Shifting Cardiovascular Care to Nurses Results in Structured Chronic Care
Published Online: August 06, 2014
Elvira Nouwens, MSc; Jan van Lieshout, MD, PhD; Pieter van den Hombergh, MD, PhD; Miranda Laurant, PhD; and Michel Wensing, PhD
Internationally, in primary care, clinical tasks have been shifting from physicians to nurses in varying degrees. Reallocation of such tasks from family physicians (FPs) to nurses has been found to be associated with improved, or at least equivalent, quality and outcomes of chronic disease care.1-5 The degree of reallocation of tasks varies across countries as a result of policy, organizational, and legal factors.6 In addition, the absence of robust evidence on the impact of such reallocation in natural settings (as opposed to controlled trials and demonstration projects) is an issue in decisions to shift clinical tasks to nonphysicians.
Better structuring of primary healthcare for patients with chronic diseases is expected to result in better outcomes for patients and societies,7 if it integrates all domains (eAppendix, available at www.ajmc.com) specified by the Chronic Care Model (CCM), a model created by Wagner.8 Allocation of tasks to nurses may contribute to better implementation of structured chronic care by increasing capacity and competence in specific areas, such as coordination and patient education, compared with healthcare delivered by physicians only.2,9
Cardiovascular disease (CVD) remains a major cause of death and disability around the world, and many preventive measures focusing on CVD have been developed for use in public health and primary care.10,11 Reassigning some tasks usually performed by physicians to nurses can be applied in cardiovascular risk management (CVRM).12,13 Nursemanaged CVRM programs have been shown to improve patients’ lifestyle, risk factor control, use of medications, and quality of life.13 These positive effects are based on the clinical knowledge and skills of nurses,14 but potentially also on their contribution to teamwork and practice organization.
Our study sought to explore the potential contribution of task allocation to primary care nurses to the implementation of structured chronic care for patients with cardiovascular conditions. The study focused on primary care practices in 7 countries to try to get meaningful results from real clinical practice settings across various healthcare systems.
Design and study population
This study was based on secondary analysis of data from the European Practice Assessment of Cardiovascular Risk Management Project (EPA Cardio Project), an observational study on CVRM in 315 primary care practices in 10 countries in 2008-2009.15 EPA Cardio focused on patients with CVD, patients at high risk for CVD, and healthy adults. Multiple measurement instruments were used, including chart reviews, patient surveys, and validated questionnaires on practice characteristics16 completed by FPs in the practices. For the present study, we included data from 202 practices in 7 countries: Austria, Belgium, Germany, the Netherlands, Slovenia, Spain, and Switzerland. Data from the remaining 3 countries were excluded from this analysis because the required data on nurse involvement in tasks concerning CVRM were missing.
For this study, we used data from the EPA Cardio Project’s FP questionnaires—specifically, the data on practice characteristics (practice size, number of full-time-equivalent FPs and nurses) and the diversity of tasks assigned to FPs and nurses. For our purposes, “nurses” included both nurse practitioners and advanced practice nurses, all of whom performed clinical tasks in primary care practices. We measured nurse and FP involvement in 5 specific tasks concerning CVRM: 1) administrative tasks, recalls, and recording (enrolling and selecting patients for periodic checkups, managing patient records, archiving laboratory results, and mail); 2) monitoring risk factors (periodic monitoring of blood pressure, smoking status, weight, cholesterol, etc.); 3) patient education and counseling (including discussion of control of diet and physical activity); 4) defining protocol and organization of CVD care (drafting protocols based on guidelines for CVD care, coordination of CVD consulting hours); and 5) diagnosis, risk assessment, and medical treatment (determining risk profile, periodic check-up for CVD patients, discussing medication).
These are tasks that the core group (consisting of researchers from the EPA Cardio research group) decided that nurses are presumably capable of performing.17 The tasks were measured on a dichotomous answering scale (yes/ no), and in addition, for each one, the degree of adoption of CCM was calculated using a list of 35 criteria (eAppendix).18 The measurements are translated into a score on a 0-to-5 scale based on the 5 practice-related domains derived from the CCM19, each of which has equal weight: 1) healthcare organization (7 items); 2) delivery system design (15 items); 3) decision support (3 items); 4) clinical information systems (6 items); and 5) self-management support (4 items). A higher total score means a greater degree of structured chronic illness care.
Practice size was determined from patient list size when available, and otherwise on yearly attending patient numbers. Results of both nurse practitioners and advanced practice nurses on the 5 specific CVRM tasks were combined.
Data analysis consisted of both descriptive and analytical methods. Random coefficient linear regression modeling was used to explore the effect of nurse involvement on structured illness care (adoption of CCM). The influence of the number of FP tasks and nurse tasks, nurse/FP ratio, and practice size on the adoption of CCM was also explored in the regression model. Two-level models were specified with practices nested in countries. Data analysis was performed using the SPSS 16.0 software package (Chicago, Illinois).
Table 1 shows characteristics of the practices in the 7 countries. The number of included primary care practices ranged from 21 in Switzerland and Germany to 36 in Spain. Spain had the largest mean practice size and Slovenia the smallest. In Austria, Germany, the Netherlands, and Spain, 100% of practices had a nurse employed. In Slovenia, 94.1% of practices employed a nurse, Switzerland 85.7%, and Belgium 33.3%. Overall, 89.6% (N = 177) of practices employed a nurse. The number of FP tasks varied minifrom 3.9 in Slovenia to 4.9 in Spain, while the number of nurse tasks varied from 1.5 in Belgium to 4.8 in Spain. The adoption of CCM was highest for practices in Spain and lowest for practices in Switzerland.
Considering nurse involvement in specific tasks, Table 2 shows nurses were mostly involved in CVRM-related administrative tasks, recalls, and recording (82.2%) and least involved in diagnosis, risk assessment, and medical treatment (34.6%). FP involvement was highest in diagnosis, risk assessment, and medical treatment (99%) and lowest in administrative tasks, recalls, and recording (68%). For administrative tasks, recalls, and recording; monitoring risk factors; and patient education and counseling, most percentages of nurse involvement were above 50%. For the tasks defining protocol and organization of cardiovascular care and diagnosis, risk assessment, and medical treatment, most percentages of nurse involvement were below 50%. Considerable variation between countries was found on involvement in specific tasks of both nurses and FPs.
Relation of nurse involvement to structured chronic illness care
Figure illustrates the number of nurse tasks related to the adoption of CCM. When more tasks were allocated to nurses, the mean adoption of CCM was higher. Table 3 shows that the mean adoption of CCM in practices without nurse involvement was 2.6 (SD 1.3; 95% CI, 1.5-3.6) and in practices with nurse involvement on all tasks, the mean adoption of CCM was 4.0 (SD 0.5; 95% CI, 3.8-4.1). Practices without nurse involvement on CVRM-related tasks showed a minimum adoption of CCM of 0.5 and a maximum adoption of CCM of 4.4. For practices with nurse involvement on all CVRM-related tasks, the minimum and maximum were 2.4 and 4.7. In practices where nurses had only 1 CVRM-related task, that task, in the vast majority, was administrative (eAppendix available at www.ajmc.com). Nurses with 2 tasks mostly had administrative tasks combined with monitoring risk factors. When nurses had 3 or 4 tasks, combination of tasks varied more. The mean adoption of CCM was lowest when nurses’ tasks skewed towards the administrative (3.3, SD 0.9) and highest when the tasks included diagnosis, risk assessment, and medical treatment (4.0, SD 0.7). The eAppendix provides further explanation.
The Figure suggests visually that with increasing involvement of nurses, the variation on the adoption of CCM decreased. Low scores on the adoption of CCM were absent when nurses had more tasks. However, some practices without nurse involvement had high adoption of CCM.
Table 4 shows the results of the regression model. Number of tasks of FP and of nurse, nurse/FP ratio, and practice size were taken into account. Practice size showed no effect on the adoption of CCM (P = .19). The analysis included 155 cases. The degree of skewness was explored for the adoption of CCM. This variable was normally distributed. Results show that the number of nurse tasks had a positive effect on the adoption of CCM. When the number of nurse tasks increased by 1, the adoption of CCM increased by 0.13 (P <.05; 95% CI, 0.03-0.22). The regression model used in this analysis explained 19% of the variance on the adoption of CCM.
Nurses were involved in the delivery of CVRM in this sample of European primary care practices, but in varying degrees. In a vast majority of practices, nurses were involved in administrative tasks, recalls, and recording, and they were least involved in diagnosis, risk assessment, and medical treatment of cardiovascular care. Involvement of nurses in more aspects of CVRM was associated with better structured chronic illness care, but some practices with limited nurse involvement, related to administrative tasks only, also provided well-structured chronic illness care.
The variation of nurse involvement in the delivery of CVRM in this study may have several determinants. Reallocation of tasks concerning CVRM could differ among types of nurses, although previous research has shown that mid-level providers, with different educational backgrounds, perform similar tasks.20 Nurses’ levels of training and experience21 could influence the outcomes, but they were not taken into account in this study. The definition of “nurse” applied in this study could also influence the outcomes.
Nurses’ responsibility for chronic disease management may vary from practice to practice, dependent on the willingness of FPs to delegate tasks.22 Of the practices in our study with low nurse involvement (less than 2 tasks on CVRM; N = 23) and with high adoption of CCM (>3.5), none employed other nonclinical staff members performing administrative tasks or other CVRM-related tasks. Overall, nurse involvement in CVRM was mixed, but limited in the sample of practices.
Our results showed that well-structured chronic care is possible in practices with little nurse involvement. With maximum nurse involvement, there is on average better chronic care with less variation in the adoption of CCM. Successfully organizing chronic illness care may well be dependent on elements of care not accounted for in this study. Practice characteristics like team size and workload could affect the organization of chronic illness care.18
Furthermore, it may be that better-organized practices are more likely to employ and retain nurses to manage chronic patients; a better-organized practice likely provides a better working climate for nurses, thus providing higher job satisfaction. Practice size could influence the organization of chronic illness care; evidence shows that in larger patient populations, more clinical tasks are performed by healthcare providers other than physicians.23 However, our study suggests practice size has no significant influence on the adoption of CCM.
PDF is available on the last page.