| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
SPECIAL SECTION |
1 Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland.
2 Johns Hopkins University School of Medicine, Baltimore, Maryland.
3 Kaiser-Permanente Mid-Atlantic, Rockville, Maryland.
Address correspondence to Chad Boult, MD, MPH, MBA, 624 North Broadway, Room 693, Baltimore, MD 21205. E-mail: cboult{at}jhsph.edu
 |
A |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. We hypothesized that GC would improve the quality of health care for this population. In 2006, we began a cluster-randomized controlled trial of GC at eight practices (n = 49 physicians). Older patients of these practices were eligible to participate if they were at risk for using health services heavily during the coming year. Teams of two to five physicians and their at-risk older patients were randomized to either GC or usual care (UC). Six months after baseline, participants rated the quality of their health care by answering validated closed-ended questions from telephone interviewers who were masked to group assignment.
Results. Of the 13,534 older patients screened, 2391 (17.7%) were eligible to participate in the study, of which 904 (37.8%) gave informed consent and were cluster-randomized. After 6 months, 93.8% and 93.2% of the GC and UC participants who remained alive and eligible completed telephone interviews. GC participants were more likely than UC participants to rate their care highly (adjusted odds ratio = 2.0, 95% confidence interval, 1.2-3.4, p =.006), and primary care physicians were more likely to be satisfied with their interactions with chronically ill older patients and their families (p <.05).
Conclusions. GC improves important aspects of the quality of health care for multimorbid older persons. Additional data will become available as this trial continues.
Key Words: Multimorbid • Care management • Randomized trial • Self-management • Quality of care
In accordance with the Chronic Care Model, we designed "Guided Care" (GC) to address these deficiencies in the quality of care by enhancing primary care. GC is primary health care infused with the operative principles of several successful innovations to improve outcomes for patients with chronic conditions and complex health care needs. A registered nurse who has completed a supplemental educational curriculum works in a practice with several PCPs to provide comprehensive chronic care to 50–60 multimorbid patients. Patients 65 years old or older are eligible to receive GC if they are in the upper quartile of risk for using health services heavily during the coming year, according to their scores on the hierarchical condition category (HCC) predictive model, which is based on diagnoses on health insurance claims submitted during the previous year (5). Using a web-accessible electronic health record (EHR), the Guided Care nurse (GCN) collaborates with the patient's PCP in performing eight clinical processes: assessing the patient and primary caregiver at home, creating an evidence-based care plan, promoting patient self-management, monitoring the patient's conditions monthly, coaching the patient to practice healthy behaviors, coordinating the patient's transitions between sites and providers of care, educating and supporting the caregiver, and facilitating access to community resources. A more detailed description of GC has been published previously (6).
In 2003–2004, we conducted a pilot test of a partial version of GC in a private practice in urban Baltimore, Maryland. This small study (n = 150) suggested that implementing GC is feasible and associated with improved quality of care (7,8) and greater efficiency in the use of health care resources (9). These encouraging results supported the investment of substantial resources in a larger, longer study of GC.
|
METHODS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
To give the study an 85% probability of detecting a 3-point difference between the groups' mean SF-36 Physical Component and Mental Component Subscale scores (10) with a <5% probability of a Type I error, a final analytic sample size of 580 participants was required.
Recruitment
We offered participation to primary care practices in three health care delivery systems in the Baltimore–Washington DC area: Johns Hopkins Community Physicians, Kaiser-Permanente, and MedStar. Practices were eligible if they cared for at least 400 patients who were 65 years old or older and could provide an on-site office for a GCN for 27 months. Eight practices were eligible, and all chose to participate. Individual PCPs were eligible if they worked at least 70% time at these practices. All eligible physicians at the eight practices were briefed on the requirements of the study; all gave written informed consent to participate (n = 49).
To recruit nurses for the GCN role, we placed advertisements in local newspapers, human resources Web sites of participating delivery systems, and a regional nursing journal. To be eligible, applicants had to be licensed registered nurses with at least 3 years of practice experience. Applicants with interest and/or experience in gerontologic nursing, enthusiasm for patient counseling, and comfort with electronic information technology and interdisciplinary practice were given preference in the hiring process. The selected nurses gave written informed consent to serve as research participants as well as providers of health care.
Recruitment of patient participants was a multistage process. Patients of the participating physicians were eligible for initial screening if they were 65 years old or older and covered by fee-for-service Medicare Parts A and B, Kaiser-Permanente Medicare Advantage, or TriCare (a Medicare-like insurance plan for military retirees). Patients' insurance claims for health care during the previous 12 months were analyzed by their insurers using the HCC model. Patients with HCC scores of 1.2 or higher or in the top 25% of the population (indicating a high risk of using health care services heavily during the following year) entered the next phase of recruitment. Beginning in December 2005, high-risk persons received introductory letters advising them that they might be eligible for the study and offering them the opportunity to "opt out" by returning a form in a pre-addressed, stamped envelope. Those who did not "opt out" received a telephone call in which a professional interviewer described the study, answered questions, and offered an in-home meeting to provide additional information. Professional interviewers then visited the homes of those who accepted to screen potential participants for eligibility, describe the study in detail, answer questions, offer participation, obtain written informed consent, and complete the baseline interview. Interviewees were ineligible if they did not have a telephone, did not speak English, were planning extended travel during the following 2.5 years, or failed a brief cognitive screen and did not have a proxy. Proxies were accepted if they were legal guardians or close family relatives. The study was approved by the Institutional Review Boards of Johns Hopkins Bloomberg School of Public Health, Kaiser-Permanente Mid-Atlantic, and MedStar.
Randomization
The unit of randomization was the "pod," that is, a colocated team of two to five PCPs and their consenting high-HCC older patients. Six practice sites housed two pods each; two practices with similar panel sizes, geographic location, and ownership housed one pod each. Fourteen pods, comprising 49 physicians and 904 consented patients, were assigned by a random allocation program operated by the study's statistician (DS) to either GC or "usual care" (UC) for 2 years. Within the six practice sites that housed two pods each, one pod was randomly assigned to GC, one to UC. Of the two practices that housed one pod each, one was randomly assigned to GC, the other to UC. For analysis purposes, these latter pods were considered as one "site." The study's coordinating center revealed the group assignments of the 14 pods (GC or UC) to the patients, nurses, and physicians after they had provided written informed consent to participate in the study.
Intervention
Before joining their pods, the GCNs completed an educational program designed to prepare them to provide all the services included in GC. Topics included comprehensive assessment, evidence-based guidelines for chronic conditions, motivational interviewing for health behavior change, collaborative care, chronic disease self-management, elder abuse, cultural competence, community resources, educating and supporting family caregivers, and using the GC EHR. The curriculum included case-based, interactive seminars and workshops, supplemented by readings and brief recorded lectures. Successful completion of the program required demonstration of GC competencies during a practicum with simulated patients.
Beginning in May 2006, the GCNs were introduced to their physician collaborators and integrated into their primary care practice sites through a process of orientation, observation, and interaction with the staff. During each week of this phase, the GCNs received the names and contact information of additional consented patients. Each GCN scheduled and conducted in-home assessments of, on average, two patients per week until a case load of 50–60 patients was developed. Following each assessment, the GCN began providing the assessed patient with the other services described previously (6).
Measures
The baseline interviews collected information about participants' sociodemographic characteristics and health insurance. The baseline and follow-up interviews inquired about patients' health and functional status, quality of health care, and satisfaction with health care. All baseline interviews were conducted face-to-face, and all follow-up interviews (6 months after each patient's start date) were conducted by telephone by rigorously trained, closely supervised professional interviewers who were masked to group assignment, used computer-assisted interviewing technology, and underwent 10% reliability testing.
The Patient Assessment of Chronic Illness Care (PACIC) is a validated measure of patients' perceptions of various qualities of their chronic care (11). It consists of 20 Likert-scaled items representing care processes (e.g., being asked about one's health habits, being given a list of things to do to improve one's health, receiving a copy of one's treatment plan) that comprise five scales (goal setting, coordination of care, decision support, problem solving, and patient activation). Respondents indicate how often, during the past 6 months, these processes occurred. We coded scores to indicate that care was either "higher quality" (the process occurred "most of the time" or "always") or "lower quality" (the process occurred "none of the time," "a little," or "some of the time").
All PCPs participating in the study were surveyed anonymously at baseline and approximately 1 year later. The surveys included questions about:
After 1 year in practice, all seven GCNs completed anonymously a validated nurses' job satisfaction instrument (13).
Analysis
We used chained equations (14) to impute values for patient item responses missing at baseline. Five imputed data sets were created, and estimates, confidence intervals, and p values were computed using Rubin's combining rules (15). We computed all scale scores as recommended by the originators of the scales and analyzed all data according to the "intention-to-treat" principle. To the extent possible, we used site-stratified testing procedures to evaluate differences between GC and UC on key baseline factors. For very rare or very common baseline factors, unstratified testing procedures were used. To estimate the effect of the intervention on the quality of care (i.e., on PACIC scores), we constructed multivariate logistic regression models that adjusted for a priori-defined participants' baseline characteristics that might be potential confounders. These factors included sociodemographic characteristics (i.e., age, race, sex, educational level, economic status, habitation status), health status (i.e., HCC score), functional ability (i.e., SF-36 Physical Component and Mental Component Subscale scores), and baseline PACIC scores and satisfaction with health care. All models included site indicators to account for the fact that patients within a site may report more similarly than patients at other sites. These analyses were performed on Stata version 9 statistical software (StataCorp, College Station, TX).
To estimate the effect of the intervention on physicians' impressions of their chronic care, we used ordinary least squares to regress baseline-to-follow-up changes in physicians' ratings on study group assignment. These analyses were run on SAS version 9.1 statistical software (StataCorp).
|
RESULTS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
DISCUSSION |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
GC also improves PCPs' satisfaction with several aspects of their care of multimorbid older patients: knowing which medications their patients are taking, communicating with patients and families, educating family caregivers, motivating patients, and referring patients and families to community resources. A small vanguard group of GCNs reports uniformly high job satisfaction.
Conclusions about the effects of GC on other realms of health care quality, such as efficiency, effectiveness, timeliness, and technical proficiency, cannot be inferred from these findings. Likewise, its effects on other important outcomes of health care, such as quality of life, functional independence, cost of care, and caregiver strain will become clear only as the trial continues and future waves of data are analyzed.
This study's limitations include the small number of nurses (seven), its geographic scope (urban mid-Atlantic), its reliance on self-reported data, and a limited consent rate. Likewise the physicians' response rate at the 1-year follow-up survey was suboptimal and could have biased the observed results. Future analyses will be needed to confirm these findings and to evaluate the effects of GC on other important outcomes.
|
A |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
We acknowledge the invaluable contributions to this study made by Johns Hopkins Community Physicians, MedStar, Battelle Centers for Public Health Research, the Centers for Medicare & Medicaid Services, Accumen, ResDAC, the University of Minnesota Survey Research Center, the study consultants (Jean Giddens, RN, PhD; Kate Lorig, RN, DrPH; Richard Bohmer, MD, MPH, MBA; Mary Naylor, RN, PhD), the nurse managers (Lora Rosenthal, RN, BA, CCM, and Carol Groves, RN, MPA), administrative assistant (Adriane King, MA), data analysts (Martha Sylvia, RN, MSN, MBA; Paula Norman, BS), and all of the participating patients, caregivers, physicians, and Guided Care nurses.
NOTE
CONSORT checklists are available from the authors.
|
F |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Received August 3, 2007
Accepted August 30, 2007
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. L. Wolff, E. Rand-Giovannetti, S. Palmer, S. Wegener, L. Reider, K. Frey, D. Scharfstein, and C. Boult Caregiving and Chronic Care: The Guided Care Program for Families and Friends J Gerontol A Biol Sci Med Sci, July 1, 2009; 64A(7): 785 - 791. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
