Use of a Medical-Alert Accessory in CKD

A Pilot Study

Abstract

Visual Abstract

Background and objectives 

Poor disease recognition may jeopardize the safety of CKD care. We examined safety events and outcomes in patients with CKD piloting a medical-alert accessory intended to improve disease recognition and an observational subcohort from the same population.

Design, setting, participants, & measurements 

We recruited 350 patients with stage 2–5 predialysis CKD. The first (pilot) 108 participants were given a medical-alert accessory (bracelet or necklace) indicating the diagnosis of CKD and displaying a website with safe CKD practices. The subsequent (observation) subcohort (n=242) received usual care. All participants underwent annual visits with ascertainment of patient-reported events (class 1) and actionable safety findings (class 2). Secondary outcomes included 50% GFR reduction, ESKD, and death. Cox proportional hazards assessed the association of the medical-alert accessory with outcomes.

Results 

Median follow-up of pilot and observation subcohorts were 52 (interquartile range, 44–63) and 37 (interquartile range, 27–47) months, respectively. The frequency of class 1 and class 2 safety events reported at annual visits was not different in the pilot versus observation group, with 108.7 and 100.6 events per 100 patient-visits (P=0.13), and 38.3 events and 41.2 events per 100 patient visits (P=0.23), respectively. The medical-alert accessory was associated with lower crude and adjusted rate of ESKD versus the observation group (hazard ratio, 0.42; 95% confidence interval, 0.20 to 0.89; and hazard ratio, 0.38; 95% confidence interval, 0.16 to 0.94, respectively). The association of the medical-alert accessory with the composite endpoint of ESKD or 50% reduction GFR was variable over time but appeared to have an early benefit (up to 23 months) with its use. There was no significant difference in incidence of hospitalization, death, or a composite of all outcomes between medical-alert accessory users and the observational group.

Conclusions 

The medical-alert accessory was not associated with incidence of safety events but was associated with a lower rate of ESKD relative to usual care.

Introduction

The approximately 30 million persons with CKD in the United States (¹) have limited disease-modifying treatments; hence, safeguarding their care is a priority for improving outcomes. CKD is often overlooked, improperly treated, and referred late for specialty care (²). As a result, patients with CKD are at higher risk of experiencing adverse safety events, defined as unintended harm from medical therapy (³). Reduced kidney function, or GFR, is a primary factor contributing to adverse safety events because kidney function affects the clearance of many drugs, and complicates therapeutic interventions (⁴). Moreover, the disease is frequently associated with comorbidities such as diabetes, hypertension, cardiovascular disease, and their complex and often conflicting treatments. Enhanced disease recognition offers the opportunity for caregivers to avoid treatment mishaps such as the administration of contraindicated or incorrectly doses medications (⁵).

Embedding note templates with clinical decision support or alerts into electronic health records to inform providers of the special care and documentation needs in CKD has modest effects on clinical care (⁶,⁷). Home medical network electronic health records alerts pertaining to CKD have limited utility because patients do not only seek care in their primary medical network but also elsewhere, and often involuntarily (⁸). Hence, portable banners indicating a patient’s medical condition to home and nonaffiliated medical providers may improve CKD recognition and the safety of care. Medical-alert accessories, typically a bracelet, can serve to transmit critical information to a health care provider, especially when the wearer is not able to reliably communicate. No studies have examined the role of a medical-alert accessory in changing providers’ CKD awareness, care delivery, or outcomes.

In this study, we report on a pilot group of individuals with CKD who were provided with a medical-alert accessory and enrolled in a longitudinal cohort observed for adverse safety events. We compared the outcomes of the pilot group with the remainder of the CKD cohort who were followed on an identical study schedule and tracked while receiving usual care.

Materials and Methods

The Safe Kidney Care study was an observational cohort enrolling 350 qualifying patients with stage 2–5 predialysis CKD. The study objective was to ascertain the frequency of adverse safety events in participants tracked prospectively for outcomes, including kidney function loss, ESKD, and death. As a pilot in the cohort we evaluated the utility of a medical-alert accessory as a novel intervention in CKD. The study was conducted in accordance with the International Conference on Harmonization Guideline for Good Clinical Practice and the Declaration of Helsinki. The protocol was approved by the University of Maryland School of Medicine Institutional Review Board and Baltimore Veterans Affairs Medical Center Research and Development Committee and registered with Clinicaltrials.gov (identifier: NCT01407367). All participants were recruited from nephrology clinics at the University of Maryland Medical Center and Baltimore Veterans Affairs Medical Center and provided written informed consent to participate. Participants were aged ≥21 years, with no exclusions on the basis of race or ethnicity. The first 108 study participants were enrolled in the pilot subcohort and provided with a medical-alert accessory. After completion of pilot group enrollment, the remaining 242 study participants were enrolled into the observational subcohort and followed with usual care. Study inclusion and exclusion criteria for all participants included the presence of CKD defined by a GFR estimated with the Modified Diet in Renal Disease equation of <60 ml/min per 1.73 m², measured on two outpatient occasions at least 90 days apart, and no more than 18 months before study enrollment. Patients were excluded if the clinical provider anticipated they would reach ESKD or death within a year of enrollment. Supplemental Figure 1 shows enrollment into the pilot and observation cohort were consecutive from the same clinic pool used for screening. The number of patients deemed eligible, and participant recruited are also depicted.

Procedures

All participants underwent annual in-center visits; 6-month telephone calls between visits until June 30, 2016; and were followed to the earliest time of ESKD, death, or study withdrawal. Additional outcome ascertainment continued until June 30, 2017 via annual telephone contact with the study participant or next of kin. At the baseline visit, pilot group participants were asked to wear the medical-alert accessory as the condition for pilot participation. At both the 6-month telephone and annual in-center visits, a range of data were collected, including concomitant medications, patient-reported safety events, and medical events (including hospitalizations), onset of ESKD, and death as reported by next of kin. Participants also underwent phlebotomy and physical measures at annual in-center visits. Kidney function was determined using serum creatinine to estimate GFR as calculated by the Modified Diet in Renal Disease equation. Fasting serum glucose, potassium, and hemoglobin were also obtained from the annual blood sample. Physical measures obtained during the in-center visit included seated BP and pulse after a 5-minute quiet period, and one standing BP and pulse after a 2-minute stand.

Medical-Alert Accessory

Pilot group participants were offered the choice of a bracelet or necklace made of either stainless steel or sterling silver. The accessory was engraved with the following statement on the visible surface: “Please consider my decreased kidney function in planning my care,” and an internet-based informational link, outlining the safe care of patients with kidney disease, on the back. This was intended to alert patients and providers of the former’s condition and offer a reference on best practices in safe CKD management. Each participant was encouraged to wear the accessory over the duration of the study and refer to the associated website. Each medical-alert accessory also had an engraved unique numeric identifier, and participants were requested to enter the identifier when accessing the website. Initial website usage by pilot group participants has been previously reported (⁹). All pilot subcohort participants were surveyed annually about their frequency of wearing the accessory.

Adverse Safety Events

We report adverse safety events recorded at annual visits including patient-reported safety incidents (class 1), as previously defined (⁷), and ascertained via participant questionnaires regarding incidents up to 6 months before the visit and attributed by the participant to a medication or treatment. Actionable safety findings (class 2) were those events detected at the corresponding annual visits from physical findings and blood testing, deemed attributable to a medication, with or without symptoms, and likely to be mitigated with medication modification. Class l events included hypoglycemia defined at serum glucose of <70 mg/dl with symptoms or <60 mg/dl with or without symptoms, and included self-report of hyperkalemia (by knowledge of laboratory value or hyperkalemic-treating intervention), dizziness, falling, and bleeding related to a medication. Class 2 events included hyperglycemia and hypoglycemia with serum glucose >250 mg/dl and <70 mg/dl, respectively. Additionally, hyperkalemia and hypokalemia were set at thresholds of serum potassium >5.5 mEq/L and <3.5 mEq/L, respectively. Anemia as an adverse safety event was defined at venous hemoglobin <9.0 g/dl. Vital sign changes included as class 2 events were sitting systolic BP of <90 mg Hg, orthostatic hypotension set at threshold of >20 mm Hg drop in systolic BP from sitting to standing position, and bradycardia defined as a sitting pulse of <50 beats per minute.

Statistical Analyses

For descriptive presentation of the pilot and observational participant characteristics and adverse safety events, continuous variables were summarized as mean±SEM and compared using the t test. Dichotomous and categorical variables were summarized as n (%) and compared using the chi-squared test. To account for the differential follow-up time in the pilot and observational groups, we tallied all events by study group and calculated the rate of any events per 100 patient-visits only including visits where specified measurement was obtained and excluding visits with missing readings.

For comparison of key outcomes in the pilot versus observation group, we used time-to-event analyses from baseline to outcome, loss to follow-up, or end of study. Kaplan–Meier curves were used to depict survival time in each group for key outcomes including onset of ESKD, the composite kidney endpoint of 50% reduction in GFR at a given annual visit from baseline or ESKD, death, and a composite of the kidney outcomes and death. Cox proportional hazards models were used for estimating hazard ratios of pilot versus observational groups with multivariable analyses. Covariates were measured at baseline and include age, sex, race, baseline GFR, presence of diabetes, hypertension, cardiovascular disease, and actively treated cancer. Additional covariates included baseline body mass index, mean arterial pressure, smoking history, level of education, employment status, and health literacy as measured by the Short Test of Functional Health Literacy in Adults. We also adjusted for baseline use of renin-angiotensin-aldosterone system (RAAS) blocker use, which was statistically different between the groups.

We assessed the proportionality assumption for each time-to-event analysis and, where necessary, examined time varying hazards associated with a specified end point. Analyses were based on the intent-to-treat assumption, with baseline assignment to the pilot versus observation group maintained over the duration of the analysis period.

Results

Baseline Characteristics

The median follow-up in the pilot and observational subcohorts was 52 (interquartile range, 44–63) and 37 (interquartile range, 27–47) months, respectively. The number of participants who withdrew or were lost to follow-up in the pilot and observational groups were four (4%) and 15 (6%), respectively. The number of expected and completed annual and telephone follow-up visits are depicted in Supplemental Table 1, with a slight preponderance of completed visits in the pilot group versus observation group and a greater anticipated drop-off in expected and completed visits for the latter reflecting its later commencement. Self-reported medical-alert accessory usage of at least 1–3 days per week was 84%, 76%, 75%, and 69% at year 1, 2, 3, and 4, respectively. Over the duration of the study, 41 participants developed ESKD and 56 others died. Table 1 depicts the demographic characteristics of each subcohort. Pilot group participants were slightly younger, more likely to be female, with a slight preponderance of participants with stage 2 CKD measured at baseline kidney function measurement, and after eligibility determination. Pilot group participants had a slightly lower proportion with cancer and slightly higher baseline use of RAAS blockers.

Table 1. - Demographic characteristics of participants by study group

Characteristic	Pilot Group	Observation Group	P Value
Participants n (%)	108 (31)	242 (69)
Age, mean±SD	65±11	67±11	0.68
≥65 yr	52 (48)	152 (63)	0.01
<65 yr	56 (52)	90 (37)
Sex
Male	70 (65)	181 (75)	0.04
Female	38 (35)	61 (25)
Black
Yes	78 (72)	165 (68)	0.57
No	30 (28)	77 (32)
BMI a at baseline, mean±SD	34±8	33±7	0.55
MAP at baseline, mean±SD	88±15	88±12	0.32
Baseline CKD
GFR, mean±SD	46±15	45±15	0.84
CKD stage 2, 60–89 ml/min/1.73m2	19 (18)	35 (15)	0.75
CKD stage 3A, 45–59 ml/min/1.73m2	36 (33)	86 (35)
CKD stage 3B, 30–45 ml/min/1.73m2	36 (33)	82 (34)
CKD stage 4, 15–30 ml/min/1.73m2	15 (14)	31 (13)
CKD stage 5, <15 ml/min/1.73m2	2 (2)	8 (3)
Hypertension
Yes	103 (95)	236 (97)	0.29
No	5 (5)	6 (3)
Cancer
Yes	16 (15)	61 (25)	0.03
No	92 (85)	181 (75)
Diabetes
Yes	68 (63)	151 (62)	0.22
No	40 (37)	91 (38)
Smoking
Yes	88 (82)	199 (82)	0.87
No	20 (18)	43 (18)
Health literacy
Adequate, S-TOFHLA ≥67	69 (64)	165 (68)	0.69
Inadequate, S-TOFHLA <67	33 (31)	71 (29)
Not performed	6 (5)	6 (3)
Education
< High school diploma	22 (20)	40 (17)	0.46
High school graduate/GED/vocational degree	64 (60)	160 (66)
College graduate/graduate degree	22 (20)	42 (17)
Employment status
Employed full or part time	19 (18)	38 (16)	0.66
Unemployed/retired/permanently disabled	89 (82)	204 (84)
Baseline RAAS blocker			0.02
Yes	90 (83)	174 (72)
No	18 (17)	68 (28)
Baseline NSAIDs			0.27
Yes	5 (5)	19 (8)
No	103 (95)	223 (92)
During the past 30 d did you forget to take a medicine, not take a medicine on purpose, or add an extra pill?			0.40
Yes	65 (60)	134 (55)
No or not reported	43 (40)	108 (45)

Data are shown as n (%) unless otherwise indicated. Pilot group includes participants provided with a medical-alert accessory. BMI, body mass index; MAP, mean arterial pressure; S-TOFHLA, Short Test of Functional Health Literacy in Adults; GED, general education diploma; RAAS, renin-angiotensin-aldosterone system blocker; NSAIDs, nonsteroidal anti-inflammatory drugs.

^aMissing three participants because of inability to take height measurements.

Outcome Assessment

Class 1 and 2 safety events ascertained at annual in-center visits are shown in Table 2. The frequency of class 1 events was 108.7 events per 100 patient-visit versus 100.6 events per 100 patient-visit in pilot and observation groups, respectively, but was not significantly different except for the higher rate of glucose <70 mg/dl with symptoms in the pilot group. The frequency of class 2 safety events was 38.3 events per 100 patient-visits compared with 41.2 events per 100 patient-visit in the pilot versus observation group, and not statistically different. From 32% to 39% of participants reported one or more hospitalization in the 6 months before their first to fourth annual visit, with no statistical difference between groups (Supplemental Table 2).

Table 2. - Count (events) and rates (per 100 patient-visits) of both class 1 and class 2 safety events in pilot and observational groups in the Safe Kidney Care cohort

Event Type	Bracelet Group (n=108 Participants)				Observation Group (n=242 Participants)
	Counts	Patient-Visits	Rate (per 100 Patient-Visits)	Events	Patient-Visits	Rate (per 100 Patient-Visits)
Class 1 events
Hypoglycemia (<70 mg/dl) with symptoms	89	355	25.1	108	548	19.7
Hypoglycemia (<60 mg/dl) with or without symptoms	58	355	16.3	78	548	14.2
Hyperkalemia causing change in medication/diet	30	355	8.5	49	548	8.9
Dizziness	45	355	12.7	74	548	13.5
Falling	13	355	3.7	16	548	2.9
Bleeding	12	355	3.4	10	548	1.8
Facial, tongue, throat swelling	4	355	1.1	7	548	1.3
Confusion	27	355	7.6	29	548	5.3
Nausea/vomiting/diarrhea	40	355	11.3	56	548	10.2
Ankle swelling	25	355	7.0	47	548	8.6
Muscle weakness or cramps	31	355	8.7	56	548	10.2
Skin rash	12	355	3.4	21	548	3.8
All events	386	355	108.7	551	548	100.6
Class 2 events
Hyperglycemia	12	326	3.7	10	527	1.9
Hypoglycemia	14	326	4.3	19	527	3.6
Hyperkalemia	11	325	3.4	16	524	3.0
Hypokalemia	13	325	4.0	25	524	4.8
Anemia	34	326	10.4	61	525	11.6
Systolic hypotension	2	325	0.6	5	527	0.9
Orthostatic hypotension a	31	315	9.8	57	517	11.0
Low pulse	8	324	2.5	25	527	4.7
All events	125	326	38.3	218	529	41.2

Class 1 events were self-reported at each annual visit from preceding 6 months by pilot group (n=108) and observation group (n=242). Class 2 patient-visits were included for a specific event type where phlebotomy or physical measurements were available for that measure.

^aOrthostatic hypotension defined as a decrease in systolic BP from sitting to standing of 20 mm Hg or more.

Figure 1 displays the Kaplan–Meier curves for each study outcome, including (1) ESKD, (2) ESKD or 50% decline in GFR, (3) death, and (4) a composite of all three outcomes. Comparing incidence of ESKD in the subgroups reveals the pilot group participants had a significantly lower incidence of ESKD at any time after baseline versus the observational group (P=0.02). Kaplan–Meier curves overlapped for all other outcomes over the duration of study observation, with no overall statistical differences.

Figure 1.: Kaplan-Meier plots demonstrating the lower incidence of (A) ESKD with medical-alert accessory (solid line) versus observational (dashed line) subgroups, and no significant overall difference between the two subgroups for the outcomes of (B) 50% reduction of GFR or ESKD, (C) death, and (D) a composite of all outcomes. Log-rank statistic P value. Obs, observational.

Table 3 depicts the crude and adjusted relative hazard of each endpoint for pilot versus observational group assignment. The crude risk of ESKD was significantly lower for pilot group with the medical-alert accessory versus the observational group (hazard ratio, 0.42; 95% confidence interval [95% CI], 0.20 to 0.89; P=0.02) with crude rates of 2.2 (95% CI, 1.2 to 3.7) and 3.3 (95% CI, 2.3 to 4.7) per 1000 patient-months in the former versus latter, respectively. The risk estimate remained significant with adjustment for demographic factors, case-mix, baseline GFR, and other characteristics (model 4 hazard ratio, 0.38; 95% CI, 0.16 to 0.94; P=0.04).

Table 3. - Cox proportional hazards models expressing hazard ratio of medical-alert accessory use (pilot) versus usual care (observation) associated with outcomes ascertained in Safe Kidney Care cohort

Outcome	Study Size	Events	Crude (95% CI)	Model 1 (95% CI) a	Model 2 (95% CI) b	Model 3 (95% CI) c	Model 4 (95% CI) d
ESKD	350	41	0.42 (0.20 to 0.89) e	0.36 (0.17 to 0.77) e	0.36 (0.16 to 0.81) e	0.36 (0.16 to 0.82) e	0.38 (0.16 to 0.94) e
ESKD and GFR decline	350	52	0.66 (0.36 to 1.22)	0.58 (0.32 to 1.07)	0.61 (0.32 to 1.15)	0.63 (0.33 to 1.18)	0.63 (0.31 to 1.20)
Death	350	56	0.95 (0.53 to 1.69)	1.04 (0.58 to 1.88)	1.00 (0.55 to 1.82)	1.05 (0.57 to 1.93)	1.03 (0.53 to 1.99)
ESKD and death and GFR decline	350	106	0.79 (0.52 to 1.20)	0.77 (0.50 to 1.18)	0.79 (0.51 to 1.23)	0.79 (0.51 to 1.23)	0.76 (0.47 to 1.23)

95% CI, 95% confidence interval.

^aModel 1 adjusted for sex, race, and age.

^bModel 2 additionally adjusted for baseline body mass index, mean arterial pressure, and GFR.

^cModel 3 additionally adjusted for smoking history and comorbidities including diabetes, hypertension, cardiovascular disease, cancer.

^dModel 4 additionally adjusted for education, employment status, health literacy, and renin-angiotensin-aldosterone system blocker use.

^eModels were significant (P<0.05).

The adjusted risk of assignment to pilot versus observation groups was not significant for any of the other outcomes; however, because the survival curves for pilot versus observation group overlapped for the outcomes other than ESKD, we assessed the interaction with time for the hazard ratios of pilot versus observation group assignment for these three outcomes. Figure 2 and Supplemental Table 3 reveal that from 6 months (when first event was recorded) to 23 months, pilot group assignment was associated with lower risk of the composite outcome of ESKD and 50% reduction in GFR, but not thereafter, and the test for a time interaction was not significant (P=0.07). No such interval of reduced risk from use of the medical-alert accessory was observed for death or the composite of kidney outcomes and death.

Figure 2.: Cox proportional hazards model adjusted hazard ratio varies over time (6–48 months) with early lower risk associated with medical-alert accessory (pilot) versus observational subcohort noted for the composite outcome of 50% reduction in GFR or ESKD. Covariates include baseline age, sex, race, GFR, body mass index, mean arterial BP, diagnosis of diabetes, hypertension, cardiovascular disease, cancer, education, employment status, health literacy, and RAAS blockers.

Discussion

This study provides preliminary evidence of the utility of a medical-alert accessory in CKD. A majority of the pilot study participants reported using the accessory over the duration of the study. The risk benefit associated with use of the medical-alert accessory and the outcome of ESKD was significant when adjusting for multiple factors including baseline kidney function. The risk estimates reported for medical-alert accessory use and other outcomes were not constant over time, but the results suggested an early benefit associated with the accessory’s use and the composite of 50% reduction of GFR or ESKD. Notably, wearing a the medical-alert accessory was not associated with a reduction in safety events because those participants in the pilot group had no significant difference in overall rates of measured safety events when compared with those of the observational group participants.

To our knowledge, this is the first study examining the relationship of a medical-alert accessory use with outcomes in predialysis CKD. Medical-alert bracelets were first manufactured and distributed in 1956 by the MedicAlert Foundation, and Dr. and Mrs. Marion Collins, after their daughter survived an episode of anaphylactic shock from a dose of tetanus antitoxin (¹⁰,¹¹). Today, many vendors manufacture medical-alert accessories from a variety of materials (¹²). Although current literature includes recommendations for the provision of medical-alert accessory, in a range of conditions such as diabetes, preserving veins for future vascular access in CKD (¹³), bleeding disorders (¹⁴), or peanut allergy (¹⁵), there are no evidence-based guidelines for their use. Notably, some studies point to a lack of benefit of medical-alert accessories in specific patient groups. One trial demonstrated no benefit from medical-alert bracelets in preventing falls among patients undergoing inpatient physical rehabilitation (¹⁶). Another study of postoperative patients having difficulty with endotracheal intubation and who were sent letters advising them to obtain a medical-alert bracelet revealed a majority did not obtain the accessory (¹⁷).

CKD is a disease state without a prototypical sign or symptom profile alerting caregivers of this population’s special management needs (²,¹⁸,¹⁹). Moreover, laboratory evidence of kidney dysfunction, which is a hallmark of CKD, may be neglected or underappreciated (¹⁸,²⁰,²¹). Even recognition of CKD does not necessarily increase the likelihood of appropriate care (¹⁹). Examination of a regional health maintenance organization’s care of enrollees identified with CKD reveals under-referral to a nephrologist and infrequent use of RAAS blockers. However, more recent implementation of a point-of-care alert to primary care providers of patients with CKD has shown demonstrable improvement in BP control, use of RAAS blockers, monitoring of GFR, and checking microalbuminuria (²²).

Patients’ awareness of their CKD is also quite low (²³,²⁴), and only marginally improved when their disease has clinical manifestations (²⁵). Notably, the US Preventive Services Task Force has elected not to endorse CKD screening for asymptomatic adults (²⁶), and this recommendation has met with dissenting opinions from the kidney community (²⁷). The failure to systematically identify patients with CKD as they engage with the health care system is a lost opportunity to minimize safety threats. New strategies are needed to augment patients’ and providers’ discovery and awareness of CKD, and to guide them to best and safe practices. Use of a medical-alert accessory offers a low-technology, inexpensive, and portable means to broadcast the special needs of this disease population.

The findings of this study need to be interpreted with certain limitations in mind. The study was not designed as a randomized trial and the potential for a sampling bias cannot be dismissed. Nevertheless, the pilot sample was drawn from the same population as the observational study participants and the baseline characteristics were comparable between the two groups including severity of kidney disease at the time of enrollment, although baseline proteinuria was not available. Participants and study staff were not blinded to the provision of the medical-alert accessory, which could have biased both the reporting and ascertainment of events; however, this did not translate into a difference in reported safety events between groups. Of note, a portion of participants who were found to be eligible on screening had improved kidney function at baseline evaluation. This inclusion of participants with stage 2 CKD is likely to increase the generalizability of the study findings. The sample size was relatively limited hindering the assessment of time-varying exposures. Moreover, the findings may be subject to a type 1 error, given the number of hypothesis tests (multiple comparisons), and should be interpreted as such. Nevertheless, the risk estimates are substantial and relatively stable with multivariable adjustment. The comparable incidence of adverse safety events, as defined in the study, between the medical-accessory pilot and observational subcohorts detracts from the a priori hypothesis that increased CKD awareness with a medical-alert accessory improves the safety of care, and thereby improves outcomes. However, the association of the medical-alert accessory appears independent of its association with adverse safety and may relate to heretofore unknown effects of the accessory on practice patterns and lifestyle effects, which need to be explored further.

The findings of this study demonstrate a potential benefit associated with use of a largely overlooked, relatively ubiquitous, and simple health information technology. One might consider medical-alert accessories as a useful solution to the problem of under-recognized CKD, but it may also serve as a behavior prompt directing patients to best practices in self-care and disease management. The results of this exploratory study call for a randomized trial to fully evaluate the promise suggested by this early pilot study of a medical-alert accessory in CKD.

Disclosures

Dr. Diamantidis, Ms. Doerfler, Mr. Farhy, Dr. J. Fink, Ms. W. Fink, and Dr. Zhan have nothing to disclose.

Funding

The study in this paper was supported by the National Institute of Diabetes and Digestive and Kidney Disease (grant number R01 DK084017), University of Maryland, Baltimore, Institute for Clinical and Translational Research, and the University of Maryland, Baltimore, School of Medicine, Summer Program in Obesity, Diabetes, and Nutrition Research Training (SPORT) (grant number T35-DK095737).