The modern haemodialysis factory: must quality improvement trump personalized care?

Haemodialysis (HD) patients expect and deserve sufficient access to nephrologists, personalized patient-directed care and effective provision of services, as our mentors and role models have taught us [ 1 ]. While this may be ideal, it may only be possible in relatively small HD facilities and in home peritoneal dialysis (PD) and HD programs. Most large, modern HD units will face enormous challenges in fulfilling this. Epidemic end-stage renal disease (ESRD) growth rates, in concert with a static number of trained nephrologists, result in an ever-increasing patient-to-physician ratio. Fortunately, technological improvements over the last 20 years have facilitated more efficient management, such that it takes less physician time per patient. However, the size of modern HD facilities often reaches 50 stations or more. This patient volume is outstripping the time efficiencies realized through technology, and causes the HD unit to become more like a factory and less like a health care clinic each year. The analogy of the HD unit as a factory might be offensive to some. We do not intend to promote the notion that our patient care processes are analogous to manufacturing widgets. Indeed, we would be delighted if it were possible to stabilize or reduce the patient-to-nephrologist ratio. Patient care should continue to be personalized, but it is just not possible to spend 10 min per patient when 50 or more patients are treated during a 4- or 5-h shift. In the face of these time constraints, perhaps we can improve high quality HD care by carefully considering the industrial factory analogy. The application of industrial models of quality control and continuous quality improvement (CQI) to our HD facilities is one strategy. Indeed, volume-driven processes of care in HD would seem to be amongst the most ideal health care models available to implement and study modified industrial solutions. Why look to industry for a new paradigm? Consider that serious error rates in ultra-safe industries like commercial aviation are one in a million [ 2 ], while more people in the USA die of medical errors than motor vehicle accidents, breast cancer or HIV [ 3 ]. CQI methods can and must be applied successfully in health care settings. Quality improvement in the nephrology world is at the toddler stage of development. For us to grow towards adolescence and mature adulthood requires a pragmatic consideration of our discipline's strengths and weaknesses. The Institute of Medicine has suggested the following changes to health care in the effort to shift to a culture of safety: the utilization of multidisciplinary teams, guidelines, performance and outcome measures to improve quality and accountability [ 4 ]. Importantly, these are already found within modern dialysis practice. However, notably lacking are suitable information systems (see below). Academic and clinical nephrologists from around the world have critically assessed the literature and have developed clinical practice guidelines concerning key topics. An initial phase of infatuation with guidelines has begun to wear off, as exemplified by the controversy surrounding the 2006 NKF/KDOQI anaemia guidelines [ 5 ], and a more realistic approach has begun to evolve. We practice in a data-impoverished area of medicine, where comprehensive evidence-based guidelines derived from randomized controlled clinical trials are not usually possible [ 6 ], but clinical practice recommendations, position statements and consensus statements built upon evidence and opinion can still be valuable. Within these documents, we implicitly recognize that there is a need to define clear targets that facilities should strive to reach. Whether the targets are ‘evidence-based guidelines’ or not, CQI requires targets, and data relative to targets, in order to be applied. It is now well known internationally that compared to the guideline targets, there are major gaps in achieving them. It is acknowledged that there are substantial variations in quality of care indicators between doctors within a facility, between facilities and between countries. It is desirable to reduce these variations and to improve performance so that more patients reach targets ( Figure 1 ). Finally, CQI works best when it focuses on the health care team and on improving the process and system of care. It does not work well by identifying individual poor performers and then naming, blaming and shaming them. Safety and quality should no longer be considered the responsibility of individuals; rather it must be built into all systems and processes of care. The goal of continuous quality improvement is to shift the mean of the performance curve towards higher quality of care and to reduce variability of care. The major limitation to the global application of CQI principles in HD facilities is the lack of electronic medical records (EMR) and suitable information systems. It is shocking that in 2007, there is no industry-wide standard information management tool, and that the health care industry spends so little money on information systems relative to other industries. Especially when individual facilities operate in isolation and face budgetary pressures to contain expenditures, there are no economies of scale, no surpluses and hence little opportunity to invest in the information technology required to do CQI in a timely manner. Often, busy clinical staff retrospectively perform time-consuming audits and quality assessment, by sifting through the ‘congestive chart failure’ of the paper record. A seamless, comprehensive, prospective EMR is an essential tool, and should be started in the early stages of chronic kidney disease. These data issues are now explicitly recognized in Europe [ 7 ]. A new initiative called Nephroquest, funded by the European Union, is addressing the problems related to data collection, standardization and benchmarking to measure, analyse and report quality indicators. This effort is to be applauded, and should become a model for non-Europeans to emulate. In 2007, most facilities track and review the common but unidimensional indicators of quality of care, such as dialysis dose (Kt/V), haemoglobin and central venous catheter (CVC) use. The facility medical record (paper and/or electronic) may allow for effective delivery of care at the patient level, but it often lacks the capability to facilitate timely and frequent grouping and analysis of data that would be required for facility level CQI. Many regional and national registries provide some useful information, but may report too late to be meaningful. These registries may focus on demographics, incidence and prevalence rather than on quality indicators. Ironically, some of this information gap, at least in the 12 participating countries, is filled by the International Dialysis Outcomes and Practice Patterns Study (DOPPS), and the investment and support of Arbor Research Collaborative for Health, Amgen and Kirin Pharma is most gratefully acknowledged [ 8 ]. However, many countries (like Canada) are left in a vulnerable situation such that without DOPPS, they would have very little idea about quality of HD care. The United States is a shining example of nationwide HD quality improvement, providing lessons to the world. Approximately 15 to 20 years ago, it was recognized that American death rates on HD were high, and that quality of care may have been sacrificed under the dominant, for-profit model of care [ 9 ]. Today, quality of care information flows to and from payers [ 10 ], networks, chains and to the superb USRDS registry [ 11 ]. The 18 ESRD networks have a mandate to monitor quality of care and to improve it. Detailed reports are sent annually to all US dialysis facilities to describe their practices and outcomes compared to regional and national reference data, while adjusting for differences in case mix. Even patients (and the press) can view key indicators of practice and outcomes for each US facility at www.cms.hhs.gov/DialysisFacilityCompare. Profound improvements in many quality indicators have already been documented [ 10,11 ]. While much can be said for and against for-profit versus not-for-profit models of care, competition and consolidation of the US market into large chains and public scrutiny has made it in the self-interest of all providers to improve quality of care. Ironically, in countries like Canada where there is more government control and limited funding, there is no incentive to publicize improvements in quality, since that would attract more patients and increase costs. Recently, a new way to measure quality in HD units has emerged. The DaVita dialysis chain in the United States has developed a dialysis quality index (DQI) that it uses to promote CQI across its facilities [ 12,13 ]. This aggregate index includes seven variables that are weighted and then summed to create a 100-point scale. Facilities are tracked on a quarterly basis, so that both internal and chain-wide comparisons can be made. Their preliminary data show that improvements in the DQI are associated with a decrease of standardized mortality rates. Furthermore, the DQI allows for the identification of poorly performing facilities that require external assistance with quality improvement. It also creates a ‘culture of quality’ that serves as a basis for pride and rewards for a facility's quality improvement activities. The former Renal Care Group developed a similar index that was based on 10 clinical quality indicators and patient satisfaction measures, aggregated to derive a Composite Facility Score (R. Hakim, personal communication). Similarly, a group of Canadian nephrologists have been working with researchers from DOPPS to develop an outcome-driven rather than opinion-based aggregate HD quality index that has been named the Practice Risk Score (PRS) [ 14 ]. The index includes four variables: Kt/V, haemoglobin, albumin and CVC utilization. Efforts to validate and improve it are ongoing. Other elements that might be added to get a more complete portrait of the multidimensional performance of an HD facility include patient and staff satisfaction variables [ 15,16 ]. This evolution from a unidimensional series of performance measures, to a more comprehensive aggregate index that inspires a culture of quality within and across facilities, would seem to be a positive step forward. Sadly, a tension exists between the empowerment of the individual doctor as an autonomous expert provider of care, empowered within the doctor–patient relationship, and the clear imperative to improve HD care systems and processes in a factory-like manner [ 1 ]. For example, protocols and algorithms have the potential to reduce variation in care and to improve performance relative to quality indicators. However, they can undermine physician independence and can even be delivered, in part, without physician input at all. Reconciling these two paradigms will be a challenge, but must be faced by the nephrology community through the prism of focusing on optimizing patient care. It is quite possible that a validated patient satisfaction instrument could be used to bridge that uncomfortable tension between personalized care and care based on systems and CQI processes. Unfortunately, there have been very few published reports of potentially valid patient satisfaction tools for dialysis populations [ 15 , 17–19 ]. Accordingly, evidence-based patient satisfaction tools have not consistently been used in CQI processes. It would be of major interest to determine empirically if there are associations between personalized care and patient satisfaction, and whether either of these is associated with hard outcomes. If they do correlate, then incorporating patient satisfaction into an aggregate quality index would be an essential new tool that will ensure preservation of valued personal aspects of care and caring to the appropriate extent. If not, then the path that we need to follow becomes nonetheless clearer. The authors wish to thank Dr Sandra Donnelly, who coined the term ‘congestive chart failure’ and allowed us to use it here, and Dr Fritz Port for his helpful review and suggestions. Conflict of interest statement. None declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.