulcers was strongly related to low toe blood pressure levels (O-15 mm Hg, 60% gangrene; 16-30 mm Hg, 42%; 31-45 mm Hg, 26%). There was no correlation between distal blood pressure levels and depth, size, and localization of ulcer with the exception of patients with multiple ulcers or patients developing gangrene having lower ankle and toe pressure levels compared with all other sites or types of lesions (p < 0.001 and p < 0.001, respectively). There was no difference in systolic ankle pressure levels comparing patients with primary healing and toe pressure less than 30 mm Hg, and those healed after amputation and toe pressure less than 30 mm Hg (68 ? 36 mm Hg versus 74 + 33 mm Hg; n = 45 and n = 27, respectively. Eleven patients healed primarily after angioplasty or vascular surgery (O-15 mm Hg, n = 5; 16-30 mm Hg, n = 2; 31-45 mm Hg: n = 4) with a mean preoperative toe pressure of 28 mm Hg, (range, O-45 mm Hg) and a mean postoperative pressure of 63 mm Hg (range, 25-90 mm Hg).
74% of the patients had no palpable pedal pulses, but the presence of pedal pulses was not related to the outcome (Table 5). Only 41 (19%) had intermittent claudication, and 72 (33%) had rest pain. Rest pain was more common among patients who healed after amputation compared with primary healed (p < O.Ol), but 18 patients (25%) with rest pain healed primarily. Patients with rest pain and primary healing had a higher toe pressure compared with those with rest pain who healed after amputation 30 +- 14 mm Hg (n = 18) versus 20 rt 14 (n = 38) mm Hg (p < 0.001). There was no difference in ankle pressure levels in the corresponding groups (82 + 42 mm Hg versus 79 + 41 mm Hg). Four of 18 patients with rest pain and primary healing had vascular surgery or angioplasty and one healed after autoamputation. Previous deep venous thrombosis was more common among patients with amputation compared with primary healed (p < 0.05). Neuropathy, Retinopathy and Nephropathy. VPT values above 30 arbitrary units were seen in 158 patients (76%), which represent 85% of patients with a measurable WT. There was no difference in presence of sensory neuropathy between primary healed patients and those who healed after amputation or died (Table 1). Severe retinopathy was seen in 60 patients (29%) and was equally present in the three groups.
Edema, Pedal Pulses, Rest Pain, and Claudication. Peripheral edema was present in almost onehalf of the patients and was most common among patients who healed after amputation or died (p < 0.001 and p < 0.001, respectively; Table 5). A total of
Systolictoe Dressure
0
1
20 I
40 I
60 I
80 I
100% I
N
hmHg) 0 - 15
57
16 - 30
78
31 - 45
73
FIGURE 1 Systolic toe blood pressure levels (mm Hg) in patients with primary healing (open area), amputation (hatched area), or death before healing occurred (solid area). N indicates number of patients.
172
APELQVIST
J Diab Camp 1992; 6:3
ET AL.
TABLE 5. EDEMA,
PEDAL PULSES, REST PAIN, CLAUDICATION, AND PREVIOUS THROMBOSIS IN RELATION TO FINAL OUTCOME Primary Healed ?I (o/o)
n
80
Edema Absence of pedal pulses Rest pain Claudication Previous deep venous thrombosis
25 56 18 20 7
(31) (70) (22) (25) (9)
DEEP VENOUS
Amputated n (o/o)
Deceased n (%I
83
45
48** (58) 62 (75) 38** (46) 17 (20 17* (20)
28** (62) 35 (78) 16 (36) 4 (9) 9 (20)
* p < 0.05, ** p < 0.001 compared with primary heaIed.
Persistent albuminuria was most common among patients who healed after amputation (n = 28) compared to patients with primary healing (n = 11; p < 0.01) (Table 1). The majority of the patients had signs of cardiovascular disease; however, there were no differences concerning such diseases or smoking habits between patients who healed primarily or healed after amputation. Nevertheless, the number of years smoked was related to the outcome, because smokers with amputation had a longer duration of smoking compared to smokers with primary healing (17 < 0.05) (data not shown). Simultaneous Analysis of Risk Factors for Amputation. To investigate the simultaneous influence of different possible risk factors for amputation logistic regression analysis were performed. First, we removed those 45 patients who died before healing. Second, the conventional explanatory variables age and sex were investigated; we found that neither age nor sex had a significant influence on the risk for amputation. As a third step for each of a long list of possible explanatory variables, a logistic regression analysis of the amputation risk was performed using only those factors showing promise of influence, the latter being defined as p < 0.15. Finally, in a stepwise manner, logistic regression analysis was performed with remaining variables simultaneously, starting with a full model and removing nonsignificant variables one by one. The result was that only three variables remained, i.e., diabetic nephropathy, rest pain, and systolic toe blood pressure (Table 6). In the present
TABLE 6. MULTIPLE STEPWISE REGRESSION ANALYSES OF RISK FACTORS FOR AMPUTATION
Diabetic nephropathy Rest pain Systolic toe blood pressure
Odds Ratio
p Value
3.0 2.4 0.95
0.048 0.037 0.001
study a 1 mm Hg increase in the systolic toe pressure would decrease the risk for amputation by 5% (odds ratio, 0.95) and a corresponding increase by 10 mm Hg gives an odds ratio of 0.951° (0.61) and a risk reduction by 39%. DISCUSSION In the present study putative medical risk factors were studied in diabetic patients with foot ulcers and severe peripheral vascular disease. Primary healing occurred in 38% of the patients. To eliminate for the physician being influenced by knowledge of clinical data in his medical decision making, the indications accepted for amputation were progressive gangrene, septic condition, and rest pain not responding to conservative treatment. All patients were followed until final outcome and were treated by the same team of physicians in order to eliminate for differences in strategy of treatment. For the same reason strict amputation criteria were maintained. The selection of patients with foot ulcer and a systolic toe blood pressure equal to or below 45 mm Hg was based on a previous study where such pressure indicated a high probability for development of gangrenes7 Furthermore, amputations above that level were seldom preceeded by development of gangrene.7 In the present study the systolic toe pressure level was strongly related to the outcome. The finding indicates that patients with a systolic toe blood pressure level below 30 mm Hg are at high risk for development of gangrene, and it might be valuable to identify diabetic patients with foot lesions in need of further vascular intervention as suggested by the European Consensus Document on Critical Limb Ischaemia.23 However, one of four patients with a toe pressure between 31 and 45 mm Hg developed gangrene . There was no difference in systolic ankle pressure levels between patients with primary healing, those that healed after amputation, or deceased patients. But an ankle pressure below 40 mm Hg was inconsis-
DIABETIC
j Diab Comp 1992; 6:3
tent with primary healing as has previously been shown. 7 In the present study one of three patients (39%) had an angiography. Of these 35% had either angioplasty or vascular surgery, and 38% of these interventions resulted in primary healing. These results might have been different if the decision for angiography had been made earlier and in more patients whose condition would allow for mechanical angioplasty and with routine visualization of the pedal arch. This statement is further stressed by the finding that although 77% of these patients had absence of both pedal pulses, absence of pedal pulses were not more common among patients who healed primarily compared with those who required an amputation. Furthermore, only 19% had intermittent claudication, and 33% had rest pain. Of those patients with rest pain, 25% healed primarily. The fact that some patients with rest pain healed primarily indicates the importance of evaluating other causes of pain than ischemia. Despite the difficulties in evaluating pedal pulses, rest pain, and intermittent claudication in these patients, the findings indicate that a clinical evaluation of peripheral vascular disease can give valuable but imperfect information concerning the need for further vascular intervention. Peripheral edema was found in almost 50% of the patients and was most common among patients who had had amputation or died which is in agreement with previous observations where edema, often from multifactorial causes,l* has been associated with the development and of importance for the outcome of foot ulcers in diabetic subjects.‘4J4*25 The most common factors associated with edema have been congestive heart failure, proteinuria, and neuropaPeripheral edema impairs wound thy. 14,24-28,26~27~28 healing probably by decreasing the microcircuIation.27 Persistent proteinuria was seen in 22% of the patients and was most common among those who healed after amputation. These findings are in agreement with the study by Nelson et al.29 and in studies of atherosclerotic occlusive disease in the lower extremity in diabetic patients.30,3* Proteinuria has also been suggested to be a predictor of cardiovascular mortality32 and an indicator for widespread cardiovascular disease in diabetic patients.33 In summary, the present study shows that diabetic patients with foot.ulcers and low systolic toe blood pressure are not excluded from the possibility of primary healing, as two of five healed primarily. The most important factor related to the outcome was a systolic toe blood pressure of less than 30 mm Hg. However, rest pain, proteinuria, and peripheral edema were also important factors related to amputation and death. ClinicaI signs of peripheral vascular
FOOT ULCERS
173
disease give important but imperfect information concerning the need for vascular intervention in these patients. ACKNOWLEDGMENT This work was supported by the Swedish Research Council (Grant No. 02872), the Hoechst Diabetes Foundation, The Swedish Diabetes Association, and The Medical Faculty, University of Lund. REFERENCES 1.
Rosenqvist LJ: An epidemiological survey of diabetic foot problems in the Stockholm County 1982. Acfa Med Stand Suppl687:55-60,
1984.
2.
Levin ME, O’Neal LW (ed), The Diabetic Foot. St. Louis MO: Mosby, 1988.
3.
Edmonds ME: The diabetic foot: Pathophysiology and treatment, in Watkins PJ (ed), Long-term complications of diabetes. Baillieres Clin Endocrinol Mefab 15: 889-916,
1986.
4.
Klamer TW, Towne JB, Bandyk DF, Bonner MJ: The influence of sepsis and ischemia on the natural history of the diabetic foot. Am Surg 53:490-494, 1987.
5.
Gutman M, Kaplan 0, Skornick Y, Klausner JM, Leicuk S, Rozin RR: Gangrene of the lower limbs in diabetic patients: A malignant complication. Am J Surg 154:305-308, 1987.
6.
Bild DE, Selby JV, Sinnock P, Browner WS, Braveman P, Showstack JA: Lower-extremity amputation in people with diabetes. Epidemiology and prevention. Dia-
7.
Apelqvist J, Castenfors J, Larsson J, Stenstriim A, Agardh C-D: Prognostic value of systolic ankle and toe blood pressure levels in outcome of diabetic foot ulcer.
8.
Holstein P, Lassen NA: Healing of ulcers on the feet correlated with distal blood pressure measurements in occlusive arterial disease. Acfa Orthop Stand 51:
betes Care 12:24-31,
1989.
Diabetes Care 12:373-378,
995-1006,
9.
1989.
1980.
Ramsey DE, Manke DA, Sumner DS: Toe blood pressure a valuable adjunct to ankle pressure measurement for assessing peripheral arterial disease. 7 Cardiovasc Surg 24143-48,
1985.
10.
Faris I, Duncan H: Skin perfusion pressure in the prediction of healing in diabetic patients with ulcers or gangrene of the foot. J Vascular Surg 4:536-540, 1985.
11.
Carter SA: The relationship of distal systolic pressures to healing of skin lesions in limbs with arterial occlusive disease with special reference to diabetes mellitus.
12.
Bone GE, Pomajzl MJ: Toe blood pressure by photoplethysmography: An index of healing in forefoot amputation. Surgery 89:569-574, 1981.
13.
Apelqvist J, Castenfors J, Stenstrijm A, Agardh C-D: Wound classification is more important that site of ulceration in the outcome of diabetic foot ulcers. Diabetic
Stand J Clin Lab lnuest Suppl 128:239-243,
Med 6:526-530, 14.
1973.
1989.
Apelqvist J, Larsson J, Agardh C-D: The importance of peripheral pulses, peripheral oedema and local pain for the outcome of diabetic foot ulcers. Diabetic Med 7: 590-594,
1990.
174
APELQVIST ET AL.
1 Diab Camp 1992; 6:3
15.
Apelqvist J, Larsson J, Agardh C-D: The influence of external precipitating factors and peripheral neuropathy on the development and outcome of diabetic foot ulcers. J Diubet Complications 4:21-25, 1990.
24.
16.
Wagner FW: The dysvascular foot: A system for diagnosis and treatment. Foot Ankle 2:64-122, 1981.
25.
17.
Myocardial Infarction Community Register (WHO). Public Health in Europe no. 5, Copenhagen, WHO, 1976.
26.
Rayman G, Hassan A, Tooke JE: Blood flow in the skin of the foot related to posture in diabetes mellitus. EMI 292:87-90, 1986.
18.
Rose G, Blackburn H: Cardiovascular Geneva: World Health Organization, ries No. 56, 1968.
survey methods. Monograph Se-
27.
Tooke JE: The microcirculation Med 4189-196, 1987.
19.
Criteria Committee, New York Heart Association, Inc. Diseases of the heart and blood vessels. Nomenclature and Criteria JOY Diagnosis, 6th ed. Boston, Little Brown and Co, 1964, p. 114.
28.
Edmonds ME: The neuropathic foot in diabetes. 1: Blood flow. Diabetic Med 3:111-115, 1986.
29.
Nelson RG, Gohdes MD, Everhart JE, Hartner JA, Zwemer FL, Pettitt DJ, Knowler WC. Lower extremity amputation in NIDDM: 12-year follow up study in Pima Indians. Diabetes Cure 11:8-16, 1988.
30.
Bendick PJ, Glover JL, Kuebler TW, Dilley RS: Progression of atherosclerosis in diabetics. Surgery 93:834-838, 1983.
20.
Colwell JA, Bingham SF, Abraira C, Anderson JS, Kwaan HC. The cooperative study group: V.A. COoperative study on antiplatelet agents in diabetic patients after amputation for gangrene. III. Definition and review of design and baseline characteristics. Horm Metub Res Suppl 15:69-73, 1985.
Edmonds ME, Blundell Ml’, Morris ME, Maelor TE, Cotton LT, Watkins PH: Improved survival of the diabetic foot: The role of a specialised foot clinic. Q 1 Med 60:763-771, 1986. Lithner F, Tijrnblom N: Gangrene localized to the feet in diabetic patients. Acta Med Stand 215:75-79, 1984.
in diabetes.
Diabetic Part
21.
Boulton AJM, Kubrusly DB, Bowker JH, Gardia MT, Quintero L, Becker DM, Skyler JS, Sosenko JM: Impaired vibration perception and diabetic foot ulceration. Diabetic Med 3:335-337, 1986.
31.
Beach KW, Strandness DE: Arteriosclerosis obliterans and associated risk factors in insulin-dependent and non-insulin-dependent diabetes. Diabetes 29882-888, 1980.
22.
Boulton AJM, Hardisty CA, Betts RP, Franks CI, Worth RC, Ward JD, Duckworth T: Dynamic foot pressure and other studies as diagnostic and management aids in diabetic neuropathy. _ _ Diabetes Cure 6:26-33, 1983.
32.
Borch-Johnsen K, Kreiner S: Proteinuria-A predictor of cardiovascular mortality in insulin-dependent diabetes mellitus. BMJ 294:1651-16.54, 1987.
33.
23.
Dormandy J (ed), European Consensus Document on Critical Limb Ischaemia March 1989. Berlin, Springer Verlag, 1989.
Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T, Kofoed-Enevoldsen A: Albuminuria reflects widespread vascular damage. The Steno hypothesis. Diabetologia 32:219-229, 1989.
