Our study provides a unique insight into our institutional practices surrounding comprehensive genetic testing and highlights a need to examine our current ES and reanalysis processes in the rapidly evolving genetics landscape. The most impactful results of our study are the emphasized importance of a genetics provider being involved in a patient’s care, lack of follow-up with genetics providers, underutilization of reanalysis for ES performed during a specific time period, and that patients with private insurance were more likely to have had reanalysis performed. This provides a chance to reflect upon possible causes of these associations and improve our systems to allow for more effective patient care.
Earlier studies have examined patient views of the use of exome sequencing in a variety of settings, cost-effectiveness of reanalysis, and other items (Baker et al., 2019; Meng et al., 2017; Robertson et al., 2022; Wenger et al., 2017). We are not aware of other studies that have examined the ES reanalysis practices at a single institution. By performing reviews such as this, we can ensure our system is as efficient as possible and improve patient care for all disciplines. When discussing the reanalysis results themselves, 14% (7/50) of our patients who underwent reanalysis received an updated positive or pathogenic result. This is in line with the reported yield of reanalysis by other studies (Baker et al., 2019; Ewans et al., 2018). In addition, 35% (18/50) received a variant of uncertain significance. These results emphasize the rapid expansion of knowledge and evolving clinical phenotypes. In situations where genetic results are uncertain, providers can use clinical judgement to provide a clinical diagnosis. In addition, 14% of reanalyses resulting in pathogenic variants emphasizes the importance of continuing to reanalyze ES even more than 3 years after the ES was performed. Last, we have not encountered other studies that investigated which patients have continued care with providers and associated characteristics. We would also re-emphasize the need for clearer guidelines for exome and variant reinterpretations that other studies have highlighted (Richardson et al., 2021; Robertson et al., 2022). As Leung et al. (2021), described, clinicians may have differing opinions and practices for implementing ES reanalysis. By having guidelines for when ES reanalysis should be initiated, we can standardize this across the field and clarify the process for patients as they navigate their genetic diagnostic journeys.
In examining the involvement of genetics providers in patient care, we found that a genetic counselor was also involved in the care of patients who underwent reanalysis more than 81% of the time. This highlights the role that geneticists and genetic counselors have in ordering ES and facilitating ES reanalysis. Given their involvement, it was interesting to see how few patients continue to follow with these providers at our institution. While understanding why these patients are not following with genetics providers is beyond the scope of our review, our data revealed that most of our patients were indeed provided specific information for when they should be following up with their ordering provider. At this time, we are not able to reanalyze results without patients being followed in our clinics. However, by adopting an automated process, we could potentially have reanalyses performed periodically. This current study will also motivate future studies comparing the 2013–2016 period to other more recent periods, especially as exome sequencing ordering has increased institutionally in the last five to six years.
In situations in which a patient was reported to have ES or reanalysis but no report was found in the EMR, we did not attempt to determine if the patient originally had a physical chart. This reveals a gap in a patient’s medical record as other providers may not be able to find the patient’s genetic testing results and/or ES reanalyses. Without access to these records, we cannot determine the patient’s original phenotype, test ordered, or test results. Furthermore, without a record of the testing results, other providers may assume the testing was not completed and could attempt to order the same test again. This is an inefficient use of resources for the patients, providers, laboratories, and healthcare system.
While we cannot conclude that providers did not discuss reanalysis with their patients, without documentation of this discussion we must assume this topic did not arise. This may indicate patients and families are being unaware of the option for reanalysis and may have contributed to the observed lack of follow-up. Documenting these discussions ultimately helps providers by ensuring best testing strategies are implemented. This is critical now that these tests are being ordered more frequently, and inefficient ordering practices place a potential strain on the healthcare system.
We also noted a statistically significant difference between the number of patients who did have a reanalysis with private vs. public insurance. Given the barriers that private insurance often presents with pre-authorization requirements and benefits investigations, this was surprising to us. In addition, this may be indicative of more subtle socioeconomic differences influencing patient care. Patients with private insurance likely acquired it through an employer, meaning they have additional access to time-off from work to attend appointments, may have a higher disposable income, and may have better access to transportation to attend appointments. While we cannot currently provide reanalyses to families who do not return to clinic, we also need to do our best to accommodate the availability of our families, and to frequently communicate the benefits and importance of following with their genetics providers.
Further, patient age was not associated with a higher likelihood of reanalysis. While our age range encompassed almost a full lifespan, the age at initial ES was not associated with an increased ES reanalysis. While 60% of those who did have reanalysis were older children, it is worth considering the potential importance of longitudinal data in performing a reanalysis. Longitudinal data provides unique information on developing phenotypes which allows for a more comprehensive review of variants. Age of onset for certain features and progression of the phenotype are extremely useful for determining a differential diagnosis. In addition, vital status was not a significant indicator of reanalysis. No patients, who had ES performed during their life, had a reanalysis of the data after their death. Despite genetic counselors being uniquely trained to approach genetic information in the context of difficult psychosocial situations, we are not seeing these families after the patient’s passing. Efforts to remain engaged with these families after a patient’s death may ultimately result in a molecular diagnosis for these patients and can provide the family with more accurate recurrence risks.
Our study provided a unique opportunity to review our ES and reanalysis process and revealed interesting results. This retrospective analysis allowed us to examine areas where our institution may improve patient care and improve efficiency in how we utilize genetic testing. We observed several interesting trends such as increased likelihood of reanalysis for those with private insurance and that a genetics provider was involved in all instances of reanalysis. However, our study was not designed to reveal the specific reasons for these trends.
Limitations of our study include the relatively small sample size, single institution resources, years examined, and unadjusted statistical analyses. While we wanted to perform a retrospective review, our timeframe includes years before genetic testing was readily covered or accessible to many of our families. This may reflect that our families have a more negative view of the genetics team due to the cost of testing or availability of testing previously compared to the ability to offer testing and get testing covered now. In addition, we could have examined data that is no longer reflective of our current institution’s practices and patient demographics. Last, by not performing adjusted analyses, we may have missed confounding factors for certain variables. Our study does have several strengths which include that a single person performed the data collection, our studied timeframe allows for adequate time for reanalysis to have been performed, and it includes multiple years. The patient demographics are reflective of our institution and the patient populations that we serve.
As research continues to move forward, we may be able to determine the causes for some of these trends. Finally, as we continue to move forward with establishing the best genetic care for patients and clinical practice, it is important to frequently re-examine clinical practices and identify areas where change should take place. Using the information gained from this study, future research focusing on establishing ES reanalysis guidelines will be important and help guide providers to best patient care practices. In addition, larger studies that can further confirm the association between the likelihood of reanalysis with variables such as patients having private insurance will be important for identifying patient populations who may need extra support for the best genomic and precision medicine care that we can offer. By performing reviews such as this, we can ensure our system is as efficient as possible and improve patient care in all disciplines.
Data Availability
A full report of our data and our collection instrument can be found in the supplemental materials. Requests for data can be reviewed by the corresponding author.