Pragmatic Randomized, Controlled Trial of Patient Navigators and Enhanced Personal Health Records in CKD

Abstract

Background and objectives 

Patient navigators and enhanced personal health records improve the quality of health care delivered in other disease states. We aimed to develop a navigator program for patients with CKD and an electronic health record–based enhanced personal health record to disseminate CKD stage–specific goals of care and education. We also conducted a pragmatic randomized clinical trial to compare the effect of a navigator program for patients with CKD with enhanced personal health record and compare their combination compared with usual care among patients with CKD stage 3b/4.

Design, setting, participants, & measurements 

Two hundred and nine patients from six outpatient clinics (in both primary care and nephrology settings) were randomized in a 2×2 factorial design into four-study groups: (1) enhanced personal health record only, (2) patient navigator only, (3) both, and (4) usual care (control) group. Primary outcome measure was the change in eGFR over a 2-year follow-up period. Secondary outcome measures included acquisition of appropriate CKD-related laboratory measures, specialty referrals, and hospitalization rates.

Results 

Median age of the study population was 68 years old, and 75% were white. At study entry, 54% of patients were followed by nephrologists, and 88% were on renin-angiotensin system blockers. After a 2-year follow-up, rate of decline in eGFR was similar across the four groups (P=0.19). Measurements of CKD-related laboratory parameters were not significantly different among the groups. Furthermore, referral for dialysis education and vascular access placement, emergency room visits, and hospitalization rates were not statistically significant different between the groups.

Conclusions 

We successfully developed a patient navigator program and an enhanced personal health record for the CKD population. However, there were no differences in eGFR decline and other outcomes among the study groups. Larger and long-term studies along with cost-effectiveness analyses are needed to evaluate the role of patient navigators and patient education through an enhanced personal health record in those with CKD.

Introduction

CKD burden continues to remain high, and health care spending for the CKD population is rising (¹,²). Innovative methods to improve the quality of care delivered to those with CKD are needed. Particularly, engaging and educating patients with CKD about their underlying disease could help them become a central player in the care of a disease that often warrants multidisciplinary care (³), which eventually could lead to an empowered patient and delivery of patient-centered care. With the increasing use of electronic health records and personal health records (PHRs), empowering patients with pertinent health-related information (i.e., providing disease-specific information) could improve self-management, promote discussions with their health care team that result in better adherence to recommended care, and ultimately, improve clinical outcomes (⁴,⁵). Patient navigator is an attractive alternative that has been used with improved patient outcomes among patients with cancer (particularly in vulnerable populations), and clinical trial data suggest their potential use to improve outcomes among patients on the transplant waitlist (⁶,⁷). These interventions could provide benefits by improving patient adherence to medications, which could help slow the progression of kidney disease through better BP control. Furthermore, through patient education, rates of follow-up with health care providers could be improved, which in turn, could also help physicians obtain blood tests at appropriate times to institute timely interventions (e.g., monitoring urine albumin-to-creatinine ratio and titrating renin-angiotensin system inhibitors).

In an era of limited resources for clinical trials and lack of an adequate number of clinical trials, larger pragmatic trials are encouraged to test various interventions in a cost-effective manner (^8–10). We hypothesized that patient navigators and enhanced personal health records (E-PHRs) could improve the quality of care provided to patients with CKD, thereby affecting their kidney and overall health. To address these hypotheses, we (1) developed a navigator program for patients with CKD adapting the use of patient navigators with the chronic care model and (2) developed an E-PHR that will use electronic communication to disseminate CKD stage–specific goals of care and CKD education modeling from the well established chronic care model to improve outcomes for patients with CKD stage 3b/4. We then conducted a randomized clinical trial using a factorial design to investigate the clinical outcomes of two interventions, the patient with CKD navigator and E-PHR, and their combination compared with usual care for patients with CKD stage 3b/4. In a previous publication (¹¹), we described the development of the patient navigator program. Herein, we describe the development of E-PHR and report the results of the randomized, controlled trial that compared these novel educational interventions among an outpatient CKD population.

Materials and Methods

Study Overview

This is a randomized clinical trial in which patients with CKD were randomized to the patient with CKD navigator, E-PHR, both interventions, or usual care and followed for up to 2 years (clinicaltrials.gov identifier: NCT01792661).

Inclusion/Exclusion Criteria

Participants were identified from a preexisting electronic health records–based CKD registry at the Cleveland Clinic (¹²). All English-speaking adults ages 18–80 years old with an eGFR=15–45 ml/min per 1.73 m² residing in northeast Ohio who were receiving care at the Cleveland Clinic Health System (July of 2012 to December of 2013) were eligible for inclusion. Patients with cancer or other terminal illness and who were on dialysis or received kidney transplant in the past were excluded. Figure 1 outlines the recruitment details of this clinical trial.

Figure 1.: Flow diagram shows recruitment process (screening, randomization, and follow-up of study participants).

Randomization

Patients who met the eligibility criteria were randomly assigned to one of the four groups using a computer-generated randomization scheme that was stratified by family health center, because Cleveland Clinic has multiple family health centers across the greater Cleveland area. Randomization allocation was concealed. For this study, participants were recruited from six Cleveland Clinic family health centers. Participants and the study personnel (study coordinator and the navigators) were aware of their assignment, but the outcome assessors were not aware of the study assignments.

Study Procedures

Usual Care Group (Standard of Care).

Patients who were assigned to the usual care group were advised to use their PHR (MyChart account via EPIC [Madison, WI]) accounts to aid in the management of their health. No specific changes to their PHR accounts were made. All patients who use the PHR can review and schedule appointments, request prescription renewals, view health summaries, access a current list of medications, review test results, and send a secure message to their physicians or health care team. Moreover, patients also receive automated important health reminders on the basis of sex- and age-based health maintenance schedules as well as chronic disease–related reminders. Links within the PHR allow patients to access reliable health information about a broad range of topics of personal interest through a third-party vendor (MedlinePlus).

Patient Navigators.

We have previously described the development of our patient navigator program (¹¹). Briefly, we followed the key steps outlined in developing a patient navigator program in the oncology field. However, important adaptations for CKD (due to its chronic nature) were made during the development of the navigator program for patients with CKD. We recruited two nonmedical or lay people (paid positions) with college degrees who underwent a training program with a special focus on general patient navigator training, CKD education, and electronic health record training. We identified barriers that they might address during their in-person meetings with patients with CKD. The 11 predefined barriers included assessing compliance, addressing insurance needs, providing disease-specific information, and helping with transportation issues, etc. (¹¹). Patients with CKD navigator patient visits were scheduled to occur every month or every quarter during patients’ scheduled office visits or other health care visits. Navigators also had follow-up phone calls with the patients depending on their needs after their in-person meetings.

Navigators’ meetings with patients depended on their concerns and how well they understood the responses. On average, they met with the patients every 2–4 weeks. Interventions also varied with the patient’s need. Both navigators addressed various patient concerns ranging from attending physician appointments to addressing other life stressors. They educated them about CKD, accompanied them to nephrologist appointments, and helped them to make provider appointments. For instance, one of our navigators accompanied a patient to his/her dialysis training, because he/she was recently divorced. For another patient, the patient navigator helped to get a patients’ spouse qualified for a certain Medicaid program that paid for adult day care several days a week.

E-PHR.

The E-PHR functionality was developed with the assistance of Cleveland Clinic’s Information Technology Division MyChart team to securely review CKD education materials. These features were in addition to the existing features available to all PHR users. CKD alert appeared only once, and when the patient clicked on the alert, it led them to the page that provided details for CKD. After the patient clicks on the alert, it closes. Supplemental Figures 1 and 2 display the various information that was incorporated into the PHR for those with stages 3b and 4, respectively. These educational materials were adapted from the educational materials developed by the National Kidney Disease Education Program, the National Kidney Foundation, and other educational organizations.

Data Collection

Most of the study data were obtained electronically from the electronic health record–derived CKD registry. We obtained baseline demographic characteristics (age, sex, race, insurance, and education), comorbidities, medication use, and smoking status during the baseline visit. Furthermore, CKD stage–specific laboratory data (i.e., whether a laboratory parameter was measured or not) were assessed 24 months before enrollment and for up to 24 months after study entry. Health literacy was assessed in person using the Short Test of Functional Health Literacy in Adults during the baseline visit and at the end of the study via phone call. Details about nephrology consultation, vascular access, dialysis and transplant referrals, hospitalizations, and emergency room visits were obtained through chart review by two members of the research team (S.E.J. and G.N.) after completion of the study. We conducted a phone exit interview survey at the end of the planned 2-year follow-up, and results are presented in Supplemental Material.

Study Outcome Measures

The primary outcome of the study was the change in eGFR over the 2-year study period. Secondary measures included (1) acquisition of appropriate laboratory measures (hemoglobin [Hb], phosphorus, 25-hydroxy vitamin D, parathyroid hormone, LDL cholesterol, HbA1c, and urine albumin-to-creatinine ratio); (2) prescription of renoprotective medications, such as angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs); (3) referral rates to nephrologists and vascular surgeons and for kidney transplantation assessment; (4) achieving BP control (<130/80 mmHg); (5) number of hospitalizations and emergency room visits; and (6) death.

Statistical Analyses

Sample Size Calculation.

For the purposes of this study, we assumed an equal effect size associated with each intervention relative to the usual care group that had access to the personalized health record only (control group). On the basis of an estimated decline in eGFR of approximately 3 ml/min per 1.73 m² per year, we proposed that each intervention would produce an attenuation of this decline by one half or 1.5 ml/min per 1.73 m² per year over the 2-year study period. We did not power the study specifically to estimate the interaction of the two interventions.

Analyses.

We tabulated baseline characteristics across groups. The primary hypotheses of the study were tested using repeated measures mixed models with subjects considered a random effect. We tested the interaction of follow-up time and study group to evaluate the association of change in kidney function over time with randomized group. We iteratively tested the covariance structure for mixed models and selected our final model on the basis of best fit as determined by the Akaike Information Criteria. An autogressive covariance structure was used for the final model. Secondary outcomes include both proportions and count data. These were tested between study arms using chi-squared tests, Fisher exact tests, and negative binomial models, respectively, with the study arm as the sole independent predictor variable. Because the study arms were randomized, we did not include any risk adjustment in the models but report a detailed table of patient demographic and comorbid conditions for each of the study groups and the full study cohort. We carried out a prespecified subgroup analysis to assess any potential differences in outcomes measures stratified by racial groups. Hypotheses were tested with two-sided type 1 error probabilities of 0.05. All analyses were carried out in SAS v.9.4 (SAS Institute, Cary, NC). The Institutional Review Board at the Cleveland Clinic approved this study, and informed consent was obtained from the study participants.

Results

Recruitment and Baseline Characteristics

Of 485 eligible participants, 276 declined to participate, and 209 patients were randomized into one of four groups (Figure 1). Table 1 shows baseline characteristics of the study participants. Median age of the study population was 68 years old, and 75% were white. At study entry, 54% of patients were followed by nephrologists, and 88% were on ACEI/ARBs.

Table 1. - Baseline characteristics of the trial participants

Variables	N	Usual Care, n=57	Patient Navigator, n=53	Enhanced Personal Health Record, n=50	Both, n=49
Age, yr, median (P25, P75)	209	68 (64, 72)	71 (64, 75)	67 (61, 72)	69 (62, 73)
Sex, N (%)	209
Women		39 (68)	33 (62)	25 (50)	21 (43)
Race, N (%)	209
White		40 (70)	46 (87)	35 (70)	36 (74)
Black		16 (28)	5 (9)	15 (30)	10 (20)
Multiracial		0 (0)	1 (2)	0 (0)	1 (2)
Asian/Pacific islander		1 (2)	1(2)	0 (0)	2 (4)
Insurance, N (%)	207
Medicaid		1 (2)	0 (0)	0 (0)	1 (2)
Medicare		45 (80)	37 (70)	28 (57)	33(67)
Private		10 (18)	16 (30)	21 (43)	15 (31)
Education, N (%)	209
Less than grade school		1 (2)	0 (0)	0(0)	0 (0)
Grade school		3 (5)	1 (2)	1 (2)	1 (2)
High school		25 (44)	21 (40)	29 (58)	17 (35)
College degree		16 (28)	19 (36)	17 (34)	18 (37)
Master’s degree		7 (12)	7 (13)	2 (4)	11 (22)
Doctorate/PhD		5 (9)	4 (8)	1 (2)	2 (4)
Unknown		0 (0)	1 (2)	0 (0)	0 (0)
Income range, N (%)	209
<$15,000		9 (16)	3 (6)	8 (16)	4 (8)
$16,000–$30,000		11 (19)	7 (13)	9 (18)	12 (25)
$31,000–$50,000		13 (23)	11 (21)	8 (16)	10 (20)
$51,000–$75,000		5 (9)	11 (21)	8(16)	9(18)
$76,000–$100,000		5 (9)	11 (21)	6 (12)	6 (12)
>$100,000		8 (14)	7 (13)	7 (14)	5 (10)
Data not available		6 (11)	3 (6)	4 (8)	3 (6)
Literacy score (S-TOHFLA), N (%)	209
Inadequate (0–16)		3 (5)	0 (0)	1 (2)	1 (2)
Marginal (17–22)		0 (0)	1 (2)	0 (0)	1 (2)
Adequate (23–36)		54 (95)	52 (98)	49 (98)	47 (96)
Access to a computer, N (%)	209	52 (91)	52 (98)	42 (84)	44 (90)
Access to MyChart, N (%)	209	40 (70)	41 (77)	34 (68)	37 (75)
eGFR, ml/min per 1.73 m2, mean±SD	209	35.7±6.3	34.0±6.7	34.0±6.2	35.9±7.1
CKD stage, N (%)	209
Stage 3a		1 (2)	0 (0)	0 (0)	0 (0)
Stage 3b		46 (81)	36 (68)	38 (76)	39 (80)
Stage 4		10 (18)	17 (32)	12 (24)	10 (20)
Proteinuria, N (%)	100	9 (33)	8 (38)	11 (44)	16 (59)
Systolic BP, mmHg, mean±SD	202	131±21	133±16	132±18	128±19
Diastolic BP, mmHg, mean±SD	202	73±14	75±10	72±11	72±9
BP<140/90 mm Hg, N (%)	202	41 (72)	32 (63)	34 (72)	34 (72)
Diabetes, N (%)	209	29 (51)	24 (45)	24 (48)	30 (61)
Coronary artery disease, N (%)	209	21 (37)	16 (30)	11 (22)	12 (25)
Congestive heart failure, N (%)	209	12 (21)	3 (6)	3 (6)	5 (10)
Malignancy, N (%)	209	8 (14)	9 (17)	5 (10)	20 (41)
Smoking (current), N (%)	209	1 (2)	3 (6)	4 (8)	1 (2)
Body mass index, kg/m2, mean±SD	208	32.1±6.4	32.7±8.2	33.9±9.0	32.4±6.1
Statins use, N (%)	209	44 (77)	46 (87)	38 (76)	37 (76)
Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers use, N (%)	209	53 (93)	47 (89)	43 (86)	41 (84)
Seen by a nephrologist before study entry, N (%)	209	25 (44)	29 (55)	30 (60)	26 (53)
Proteinuria among those seen by a nephrologist, N (%)	63	7 (47)	7 (50)	9 (56)	11 (61)
Proteinuria among those not seen a nephrologist, N (%)	37	2 (17)	1 (14)	2 (22)	5 (56)

N, number; P, percentile; S-TOHFLA, Short Test of Functional Health Literacy in Adults.

Utilization of Study Tools

Table 2 outlines the utilization of study tools during the study period. Median number of meetings between the navigator and study participant was seven (range =1–26). We have described the barriers addressed by the navigators in Supplemental Material and Supplemental Table 1. During the study period, median number of days that patients logged into their PHR account was 45 in the usual care group versus 70 in the patient navigator group versus 34 in the E-PHR group. Furthermore, the number of clicks in the PHR account was lower for the E-PHR group versus the personal navigator/E-PHR group.

Table 2. - Number of navigator visits and details of personal health record use by participants

Encounter Details Median (P25, P75)	Usual Care, n=57	Patient Navigator, n=53	Enhanced Personal Health Record, n=50	Both, n=49	P Value a
Navigator visits		7 (3, 12)		6 (2, 14)	0.52
No. of days logged into personal health records	45 (10, 82)	70 (30, 87)	34 (2, 85)	47 (19, 84)	0.10
No. of times logged into personal health record, N (%)					0.04 b
≤12	17 (30)	5 (9)	16 (32)	7 (14)
13–24	5 (9)	6 (11)	6 (12)	10 (20)
25–48	9 (16)	5 (9)	5 (10)	8 (16)
>49	26 (46)	37 (70)	23 (46)	24 (49)
No. of clicks into enhanced personal health record			10 (1, 20)	31 (18, 63)	<0.001
No. of clicks into personal health record activity	269 (113, 454)	427(283, 633)	253 (46, 488)	367 (145, 529)	0.08

P, percentile.

^aNegative binomial test for overall four-group comparison except for number of navigator visits and number of clicks into enhanced MyChart, which are only applicable to two groups.

^bChi-squared test.

Study Outcomes

eGFR Decline.

Figure 2 shows the decline in eGFR during the study period. No significant differences in the slope of eGFR were noted among the groups (P=0.19). Compared with the usual care group, the differences in eGFR (milliliters per minute per 1.73 m²) slope (annualized) during the 2-year study period were: navigator only group: 0.4 (95% confidence interval [95% CI], −2.2 to 3.1); E-PHR: −0.3 (95% CI, −2.7 to 2.1); and combined: −0.3 (95% CI, −2.7 to 2.1). In further analysis, the following factors were associated with a lower eGFR: men, blacks, and diabetes (P<0.001).

Figure 2.: Slope of decline in eGFR among the four study groups during follow-up.

Laboratory Measures, Referrals, and Hospitalizations.

Rates of relevant laboratory measures performed in the subgroups were similar, except for serum creatinine, Hb, and HbA1c, for which the usual care group had higher rates (Table 3). During the 24-month follow-up, there were no differences between the groups for the following outcomes (Table 4): proportion of patients who reached the next stage of CKD, referred to specialists (nephrology, vascular surgery, and kidney transplantation), received dialysis education, hospitalizations or emergency department visits, underwent kidney transplantation, or died.

Table 3. - Laboratory measurements 24 months before and after follow-up among the study groups

Variables	N	Usual Care, n=57	Patient Navigator, n=53	Enhanced Personal Health Record, n=50	Both, n=49	P Value a
Serum creatinine, N (%)
24 mo before	209	56 (98)	51 (96)	48 (96)	48 (98)
24 mo after	209	57 (100)	52 (98)	42 (84)	48 (98)	<0.001 b , c
Hemoglobin, N (%)
24 mo before	209	52 (91)	47 (89)	43 (86)	42 (86)
24 mo after	209	54 (95)	51 (96)	41 (82)	48 (98)	0.02 b , c
Serum phosphorus, N (%)
24 mo before	209	20 (35)	22 (42)	20 (40)	20 (41)
24 mo after	209	39 (68)	35 (66)	28 (56)	29 (59)	0.52
25-Hydroxy vitamin D, N (%)
24 mo before	209	35 (61)	27 (51)	30 (60)	33 (67)
24 mo after	209	37 (65)	38 (72)	28 (56)	28 (57)	0.31
Parathyroid hormone, N (%)
24 mo before	209	21 (37)	17 (32)	19 (38)	21 (43)
24 mo after	209	33 (58)	32 (60)	22 (44)	25 (51)	0.34
LDL cholesterol, N (%)
24 mo before	209	48 (84)	49 (93)	38 (76)	36 (74)
24 mo after	209	48 (84)	48 (91)	39 (78)	40 (82)	0.36
HbA1c, N (%)
24 mo before	107 d	29 (100)	23 (96)	21 (88)	28 (93)
24 mo after	107 d	29 (100)	23 (96)	19 (79)	25 (83)	0.02 b , e
Urine albumin-to-creatinine ratio, N (%)
24 mo before	209	19 (33)	18 (34)	23 (46)	25 (51)
24 mo after	209	25 (44)	19 (36)	19 (38)	28 (57)	0.13
BP<130/80 mmHg during follow-up, N (%)	208	53 (93)	48 (91)	45 (92)	45 (92)	0.98 b

HbA1c, hemoglobin A1c.

^aChi-squared test for overall four-group comparison unless otherwise noted.

^bFisher exact test for overall four-group comparison.

^cAll other groups significantly different from enhanced personal health record group (P<0.05).

^dApplies only to patients with diabetes.

^eUsual care significantly different from enhanced personal health record group (P<0.05).

Table 4. - Outcomes during study follow-up (after 24 months)

Variables	Usual Care, n=57	Patient Navigator, n=53	Enhanced Personal Health Record, n=50	Both, n=49	P Value
Reached next CKD stage, started dialysis, or got transplant	23 (40)	18 (34)	20 (40)	20 (41)	0.87 a
Nephrology referral b	12 (38)	6 (25)	4 (20)	7 (30)	0.55 a
Dialysis education and/or vascular surgery and/or transplant referral	5 (9)	7 (13)	8 (16)	8 (16)	0.59; c 0.27 d
No. of hospitalizations and/or emergency department visits	1 (0, 9)	0 (0, 6)	1 (0, 14)	1 (0, 7)	0.24; e 0.17 f
Dialysis initiation or transplant	1 (1.8)	1 (1.9)	4 (8.0)	1 (2.0)	0.36 g
Death	4 (7.0)	1 (1.9)	2 (4.0)	3 (6.1)	0.61 g

^aChi-squared test for overall four-group comparison.

^bApplies only to 99 patients with no prior nephrology visit.

^cChi-squared test comparing navigator group with no navigator group.

^dChi-squared test comparing enhanced personal health record group with no enhanced personal health record group.

^eNegative binomial model comparing navigator group with no navigator group.

^fNegative binomial model comparing enhanced personal health record group with no enhanced personal health record group.

^gFisher exact test for overall four-group comparison.

Discussion

We successfully developed a patient navigator program and an E-PHR to address barriers and provide education for patients with stage 3b/4 CKD. In this randomized clinical trial, we compared patients using E-PHR, patient navigator, or their combination with those who used PHRs alone. Overall, acceptance of a patient navigator to assist them with their care among patients with CKD was high. After 24 months of follow-up, the rate of eGFR decline was similar among the study groups. Furthermore, measurement of CKD-related laboratory parameters did not differ significantly, and there were no differences in referral for dialysis education and vascular access placement, emergency department visits, and hospitalization rates among the four study groups.

Pragmatic clinical trials adapting T2 translational research (moving findings from clinic-based interventions to community-based interventions) have shown success in other disciplines (¹³,¹⁴). Except for a few recently reported studies, such pragmatic trials are sparse in CKD (^15–19). Ishani et al. (¹⁹) examined the effect of delivering telehealth by an interdisciplinary team in the Veterans Affairs health care system in the Minneapolis area. Although they showed feasibility of this intervention, no significant differences in the composite end point of death, hospitalization, emergency department visits, or admission to skilled nursing facilities were noted between the intervention and control groups. Similarly, Tuttle and colleagues (¹⁵) developed a pharmacist-led, home-based medication management intervention to prevent readmission among those with CKD stages 3–5. Using the T2 translational research model, we successfully used some of the key elements, such as health information technology and a multidisciplinary team to address CKD (²⁰). The lack of benefit noted with our primary end point (eGFR decline) could be attributed to the following reasons: (1) slower than anticipated decline in eGFR during the 2-year follow-up across all groups as these patients were followed in a health care system and (2) a higher quality of care in all groups and in particular, the usual care or control group, even at the onset of the trial. These could have led to our study being underpowered to detect significant changes in kidney function decline. In addition, lack of benefits with respect to referral (for kidney transplantation) and hospitalization could be attributed to the small number of patients with advanced CKD in this study.

Despite the lack of significant results, several findings are worth noting and would inform the scientific community in designing larger pragmatic trials, which are in alignment with the objective of funding agencies to support studies like this. We hypothesized that the use of E-PHR and patient navigators would contribute to improvement in knowledge about kidney disease, thereby empowering them to assume self-care. Navigators not only educated patients about their disease and assisted them with obtaining appointments, etc. (measurable benefits) but also, helped coordinate various aspects of their care, including securing several daily livelihood needs (unmeasured benefits). For instance, one of the major barriers that was addressed included transportation needs, which not only help them with physician visits but also, indirectly provide benefits that could affect their health (access to grocery stores and laboratory visits). Such indirect benefits obtained with the interventions should be taken into consideration while designing studies.

Furthermore, it is important to note that the mean age of the study population (including those enrolled in the E-PHR group) was >65 years old and that adoption of the E-PHR by this group is encouraging. This is evident by the fact that they could navigate the E-PHR as shown by the personalized health record click data (Table 2). Cumulatively, these data suggest that older adults could use PHRs and also support testing interventions of PHRs as a tool to improve care in a geriatric population. In the E-PHR group, we tailored information on the basis of the stage of kidney disease but not on the basis of their observed parameters (such as anemia and uncontrolled hypertension) during the study. Such real-time clinical decision support, which provides personalized medical information, would be the ideal intervention and could deliver promising results. Electronic health record technology is evolving rapidly, and the frameworks for implementing such personalized recommendations are being developed (²¹).

Although we examined the effect of these interventions on both laboratory and other patient end points, whether they are cost effective is unclear. Patient navigators’ salaries would vary between centers and cities and had to be taken into consideration during development of a navigator program. It is important to note that navigators not only identify barriers for patients but also, help them receive appropriate care in a timely manner, which could prevent unwanted emergency room visits and hospitalizations. Cost-effectiveness analyses data from the University of Alabama Cancer community network reported that the patient navigator program for those with cancer saved approximately $19 million/yr in patient-related expenditures (²²,²³). Patient navigators have become an integral part of cancer centers across the country. Whether such cost-savings could be translated to those with CKD merits future studies, because this would determine acceptance of patient navigators in the CKD clinics. We used the categories recommended by Freund et al. (²⁴) to document various barriers addressed by the navigators. These were originally developed to care for those with cancer. Although CKD and cancer are both chronic disease states, some unique aspects of CKD should be taken into account (such as coordinating multidisciplinary care among endocrinologists and dietitians) while considering barriers to be addressed. It is unclear if some critical barriers that were not addressed by the navigators during their interactions could have influenced the results of our study.

Although we conducted a randomized, controlled trial, our study is subject to limitations. Our patients were followed in a single large health care system, had higher literacy scores, and were engaged in their health care. We were not able to obtain health care received outside the Cleveland Clinic health system. It is important to note that nearly all (93%) of the usual care group were on an ACEI or ARB and that almost one half (44%) of them had seen a nephrologist at the onset of the study. Thus, whether similar findings would be evident among a community-based population with less interaction with the health care system or a more vulnerable population is unclear. This is vital, because an electronic health record portal adoption has been reported to be low among a vulnerable population with kidney disease (²⁵). Also, we did not examine whether more patients achieved recommended laboratory and BP targets in the intervention groups than the control group. We did not have predefined visits other than the requirement of those placed in one of the two groups with a patient navigator to have one face to face meeting or an in-person exit interview. Longer follow-up may be needed, which was seen in other educational or health care system redesigning, where it took 5–10 years to see an improvement in the intervention groups (²⁶,²⁷). Furthermore, it is unclear if the interventions led to unintended effects or effects that might affect the outcomes in an opposite direction that could have led to the null result.

In summary, we showed that the development of the patient navigator program and E-PHR portal for those with CKD is feasible. However, in a clinical trial, the rate of decline in eGFR and the processes of care delivered during the 2-year study period did not differ between the study groups. Future studies should study the feasibility of these interventions in other settings, such as those serving underserved populations. More importantly, whether implementation of these interventions would translate into long-term health benefits and is cost effective warrant additional investigations before it can be adopted in clinical practice.

Disclosures

None.