Single center experience developing sustainable genetics clinical care: a model to address workforce challenges in medical genetics

Abstract

Purpose of review

The national workforce shortage in genetics is being evaluated in order to identify a sustainable solution to the increasing demand for genomic services. An innovative solution to the short term needs is to integrate advanced practice providers (APPs) and embed genetic counselors into both outpatient and inpatient specialty care. Incorporating APPs into a genetic service is not unique in itself, but the method of implementation at Medical College of Wisconsin (MCW) was, at the time, unchartered.

Recent findings

There are >100 vacancies for board certified medical geneticists across the nation, training programs are not enrolling sufficient trainees to meet demand and more than a third of the current workforce plan to retire within the next 10 years.

Summary

The integration of advanced practice providers (nurse practitioners, midwives, physician assistants etc.) into both primary and specialty care has been an evolving practice since the mid-1900s and incorporating APPs into a genetic service was not unique in itself but the method of implementation was new at that time. This is a model to successfully develop a clinical practice model around a team-based structure incorporating nurse clinicians, advanced practice providers, genetic counselors, nutritionists, and physicians into an academic clinical genetics practice.

INTRODUCTION

The general state of clinical genetics practice, and delivery of medical genetics services across the United States is challenged by a limited workforce. This national concern is well recognized and coordinated workgroups from several representative organizations are evaluating various solutions to an ever-expanding problem. A 2020 GAO (government accountability office) report estimated an average of two medical geneticists per 500 000 people but the actual numbers varied by state and geographic location [1]. There is no universal standard for comparison but these numbers are significantly lower than any other specialty surveyed over the past three decades [2].

Part of the challenge in estimating need is the lack of adequate data: “It is difficult to identify changes in the size of the medical geneticist workforce over time, because there are limited data. Due to the small size of the workforce, BLS and HRSA do not collect data on the size and growth of this workforce” (GAO report [1]).

The most recent workforce data from the American Board of Genetics and Genomics, published in 2021, assessed the state of needs across the nation as of 2019 [3^▪]. This data aggregated responses from 491 medical genetics diplomates. They reported an average work week of 50 h and 58% of respondents spent more than half of their time in clinical duties. An average of 13 new and 10 follow-up patient visits per week were reported. Wait times were aggregated at three days or more for new emergency referrals, and the majority of patients wait more than three months for an appointment. Internal communication within the directors workgroup report having extended wait times of >6 months for new appointments (personal communication – unpublished data). Many respondents reported faculty vacancies at their institutions that had been recruiting for >3 years. A key metric identified in this study was that the average age of respondents was 51.4 years and that 36% had been practicing for >21 years with almost a quarter planning to retire within 10 years of the date of the survey. Prior workforce studies highlight the variability amongst practices with average work weeks ranging from 52 to 60 h/week, clinic visit times ranging from 60 to 120 min for new patients and 30 to 60 min for follow up and an average of 570 patients seen annually for a full time full-time equivalent (FTE) [4–7] (further details of these studies in Appendix 2) The extended wait times for patient appointments emphasize the limited access to medical geneticists. The pending impact of an aging workforce with insufficient trainees to meet the practice needs as well as the growing demands for genetic experts as genomic technologies impact all fields of medical practice is not limited to the United States. There have been several editorials highlighting this global phenomenon, and coordinated efforts are underway to limit the effect on patient care [3^▪,6–9]. 

Box 1.

no caption available

“Necessity is the mother of invention”

Our clinic story, not unlike others, begins with untimely attrition in both its physician and genetic counselor workforce in 2015. Faced with greater than a 50% reduction of total workforce, clinic wait times extended to 18 months and the waitlist grew to more than 1000 patients. These dire circumstances forced an internal analysis and inspired a complete revision of the clinical genetics workflows for the practice needs, and it was imperative to establish a sustainable implementation for integrating genomic medicine at a systems level.

The division of genetics at the Medical College of Wisconsin/Children's Wisconsin (MCW/CW) had successfully incorporated a Nurse Practitioner into clinical workflows in 2009 and a decision was made to rebuild the clinical practice model around a team-based structure incorporating nurse clinicians, Advanced practice providers (APPs), genetic counselors (GCs), nutritionists, and physicians. Incorporating APPs into a genetic service was not unique in itself but the method of implementation was, at that time, unchartered. A thorough workforce analysis was undertaken through the combined efforts of the hospital and Department of Pediatrics administration. A short-term cost neutral plan was developed and a longer vision was proposed for implementing genetic services across the campus and hospital system.

Principal genetics clinic short term “rescue plan”

The practice model prior to this climactic change had revolved around the classic structure in which the full time medical geneticist/genetic counselor dyad would see an average of 14 patients weekly, 2/3 new and 1/3 follow up with appointment times ranging from 90 to 120 min. This required a significant amount of time on patient related activities, including tracking down records, patient care coordination, insurance preauthorization and all the other time-consuming tasks which reduce overall efficiencies in a clinical genetics practice. A noted negative consequence of this model was the significant amount of time spent on non-face-to-face and nonbillable activity from a hospital system perspective. To adjust this model, two of the senior geneticist positions (vacated and unfilled) were replaced with four APP positions for an overall cost neutral proposition for the department. The service was additionally supported by two senior nurse clinicians whose experience in patient care coordination and hospital systems navigation enabled the GCs to redefine their clinical work role to more appropriately align with their scope of practice. The calculated workforce transition equated to 4.6 combined clinical FTE (cFTE) with a patient volume of 2300 for a 46 week work year prior to the transition and 7.1 cFTE with a projected patient volume of 3900 when fully staffed. Figure 1 captures this transitional step.

FIGURE 1.

Representation of classic genetic clinical dyad (a) with geneticist working closely with a genetic counselor. Proposed model for an integrated team based model (b) creating working collaborative partnerships between APPs and physicians, supported by nurse clinicians. The model change predicted a change in patient volumes from 2300 patients with 4.6 physician cFTE to an anticipated 3900 patients with 7.1 provider cFTE. APP, advanced practice provider.

Additional challenges addressed due to the limited workforce related to the acute medical management requirements of the inpatient service and metabolic newborn screening program. The genetics practice at Children's Wisconsin admitted and directly managed the patients on the inpatient wards. An agreement was undertaken with the hospitalist service that they would serve as primary attendings for all admissions with the genetics physicians serving on a consultative basis. Several education seminars were held with the hospitalist team to provide a foundation for metabolic disorders and detailed management guidelines were developed for commonly encountered disorders. This relieved a significant burden on the on-call genetics service and improved patient safety during admissions. In addition to this collaborative agreement, several tiered evaluation guidelines were developed to assist the frontline hospital team to initiate diagnostic workups in patients who fell into a predefined cohort. These included children admitted for: developmental delays, seizures, congenital cardiac defects and failure to thrive. These guidelines were additionally implemented for nonurgent referrals from primary care. Overall the inpatient and outpatient process were aimed at improving time efficiencies for the patient and available geneticist. A full-time inpatient GC was established to facilitate the service and support the frontline genetic test orders and results delivery with the Geneticist attending available as needed. APPs were not initially integrated into the inpatient service, but as our overall systems integration of genomic services has evolved, there are now two APPs and four GCs that support the inpatient service with the collaboration of a single on call Medical Geneticist. Hospital administrative support through the financial clearance service was initiated for preauthorization of genetic tests and expanded later as part of the laboratory utilization initiative. The newborn screening program remained the primary responsibility of the medical geneticist but additional members (nurse clinicians, GC, APP, nutritionist) were brought into the team for all follow up monitoring and care. Comprehensive disorder and nutrition guides were developed to aid management recommendations. These processes were not designed to remove physician responsibility but to enable shared coordination of care, thus relieving overall burden of care from a single provider.

IMPLEMENTATION

Advanced practice provider workforce

The role of the Nurse Practitioner evolved in 1965 when Dr Loretta Ford and Dr Henry Silver established the first training program at the University of Colorado [10]. Dr Silver went on to create a new health professional when he established the physician assistant training program in 1969. Nurse practitioners and physician assistants form the majority of APPs in clinical practice today. Genetic Counselors would naturally fall into this professional body except that many states do not recognize professional licensure to practice and CMS (Centers for Medicare & Medicaid Services) has not recognized GCs as billing providers at the time of writing.

Despite the importance of genomic medicine, most practicing APPs receive little to no formal training in genetics. In a response to the Institute of Medicine 2011 report, many of the governing accreditation bodies for nursing and nurse practitioners implemented core competencies in genetics and genomics as a standard curriculum [10,11,12^▪,13–15]. A review of PA training programs identified that less than half of the surveyed programs included as little as 1 to 10 contact hours of genetics throughout the training syllabus, the vast majority of this being didactic in nature [16]. The challenge remains that there is a large body of APPs who have not had an opportunity for primary training in genetics, let alone training required to function as an effective provider in a clinical genetics practice.

Implementation of an advanced practice provider model

Careful thought and planning was required in order to successfully implement such a significant clinical change within the context of a limited trained workforce. Selection of the APPs was a critical component of this process. Individuals who were clearly committed to a career in genetics and were prepared to apply themselves to an intense training period were selected from the candidate pool of applicants. A condensed “resident competency training” model was put into place with a combination of in person didactics, online self-study, one on one mentoring through a paired GC-APP tutor/learner dyad, and hands on training overseen by the attending physician. Clear clinical work expectations were established (Table 1) in addition to core competencies which were modeled after the Medical Genetics Residency training and the guidelines on G2C2 (Genetics and Genomics Competency Center – this resource is no longer available, outlines detailed in Appendix 1, [14,15,17–20]), but the focus was placed on collaborative patient care instead of independent practice. A time efficient clinical work flow was established to optimize collaborative decision making (Fig. 2). Clinic visits graduated from job shadow, followed by supervised visits until competencies were attained at which point a physician was available for every visit to advise as needed. This need was determined by either the GC or APP who worked as a collaborative team as well as competency review from personal discussions and asynchronous chart review. This model required a high degree of personal time investment from all members of the team as well as a supportive, collegial work environment. Key professional workshops and team building exercises were built into the development of this model as it was recognized that this change in the work environment required additional support. The GC-APP paired mentorship model continued through the first year of practice, similar to our practice of supporting new GC graduates through their first year in the workplace.

Table 1.

Clinical work expectations for full time faculty and advanced practice providers

Clinical work week
Each day divided into two clinic blocks a.m./p.m. on M/Tu/W & a.m. only F. Thursday is an administrative day
Outpatient
- MD: 1.0 FTE = 4/7 clinic blocks with four patients per block (RVU target of 2240)
- APP: 1.0 FTE = 5/7 clinic blocks with three patients per block (RVU not currently in use)
Inpatient
- MD cover one week at a time, shared equally based on total FTE
- 2 APPs share cover for 1.0 FTE – take first pager call during work hours (8 a.m.–4:30 p.m.) and staff all consults
- 1½ GC FTE (shared by 3 GCs) – a subset of patients are GC only

APP, advanced practice provider; GC, genetic counselor.

FIGURE 2.

The clinic work flow for new patients was established in such a way as to allow time for diagnostic consideration during the patient encounter. The APP was able to consult with the MD or utilize decision support tools while the GC was obtaining the family heath history and together, a management plan could be formed to optimize the patient experience and limit opportunity for medical error. APP, advanced practice provider; GC, genetic counselor.

Additional considerations for success

Given that this new workflow was a significant departure from the historical norm, several change management strategies had to be implemented to ensure buy in from all involved and a system review process was established to assess if there were challenges or harms that needed to be addressed. In the initial strengths, weaknesses, opportunities, and threats analysis, some concerns were raised about patient safety, patient trust and satisfaction as well as referring provider confidence. Aside from initiating a plan-do-study-act (PDSA)-like quality/safety evaluation process, the key driver for successful implementation was transparency into the process for oversight and individualized patient review by experienced team members as well as open communication. This instilled the necessary confidence in our patients and referring specialist providers which was apparent in the NRC feedback scores and personal communication received during that time.

The significant waitlist and extended wait times demanded a reconsideration for the existing triage system. Patients were categorized and triaged based on their acuity and complexity. The primary aim was to identify patients prioritized to needing to be seen by the attending geneticist. The acuity score sorted patients to be seen within: 1 month, 3 months, 6 months, or 1 year. The final category of patients were deemed inappropriate for genetics and referred back to their referring provider with additional recommendations. The details of this scoring scheme are available in the appendix.

Care was taken to minimize the possibility of medical errors. The IOM (Institute of Medicine) report on medical errors has had a significant impact on the general approach to clinical practice and training. The implementation of this new practice model was thoughtful of the risk for diagnostic error, and although Olson's foundational work on competencies at the individual, team and system levels had not yet been published [21], the premise of these principles were considered, and strategies were put into place to minimize any potential patient harm. At the individual level, aside from the education around the various competencies, every patient was subject to additional scrutiny through the application of advanced decision support databases. This included specific training on utilizing Human Phenotype Ontological (HPO) terminology to search OMIM (Online Mendelian Inheritance in Man), POSSUM (Pictures of Standard Syndromes and Undiagnosed Malformations) and LDB (London Dysmorphology Database) to create rank listed differential diagnoses. SimulConsult has been added as a tool utilizing HPO for focusing diagnostic considerations and guiding test decision making. Face2Gene was additionally implemented as standard of care for every patient. In this way, point of care tools were optimized for effective patient care delivery. Prior to the patient visit, detailed chart preparations were reviewed so that the fundamental problems were prioritized as much as possible. The clinic workflow was adapted to improve efficiencies during the clinic visit itself and implemented in a way to encourage collaborative team diagnostics (Fig. 2). The team element of analysis and patient care was further braced through the introduction of a revitalized weekly “clinical case conference”. All clinical team members were required to participate in this weekly 1½ h clinical care conference at which every patient from the prior week was reviewed. The key patient images were projected on a large screen, the APP or GC presented a short clinical vignette, then rank listed differentials from the database analysis were reviewed along with the Face2Gene analysis. Management recommendations were reviewed and adjusted if necessary. The outcome of this meeting was documented in the patient chart as part of their clinical care. This process offered broader exposure to practical patient care, improving the clinical care and providing an educational opportunity (Fig. 3).

FIGURE 3.

Overview of process implementation with weekly review, continuing education, process evaluation and adjustment to improve the model of care.

Outcomes

The best laid plans are only as good as hindsight. Several adjustments have been made to the initial implementation as a result of our ongoing program evaluation. This would be expected in the context of the evolving nature of genomics, not to mention a pandemic which impacted all of medical practice. The integration of APPs into our genetics service has been successful, patient satisfaction reached an all-time high at the 3-year mark and has remained above system average throughout the 9 years since the change was introduced. Referring provider satisfaction scores are very good and wait times are down to average around a month while maintaining high engagement from the genetics team and a satisfied hospital administration. But, it has not been without a few unanticipated challenges. The data from the initial implementation role out is detailed in appendix 2. The initial intent of APPs seeing primarily low complexity patients without the need for genetic counselor supporting every visit was challenged by the limitations of the triage process which, more often than not, underestimated the complexity of the patients’ problems, requiring more GC input (and physician collaboration) for all the visits. The concurrent limitations in GC workforce at that time thus limited the total capacity of patients able to be seen during this time. As a component of the initial consolidation of services, all off-campus clinical outreach services were relocated to the Genetics Clinic location. As additional GC and medical doctor (MD) providers were recruited, available clinic rooms limited capacity further and even with re-establishing an off-site clinic, two maternity leaves challenged the ability to fully attain predicted targets in the short term. The current state is that we have overshot the predicted volumes with less clinical FTE (4587 patient visits with 7.0 cFTE relative to a predicted 3900 with 7.1 cFTE).

The successful integration of such a model where none exists requires significant resources and institutional support. Without the ability to adapt to key drivers that emerged through the PDSA cycle, it would not have been possible to sustain the early iteration of the model. Additionally, had the institution not recognized that the overseeing physician would not be generating relative value unit (RVUs) for their own clinical work but facilitating the RVUs of five APPs, it would not have been possible to sustain the clinical work expectations. While the scenario of onboarding four APPs simultaneously under the context of extremely limited attending geneticist availability will likely not be the norm for many practices looking to include APPs within their clinic workflows, the general themes outlined and the principals for success however should be similar.

Concluding remarks: new frontiers/integrating systems genomics

The ongoing need for medical geneticists and the implementation of “integrated genomic medicine” throughout the healthcare system requires additional administrative support to be successful. As part of the original planning, a 5- to 10-year vision was to develop some of these ancillary services at a systems level to reduce the workforce demand placed on the designated genetics clinical practice. This took the form of education around use and interpretation of genomics data, developing an “embedded” clinical model in which GCs are integrated into specialty practices and the specialty specific genomics expert (frontline content expert) is supported by the GC with additional support available on a consultative basis. Genetic counselors have additionally been integrated into the hospital laboratory services under the PLUGS (Patient-centered Laboratory Utilization Guidance Services) model. All of these efforts are driven through the mission of providing the best and safest care for patients while allowing for more efficient utilization of the attending geneticist.

One aspect that has become apparent for sustainable success is the need for autonomous practice and recognition for the value APPs bring to the team-based practice (personal communication from national survey data collected through CGAPP [Clinical Genetics Advanced Practice Provider Conference]). To achieve this more universally, APP education is going to have to be addressed at a national organizational level. CGAPP was developed by our team after recognizing the need for APPs practicing in clinical genetics to have an opportunity for CME related activity that specifically pertained to their clinical and on-going educational needs. CGAPP is now an annual education meeting and work is underway to develop this group as a national representative organization for APPs practicing medical genetics. There are groups working on establishing clear education needs for APPs, establishing Entrustable Professional Activities (EPAs) in medical genetics and the core competencies required to attain them. The challenge remains for certification/credentialing which allows for both nurse practitioners and physician assistants to certify under a unified body for a recognized standard. This is one way forward to ensure a sustainable workforce that will meet the growing needs of genetics and genomics in both the clinic and community.

Acknowledgements

We would like to acknowledge the many administrators and team members who enabled the success of implementing a model of this nature.

Financial support and sponsorship

None.

Conflicts of interest

There are no conflicts of interest.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

• ▪ of special interest

• ▪▪ of outstanding interest