A cohort study of diabetic patients and diabetic foot ulceration patients in China Yufeng Jiang, M.D, Ph.D*, Xuemei Wang, M.S*, Lei Xia, Ph.D*, Xiaobing Fu, M.D., Ph.D, Zhangrong Xu, M.D., Xingwu Ran, M.D., Li Yan M.D., Qiu Li, M.D., Zhaohui Mo, M.D., Zhaoli Yan, M.D., Qiuhe Ji M.D. ,Qiang Li, M.D.
From Wound Healing Unit, Trauma Center of Postgraduate Medical School, Chinese PLA General Hospital, Beijing (Y.J, L.X, X.F,); The Key Laboratory of Wound Repair and Regeneration of PLA, The First Affiliated Hospital of Chinese PLA General Hospital, Beijing(Y.J, X.F); The School of Public Health of Inner Mongolia Medical College, Hohhot (X.W.); The Diabetic Center, 306th Hospital of Chinese P.L.A, Beijing (Y.J, Z.X); Huaxi Hospital of Sichuan University, Chengdu(X.R.); Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou(L.Y);People’s Hospital of ShanDong Province, JiNan(Q.Li); Third Xiangya Hospital, Central South Univety, Changsha(Z.Mo); The Affiliated Hospital of Inner Mongolia Medical College, Hohhot (Z.Y.) Xijing Hospital of Fourth Military Medical University, Xi’an (Q.Ji.); The Second Affiliated Hospital of Harbin Medical University, Harbin (Q.Li.);all in China. * contributed equally to this paper. Correspondence should be addressed to: Xiaobing Fu, MD, Ph.D Professor General Hospital of Chinese PLA Trauma Center of Postgraduate Medical College 28 Fu Xing Road, Beijing 100842 P. R. CHINA Fax: 0086 10 68989955 e-mail: [email protected]
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1111/wrr.12263 This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
ABSTRACT Aims: To determine the annual incidence and clinically relevant risk factors for foot ulceration in a large cohort study of diabetic foot ulcer (DFU) patients and diabetes mellitus (DM) patients in China. Methods: To investigate a cohort of 1333 patients comprising 452 DFU patients and 881 DM patients, who underwent foot screening, physical examination and laboratory tests in eight hospitals. The patients were assessed at baseline in terms of their demographic information, medical and social history, peripheral neuropathy disease (PND) screening, periphery artery disease (PAD) screening, assessment of nutritional status and diabetic control. One year later, the patients were followed up to determine the incidence of new foot ulcers, amputation and mortality. By univariate analysis, statistically significant differences were found in age, location, gender, living alone (yes/no), occupation, smoking, hypertension, PND, PAD, nephropathy, retinopathy, cataracts, duration of diabetes, Glycosylated Hemoglobin A (HbA1c), fasting plasma glucose level, postprandial blood glucose level, insulin level, blood urea nitrogen, creatinine, cholesterol, triglyeride, high density lipoprotein (HDL), serum albumin, white blood cell, and body mass index (BMI). A binary logistic regression model was used to examine which of these risk factors were independent risk factors for foot ulceration. Results: A total of 687 (51.5%) of the 1333 patients were followed up for an average of 12 months; there were 458 DM patients and 229 DFU patients. A total of 46 patients died during the follow-up period; 13 were DM patients, and 33 were DFU patients. Of the 641 patients, 445（69.4%） patients were DM patients, and 196 （30.6%）were DFU patients. At follow-up, 36/445 DM patients (8.1%), and 62/196 DFU patients (31.6%), developed new ulcers; 10/196 DFU patients underwent an amputation. The annual incidence of ulceration for DM patients and amputation for DFU patients were 8.1% and 5.1%, respectively. The annual mortality of the DM patients and DMF patients were 2.8% and 14.4%, respectively. A binary logistic regression model was used to examine which risk factors were independent risk 2
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 2 of 25
Page 3 of 25
Manuscript under review - CONFIDENTIAL
factors for foot ulceration during the follow-up period, and the final results showed that nephropathy (odds ratio 2.32), insulin level (odds ratio 3.136, 2.629), and decreased HDL (odds ratio 0.427) were associated with increased risks for foot ulceration. Conclusions: Complications of diabetes affecting the feet represent a serious problem in China. The incidence of foot ulcers and amputation are much higher than that of Western countries. More intensive surveillance and aggressive care following a diagnosis of DFU and earlier referral to specialty care might improve the patient outcome.
3
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
Page 4 of 25
A recent World Bank report, Toward a Healthy and Harmonious Life in China, urged China
to
intensify
its
efforts
to
address
the
increasing
prevalence
of
non-communicable diseases (e.g., cardiovascular disease, chronic obstructive pulmonary disease, diabetes, and lung cancer), warning of the social and economic consequences of inaction.1 The estimated prevalence of diabetes mellitus (DM) in a representative sample of Chinese adults was 11.6%, and the prevalence of prediabetes was 50.1%. Projections based on sample weighting suggest that these percentages might represent up to 113.9 million Chinese adults with diabetes and 493.4 million with prediabetes. These findings indicate the importance of diabetes as a public health problem in China.2 Diabetic foot ulcer (DFU) and lower extremity amputation (LEA) are common complications of diabetes that have dramatic effects on the health and general well-being of patient as well as being expensive to treat.3There are few published data on the complications of diabetes affecting the feet and on the annual incidence rate of diabetic foot ulcers in particular. The clinically relevant risk factors for foot ulceration in diabetic patients have not been determined in China. We conducted a prospective analysis of a very large cohort of diabetic patients and DFU patients that aimed to determine the most effective factors for predicting the risk of diabetic foot ulceration in diabetic patients. (Figure1) Patients and methods Type 2 diabetes patients diagnosed with diabetic foot ulcers in the diabetic centers of eight hospitals were targeted in. Eight hospitals in eight administrative divisions of China (East, North, South, Central, Northeast, Southwest, Northwest, and the capital, Beijing) represented nearly all the geographic regions of China (Supplement 1). All adults diagnosed with Type 2 diabetes, according to the WHO Criteria of 1999, inpatients of these diabetic centers, were targeted as the control group. The cases were defined as the subjects who were admitted to hospitals for a diabetic foot ulcer whereas the controls were inpatients from the identical outpatient population who were diabetic without foot ulcer. A foot ulcer was defined as a full-thickness skin break at least of Wagner stage 1, occurring distal to the malleolus.4To avoid sampling errors and to provide useful prevalence data, all 8 hospitals were Grade III-A teaching 4
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 5 of 25
Manuscript under review - CONFIDENTIAL
hospitals of medical colleges or universities, indicating that the data were more likely to be reliable. Multidisciplinary diabetic foot care teams in each hospital were invited to become involved in this project, which provided standardized foot screening, a physical examination and medical care for all of the inpatients on their lists. One full-time research nurse was trained to assist with the data recording and abstraction in each hospital. Screening began on June 1, 2011, lasted for 1 year, and was extended for an additional year as the follow-up period; 687 patients were randomly selected for follow-up one year after the day they were discharged from the hospital. Prior to screening, the staff members involved in this project attended a series of training sessions at the study headquarters to standardize screening and physical examination techniques and medical care. Screening measurements At baseline, the following variables were recorded for all of the patients: gender, age, ethnic origin (based upon the patients’ ID card), occupations, habitations, and whether the patient lived alone (yes/no). The medical history was taken including the duration and methods of diabetes control, the social history of the patient including a detailed history of smoking and evidence of regular alcohol consumption (> 7 units/week). The medical documentation was examined to determine whether patients had diabetic complications including retinopathy, which was diagnosed in cases in which nonproliferative or proliferative retinopathy was observed at the fundus by an ophthalmological examination; the documentation determined the following: whether the patient was blind/partially sighted or had impaired vision and, therefore, unable to see their feet clearly; nephropathy; peripheral neuropathy disease (PND) and peripheral vascular disease (PVD); and the presence of co-morbidities such as hypertension, cerebrovascular diseases and ischemic heart disease (IHD). The details of past or present foot ulcers were documented via examination and accessing the medical notes. The baseline laboratory data including Hemoglobin A-Glycosylated (HbA1c), full blood counts, serum lipids, serum albumin, and creatinine, which predominantly indicated the blood glucose level, nutritional status and metabolism, were recorded. Nephropathy was diagnosed in cases in which the serum creatinine 5
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
was above 1.5 mg/dl, there was persistent proteinuria of 0.5g, or the urine albumin/excretion rate (UAER) was above 30mg/g at least twice. Peripheral neuropathy was considered in cases in which the patients had neuropathic symptoms and signs or objectively abnormal results with the Semmes-Weinstein 5.07/10 g monofilament or 128-Hz tuning-fork test and without other significant disease. PVD was diagnosed by a history of intermittent claudication and rest pain, a history of previous lower limb vascular surgery, absent or reduced pedal pulses, ankle-brachial pressure index (ABI) less than 0.9, or angiography showing significant stenosis in low extremity arteries. Personal identifiers were removed from the completed data collection sheets before entry into an Epidata 3.0 database. The entering and editing of the tool data was accomplished by an experienced wound care specialist and a diabetologist, who verified the accuracy and authenticity of the data; any disagreements were resolved by discussion. Prospective follow-up One year after discharge day, 687 patients who were selected randomly from 1333 patients were re-contacted via telephone to assess the incidence of new foot ulcers occurring since the baseline. The first new foot ulcer event that occurred within the follow-up period was recorded as the final outcome for each patient to calculate the annual incidence of foot ulcers. The returned questionnaires that reported a positive new ulcer event were rigorously crosschecked, i.e., the patients were re-interviewed by principle investigators at every center. Thus, the reported new ulcer events were confirmed, and the reported false-positive events were eliminated. Ethical approval was obtained from the research ethics committees for this study. Statistical Analysis The statistical analyses were performed using IBM SPSS, version 17.0, for Windows® (SPSS,® Inc., Chicago, Illinois, USA). The statistical tests and confidence intervals were two-sided, with a significance level of 0.05. We conducted a baseline univariate analysis of the data on the participants’ demographic, social, and medical characteristics between the DM and the DFU patients. The Chi-square test was applied for the comparison of the categorical 6
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 6 of 25
Page 7 of 25
Manuscript under review - CONFIDENTIAL
variables. The normally distributed continuous variables were tested using independent sampled T-tests, whereas the non-normally distributed variables were analyzed using the two-independent sample Mann-Whitney U test. The variables that were significant at the 0.10 level were included in the multivariable analysis. We applied two standard binary logistic-regression models for the multivariable analysis. One model was performed to find a cross-sectional association between the DM patients and the DFU patients. The other model was for the longitudinal analysis to predict the risk factors and incidence probability for foot ulcers in the DM patients. The variables were retained in the logistic regression model if they were significant at the 0.05 level. The adjusted odds ratio and 95% confidence intervals were calculated with the logistic regression to determine the associations between DM and new ulcers as the outcome. RESULTS This work is a multi-center prospective cohort study of 1333 inpatients (452 DFU patients and 881 DM patients), who were treated for diabetic foot problems (DFP) by a multi-disciplinary team from June 1, 2011 to May 30, 2013; 687 patients were randomly selected for the follow up. Demographics A total of 881 DM patients and 452 DFU patients were investigated in the eight hospitals over 2 years. The demographic details of the patients are displayed in Table 1. Of the 881 DM patients, 490 were male and 391 were female. Of the 452 DFU patients, 294 were male and 157 were female. The two gender ratios have great statistical variability (χ2=11.282, P=0.001). The median age of the DM patients and DFU patients was 56.0 years (47, 66; 25%, 75%) and 64.0 years (55, 72; 25%, 75%) (p
From Wound Healing Unit, Trauma Center of Postgraduate Medical School, Chinese PLA General Hospital, Beijing (Y.J, L.X, X.F,); The Key Laboratory of Wound Repair and Regeneration of PLA, The First Affiliated Hospital of Chinese PLA General Hospital, Beijing(Y.J, X.F); The School of Public Health of Inner Mongolia Medical College, Hohhot (X.W.); The Diabetic Center, 306th Hospital of Chinese P.L.A, Beijing (Y.J, Z.X); Huaxi Hospital of Sichuan University, Chengdu(X.R.); Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou(L.Y);People’s Hospital of ShanDong Province, JiNan(Q.Li); Third Xiangya Hospital, Central South Univety, Changsha(Z.Mo); The Affiliated Hospital of Inner Mongolia Medical College, Hohhot (Z.Y.) Xijing Hospital of Fourth Military Medical University, Xi’an (Q.Ji.); The Second Affiliated Hospital of Harbin Medical University, Harbin (Q.Li.);all in China. * contributed equally to this paper. Correspondence should be addressed to: Xiaobing Fu, MD, Ph.D Professor General Hospital of Chinese PLA Trauma Center of Postgraduate Medical College 28 Fu Xing Road, Beijing 100842 P. R. CHINA Fax: 0086 10 68989955 e-mail: [email protected]
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1111/wrr.12263 This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
ABSTRACT Aims: To determine the annual incidence and clinically relevant risk factors for foot ulceration in a large cohort study of diabetic foot ulcer (DFU) patients and diabetes mellitus (DM) patients in China. Methods: To investigate a cohort of 1333 patients comprising 452 DFU patients and 881 DM patients, who underwent foot screening, physical examination and laboratory tests in eight hospitals. The patients were assessed at baseline in terms of their demographic information, medical and social history, peripheral neuropathy disease (PND) screening, periphery artery disease (PAD) screening, assessment of nutritional status and diabetic control. One year later, the patients were followed up to determine the incidence of new foot ulcers, amputation and mortality. By univariate analysis, statistically significant differences were found in age, location, gender, living alone (yes/no), occupation, smoking, hypertension, PND, PAD, nephropathy, retinopathy, cataracts, duration of diabetes, Glycosylated Hemoglobin A (HbA1c), fasting plasma glucose level, postprandial blood glucose level, insulin level, blood urea nitrogen, creatinine, cholesterol, triglyeride, high density lipoprotein (HDL), serum albumin, white blood cell, and body mass index (BMI). A binary logistic regression model was used to examine which of these risk factors were independent risk factors for foot ulceration. Results: A total of 687 (51.5%) of the 1333 patients were followed up for an average of 12 months; there were 458 DM patients and 229 DFU patients. A total of 46 patients died during the follow-up period; 13 were DM patients, and 33 were DFU patients. Of the 641 patients, 445（69.4%） patients were DM patients, and 196 （30.6%）were DFU patients. At follow-up, 36/445 DM patients (8.1%), and 62/196 DFU patients (31.6%), developed new ulcers; 10/196 DFU patients underwent an amputation. The annual incidence of ulceration for DM patients and amputation for DFU patients were 8.1% and 5.1%, respectively. The annual mortality of the DM patients and DMF patients were 2.8% and 14.4%, respectively. A binary logistic regression model was used to examine which risk factors were independent risk 2
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 2 of 25
Page 3 of 25
Manuscript under review - CONFIDENTIAL
factors for foot ulceration during the follow-up period, and the final results showed that nephropathy (odds ratio 2.32), insulin level (odds ratio 3.136, 2.629), and decreased HDL (odds ratio 0.427) were associated with increased risks for foot ulceration. Conclusions: Complications of diabetes affecting the feet represent a serious problem in China. The incidence of foot ulcers and amputation are much higher than that of Western countries. More intensive surveillance and aggressive care following a diagnosis of DFU and earlier referral to specialty care might improve the patient outcome.
3
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
Page 4 of 25
A recent World Bank report, Toward a Healthy and Harmonious Life in China, urged China
to
intensify
its
efforts
to
address
the
increasing
prevalence
of
non-communicable diseases (e.g., cardiovascular disease, chronic obstructive pulmonary disease, diabetes, and lung cancer), warning of the social and economic consequences of inaction.1 The estimated prevalence of diabetes mellitus (DM) in a representative sample of Chinese adults was 11.6%, and the prevalence of prediabetes was 50.1%. Projections based on sample weighting suggest that these percentages might represent up to 113.9 million Chinese adults with diabetes and 493.4 million with prediabetes. These findings indicate the importance of diabetes as a public health problem in China.2 Diabetic foot ulcer (DFU) and lower extremity amputation (LEA) are common complications of diabetes that have dramatic effects on the health and general well-being of patient as well as being expensive to treat.3There are few published data on the complications of diabetes affecting the feet and on the annual incidence rate of diabetic foot ulcers in particular. The clinically relevant risk factors for foot ulceration in diabetic patients have not been determined in China. We conducted a prospective analysis of a very large cohort of diabetic patients and DFU patients that aimed to determine the most effective factors for predicting the risk of diabetic foot ulceration in diabetic patients. (Figure1) Patients and methods Type 2 diabetes patients diagnosed with diabetic foot ulcers in the diabetic centers of eight hospitals were targeted in. Eight hospitals in eight administrative divisions of China (East, North, South, Central, Northeast, Southwest, Northwest, and the capital, Beijing) represented nearly all the geographic regions of China (Supplement 1). All adults diagnosed with Type 2 diabetes, according to the WHO Criteria of 1999, inpatients of these diabetic centers, were targeted as the control group. The cases were defined as the subjects who were admitted to hospitals for a diabetic foot ulcer whereas the controls were inpatients from the identical outpatient population who were diabetic without foot ulcer. A foot ulcer was defined as a full-thickness skin break at least of Wagner stage 1, occurring distal to the malleolus.4To avoid sampling errors and to provide useful prevalence data, all 8 hospitals were Grade III-A teaching 4
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 5 of 25
Manuscript under review - CONFIDENTIAL
hospitals of medical colleges or universities, indicating that the data were more likely to be reliable. Multidisciplinary diabetic foot care teams in each hospital were invited to become involved in this project, which provided standardized foot screening, a physical examination and medical care for all of the inpatients on their lists. One full-time research nurse was trained to assist with the data recording and abstraction in each hospital. Screening began on June 1, 2011, lasted for 1 year, and was extended for an additional year as the follow-up period; 687 patients were randomly selected for follow-up one year after the day they were discharged from the hospital. Prior to screening, the staff members involved in this project attended a series of training sessions at the study headquarters to standardize screening and physical examination techniques and medical care. Screening measurements At baseline, the following variables were recorded for all of the patients: gender, age, ethnic origin (based upon the patients’ ID card), occupations, habitations, and whether the patient lived alone (yes/no). The medical history was taken including the duration and methods of diabetes control, the social history of the patient including a detailed history of smoking and evidence of regular alcohol consumption (> 7 units/week). The medical documentation was examined to determine whether patients had diabetic complications including retinopathy, which was diagnosed in cases in which nonproliferative or proliferative retinopathy was observed at the fundus by an ophthalmological examination; the documentation determined the following: whether the patient was blind/partially sighted or had impaired vision and, therefore, unable to see their feet clearly; nephropathy; peripheral neuropathy disease (PND) and peripheral vascular disease (PVD); and the presence of co-morbidities such as hypertension, cerebrovascular diseases and ischemic heart disease (IHD). The details of past or present foot ulcers were documented via examination and accessing the medical notes. The baseline laboratory data including Hemoglobin A-Glycosylated (HbA1c), full blood counts, serum lipids, serum albumin, and creatinine, which predominantly indicated the blood glucose level, nutritional status and metabolism, were recorded. Nephropathy was diagnosed in cases in which the serum creatinine 5
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Manuscript under review - CONFIDENTIAL
was above 1.5 mg/dl, there was persistent proteinuria of 0.5g, or the urine albumin/excretion rate (UAER) was above 30mg/g at least twice. Peripheral neuropathy was considered in cases in which the patients had neuropathic symptoms and signs or objectively abnormal results with the Semmes-Weinstein 5.07/10 g monofilament or 128-Hz tuning-fork test and without other significant disease. PVD was diagnosed by a history of intermittent claudication and rest pain, a history of previous lower limb vascular surgery, absent or reduced pedal pulses, ankle-brachial pressure index (ABI) less than 0.9, or angiography showing significant stenosis in low extremity arteries. Personal identifiers were removed from the completed data collection sheets before entry into an Epidata 3.0 database. The entering and editing of the tool data was accomplished by an experienced wound care specialist and a diabetologist, who verified the accuracy and authenticity of the data; any disagreements were resolved by discussion. Prospective follow-up One year after discharge day, 687 patients who were selected randomly from 1333 patients were re-contacted via telephone to assess the incidence of new foot ulcers occurring since the baseline. The first new foot ulcer event that occurred within the follow-up period was recorded as the final outcome for each patient to calculate the annual incidence of foot ulcers. The returned questionnaires that reported a positive new ulcer event were rigorously crosschecked, i.e., the patients were re-interviewed by principle investigators at every center. Thus, the reported new ulcer events were confirmed, and the reported false-positive events were eliminated. Ethical approval was obtained from the research ethics committees for this study. Statistical Analysis The statistical analyses were performed using IBM SPSS, version 17.0, for Windows® (SPSS,® Inc., Chicago, Illinois, USA). The statistical tests and confidence intervals were two-sided, with a significance level of 0.05. We conducted a baseline univariate analysis of the data on the participants’ demographic, social, and medical characteristics between the DM and the DFU patients. The Chi-square test was applied for the comparison of the categorical 6
Wound Repair and Regeneration This article is protected by copyright. All rights reserved.
Page 6 of 25
Page 7 of 25
Manuscript under review - CONFIDENTIAL
variables. The normally distributed continuous variables were tested using independent sampled T-tests, whereas the non-normally distributed variables were analyzed using the two-independent sample Mann-Whitney U test. The variables that were significant at the 0.10 level were included in the multivariable analysis. We applied two standard binary logistic-regression models for the multivariable analysis. One model was performed to find a cross-sectional association between the DM patients and the DFU patients. The other model was for the longitudinal analysis to predict the risk factors and incidence probability for foot ulcers in the DM patients. The variables were retained in the logistic regression model if they were significant at the 0.05 level. The adjusted odds ratio and 95% confidence intervals were calculated with the logistic regression to determine the associations between DM and new ulcers as the outcome. RESULTS This work is a multi-center prospective cohort study of 1333 inpatients (452 DFU patients and 881 DM patients), who were treated for diabetic foot problems (DFP) by a multi-disciplinary team from June 1, 2011 to May 30, 2013; 687 patients were randomly selected for the follow up. Demographics A total of 881 DM patients and 452 DFU patients were investigated in the eight hospitals over 2 years. The demographic details of the patients are displayed in Table 1. Of the 881 DM patients, 490 were male and 391 were female. Of the 452 DFU patients, 294 were male and 157 were female. The two gender ratios have great statistical variability (χ2=11.282, P=0.001). The median age of the DM patients and DFU patients was 56.0 years (47, 66; 25%, 75%) and 64.0 years (55, 72; 25%, 75%) (p
