National guidelines recommend a team model of care to facilitate adherence to evidence-based practices; however, previous studies suggesting benefit may have limited generalizability. The aim of this study was to examine the influence of advanced practice nurse (APN) and physician assistant (PA) staffing on the delivery of guideline-recommended therapies for outpatients with heart failure (HF). The Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF), a prospective cohort study, enrolled 167 cardiology practices to characterize outpatient management of 15,381 patients with chronic HF and left ventricular ejection fractions ≤35%. Adherence to guideline-recommended HF therapies was recorded, and the presence of APN and PA staffing was assessed by survey. Multivariate models identified contributions to the delivery of guideline-recommended HF therapies. Of cardiology outpatient practices, 66.0% had APNs and PAs. Practices with 0, >0 to <2, and ≥2.0 APN and PA staffing had similar adherence to the 7 guideline-recommended HF therapies. After adjustment, staffing with ≥2 APNs or PAs was associated with greater conformity with 2 of 7 measures (implantable cardioverter-defibrillator therapy and delivery of HF education, p ≤0.01 for both) and similar conformity to angiotensin-converting enzyme inhibitor or angiotensin receptor blocker therapy, β-blocker therapy, aldosterone antagonist therapy, anticoagulation for atrial fibrillation, and cardiac resynchronization therapy. In conclusion, staffing with APNs and PAs varied in cardiology outpatient practices. Compared to no APNs or PAs, ≥2.0 APNs or PAs per cardiology practice was associated with the greater use of implantable cardioverter-defibrillator therapy and delivery of HF education and equivalent use of drug and cardiac resynchronization therapies.
National heart failure (HF) guidelines recommend a team model of care for all patients who are at high risk for hospital admission or who have clinical deterioration as a means to provide education, manage chronic symptoms, and facilitate the implementation of evidence-based practices. In American studies, a team model of care most often involved nurse-led clinics that were carried out by nurses with advanced practice nurse (APN) credentials; however, physician assistant (PA) services have also been used in managing ambulatory patients with HF. In American nurse-managed clinics, emphasis was placed on reporting patient functioning and hospitalization rates and financial outcomes. Relatively little is known about the effectiveness of HF-devoted APNs and PAs on the delivery of recommended HF therapies in outpatient cardiology practices beyond artificial research settings, in which APNs and PAs are more likely to be highly trained and experienced in managing HF and in which they work closely with cardiologists who are also HF specialty trained. Thus, the objective of this study was to examine the influence of APN and PA staffing on the delivery of guideline-recommended therapies for patients with HF in outpatient cardiology practices.
Methods
The Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF) is a prospective, longitudinal cohort study designed to characterize the current management of patients with chronic HF or previous myocardial infarctions and left ventricular systolic dysfunction in outpatient cardiology practice settings. The overall study objectives, design, and methods, including definitions of 7 process measures, were described in detail previously. The analyses examining effect of APN and PA staffing on the delivery of guideline-recommended therapies for HF was prespecified in the study protocol. Those who were invited to participate in IMPROVE HF were from community-based practices with no academic affiliations, academically affiliated university settings, and nonuniversity settings with single-specialty or multispecialty cardiology outpatient practices from all regions of the United States. Chronicity of HF was assessed by physician documentation on ≥2 separate visits for HF treatment in the current practice setting during the 2-year period preceding study initiation. Left ventricular systolic dysfunction was confirmed by a quantitative left ventricular ejection fraction ≤35% measured by the most recent echocardiographic study, nuclear multiple gated acquisition scan, contrast ventriculographic study, or magnetic resonance imaging scan or a qualitative assessment of left ventricular function indicative of moderate to severe dysfunction with stage C HF or post–myocardial infarction without HF (stage B).
In this analysis, we used baseline data collected from medical chart reviews and entered into the IMPROVE HF registry from late 2005 through early 2007. Data were collected using a standardized case report form. Documented contraindications, intolerance, or economic, social, religious, refusal, or nonadherence reasons for not prescribing evidence-based HF therapies were also collected and used in patient eligibility for inclusion of quality improvement measures. The most recent electrocardiographic computerized reading or physician measurement was used to obtain QRS duration. During baseline assessment, a median of 90 patients with HF from each practice (interquartile range 58 to 107) were included, providing a representative sample of patients per practice.
Practice characteristics were collected by survey at baseline and included the number of APNs (defined as nurse practitioners or clinical nurse specialists) or PAs dedicated to HF management; geographic location; practice type; the numbers of cardiologists, electrophysiologists, and non-APN nurse clinicians; affiliation with a hospital or transplantation center; the presence of device-based and dedicated HF clinics; and annual average number of patients.
All practices participating in IMPROVE HF were approved by a local institutional review board or central institutional review board or received institutional review board waivers. Data were collected by highly trained chart review specialists who received centralized retraining and testing to ensure accuracy in data abstraction and who used prespecified definitions for each variable. Average interrater reliability (κ) was 0.82. To further ensure the completeness and accuracy of collected data, 1.7 automated quality checks per data field were performed, and reports were generated monthly. The registry coordinating center was Outcome Sciences, Inc. (Cambridge, Massachusetts).
Quality improvement measures were prospectively selected by the IMPROVE HF Steering Committee to quantify the quality of outpatient delivery of guideline-recommended therapies for HF. All 7 measures were therapies designated as class 1 (useful and effective) in the American College of Cardiology and American Heart Association guidelines. Four of the 7 measures involved drug therapies: the use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, β blockers, anticoagulation for atrial fibrillation, and aldosterone antagonists for eligible patients without documented contraindications or intolerance. Two measures assessed the use of cardiac devices: cardiac resynchronization therapy (CRT) and implantable cardioverter-defibrillator (ICD) therapy for eligible patients without documented contraindications or intolerance. The final measure was documentation that HF education (including discussion of a salt-restricted diet, monitoring daily weight, warning signs of worsening HF, and activity recommendations) was provided to eligible patients. The 7 measures were selected through a process that was independent of the study sponsor and described in the design and baseline findings reports. Documentation of New York Heart Association functional status was a prerequisite for eligibility for aldosterone antagonist, CRT, and ICD measures. Only patients with quantitative or qualitative documentation of functional status consistent with prespecified definitions were included in analyses of these 3 care measures.
All statistical analyses were performed by independent biostatisticians contracted by Outcome Sciences, Inc. Descriptive statistics for patient and practice characteristics were calculated and reported for all practices in the registry that completed practice surveys at baseline. The proportions and 95% confidence intervals or medians and interquartile ranges for patient and practice characteristics and the 7 measures used to define guideline-recommended HF therapies were based on the presence and level of use of APNs and PAs dedicated to HF care (categorized as 0, >0 to <2, and ≥2 full-time equivalent positions). General estimating equations were used to adjust for the clustering effect within practices. Univariate general estimating equation models were first calculated for patient clinical and demographic characteristics and practice characteristics that might be associated with ≥2 versus 0 APNs or PAs and >0 to <2 versus 0 APNs or PAs. Multivariate general estimating equation models, based on variables that were statistically significant at the 0.10 level in the univariate general estimating equation models, were then calculated to identify factors independently associated with APN and PA staffing. Analyses were completed using SAS version 9.1 (SAS Institute Inc., Cary, North Carolina). Statistical tests were 2 sided, and p values <0.05 were considered statistically significant.
Results
The analysis included medical records of 15,381 patients receiving care at 167 outpatient cardiology practices. On the basis of the completeness of responses to practice survey questions, 14,891 patients were included in the analyses. Of IMPROVE HF practices, 162 (97%) completed surveys about specific service components, and 66% used HF APN and PA health care providers. Some characteristics of patients ( Table 1 ) and practices ( Table 2 ) varied on the basis of the use of APN and PA staffing.
| Characteristic | APN/PA Staffing Level | p Values | |||||
|---|---|---|---|---|---|---|---|
| 0 (n = 4,381) | >0 to <2 (n = 4,394) | ≥2 (n = 6,116) | Overall | 0 vs >0 to <2 | 0 vs ≥2 | >0 to <2 vs ≥2 | |
| Age (years) | 68.9 ± 13.2 | 68.4 ± 13.3 | 68.6 ± 13.2 | 0.099 | 0.033 | 0.192 | 0.307 |
| Men | 71.2% | 70.4% | 71.9% | 0.265 | 0.404 | 0.468 | 0.103 |
| Race | |||||||
| White | 38.9% | 45.0% | 40.5% | 0.005 | 0.483 | 0.077 | <0.001 |
| African American | 8.8% | 10.8% | 7.8% | ||||
| Not documented or missing | 50.5% | 42.5% | 50.0% | ||||
| Ischemic cause of HF | 67.2% | 65.7% | 63.6% | <0.001 | 0.217 | <0.001 | <0.001 |
| Previous atrial fibrillation | 29.7% | 29.9% | 32.2% | 0.007 | 0.868 | 0.006 | 0.010 |
| Diabetes mellitus | 33.7% | 34.8% | 33.2% | 0.200 | 0.256 | 0.581 | 0.075 |
| Hypertension | 63.0% | 63.1% | 59.8% | <0.001 | 0.933 | <0.001 | <0.001 |
| Previous myocardial infarction | 39.1% | 39.3% | 39.6% | 0.875 | 0.863 | 0.614 | 0.751 |
| Chronic obstructive pulmonary disease | 16.7% | 17.4% | 15.6% | 0.043 | 0.398 | 0.125 | 0.014 |
| Previous coronary bypass | 31.5% | 31.2% | 30.3% | 0.395 | 0.799 | 0.206 | 0.322 |
| Peripheral vascular disease | 12.6% | 11.7% | 10.3% | <0.001 | 0.195 | <0.001 | 0.023 |
| Depression | 7.6% | 8.7% | 9.7% | 0.001 | 0.052 | <0.001 | 0.107 |
| Left ventricular ejection fraction | 25.7 ± 6.9% | 25.6 ± 6.9% | 25.2 ± 7.1% | 0.008 | 0.802 | 0.006 | 0.013 |
| Systolic blood pressure (mm Hg) | 121.2 ± 18.6 | 120.5 ± 19.0 | 119.9 ± 18.9 | 0.001 | 0.030 | <0.001 | 0.216 |
| Diastolic blood pressure (mm Hg) | 70.7 ± 11.1 | 70.2 ± 11.5 | 70.2 ± 11.2 | 0.012 | 0.004 | 0.025 | 0.374 |
| Heart rate at rest (beats/min) | 72.0 ± 11.7 | 72.5 ± 11.4 | 71.8 ± 11.5 | 0.010 | 0.016 | 0.806 | 0.005 |
| Rales on most recent examination | 3.1% | 4.1% | 3.8% | <0.001 | <0.001 | <0.001 | 0.646 |
| Edema on most recent examination | 19.7% | 20.5% | 19.3% | 0.012 | 0.175 | 0.288 | 0.002 |
| Sodium (mEq/L) | 139.3 ± 3.9 | 139.0 ± 4.0 | 139.3 ± 4.5 | <0.001 | <0.001 | 0.616 | <0.001 |
| Blood urea nitrogen (mg/dl) | 25.1 ± 13.8 | 26.0 ± 15.7 | 26.0 ± 15.0 | 0.036 | 0.247 | 0.010 | 0.184 |
| Creatinine (mg/dl) | 1.4 ± 0.7 | 1.4 ± 0.9 | 1.4 ± 0.8 | 0.157 | 0.085 | 0.817 | 0.103 |
| Potassium (mEq/L) | 4.5 ± 2.99 | 4.4 ± 0.99 | 4.5 ± 1.58 | <0.001 | 0.001 | 0.961 | <0.001 |
| B-type natriuretic peptide (pg/ml) | 992 ± 881 | 1,567 ± 713 | 1,537 ± 730 | 0.021 | 0.271 | 0.07 | 0.069 |
| QRS duration (ms) | 128.4 ± 37.9 | 127.9 ± 40.7 | 130.4 ± 41.1 | 0.186 | 0.491 | 0.072 | 0.274 |
| NYHA functional class | <0.001 | <0.001 | <0.001 | <0.001 | |||
| I and II | 73.5% | 65.7% | 68.4% | ||||
| III and IV | 26.5% | 34.3% | 31.6% | ||||
| Characteristic | Number of APN/PA HF Staff Members | p Value | ||
|---|---|---|---|---|
| 0 (n = 55) | >0 to <2 (n = 49) | ≥2 (n = 58) | ||
| Region | 0.814 | |||
| South | 43.6% | 38.8% | 34.5% | |
| Northeast | 30.9% | 36.7% | 31.0% | |
| Central | 14.5% | 12.2% | 20.7% | |
| West | 10.9% | 8.2% | 13.8% | |
| Missing | 0.0% | 4.1% | 0.0% | |
| Practice setting | 0.038 | |||
| University, teaching | 5.5% | 2.0% | 15.5% | |
| Nonuniversity, teaching | 14.5% | 28.6% | 22.4% | |
| Nonuniversity, nonteaching | 72.7% | 65.3% | 56.9% | |
| Missing | 7.3% | 4.1% | 5.2% | |
| Hospital based | 20.0% | 24.5% | 39.7% | 0.059 |
| Transplantation center | 3.6% | 8.2% | 17.2% | 0.049 |
| HF clinic in practice | 12.7% | 53.1% | 60.3% | <0.001 |
| Multispecialty practice | 23.6% | 16.3% | 32.8% | 0.173 |
| Cardiologists in practice | 9.4 ± 6.6 | 11.4 ± 7.4 | 15.2 ± 15.0 | 0.023 |
| HF patients seen annually | 3,135 ± 3,600 | 2,474 ± 2,987 | 3,782 ± 4,748 | 0.416 |
| Medical record system | 0.364 | |||
| Electronic | 27.3% | 26.5% | 43.1% | |
| Paper only | 45.5% | 49.0% | 36.2% | |
| Mixed | 18.2% | 20.4% | 15.5% | |
| Missing | 9.1% | 4.1% | 5.2% | |
Level of APN and PA staffing in practices (0 [n = 55], >0 to <2 [n = 49], and ≥2 [n = 58]) was not significantly associated with the delivery of guideline-recommended HF therapies for any of the 7 measures studied ( Figure 1 ); however, β-blocker therapy trended toward being higher in practices with >0 to <2 APNs or PAs (p = 0.08), and documentation of HF education (p = 0.08) and ICD therapy use (p = 0.11) trended toward being higher in practices with ≥2 APNs or PAs ( Figure 1 ). Importantly, there were no significant negative associations in the delivery of guideline-recommended HF therapies on the basis of APN and PA staffing.
