Development of a hemodialysis safety checklist using a structured panel process

Errors during in-center hemodialysis are common, and patients often worry that an error will occur during their hemodialysis session. Checklists are a useful patient safety strategy that have improved care across a number of medical disciplines.

The first safety checklist created specifically for the hemodialysis procedure, using a proven technique for developing quality and patient safety measures in healthcare. This provides other dialysis facilities with an example to produce their own patient safety tools, and may catalyze local quality improvement efforts to enhance the safety of dialysis patients.

The “To Err is Human” report published by the Institute of Medicine estimated that 98,000 patients die each year from preventable medical errors in the United States [1]. Similar rates of preventable adverse events have been reported in Canada [2], Australia [3], New Zealand [4,5], and England [6]. The key lesson from these studies is that errors are usually a result of the system of care, rather than the individual.

Patients with end-stage renal disease (ESRD) are at high risk for medical errors given their frequency of complex treatment, comorbidities, polypharmacy, physiological consequences of ESRD, and coordination with other hospital departments to provide care [7,8]. These factors all contribute to a stressful and busy hemodialysis (HD) unit, so it is not surprising that medical errors are common in the ESRD population.

A 2006 survey documented provider and patient opinions on safety in the HD unit [9]. Frequent problems included greater than 2 vascular access needle insertion attempts (30%), access clotting (20%), machine difficulties leading to early treatment stoppage (15%), failure to record pre-dialysis blood pressure or weight (13%), and access needle disconnection (5%) [9]. Almost half (49%) of the patients who responded indicated that they are sometimes or always worried that a mistake will occur during their HD treatment [9].

Subsequent studies have yielded further insight into the unique HD safety issues. A review of 526 HD incident reports to the Pennsylvania Patient Safety Authority over a 12 month period identified medication errors (29.0%), protocol violations (12.9%), and falls (5.9%) as common errors during HD treatment [8]. Of these events, 87.6% reached the patient, and 5.5% resulted in patient harm [8]. More recently, a Scottish retrospective study found that in 3.5% of ESRD patient deaths there was an area of concern identified that likely contributed to the death, and 2.1% of patient deaths were a direct result of a dialysis complication [7]. Given the frequency and severity of errors associated with dialysis, multiple experts have called for action [10,11].

Checklist utilization is one patient safety strategy that may be effective. Checklists have been shown to improve patient safety, adherence to protocols/policies, communication, teamwork, and consistency of care by standardizing procedures [12]. The pause or “time out” discussion prior to an invasive procedure is an important component [12]. Checklists have been effective across several disciplines, most notably surgery and central line insertion [13,14]. The HD unit may be a particularly appropriate venue for a safety checklist, where it may prevent errors under stressful conditions and maintain staff precision, focus, clarity, and memory recall [15].

At present, a HD safety checklist does not exist. Therefore, our objective was to develop a HD safety checklist using a structured, consensus-based, panel process. Our focus was to develop a checklist that would improve consistency of care and provider/patient communication, as well as be feasible to implement in clinical practice.

Through a structured Delphi panel process, a HD safety checklist (Hemo Pause) was developed using evidence-based patient safety parameters and frontline staff experience. The decision to evaluate the checklist with a panel of HD nurses strengthened the utility of the document, since the intention of the checklist was for it to be used routinely by nurses and patients during HD. This approach helped ensure that the selected patient safety parameters not only adhered to the general principles of being scientifically sound, but also were clinically relevant and feasible to implement.

The final list of 19 patient safety parameters that received ≥75% consensus included actions from different time points of the HD procedure. Several attempts were made by the development team to incorporate all of these parameters into the final checklist. However, human factors engineering concepts limited the inclusion of some parameters (hand-washing, correct needle insertion, and falls). As such, some parameters were excluded to ensure that checklist tasks were clear and could be properly integrated by frontline staff into their normal workflow [33].

The feasibility of widely applying the HD safety checklist was a prime consideration. It was designed to be used collaboratively by nurses and patients at every HD session. Not surprisingly, duplication of work and time pressures were perceived by the panel as barriers to checklist implementation, even prior to piloting the checklist with patients. Several nurses also commented that many of the checklist items are already completed as part of usual practice. This is not surprising, since one of the purposes of a safety checklist is to standardize usual care, providing a barometer from which to measure errors and identify quality gaps. While these workload concerns can only be established by piloting the checklist in real practice, these observations underscore the importance of context in patient safety and quality improvement [16,34]. Frontline staff noted that the checklist needed to provide them with an immediate advantage, either in the form of reduced documentation or improved patient satisfaction in order to offset the upfront time investment required for checklist completion. These deliverables to frontline staff will be needed to sustain future checklist quality improvement efforts.

To our knowledge, this is the first reported safety checklist that is uniquely geared to HD treatment. In a systematic review of published renal replacement therapy quality improvement initiatives, none of the 93 studies focused on patient safety [35]. This is an important gap that must be filled. Previous safety efforts in dialysis have focused on measuring the rate of preventable adverse events, primarily involving voluntarily surveys [9] or incident reporting [8]. Reporting bias is a clear limitation to these methods, with variations over time more likely to reflect changes in reporting patterns than changes in patient safety [36]. A recent study by Bray et al. retrospectively reviewed all dialysis deaths over a 3.5 year period to identify preventable factors that may have contributed to mortality [7]. They identified preventable factors that may have or did contribute to death in 3.5% of deaths. These included errors related to communication, organization, and human factors, and were due to five main causes: management of hyperkalemia, prescribing, out of hours care, infection, and vascular access. An accompanying editorial called upon nephrologists “to take clearly articulated steps to improve the safety of the patients who trust them to provide care [11]”. Examples listed included standardization, teamwork, and culture change.

Our Hemo Pause Checklist may be able to improve all of these components. Checklists provide a reminder and cognitive aid for tasks, such as the multistep HD procedure [15]. They promote teamwork and communication, such as those between dialysis nurses and patients [12], which is the purpose of the pause or “time out” discussion before connecting the patient to the HD machine. Lastly, checklists highlight the importance of patient safety, where errors represent a failure of the system rather than an individual. Despite these advantages, the Hemo Pause Checklist should not be viewed as a universal solution. The limitations of checklists in medicine have been well documented, and include over-reliance, non-adherence, and implementation challenges [15,37,38]. A recent study of mandatory WHO Surgical Checklist use in Ontario, Canada demonstrated no province-wide effect on morbidity and mortality, suggesting different adherence and implementation strategies between the 101 hospitals [38,39]. On the other hand, the success of the Michigan Keystone Project on reducing rates of catheter-related bloodstream infection by 66% in 108 intensive care units was not only the result of a checklist, but also due to the system change and frontline engagement that accompanied its adoption [14,37,38].

Therefore, the Hemo Pause Checklist must be piloted in the HD unit using quality improvement and change management methods before it can be spread more widely. This will allow for the determination of checklist feasibility, measurement of performance gaps between checklist and usual care, and local checklist modifications to suit the patient safety concerns and workflow of individual HD units. We have initiated such a study in our dialysis unit, and encourage others to evaluate Hemo Pause in their local environments or undertake a structured panel process to develop their own patient safety tools.

Some limitations to our checklist development and modified Delphi panel process require discussion. First, the review of literature was not a systematic review, rather it was a targeted review designed by the development team. While it is possible that some evidence-based safety parameters were missed, this is unlikely given the paucity of literature that was identified. Second, the lack of patient safety literature in dialysis means that individual items on the checklist may not have an extensive evidence base; however, all elements on the checklist have clinical face validity. Third, the Delphi panel consisted only of HD nurses at a single-center to maximize feedback from the frontline workers for whom the checklist was designed. Although this excluded other important perspectives and may limit generalizability, the HD treatment constitutes a fairly stereotyped process with much similarity among different HD units. Many successful checklists have been designed in single-centers, including those from the Michigan Keystone Project [40-43]. Moreover, to minimize selection bias, nephrologists and administrators reviewed the final version of the checklist, and included 3 parameters (HD access, duration, and anticipated adverse events) that did not reach ≥75% panel consensus but had either a strong evidentiary basis or clinical face validity. We were unable to obtain patient and caregiver representation on either the development or evaluation team. Fourth, the standardized questions contained a large number of variables available for selection. While it was imperative to include all of the major patient safety parameters that emerged from the literature review, this has the potential to compromise the accuracy of the Delphi method [44]. This could have contributed to a “serial position effect”, meaning that the first factors listed were treated differently than factors listed further down [45]. We attempted to minimize this effect by ordering the parameters according to their temporal relationship in the HD procedure. Lastly, the potential for strong personalities and opinions to dominate the direction of discussion is always a concern. To combat this, we completed the anonymous survey twice to ensure that each participant had an opportunity to express their view. The final survey confirmed that all participants were satisfied with the process.

Our findings have important implications. Firstly, it highlights parameters that are needed for a safe HD session, both according to the literature and a multidisciplinary team of Nephrology personnel. Secondly, it outlines the first HD safety checklist to standardize care and communication in the HD unit. Lastly, our study identifies barriers to checklist implementation to inform patient safety and quality improvement efforts.

Patient safety is an emerging field, and there are many opportunities in Nephrology for improvement. The modified Delphi consensus panel approach applied in this study enabled us to develop and evaluate the first HD safety checklist (Hemo Pause). This methodology and the multidisciplinary representation provide strong face validity to the patient safety parameters included on the checklist. We believe that these findings provide other dialysis facilities with a roadmap to produce patient safety tools and catalyze local quality improvement efforts to enhance the health and safety of dialysis patients.

More research and testing of dialysis safety measures are clearly needed to determine their usefulness and feasibility. At our center, quality improvement efforts are underway to answer these questions, as well as determine if the Hemo Pause Checklist can reduce adverse events and strengthen the safety culture in the HD unit.