Minding the gap and overlap: a literature review of fragmentation of primary care for chronic dialysis patients

Overall and disease-specific management and care coordination is a challenge for patients with complex and multiple chronic conditions (MCC), who typically see multiple providers to manage one or more co-morbid conditions. This problem is amplified for patients with kidney disease, whose condition is associated and co-morbid with highly prevalent conditions such as hypertension, diabetes, and heart disease. Much of the literature and intervention efforts have focused on primary care and nephrology physician collaborative care in the early stages of chronic kidney disease (CKD), with an emphasis on primary care physician (PCP) intervention to prevent, detect, treat, and slow the progression of kidney disease. Studies have generally found lack of clarity on the roles and responsibilities of clinical management among providers [1], poor communication and coordination between primary care physicians and nephrologists [2] and suboptimal general and CKD-related care in patients [3, 4].

However, less is known about primary care physician (PCP) and nephrologist co-management of patients in the end-stage of renal disease. Specifically, most patients with advanced kidney failure undergo chronic dialysis treatments and are clinically managed by clinical staff at dialysis facilities on a frequent basis (e.g., thrice weekly for in-center hemodialysis or monthly for home-based dialysis). Due in part to the rigidity of the dialysis treatment schedule, management of co-morbid illness and their complications by providers outside of the dialysis unit is challenging. Patients commonly receive supervision of non-renal health needs during their dialysis treatment visits due to convenience and familiarity. As a result, primary care management may default to renal providers. Compared to early-stage CKD care, this observed focus – of nephrologist involvement in treating primary care needs of patients with end-stage renal disease (ESRD) – raises important questions about the role of the PCP, traditionally considered the patient’s medical home for continuous comprehensive care and the “quarterback” responsible for assessing, balancing, and coordinating the care of patients’ multiple competing conditions [5, 6]. Who is responsible for and actively managing the primary care needs of these dialysis patients? Discrepant expectations and subsequent provision of primary care may exacerbate care fragmentation in this highly complex and vulnerable patient population, increasing the potential for unnecessary duplication of care or adverse outcomes. To date, PCP and nephrology roles in the actual provision of primary care services for chronic dialysis patients is not well-defined. We conducted a systematic review of primary care service provision for dialysis patients, in order to assess 1) patient and provider perceptions of PCPs and nephrologists roles; 2) the extent to which PCPs and nephrologists deliver primary care services to chronic dialysis patients; 3) reported barriers to patients’ receipt or physicians’ delivery of primary care services; and 4) the measures used to assess provider provision and primary care outcomes in dialysis patients.

Understanding the provision of primary care among nephrologists and PCPs, fragmentation of care, and outcomes in dialysis patients is an emerging and important concern [7]. The anticipated growth and surveillance of patients with CKD [8] and the diminishing supply of practicing nephrologists [9, 10] suggests that nephrologists may be unable to continue or increase their provision of primary care services to dialysis patients, and it remains to be seen if nephrologists’ provision of primary care services is adequate or appropriate. In addition, recent efforts to organize chronic disease management, such as multidisciplinary care teams and the patient-centered medical home, commonly exclude patients with ESRD who may also benefit from models designed to mitigate care fragmentation and improve coordination. Elucidating the types of services and providers managing dialysis patients’ primary care needs and their effectiveness will illuminate the gaps in our understanding of primary care and nephrologist collaboration to guide further research and intervention to improve care for this overlooked, high healthcare utilizing, co-morbid dialysis patient population.

This literature review is guided by conceptual frameworks for care provided to individuals with MCC [11] and interprofessional and co-managed care [12]. Together, these frameworks illustrate the complexity of care for patients receiving maintenance dialysis and is a valuable lens to frame our understanding of how their specialized and general healthcare needs are met. Briefly, the National Quality Forum’s model for MCC care recognizes the various ways in which patient preferences, care settings and providers, and types of healthcare services interact to impact health outcomes (e.g., care coordination, prevention of disease, cost). The interaction between these domains of care is dynamic, as patients’ healthcare needs evolve over time [11]. Our review of the empirical literature focuses on the domains regarding dialysis patients’ providers, their use, and measures used to assess their primary care. We also incorporate perspectives from Retchin’s collaborative model of interprofessional and co-managed care, which posits that interaction and coordination between MCC provider types is influenced by temporality (e.g., concurrent vs. sequential care), urgency, and delineation of authority in patient care [12]. To this end, our review also considers the evidence on perceptions regarding provider roles and skills in primary care delivery.

In this review, we found a limited evidence base to inform our understanding of primary care services for chronic dialysis patients. From our systematic search of the healthcare literature, we found 14 studies spanning approximately three decades of research. Overall, we identified three themes related to primary care of dialysis patients: perceived roles of providers, estimated time in providing primary care, and the extent of dialysis patients’ use of primary care services (e.g., referrals, prevention, acute illness and disease management).

Among the findings in this literature, we found overall general agreement among providers that PCPs should be involved in primary care of dialysis patients. Despite this agreement, a potential barrier to realizing this may be the adequacy of time for PCPs to provide good primary care to dialysis patients and subsequent incongruent expectation and coordination between provider specialties. At the same time, the logistical barriers to accessing care and time burden of dialysis care for patients are a significant factor, which may also attribute to the majority of dialysis patients relying on primary care services from their nephrologists, despite their report of greater confidence in primary care delivered by non-nephrologist PCPs. These findings generally apply to all dialysis patients, as found in the few studies that explicitly assessed differences between in-center hemodialysis and home-based peritoneal dialysis patients. Although more PD than HD patients report having family practitioner [19], PD patients were less likely to see them over time [19] and have similarly low vaccination rates as HD patients [21]. With increasing use of home-based dialysis [8], future research should explore differences in patient care perceptions and utilization by treatment modality.

Across studies reported from a variety of patient and provider perspectives, we found varying degrees of consistencies of primary care service delivery and use with important implications on patient outcomes. On one hand, consistent reporting of delivery and receipt of primary care services may reflect accurate reporting across survey respondent types. On the other hand, high provision of care reported by nephrologists and PCPs may also suggest duplicative care. The pooled findings related to overall delivery and receipt of preventive care suggest potential gaps in care, where studies of administrative data found lower rates of preventive care services than found in studies relying on provider or patient self-report. The mechanisms underlying these inconsistent findings is unclear: this may be due to recall bias in self-reported data and potentially incomplete documentation of care in administrative data, but may also highlight discordant expectations and communication between patients, nephrologists, and PCPs that lead to gaps in care. The difficulty in communication between providers [17] is also exacerbated by a lack of streamlined electronic health records between health systems and dialysis units, resulting in a figurative chasm of information exchange. Further, the included studies were limited in their ability to assess the contribution of patient non-adherence to provider recommendations for preventive care, which may reflect patient behavior as a barrier to receipt of primary care services rather than inappropriate, duplicative, or missing care patterns by providers.

Although providers generally agreed that PCPs should be involved in primary care of dialysis patients, the optimal extent of this involvement is unclear. For example, many patients report having a non-nephrologist PCP, but simultaneously report reliance on nephrologists for referrals, counseling, preventive care, symptom and disease management; services that traditionally fall under the expertise of a PCP in the non-dialysis setting. There are no formal guidelines on the provision of primary care for individuals with ESRD, and the clinical benefit and cost-effectiveness of traditional health maintenance in this high-risk population remains controversial [28, 29]. While recent literature and consensus statements support reductions in cost-ineffective use of cancer screening in dialysis patients [30‐32], the extent to which PCPs and nephrology providers are aware and/or agree with these recommendations is unclear.

Altogether, these findings highlight opportunities for improvement and have important implications on the amount and quality of primary care that dialysis patients receive and on healthcare policy, as depicted in Fig. 3. This figure illustrates the Swiss Cheese Model of accident causation, used in engineering safety and healthcare and described here in the context of provider care and coordination of dialysis patients. Although dialysis patients likely see more providers than depicted here, our example presents a simple model of just PCPs and nephrologists. Each slice represents care delivered by a provider, where the holes reflect unprovided care. Patient care is completed neglected when holes between the slices of cheese align. For patient care, there is potential overlap in care that may be unnecessary and duplicative, placing time burdens on patients and their families and providers and financial burden on the healthcare system. In addition, the pooled, complementary findings suggest the possibility of a distinction in roles and responsibilities for some primary care services that result in non-duplicative, singular care. Importantly, this evidence base suggests the greater likelihood of gaps in care, leaving missed opportunities for clinical intervention and increasing the risk of adverse events, exacerbation of symptoms and disease, and increased healthcare utilization and costs.

While this is an emerging area of concern, it is useful to consider its importance and the findings from the extant literature in the context of changes in the healthcare delivery system for ESRD patients. Aided by improving survival rates, the US is experiencing growing ESRD prevalence and a shrinking nephrology and primary care workforce. Recent system-level interventions, such as Medicare’s managed care program demonstration for ESRD patients and the Comprehensive ESRD Care model aim to reorganize ESRD patient care through the collaboration of dialysis clinics, nephrologists and other US providers. This latter effort has resulted in Medicare-funded ESRD Seamless Care Organizations (ESCO) to improve care coordination and outcomes for Medicare beneficiaries. Thus, the research synthesized here foreshadows the challenges and opportunities for ESCOs to retool structures and processes of care across provider types to improve clinical and logistical complexity of care for patients on chronic dialysis.

Future research will want to examine the extent to which singular care is coordinated and communicated across patients, nephrologists and PCPs, as well as ways to optimize such multidisciplinary care for dialysis patients and improve outcomes. The extant literature presented here reveal gaps and opportunities for future research to augment and improve the evidence base. First, there is a paucity of claims-based research. The differential findings of studies based on administrative data versus individual self-reported observational studies, suggest that the truth is probably somewhere in between. Well-designed mixed methods research can examine the impacts of structures and processes of care coordination in a more nuanced way than quantitatively or qualitatively-based studies alone. Second, the studies examined in this review examined a limited set of outcomes in dialysis patients’ primary care. More studies linking the relationship between providers and patients’ use and receipt of primary care to outcomes (e.g., physical function, morbidity, hospitalization, mortality) will serve as important comparisons to the nascent ESCOs, whose value is yet to be determined.