Improving Practice: Reporting Quality Improvement Activities

There is a gap between what we know and what we do.1 Hearing and reading about new scientific knowledge alone is often not sufficient to make us adopt new practices. Yet, we do not clearly understand which approaches and methods most effectively change our behavior as health care providers. Furthermore, in everyday clinical practice each of us encounters system problems that cannot be solved simply by applying new knowledge. Instead, these system problems require that we collectively develop practical solutions by trial and error. Either way, we can be more efficient in our learning when we share and critically review what others have already tried and what has or has not worked.

In this issue of AJKD, we launch a new category of original investigation for publication of quality improvement reports. This new article type provides a forum to discuss Quality Improvement Activities (QIAs) in the care of patients with kidney disease and related disorders. We invite studies that report the development and testing of strategies for quality improvement and their results. The inaugural article by Li et al2 discusses the effective delivery of rehabilitation services for elderly hemodialysis patients who have suffered acute disability, with the intent to enable them to return to independent living. The article describes how a new program was developed to combine inpatient rehabilitation, interdisciplinary care, and daily dialysis, and evaluates its success over the first 3 years by counting discharges to home, change in functional independence scores between admission and discharge, and achievement of rehabilitation goals set on admission.

QIAs have become routine in dialysis care. The coordination of QIAs in long-term dialysis units was accelerated by the concentration of dialysis care in the hands of a small number of providers with economies of scale, the practice of collecting data on most dialysis patients in centralized data registries, development of clinical practice guidelines and their endorsement by professional societies, and policies by the government as the major payer for dialysis services. QIAs have been linked to improvements in quality of care and clinical outcomes.3 But without proper study, we cannot be sure whether and to what degree improvements in care are attributable to QIAs.4 There are many other promising opportunities for quality improvement. These opportunities lie in the management of earlier stages of chronic kidney disease (CKD), acute kidney injury, and other acute kidney diseases, in transplantation, and in the early detection and prevention of acute and chronic kidney diseases in all health care settings. However, many logistic barriers remain to be overcome, in the first instance the fragmentation of providers, information systems, and payers which is characteristic of venues other than outpatient dialysis units.

A recent report by an expert panel on the topic defined QIAs as “systematic data-guided activities designed to bring about immediate improvements in health care delivery in particular settings.”5 This report highlights similarities and differences between QIAs and research, and the corresponding ethical issues and necessary oversight by regulatory and administrative authorities for each. The report considers quality improvement as an intrinsic part of normal health care operations, in which clinicians and patients have an ethical obligation to participate. Oversight for QIAs resides within the health care organization’s clinical leadership to ensure that the design and conduct of QIAs protect patients’ privacy, interests, and safety. Health care organizations differ widely in the delegation of these responsibilities to system-wide committees, practice groups, or individuals. In contrast, research is a voluntary activity by health care professionals and patients to “develop or contribute to generalizable knowledge about the nature and function of human beings and their environment.”5 Clinicians and patients have no ethical obligation to participate in research, but may choose to do so voluntarily. Oversight for research resides within federally mandated institutional review boards (IRBs), which approve and monitor specific protocols for the protection of human subjects.

The table lists features of QIAs, which are exemplified by our inaugural report by Li et al. These are contrasted with features of traditional research studies that test interventions or describe treatment effects. For our new article type, we chose the term “quality improvement report” to describe a report of a QIA that was conducted as an initiative to improve quality of care and that does not follow the design of a prospective research study. Most QIAs are not research, but share features of prospective studies. As in an uncontrolled trial or a cohort study, deliberately planned action, and prospective monitoring and data collection may overcome some of the shortcomings of retrospective studies, but usually the absence of an appropriate control group nonetheless limits inference. In some situations, QIAs will directly overlap with research. For example, a clinical trial or observational study of a QIA may permit more rigorous evaluation of the efficacy or effectiveness of that activity. Other research designs such as case-control studies, retrospective cohort studies, or case series examining treatment efficacy or effectiveness may also overlap conceptually with QIAs. Cross-sectional studies and analyses or trend studies examining treatment or policy interventions’ effectiveness over time are not shown in the table. However, they may generate valuable descriptive information of current performance and practice variability and identify areas for improvement for subsequent QIAs.

Table. Comparison of Quality Improvement Activities With Clinical Research Studies of Interventions or Treatments⁎

		Prospective study of intervention: experimental study (clinical trial) or observational study (prospective cohort study)	Retrospective observational study of treatment exposure (retrospective cohort study, case-control study, or case series)	Quality Improvement Activity
	Goal	Develop or contribute to generalizable knowledge about the nature and function of human beings and their environment.	Develop or contribute to generalizable knowledge about the nature and function of human beings and their environment.	Immediately improve health care delivery in particular settings by conducting systematic, data guided activities.
	Aim	Clinical trial: Test efficacy or effectiveness of an intervention. Observational study: Observe relationship between treatment and outcome.	Retrospectively describe temporal relationship between exposure to treatment and outcome.	Overcome perceived or previously identified barriers to care by implementing a change.
	Review and oversight/Regulatory authorities	Review by IRB based on ethical and regulatory requirements governing human subject research.	Review by IRB based on ethical and regulatory requirements governing human subject research, or no review.	Ethical review and supervision appropriate to level of potential risk and project worth. For projects with primary goal of quality improvement and low risk: review and supervision by local practice organization to ensure accountability for ethical conduct. For projects with goals and design features that overlap with other types of human research studies or where activity carries higher level of risk, review also by IRB.
	Patient’s role and responsibility	Patients are recruited. Patient participation is voluntary and consent must be fully informed.	Patients are members of the observed cohort or cases.	Patients are clients of a practice and are considered to have an ethical responsibility to participate. Consent to receive care includes consent to participate in routine, minimal risk QIAs.
	Design of intervention	Clinical trial: Intervention is designed or specified by researchers. Intervention may be “experimental.” Observational study: Not applicable; treatment is provided by patients’ regular providers. Intervention is not “experimental.”	Not applicable. Investigators record treatments that were delivered by patients’ regular providers.	Intervention entails a deliberate action that is integrated into ongoing service delivery. Care delivered within QIA is thought to be within the range of the standard of care, or no worse. It often consists of repeated cycles of measurement and change.
	Investigator role	Key investigators may not have an ongoing commitment to improvement of the local care situation.	Key investigators may not have an ongoing commitment to improvement of the local care situation.	Project originates in practice setting. Key project managers have a commitment to improving care in the specific setting and interact with practice administration.
	Funding	External or internal funding, often through resources earmarked for study conduct.	External or internal funding or no funding.	Internal funding from parties in the clinical setting or organization, in which the activity takes place. May be funded by revenue from routine heath care. May receive external funding.
	Control group	Clinical trial: May entail random allocation of patients to treatment and control group to reduce bias. Observational study: May entail comparison of groups based on differences in exposure to treatment.	May entail comparison of groups based on differences in exposure to treatments.	Often no concurrent control group. May involve a historical control group and compare care “before” and “after” the QIA.
	Feedback	Clinical trial: May entail deliberate masking of patients, providers and outcomes assessors to reduce bias in interpretation of results. Observational study: No masking of patients or providers. Study personnel may be masked.	Usually none.	No masking. May contain multiple purposeful cycles of feedback from results of data analyses and actual clinical experience to adjust project goals and approaches.
	Type of outcomes	Any (clinical or surrogate outcomes).	Any (clinical or surrogate outcomes).	Any (clinical or surrogate outcomes), often process measures.
	Outcome ascertainment	Prespecified, prospective, with aim to be complete.	Not prespecified; usually not complete.	Maybe prespecified, prospective, may be complete.
	Generalizability	Limitations due to inclusion and exclusion criteria (explicit).	Limitations due to inclusion and exclusion criteria (generally not explicit).	Limitations due to unique features of context (personnel, process or system).

Note: This table incorporates and builds on concepts from Lynn et al.5 Abbreviations: IRB, Institutional Review Board; QIA, Quality Improvement Activity.

⁎ Interventions or treatments refer to drugs or devices; nutritional, behavioral, or educational interventions; or a process of care at the patient, provider, or practice level.

In the past, publication and sharing of reports of QIAs have been hampered by several factors. First, editors and peer reviewers lack a template for appraisal of reports on QIAs or may not see the scientific merit of these reports.6 We address this by offering our new article category in AJKD for QIA reports. Regardless of the form, we will consider submissions of reports on QIAs, although we will favor studies with methodologically sound designs and rigorous conduct that strengthen the conclusions that can be drawn. Second, individuals undertaking QIAs may identify themselves as practical problem solvers rather than researchers.6 We encourage the teams and individuals conducting QIAs to report their experiences so that others can learn from them. Authors should follow our recommended abstract format7 and a checklist of items for the reporting of QIAs.8 Third, the uncertainty surrounding review and oversight of QIAs may discourage their conduct and publication. New approaches to address this problem need to be developed and evaluated going forward.5, 9 We expect that before the submission of a manuscript reporting QIAs, the plan for the QIA will have been reviewed by the clinical leadership of the organization whose experience is reported, just as we require the research design for a human subject study to have been approved by the IRB. We ask authors to indicate whether this has been the case when submitting a manuscript as a quality improvement report to AJKD.

Finally, we ask all to engage in an effort to improve the care of our patients. This requires each of the following steps: conducting clinical research studies, translating findings into practice, and continually striving to improve quality of care.