Scholarly Report submitted in partial fulfillment of the MD Degree at Harvard Medical School 22 February 2017 Stephanie Lin RISK FACTORS FOR KIDNEY DISEASE IN CHINESE AMERICANS: DATA FROM THE KIDNEY DISEASE SCREENING AND AWARENESS PROGRAM (KDSAP) Mentor: Li-Li Hsiao, M.D., Ph.D. Department of Renal Medicine, Brigham and Women’s Hospital 1 Abstract Title: Risk Factors for Kidney Disease in Chinese Americans: Data from the Kidney Disease Screening and Awareness Program (KDSAP) Authors: Stephanie Lin and Li-Li Hsiao, M.D., Ph.D. Purpose: To characterize the Chinese-American population screened through the Kidney Disease Screening and Awareness Program (KDSAP) and identify risk factors for proteinuria. Methods: Cross-sectional data were collected through KDSAP screening sites in the United States (Massachusetts, New Jersey, and Maine) and Canada (Toronto) from 2011-2016. Data were collected in the form of surveys assessing demographics, medical history, and healthcare access as well as biometric measurements (body mass index, waist-hip-ratio, random finger-stick blood glucose, and blood pressure) and urinalysis. Risk factors for proteinuria were identified through multivariable logistic regression. Results: In total, 510 participants were screened, with a prevalence of proteinuria of 14.1%. 30% of the total screened population was uninsured, while among subjects found to have proteinuria, 22.2% were uninsured. Independent risk factors for proteinuria included untreated diabetes (OR 2.38, p<0.05), history of proteinuria (OR 2.76, p<0.01), and increased diastolic blood pressure (OR 1.05, p <0.01). Less frequent medical visits were negatively correlated with proteinuria (OR 0.22, p<0.05). Conclusions: The data demonstrate that KDSAP is identifying individuals with proteinuria in an at-risk population with higher-than-average uninsured rates. The typical risk factors for kidney disease previously identified in native Chinese and non-Asian populations such as age, gender, and cardiovascular disease were not statistically significant in this dataset, while untreated diabetes, history of proteinuria, and increased diastolic blood pressure were significant. Despite the sample size limitation, this study provides a preliminary analysis of a unique dataset from a population that has been understudied in the kidney disease literature. 2 Table of Contents GLOSSARY OF ABBREVIATIONS 4 SECTION 1: INTRODUCTION 5 SECTION 2: STUDENT ROLE 7 SECTION 3: METHODS 7 SECTION 4: RESULTS 9 Population Characteristics 9 Multivariable Logistic Regression 12 SECTION 5: DISCUSSION, LIMITATIONS, CONCLUSIONS, AND FUTURE DIRECTIONS 12 Discussion 12 Limitations 13 Conclusions 14 Future Directions 15 SECTION 6: ACKNOWLEDGEMENTS 15 REFERENCES 16 3 Glossary of Abbreviations ARIC: Atherosclerosis Risk in Communities BMI: body mass index BP: blood pressure CAD: coronary artery disease CHF: congestive heart failure CHS: Cardiovascular Health Study CKD: chronic kidney disease CVD: cardiovascular disease ESRD: end-stage renal disease FHS: Framingham Heart Study FSBG: finger-stick blood glucose GFR: glomerular filtration rate HbA1c: hemoglobin A1c HDL: high-density lipoprotein HTN: hypertension KDSAP: Kidney Disease Screening Awareness and Prevention NHANES: National Health and Nutrition Examination Survey PCP: primary care provider PVD: peripheral vascular disease SD: standard deviation U.S.: United States UTI: urinary tract infection WHR: waist-to-hip ratio 4 Section 1: Introduction Chronic kidney disease (CKD) is the progressive loss of renal function most often associated with aging, smoking, diabetes, hypertension, obesity and cardiovascular disease1 and affects approximately 1 in 10 Americans2 or over 20 million Americans as of 2014.3 The asymptomatic nature of early renal disease requires diagnosis by findings on laboratory tests, specifically proteinuria and reduced GFR, which may be detected on routine lab tests. However, as many as 50% of at-risk patients do not undergo testing.4 Late detection as well as late referrals to nephrologists are associated with worse outcomes such as increased mortality and hospitalizations.5,6 Approximately 16 out of every 1000 cases of CKD progress to end-stage renal disease (ESRD), a highly morbid condition without simple treatments.7 Dialysis and kidney transplant are the mainstays of ESRD therapy and require significant time and resources while being incredibly disruptive to the patient’s way of life. Furthermore, the progression of CKD is associated with high cardiovascular morbidity and mortality.8 The cost burden on our healthcare system for these life-saving treatments is also another consideration. In 2014, spending on CKD patients constituted 20% of all Medicare expenditures among beneficiaries aged 65 and older, and 44% among beneficiaries younger than age 65.9 Fortunately, there are several interventions with good evidence for delaying the progression to ESRD including glycemic control,10 angiotensin-converting-enzyme inhibitors,11 and early referral to nephrologists.5,6 In hopes of developing better tools for early detection, several groups have published predictor tools that estimate risk of progression of kidney disease utilizing patient demographics and lab data. Predictors of new-onset kidney disease identified from the Framingham Heart Study include older age, high body mass index (BMI), diabetes, smoking, and hypertension1 which is consistent with several other predictor tools (Table 1). Models that incorporate urinalysis data have also consistently shown that proteinuria, a key feature of CKD, is predictive of disease development or progression.12–14 Interestingly, though urinalysis is a cheap and quick test, it was found to be cost-ineffective when used alone as a screening tool for CKD in the general adult population older than 50, though there may be more benefit in high- risk populations.15 Nevertheless, many of the predictor tools were developed utilizing urinalysis 5 data in addition to baseline patient characteristics readily available from primary care records. A summary of predictor tools is shown in Table 1. It is important to note almost all of these tools have only been verified in Caucasian populations14,12,16,17 despite higher prevalence of CKD in Asian populations. In 2014, the prevalence of CKD among Asians in the Medicare population older than 65 was 11.7 compared to 10.7 among Caucasians.18 While risk factors for CKD are likely similar in both groups, no studies have verified this to be true. Furthermore, diseases may have heterogeneous courses for different ethnicities even when exposed to the same environment. For example, the risk of developing type 2 diabetes was found to be much higher among Asian Americans when adjusted for BMI, providing the impetus for lowering BMI cutoffs in obesity screening and risk factor identification in this population.19,20 This is also significant for kidney disease given the high correlation between diabetes and CKD. There has been one predictor tool specifically developed using data from Taiwanese patients13 (Table 1), but the generalizability of this data to Chinese communities living in North America is dubious, as it does not address the contribution of dramatically different environmental variables. Other groups have also studied factors associated with kidney disease, as measured by GFR, in native Chinese populations without developing validated predictor tools (summarized in Table 2). These studies also identified age, diabetes, hypertension, and cardiovascular disease as the main risk factors. It is important to note that the studies that excluded proteinuria in their final model lacked urine studies. Given the scarcity of CKD data from the Asian-American population, this project aims to provide a preliminary analysis of the data collected from the Kidney Disease Screening and Awareness Program (KDSAP), which targets predominantly Asian-American communities across the U.S and Canada. KDSAP was developed by the Asian Renal Clinic of the Renal Division at Brigham and Women’s Hospital and implemented at Harvard College in 2008 to foster career interests in medicine (particularly nephrology) and to provide free kidney disease screening in Asian-American, Hispanic/Latino, and African-American communities. Over the 6 years, it has grown to 15 chapters across the U.S. with occasional screening visits to Canada, and has expanded its outreach efforts to other minorities and underserved groups. Since Chinese subjects were the most heavily represented group in our dataset, we focused our analysis on this population. Our aim was to determine the factors that correlate with detection of kidney disease, as measured by proteinuria, within Chinese communities in North America with the future goal of developing a predictor tool for this population. Section 2: Student Role The data was collected by KDSAP volunteers under the direction of my mentor, Dr. Li-Li Hsiao. The data was provided to me for analysis. I performed all the data entry and statistical analysis and authored the entirety of this report with feedback from Dr. Hsiao. Section 3: Methods The current available dataset contains approximately 900 patients screened in the Greater Boston area, New Jersey, Maine, and Toronto by the Harvard College chapter of KDSAP between 2011 and 2016. Participant age ranged from 18 to 95. While the subjects were predominantly Asian- American, other racial groups and ethnicities were also represented. We focused only on the Chinese population as they were most heavily represented. Subjects were selected based on having selected “Asian” as their race, with some specifying “Chinese” as their ethnicity and others specifying one of several Chinese dialects as their primary spoken language (e.g. Mandarin, Cantonese, Southern-Min/Taiwanese, Taishanese/Hoisanese, etc.). Participants were asked to fill out surveys containing questions about their past medical history including history of diabetes, hypertension, hyperlipidemia, tobacco and alcohol use, coronary artery disease (CAD), malignancy, anemia, glaucoma, autoimmune disease, as well as use of prescription medications and herbal remedies. The participants were also asked to indicate if there was family history of the medical conditions listed. Healthcare utilization and socioeconomic circumstances were also assessed with questions inquiring insurance status, access to primary care doctors or specialists, education level, and language barrier when interacting with healthcare providers. Quantitative data collected included BMI, waist-to-hip ratio (WHR), blood pressure (BP), 7 Table 1: Significant Variables in Published Prediction Models for CKD Fox et al.1 O’Seaghdha et al.14 Bang et al.12 Kshirsagar et al.16 Alssema et al.17 Chien et al.13 Studied Not specified but “not White (100%, 76.1%), White (72%), Hispanic White (78.6%), not White (100%) Native Taiwanese Population nationally Black (23.9%) (14%), Black (10%) specified (100%) representative or ethnically diverse” (ARIC and FHS) (NHANES) (FHS and CHS) Significant Age Age Age Age Age Age Variables Smoking Female Female Smoking Diabetes Diabetes HTN HTN HTN BP meds Stroke Diabetes Diabetes Diabetes First-degree relative PVD PVD with MI, stroke, BMI History of CHF and History of CHF diabetes Proteinuria BMI CVD Diastolic BP Baseline GFR Uric acid HTN Proteinuria Proteinuria Anemia BMI Postpradial glucose HDL Baseline GFR Anemia Waist circum. HbA1c *No urine data *No urine data *No lab data Table 2: Summary of Factors Associated with Kidney Disease in Chinese/Taiwanese Populations Zhang et al.21 Pan et al. 22 Zhang et al.23 Xue et al. 24 Liu et al.25 Shan et al.26 Studied Native Chinese Native Chinese- Native Chinese - Native Chinese – Native Chinese – Native Chinese – Population Zhuang ethnicity in Beijing Guangxi Jing Southern rural Henan province in Southwest region community population Central China (rural) Significant Age Age Age Age Age# Variables Female Female Female HTN HTN HTN > 10yrs HTN HTN# HTN Diabetes CVD Diabetes Diabetes# Diabetes CVD HDL < 40mg/dL Hyperlipidemia CAD Nephrotoxic meds. Lower education Hyperlipidemia Hyperuricemia Rural area level Hyperuricemia Nephrotoxic meds. #Associated with albuminuria Abbreviations: ARIC: Atherosclerosis Risk in Communities; BMI: body mass index; BP: blood pressure; CHF: congestive heart failure; CHS: Cardiovascular Health Study; CVD: cardiovascular disease; FHS: Framingham Heart Study; GFR: glomerular filtration rate; HbA1c: hemoglobin A1c; HDL: high-density lipoprotein; HTN: hypertension; NHANES: National Health and Nutrition Examination Survey; PVD: peripheral vascular disease 8 random finger-stick blood glucose (FSBG), and urine dipstick results (including proteinuria and hematuria). Proteinuria was graded trace, 30 mg/dL, 100mg/dL, 300 mg/dL, or >300mg/dL. This study was originally designed to utilize the split-sample method, using half the sample to develop a predictor equation for CKD, which would be tested and validated in the remaining half of the sample. However, due to unforeseen issues with data entry, our sample size was reduced to half of the expected number. Thus, only multivariable logistic regression was performed to identify correlates of proteinuria. Since blood samples were not collected at these screening sessions, proteinuria served as the outcome measure for kidney disease. Given the limited sample size, proteinuria was designated as a dichotomous variable, where “no proteinuria” reflected a negative urine dipstick result and “proteinuria” reflected combined grades of proteinuria (from trace to >300mg/dL). Covariate predictors with the largest p-values greater than 0.05 were eliminated step-wise from the final model until the remaining covariates were all statistically significant at p < 0.05. Interactions between independent variables in the final model and age and gender were assessed and were not found to be statistically significant. Section 4: Results Population Characteristics The final study sample consisted of 510 Chinese subjects who provided urinalysis data. Based on home addresses provided, the most geographically represented places were Massachusetts (N=261), New Jersey (N=119), and Maine (N=44) in the U.S. and Ontario (N=54) in Canada. The rates of detected proteinuria were 13.8%, 17.6%, 9.0%, and 16.7% in each region respectively. Population characteristics are summarized below in Table 3. The screening population was predominantly female (61%) with mean age of 61. The prevalence of diabetes was 18.2%, slightly lower than the 20.6% prevalence among non-Hispanic Asian participants studied in the National Health Nutrition Examination Survey (NHANES) between 1988 and 2012.27 Hypertension was much more prevalent than diabetes in our dataset at 35.7% among all participants and 44.4% in the proteinuria group. 9 Only 25.1% of all subjects were English-speaking though only 24.5% expressed difficulty communicating with their healthcare provider due to language barrier. 54.3% had participated in higher education (defined as “some college” to “post-graduate degree”). 66.5% identified a primary care provider, and 71.0% had health insurance (similar rates in both U.S. and Canada groups). 19.6% reported that lack of insurance prevented them from accessing care, whereas 28.0% reported that obtaining medical care was somewhat to extremely difficult. Notably, our screening measures detected trace or greater proteinuria in 14.1% of subjects. Of these individuals, only 15.3% self-reported a history of kidney disease and 18.1% had prior episodes of proteinuria. The frequencies of the degrees of proteinuria are shown in Table 4. Of those found to have proteinuria, 22.2% were uninsured and 25% expressed difficulty communicating with their healthcare providers. Remarkably, there were more English- speakers and individuals with higher education in the proteinuria group compared to the no proteinuria group (30.6 % vs. 24.2% and 59.7% vs. 53.4% respectively). Table 3: Population Characteristics Variable All (N = 510) No Proteinuria Proteinuria (N = 438) (N = 72) Proteinuria, ≥ trace (%) 14.1 - - Mean Age (SD) (years) 61 (15) 61 (15) 61 (18) Female gender (%) 61.0 61.9 55.6 Higher Education, ≥ college (%) 54.3 53.4 59.7 English speaker (%) 25.1 24.2 30.6 History of diabetes (%) 18.2 17.1 25.0 History of hypertension (%) 35.7 34.2 44.4 Other Cardiovascular Risk History of (%): - High cholesterol 38.2 37.9 40.3 - CAD 6.5 6.4 6.9 - Arrhythmia 16.7 16.2 19.4 - Heart failure 1.6 1.8 0 - Stroke 1.6 1.4 2.8 Kidney-related Diseases History of (%): - Kidney disease, unspecified 8.8 7.8 15.3 - Proteinuria 9.2 7.8 18.1 - Hematuria 8.6 8.7 8.3 - Nephrolithiasis 8.6 8.0 12.5 - Frequent bladder infection or UTI 11.4 11.9 8.3 - Gout 6.3 6.4 5.6 10 Other Medical History History of (%): - Anemia 12.9 13.2 11.1 - Autoimmune disease, unspecified 1.4 1.6 0 - Glaucoma 4.9 5.0 4.2 - Cancer 4.7 5.0 2.8 Last seen doctor (yr) - <1 year 78.3 76.6 88.4 - 1-2 years ago 10.7 11.9 2.9 - More than 2 years ago 11.0 11.5 8.7 Current Prescription Medications # (%) - 0-3 79.1 79.0 79.7 - 4-6 16.8 17.5 13.0 - >7 4.0 3.5 7.3 Access to Care - Has PCP 66.5 65.5 72.2 - Language barrier (%) 24.5 24.4 25.0 - Medication Insurance (%) 70.4 69.2 77.8 - Health insurance (%) 71.0 69.9 77.8 - Lack of insurance preventing access to 19.6 20.1 16.0 care (%) - Somewhat to extreme difficulty 28.0 29.3 20.0 obtaining medical care (%) Alternative Medicine - Acupuncture (%) 7.1 7.5 4.2 - Herbal medicines (%) 20.4 21.2 15.3 - Over-the-counter supplements (%) 61.0 61.2 59.7 Substances - Current smoker 5.3 5.5 4.2 - Former smoker 9.8 9.1 13.9 - Alcohol drinker 23.1 23.7 19.2 Body measurements - BMI, mean (SD) 24.3 (7.6) 24.3 (8.0) 24.4 (3.2) - WHR, mean (SD) 0.907 (0.061) 0.906 (0.060) 0.913 (0.065) - Random FSBG, mean (SD) 117 (43.9) 116 (43.8) 121 (44.8) - Systolic BP, mean (SD) 131 (19) 131 (18) 135 (24) - Diastolic BP, mean (SD) 77 (10) 77 (9) 81 (12) Abbreviations: BMI: body mass index; BP: blood pressure; CAD: coronary artery disease; FSBG: finger-stick blood glucose; PCP: primary care provider; SD: standard deviation; UTI: urinary tract infection; WHR: waist-hip ratio Table 4: Proteinuria grade frequency (%) (N = 72) Proteinuria Grade Freq. (%) Trace 6.9 30 mg/dL 5.1 100 mg/dL 1.8 300 mg/dL 0.4 ≥2000 mg/dL 0 11 Multivariable Logistic Regression The final regression model represented by odds ratios and p-values is shown below in Table 5. Risk factors associated with proteinuria included untreated diabetes (p<0.05), history of proteinuria (p<0.01), and increased diastolic blood pressure (i.e. for every 1 mmHg increase, the odds of developing proteinuria increased by 5%, p<0.01). Interestingly, after removing the interaction variable of diabetes and diabetes treatment (defined here as oral medications or insulin), diabetes alone was not statistically significant in our model (p=0.24). Having seen a doctor 1-2 years ago was associated with lower odds of proteinuria compared to having seen a doctor within the last year (p<0.05). To parse this, the frequency of visits was interacted with treated or untreated diabetes. When compared to treated diabetics who had visited the doctor <1 year ago, treated diabetics who visited the doctor 1-2 years ago had decreased odds of having proteinuria (OR 0.12, p = 0.035) while untreated diabetics who visited the doctor >2 years ago had dramatically increased odds of developing proteinuria, though this was only significant at the 10% level (OR = 5.67, p = 0.086). Table 5: Risk Factors Associated with Proteinuria Variables OR (95% CI) p-values Untreated diabetes 2.38 (1.04-5.45) 0.04 History of proteinuria 2.76 (1.30-5.83) 0.008 Last doctor visit 1-2 years ago (compared to within the last year) 0.22 (0.05-0.96) 0.04 Diastolic blood pressure 1.05 (1.02-1.07) 0.002 Section 5: Discussion, Limitations, Conclusions, and Future Directions Discussion Our data suggest that diabetes alone may not be as strong of a predictor for proteinuria among Chinese Americans as it may be in other populations, as this relationship was only statistically significant when excluding treated individuals. Interestingly, age and gender were not statistically significant correlates of proteinuria in this group as commonly seen in other populations. Cardiovascular disease processes implicated in previous studies of CKD risk factors 12 such as hypertension, hyperlipidemia, CAD, and stroke were also not statistically significant in this study population, though increases in diastolic blood pressure had modest effects. Lifestyle factors (use of tobacco, alcohol, or alternative medicines) were also not statistically significant. As expected, a history of proteinuria resulted in increased odds of present proteinuria. Interestingly, having visited the doctor 1-2 years ago was associated with lower odds of proteinuria when compared to having a visit within the last year. The analysis of the interaction between diabetes treatment status and frequency of visits suggests heterogeneity in the diabetic group – likely consisting of those with well-controlled, treated disease requiring less frequent follow-up (hence the decreased odds of developing proteinuria) and those with untreated disease who have poor follow-up (hence the increased odds of developing proteinuria). Approximately 30% of the screening population reported not having health insurance with the rate dropping to 22% in the proteinuria group. These values were consistent when excluding Canadian participants. These uninsured numbers are higher than the estimated U.S. national average of 11-17% from 2010 through 2016 (the lower end of this range reflects the effects of the Affordable Care Act).28 The uninsured rate combined with the statistic of 28% who reported difficulty accessing medical care suggests that KDSAP is reaching medically underserved Chinese communities. Limitations There are several limitations to this study. First, blood samples were not collected from study participants and therefore we could not diagnose CKD using creatinine (and derived GFR). However, proteinuria has been shown to be predictive of progression to ESRD independently of creatinine.29 Due to our limited sample size, we were also unable to regress different levels of proteinuria separately, with the understanding that clinically significant proteinuria corresponds to a dipstick result of ≥ 300 mg/dL, while anything less than that may represent benign transient proteinuria and require follow-up measurements.30 Repeat urinalysis was impossible given the cross-sectional nature of this study. While not all of the positive dipstick results may be indicative of true underlying kidney disease, we believe it was valuable to share these results with our participants to ensure appropriate follow-up with healthcare providers. Importantly, 13 even mild (trace to 30 mg/dL) proteinuria has been linked to increased rates of ESRD, doubling of serum creatinine, and increased all-cause mortality independently of decreases in GFR.31 While KDSAP volunteers make every effort to accommodate language barriers and different levels of education (for example, surveys are translated into Chinese and translators are often on site), there were likely inaccuracies in the data collected through surveys which reflect errors in self-reporting. The current data also excludes immigration history which will make it difficult to discern the effect of immigration on health outcomes should this data be compared to data from the native Chinese population. Due to unforeseen issues with data entry within the allotted research timeframe, the sample size was limited relative to the number of variables studied. This is the most significant limitation as it increases the risk of type I error and decreases the ability to detect significant variables with more modest effects (which may include variables found to be significant in other studies such as age, gender, hypertension, cardiovascular disease, etc.). Moreover, the confidence intervals of the odds ratios approach 1 at the extremes, which indicate the possibility that the risk factors identified have very small effects on proteinuria. Finally, the pseudo-R squared value of 0.072 of the final model is quite low and suggests our model explains only a small proportion of the variability of the outcome. Thus, the results of the regression should be interpreted with caution. Because of the sample size limitation, we were also unable to proceed with developing and validating a predictor tool as originally planned. However, we are continuing to work with other screening sites to obtain data and will plan additional screening sessions. In the interim, this study provides a preliminary analysis of the data. Conclusions This study provides a valuable initial assessment of kidney disease data gathered from Chinese communities in North America who are screened through KDSAP efforts. Our unique, though limited, dataset revealed that untreated diabetes was the strongest risk factor for proteinuria among Chinese Americans screened, whereas other typical predictors of kidney disease identified in other populations such as age, gender, and cardiovascular disease (hypertension, 14 hyperlipidemia, CAD, and stroke) did not demonstrate any statistically significant relationship. Between 2011 and 2016, the uninsured rate of the screening population was 30%, which is higher than the U.S. national average. While individuals found to have proteinuria were more likely to be insured, the uninsured rate is still higher than the national average at 22%. This suggests that KDSAP is reaching a vulnerable population in terms of healthcare access. Future Directions Other KDSAP chapters across the country have also started collecting data, and we are currently working to have their data available to us. As KDSAP continues its outreach efforts, we also expect the sample size to grow, which will allow for more analysis in the future. Ideally, we would have enough data to regress each grade of proteinuria independently. We also hope to develop a predictor tool that may be used to identify Chinese patients at risk for developing CKD. This project may also facilitate the comparison of Chinese Americans to their counterparts in their native country to allow further analysis of the effects of immigration and the adaptation of Western diet and habits on kidney disease risk. Our data also demonstrated high levels of hematuria (27.7%), after excluding concurrent pyuria, females younger than 51 (as menstruation was not consistently documented and the average age of menopause is 51), and all individuals reporting history of nephrolithiasis or frequent bladder/urinary tract infections. As the incidence of kidney disease due to chronic glomerulonephritis is higher in Asian countries with evidence of genetic susceptibility in these populations,32 future directions should include studying the characteristics of individuals found to have hematuria through our screenings. Section 6: Acknowledgements I would like to thank my mentor, Dr. Li-Li Hsiao, and her research assistants Allison Wu, Jennie Kuo, and Laura Polding for their support. I would also like to thank the multiple KDSAP chapters for their efforts in screening patients and collecting data. Finally, I would like to thank my family and my husband, Tom Beckford, who have always surrounded me with love. 15 References 1. Fox CS, Larson MG, Leip EP, Culleton B, Wilson PWF, Levy D. Predictors of new-onset kidney disease in a community-based population. JAMA J Am Med Assoc. 2004;291(7):844-850. doi:10.1001/jama.291.7.844. 2. Coresh J, Selvin E, Stevens L a, et al. Prevalence of chronic kidney disease in the United States. JAMA. 2007;298(17):2038-2047. doi:10.1001/jama.298.17.2038. 3. Centers for Desease Control and Prevention. National Chronic Kidney Disease Fact Sheet: General Information and National Estimates on Chronic Kidney Disease in the United States. Atlanta, GA; 2014. 4. Stevens L a, Fares G, Fleming J, et al. Low rates of testing and diagnostic codes usage in a commercial clinical laboratory: evidence for lack of physician awareness of chronic kidney disease. J Am Soc Nephrol. 2005;16:2439-2448. doi:10.1681/ASN.2005020192. 5. Smart NA, Titus TT. Outcomes of early versus late nephrology referral in chronic kidney disease: a systematic review. Am J Med. 2011;124(11):1073-80 e2. doi:S0002-9343(11)00412-8 [pii]\r10.1016/j.amjmed.2011.04.026. 6. Kinchen KS, Sadler J, Fink N, et al. The timing of specialist evaluation in chronic kidney disease and mortality. Ann Intern Med. 2002;137(6):479-486. 7. Hsu C, Vittinghoff E, Lin F, Shlipak MG. The Incidence of End-Stage Renal Disease Is Increasing Faster than the Prevalence of Chronic Renal Insufficiency. Ann Intern Med. 2004;141(2):95-101. doi:10.7326/0003-4819-141-2-200407200-00007. 8. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004;351(13):1296-1305. doi:10.1056/NEJMoa041031. 9. United States Renal Data System. 2016 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States. Bethesda, MD; 2016. https://www.usrds.org/2016/view/Default.aspx. Accessed January 26, 2017. 10. Turner R, Matthews D, Neil a., Mcelroy H. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. Br Med Journaledical. 1998;317:703-713. doi:10.1136/bmj.317.7160.703. 11. Giatras I, Lau J, Levey AS. Effect of angiotensin-converting enzyme inhibitors on the progression of nondiabetic renal disease: a meta-analysis of randomized trials. Angiotensin-Converting- Enzyme Inhibition and Progressive Renal Disease Study Group. Ann Intern Med. 1997;127(5):337-345. http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=med4&NEWS=N&AN=927382 4. 12. Bang H, Vupputuri S, Shoham DA, et al. SCreening for Occult REnal Disease (SCORED): a simple prediction model for chronic kidney disease. Arch Intern Med. 2007;167(4):374-381. doi:10.1001/archinte.167.4.374. 13. Chien K-L, Lin H-J, Lee B-C, Hsu H-C, Lee Y-T, Chen M-F. A prediction model for the risk of incident chronic kidney disease. Am J Med. 2010;123(9):836-846.e2. doi:10.1016/j.amjmed.2010.05.010. 14. O’Seaghdha CM, Lyass A, Massaro JM, et al. A risk score for chronic kidney disease in the general population. Am J Med. 2012;125(3):270-277. doi:10.1016/j.amjmed.2011.09.009. 15. Boulware LE, Jaar BG, Tarver-Carr ME, Brancati FL, Powe NR. Screening for proteinuria in US adults: a cost-effectiveness analysis. JAMA. 2003;290(23):3101-3114. doi:10.1016/S0022- 5347(05)61793-4. 16. Kshirsagar A V, Bang H, Bomback AS, et al. A simple algorithm to predict incident kidney disease. Arch Intern Med. 2008;168(22):2466-2473. doi:10.1001/archinte.168.22.2466. 17. Alssema M, Newson RS, Bakker SJL, et al. One risk assessment tool for cardiovascular disease, 16 type 2 diabetes, and chronic kidney disease. Diabetes Care. 2012;35(4):741-748. doi:10.2337/dc11-1417. 18. United States Renal Data. 2015 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States. Bethesda, MD; 2015. 19. Lee JWR, Brancati FL, Yeh H-C. Trends in the prevalence of type 2 diabetes in Asians versus whites: results from the United States National Health Interview Survey, 1997-2008. Diabetes Care. 2011;34(2):353-357. doi:10.2337/dc10-0746. 20. McNeely MJ, Boyko EJ. Type 2 diabetes prevalence in Asian Americans: Results of a national health survey. Diabetes Care. 2004;27(1):66-69. doi:10.2337/diacare.27.1.66. 21. Zhang L, Wang F, Wang L, et al. Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet. 2012;379(9818):815-822. doi:10.1016/S0140-6736(12)60033-6. 22. Pan L, Ma R, Wu Y, et al. Prevalence and risk factors associated with chronic kidney disease in a Zhuang ethnic minority area in China. Nephrology (Carlton). 2015;20(11):807-813. doi:10.1111/nep.12510. 23. Zhang L, Zhang P, Wang F, et al. Prevalence and factors associated with CKD: a population study from Beijing. Am J Kidney Dis. 2008;51(3):373-384. doi:10.1053/j.ajkd.2007.11.009. 24. Xue C, Ye XD, Li W, et al. Prevalence of chronic kidney disease in Jing adults in China: a village- based study. Clin Nephrol. 2013;79(1):50-56. doi:10.5414/CN107511. 25. Liu Q, Li Z, Wang H, et al. High Prevalence and Associated Risk Factors for Impaired Renal Function and Urinary Abnormalities in a Rural Adult Population from Southern China. Burdmann EA, ed. PLoS One. 2012;7(10):e47100. doi:10.1371/journal.pone.0047100. 26. Shan Y, Zhang Q, Liu Z, Hu X, Liu D. Prevalence and risk factors associated with chronic kidney disease in adults over 40 years: a population study from Central China. Nephrol. 2010;15:354-61 ST-Prevalence and risk factors associate. doi:NEP1249 [pii]. 27. Menke A, Casagrande S, Geiss L, Cowie CC. Prevalence of and Trends in Diabetes Among Adults in the United States, 1988-2012. Jama. 2015;314(10):1021-1029. doi:10.1001/jama.2015.10029. 28. Auter Z. US Uninsured Rate Holds at Low of 10.9% in Fourth Quarter | Gallup. Gallup. http://www.gallup.com/poll/201641/uninsured-rate-holds-low-fourth-quarter.aspx. Published 2017. Accessed January 26, 2017. 29. Iseki K, Ikemiya Y, Iseki C, Takishita S. Proteinuria and the risk of developing end-stage renal disease. Kidney Int. 2003;63(4):1468-1474. doi:10.1046/j.1523-1755.2003.00868.x. 30. Carroll MF, Temte JL. Proteinuria in adults: A diagnostic approach. Am Fam Physician. 2000;62(6):1333-1340. 31. Hemmelgarn BR, Manns BJ, Lloyd A, et al. Relation between kidney function, proteinuria, and adverse outcomes. Jama. 2010;303(5):423-429. doi:10.1001/jama.2010.39. 32. Jha V, Garcia-Garcia G, Iseki K, et al. Chronic kidney disease: Global dimension and perspectives. Lancet. 2013;382(9888):260-272. doi:10.1016/S0140-6736(13)60687-X. 17