Interest in healthcare outcomes research has increased tremendously since the early 1990s. Many factors have contributed to the growth of this type of research, including the demand for more accountability and concurrent development of quality improvement programs, the need to examine outcomes beyond morbidity and mortality, and the challenge to provide higher quality care using more cost-effective approaches.1
The study of healthcare outcomes includes a broader range of outcomes than does the study of traditional clinical outcomes (eg, survival, benefits and complications of treatments) and is much more patient centered. Healthcare outcomes, such as physical function, health status, psychological status, social interaction, symptom distress, quality of life, patient satisfaction, occupational function, behavior, etc, are being measured in a variety of contexts. These outcomes are being measured in patients who undergo medical and/or surgical therapies to understand the relationships between treatments and outcomes and in patients with chronic illnesses to explore changes in outcomes over time. Outcomes research is also being conducted to identify patients at risk for poor health-related outcomes so that targeted interventions can be developed and tested to assist patients to improve these outcomes.
Unfortunately, there is no simple, summary instrument that can be used to measure all human suffering, which will result in a single, easily interpretable statistic.1,2 Human life is complex and multifaceted. Therefore, measurement of the impact of illness and its treatment on outcomes relevant to the human situation is also complicated. Methodologic issues that have challenged researchers include the need to appropriately define the outcomes being studied, use relevant theoretical frameworks, and select adequate study designs, instruments, and statistical techniques within the context of trying to understand these interdependent, sometimes nonlinear, and time-dependent changes.1 Nurse researchers have accepted the challenge and have increasingly contributed to the body of healthcare outcomes research.
Implicit in the challenge of developing scientifically sound healthcare outcomes research is the need to carefully link conceptual and methodologic components within each study. Unfortunately, sometimes both the meaning and the measurement of outcomes (eg, quality of life) in studies found in the literature have not been clear, let alone consistent with one another.3 Definitions of outcomes are essential. Conceptual frameworks must guide the selection of study design, instruments, and statistical techniques and be relevant to the population of patients from which the sample is drawn. While randomized clinical trials are the gold standard of research designs, well-designed quasi-experimental studies can also contribute important information to the healthcare outcomes research literature.1 Cross-sectional designs are frequently used; however, longitudinal designs are often more appropriate given the temporal nature of outcomes.
The selection of instruments to measure healthcare outcomes is complex. Decisions must be made regarding (1) the healthcare outcome to be studied, (2) the use of generic versus disease-specific scales and single, global questions versus multiple-item tools, and (3) whether to use subjective and/or objective instruments.4 It is also critical to report how well an instrument performs (and therefore how useful it is in measuring a given outcome) by assessing reliability, validity, responsiveness, and sensitivity.5 Furthermore, respondent (ie, patient, family, healthcare worker), setting (ie, in-hospital or home), and technique used to assess healthcare outcomes (ie, interview, questionnaire) must be considered when selecting instruments.
Lastly, statistical techniques used for analysis of healthcare outcomes research are necessarily complex. During the last few years, statistical techniques that account for these complexities have been developed, including techniques for adjusting for covariates, accounting for missing data, and dealing with multiple endpoints.6,7 An example of these techniques is mixed-effects modeling, which was developed for both continuous and ordinal data sets.7
During the last 2 decades, nurses have contributed significantly to healthcare outcomes research. They have identified relevant outcomes to study using appropriate conceptual frameworks and sound scientific methods. They have developed outcomes-related research protocols and quality improvement programs and designed nursing interventions to improve patient care outcomes that are adversely affected by illness, disease, and associated therapies. Nursing outcomes research has been conducted with many patient populations including those with cardiovascular disease.
The purpose of this issue of the Journal of Cardiovascular Nursing is to provide the reader with a state-of-the-art overview of cardiovascular nurse-sensitive outcomes (ie, outcomes partially or wholly influenced by nursing care) within each one of these topical areas: ischemic heart disease, ventricular dysrhythmias, stroke, metabolic syndrome, interventional cardiology, cardiac surgery, and heart failure. Generally, within each of the areas, authors have described and critiqued the nursing research, identified literature gaps, and provided recommendations for future research.
A search for appropriate articles was undertaken by each author(s) using Medline, Citations in Nursing and Allied Health Literature (CINAHL), and Healthstar from 1990 to present, with additional searches conducted using references from papers and scans of appropriate journals. Articles were found using appropriate keywords for each topical area cross-referenced with outcomes such as psychological status, quality of life, symptom distress, etc. Titles, authors, and articles were reviewed to determine if they met prespecified criteria (ie, a nurse was an author or the focus was on patient outcomes that could be influenced by nurses). Some of the articles were not written by nurses, but reported on outcomes that were nurse sensitive; other articles included multidisciplinary authorship. Articles were then reviewed for possible inclusion in the papers and grouped into outcome categories.
This collection of articles was not intended to provide a meta-analysis, but rather to provide a road map of areas where investigators have begun to identify outcomes that are nurse sensitive, which nursing interventions influence patient care, and future directions for nursing research. As we enter the age of public reporting of outcomes, it is essential that outcomes that nurses can influence are identified, reported, and refined. Our hope is that this work will assist nurses to (1) understand the current body of research and its limitations in order to provide evidence-based patient care and (2) develop new outcomes studies.
REFERENCES