Hospital-based care of congenital heart disease patients continues to evolve to meet the increasing demands of volume and acuity. In particular, coordination of complex care systems, prioritisation of efficient resource utilization, and re-evaluation of traditional care delivery systems, including provider team composition, have become common considerations of modern practice. Reference Hoerst, Bakar and Cassidy1,Reference Mott, Neish, Challman and Feltes2
Advanced practice providers, including nurse practitioners and physician assistants, were first employed to address the shortage of paediatric rural health providers in 1965. Reference O’Brien3 Neonatal nurse practitioners, employed in the 1970s, were the first group of nurse practitioners to work in critical care. Reference O’Brien3 Resident work-hour restrictions and a decrease in subspecialty training hours have more recently increased the utilisation of advanced practice providers in subspecialty and critical care areas within paediatric hospital systems. Reference O’Brien3,Reference Freed, Dunham, Moran and Spera4 Several examples exist demonstrating the utility of the advanced practice provider care model, including adult cardiology practices which have shown an improvement in access to care, productivity, and efficiency. Reference Rodgers, Conti and Feinstein5,Reference Lambrew, Dove and Friday6 Implementation of advanced practice providers in both emergency medicine and intensive care units has been associated with improved patient outcomes. Reference Woo, Lee and Tam7,Reference Schulman, Lucchese and Sullivan8 Similarly, Newhouse and Stanik-Hutt performed an 18-year systematic review of advanced practice provider-patient care involvement and found that patient outcomes were similar, if not better, than when physicians alone were responsible. Reference Newhouse, Stanik-Hutt and White9 Lastly, nurses have acknowledged the value of advanced practice providers in regard to assessing the needs of patients and their families, communicating with multidisciplinary teams, and acting as a resource for their nursing colleagues. Reference Knaus, Felten, Burton, Fobes and Davis10
The variable composition of the paediatric acute care cardiology team (acute care should be interpreted as synonymous with inpatient or ward-based care) has only recently been described, and to date, analysis of the impact of advanced practice provider presence on practice patterns or outcomes has not been performed. Reference Hoerst, Bakar and Cassidy1,Reference Mott, Neish, Challman and Feltes2 Our study aims to evaluate the role of advanced practice providers as part of the care team in the emerging and distinct subspecialty of paediatric acute care cardiology utilising the Paediatric Acute Care Cardiology Collaborative (PAC3) as a data resource. PAC3 was founded in 2014 with the goal of improving outcomes in the paediatric acute care cardiology population, and with this aim in mind, member sites recently conducted a hospital survey to assess practice variation and acute care cardiology unit team composition. Reference Hoerst, Bakar and Cassidy1,Reference Kipps, Cassidy and Strohacker11 This hospital survey developed an inventory of the specific team members and their respective roles in the acute care cardiology unit and provides an opportunity to measure the associations between the advanced practice provider service lines and the desired patient outcomes affecting this discrete population.
Materials and methods
PAC3 is a voluntary quality improvement organisation comprised of 37 institutions primarily focused on improving acute care cardiology outcomes delivered on the acute care cardiology unit. Reference Kipps, Cassidy and Strohacker11 In 2017, a PAC3 member survey examined practice and environment variation across participating institutions. Data elements included hospital demographics, staffing, available resources and therapies, use of standard care practices, and patient transfer and discharge practices. The survey and its results have been reported previously. Reference Hoerst, Bakar and Cassidy1 The survey included 12 stem questions pertaining to staffing by advanced practice providers, cardiology fellows, and residents, with numerous branching questions following positive responses. The survey data was collected and managed using Research Electronic Data Capture (REDCap) tools hosted at the Cincinnati Children’s Hospital Medical Center. REDCap is a secure, web-based application designed to support data capture for research studies. Reference Harris, Taylor, Thielke, Payne, Gonzalez and Conde12 The study was approved by the Institutional Review Board at the Cincinnati Children’s Hospital Medical Center.
Surgical volume per institution was obtained utilising The Society of Thoracic Surgeons (STS) database, and was previously found to correlate with unit size and census. Reference Hoerst, Bakar and Cassidy1 A resource acuity score of 0–4 was derived for each unit, calculated as the sum of up to four potential therapies utilised on the acute care cardiology unit (each deemed by expert opinion amongst PAC3 clinical champions to be uniquely representative of elevated acuity): inhaled nitric oxide, high-flow nasal cannula, ventricular assist devices, and intravenous antiarrhythmic agents. Pearson’s correlation coefficients were used to evaluate the relationship between staffing models across the collaborative and centre volume, as well as unit acuity.
Eighteen PAC3 survey sites also submitted data to the Paediatric Cardiac Critical Care Consortium (PC4) database, which was used to determine post-operative length of stay analysis. PC4 is also a voluntary quality improvement organisation comprised of nearly 50 institutions primarily focused on improving cardiac critical care. All cardiac intensive care unit encounters at each participating centre are submitted to the PC4 database. Each case record includes patient demographics, clinical characteristics, data on cardiac surgical procedures and other interventions that occurred, critical care therapies, and complications, all with standard definitions. The registry shares common terminology and definitions with the International Paediatric and Congenital Cardiac Code, The Society of Thoracic Surgeons Congenital Heart Surgery Database, and the American College of Cardiology Improving Paediatric and Adult Congenital Treatment Registry, as previously described. Reference Gaies, Cooper and Tabbutt13,Reference Franklin, Beland and Colan14 Trained data managers who have passed a certification examination abstract the data. The PC4 data auditing process has been previously described, with the most recent audit results suggesting the complete, accurate, and timely submission of data across centres, and a major discrepancy rate of 0.6% across 29,476 fields. Reference Gaies, Donohue and Willis15 The University of Michigan Institutional Review Board oversees the PC4 Data Coordinating Center and this study was approved with a waiver of informed consent.
A previous study has successfully combined hospital-level factors from the PAC3 survey with PC4 outcomes data with specific attention to post-operative length of stay. Reference Hart, Tanel and Kipps16 In summary, a case-mix adjustment model was developed for the total post-operative length of stay based on pre-operative and operative variables. The PAC3 survey data was categorised into predictor variables felt to impact post-operative length of stay. For our analysis, the percentage of patients covered by advanced practice providers on the acute care cardiology unit was categorised as low (<50%), moderate (51–75%), and high (>76%). These categories were selected based on the distribution of responses. Candidate variables were included in the final model if their association with total post-operative length of stay had a p-value less than 0.05. A complete list of categorised survey responses used in the multivariate analysis are provided in Supplementary Table S1. A multinomial regression model was then used to evaluate the effect of hospital-specific survey variables with an observed-to-expected post-operative length of stay at the patient level defined as a categorical variable: ≥25% shorter-than-expected post-operative length of stay, as-expected, or ≥25% longer-than-expected post-operative length of stay. Standard error adjustment for centre clustering was not performed since the survey variables for all patients at a particular hospital were the same.
All analyses were performed using SAS Version 9.4 (SAS Institute, Cary, NC) or STATA Version 14 (Stata Corp, College Station, TX).
Results
PAC3 surveys were completed by 31 (91%) of the 34 member centres at the time of survey distribution. Nearly all centres (94%) employ advanced practice providers, with a mean of 1.7 (range 0–5) advanced practice providers present on weekday rounds. These advanced practice providers cover ≤50% (n = 6), 50–75% (n = 6), and >75% (n = 6) of acute care cardiology unit patients. Advanced practice providers provide weekend coverage in most centres but less often provide overnight coverage (Table 1). Most (90%) centres have a cardiology fellow on the acute care cardiology unit team, but fellows are often involved with other responsibilities, including weekday consults (58%), night and weekend consults (94%), and performing after-hours echocardiograms (61%). Paediatric residents are involved with the care of acute care cardiology unit patients at 81% of centres, with a mean of 2 (range 0–5) residents present on weekday rounds. Only half of the 31 centres have paediatric interns (first-year residents) on the care team, and a majority have a junior or senior resident present. Residents provide weekend coverage as often as advanced practice providers (77%), but provide more frequent overnight coverage (71%) than advanced practice provider colleagues (26%).
Table 1. Provider type present in the acute care cardiology unit.
*Paediatric interns present at 15 centres (48%), 12 of which had supervising residents (80%).
Unit bed size, as well as surgical volume obtained from STS and the calculated acute care cardiology unit resource acuity score were assessed in relation to the categories of advanced practice provider practice (Table 2). The number of advanced practice providers present on weekday rounds has a positive correlation with surgical volume (p < 0.01, Table 2). The number of advanced practice providers present on weekday rounds also had a positive correlation with the calculated unit acuity score (p = 0.03), but not unit bed size.
Table 2. Advanced practice provider coverage and correlation with surgical volume and acuity.
Abbreviations: ACCU = acute care cardiology unit; APP = advanced practice provider; r = correlation coefficient.
At the time of the survey, there were 18 hospitals that provided both PAC3 survey data and PC4 clinical registry data. A total of 20,935 clinical encounters from PC4 were included in the development of the case-mix adjustment model for total post-operative length of stay. In the multivariate model (Table 3), as coverage by advanced practice providers increased from low to moderate or high, there was a greater likelihood of having a shorter-than-expected post-operative length of stay (p < 0.001).
Table 3. Multivariable analysis: association of patient coverage by advanced practice providers with post-operative length of stay.
*Relative risk ratio as compared to ≥25% longer-than-expected LOS.
Discussion
Utilising data from the North American inpatient collaboratives (PAC3 and PC4), we have demonstrated that an increased percentage of care provided by advanced practice providers on the acute care cardiology unit is associated with having a shorter-than-expected post-operative length of stay, as well as a capacity to provide care utilising more complex mechanisms as required by the higher acuity patient. Specifically, it is an increased overall presence of advanced practice providers that was associated with these benefits rather than the mere presence of advanced practice providers or a number of patients covered by advanced practice providers.
This study is the first to describe in detail the role of various providers in the paediatric acute care cardiology environment. Certainly, there are the traditional members of the care team, including interns, residents, cardiology fellows, and cardiology attending faculty. However, there is a significant variation amongst PAC3 programmes, partly owing to the incorporation of advanced practice providers. Advanced practice providers have been contributing to inpatient care on many paediatric services over the last several decades, but their presence in the acute care cardiology unit has not been previously described to this extent. Reference Mott, Neish, Challman and Feltes2,Reference O’Brien3
The centres that responded to the PAC3 hospital survey represent diversity in geography, unit structure, and surgical volume. These factors underlie some of the diversity that was observed in the responses of the survey. A notable finding in this analysis was that the number of advanced practice providers employed by a centre correlated with surgical volume and a higher level of acuity handled by a unit, based on four representative advanced care therapies. Future analysis with additional contributing centres may help to elaborate on these findings.
The changing structure of the acute care cardiology team is consistent with some of the changes in health care delivery over recent years, but is also objectively validated by the results of this survey. Other subspecialties in paediatrics have utilised advanced practice providers for decades. For example, advanced practice providers have been central members of the neonatal intensive care unit team since the 1970s. Reference O’Brien3 We show that many centres seem to be appreciating the dedication and expertise of this provider group, as well as their measurable direct patient benefit.
As noted in this survey, nearly half of centres have advanced practice providers covering greater than 75% of acute care cardiology unit patients. This shift in acute care cardiology unit staffing with advanced practice providers involved with the majority of care, rather than the traditional team consisting of paediatric interns and residents, recognises that specialised training and experience is necessary to care for complex paediatric cardiology patients. The seasoned advanced practice provider undoubtedly has the opportunity to develop greater familiarity and knowledge of the nuanced diagnoses and physiology of congenital heart disease than the rotational resident. In addition, advanced practice providers who are dedicated to the care of the acute care cardiology patients can be responsible when cardiology fellows are attending to tasks that take them off of the unit. We have observed that cardiology fellows are often also responsible for cardiology consultations, performing urgent echocardiograms, and addressing other acute cardiac issues in other areas of the hospital while on service or on call, potentially leaving a void of expertise on the acute care cardiology unit.
There are some important limitations to consider in regard to this study. The PAC3 hospital survey provides a look into the unit structure and practice patterns present at each of the representative PAC3 member institutions, and was intentionally administered before the beginning of acute care cardiology unit-specific registry data collection. As a result, there is a limited amount of granular acute care cardiology outcome data to examine at this time. However, our partnership with PC4 allowed for a robust analysis of a critical variable of interest, post-operative hospital length of stay. In addition, the model to determine higher acuity forms of acute care cardiology unit care is built on the consensus of experts and is not a validated tool at this time. The aim is to refine the PAC3 models of case-mix analysis over time as the data from the PAC3 acute care cardiology unit-specific registry becomes available. Lastly, given the retrospective design of this study, the length of stay finding could be explained by other unmeasured factors that confounded the exposure of interest (the increasing presence of advanced practice providers). Further studies are necessary to confirm the association described here, and if confirmed, to then determine the precise nature of how advanced practice providers might facilitate shorter-than-expected post-operative length of stay.
The results of this study support the notion that acute care cardiology is a distinct subspecialty within paediatric cardiology by demonstrating that a dedicated front-line workforce is associated with improved outcomes. This work is central to the mission of PAC3 and would not have been possible without the collaborative efforts of all participating centres. Reference Kipps, Cassidy and Strohacker11 Future improvement efforts and research amongst PAC3 centres will undoubtedly lead to improved outcomes and patient experiences. This study also demonstrates substantial variation between centres in front-line providers. This variation may provide important insights into how advanced practice providers impact patient care and outcomes in future studies.
Acknowledgements
None.
Financial support
The authors have indicated they have no financial relationships relevant to this article to disclose. No external funding for this manuscript was received.
Conflicts of interest
None.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1047951120003789
