Improving access to specialty care: Analyzing the effects of a telehealth intervention on timely access, patient loyalty, and provider productivity in a rural vascular practice

A small, specialty vascular clinic with offices in the Mountain States region faced increasing wait times for new patient evaluation at one location. In this study, the implementation of a telehealth program is examined for its effectiveness in increasing patient access, improving patient loyalty, and reducing the organization’s gap in provider productivity.

Diagnosing the problem

The inability of the northern location to meet the increasing demand for specialty vascular services negatively impacted timely patient access to care, measured by patient wait times to schedule a new patient consultation. The clinic’s mean patient wait time for a new patient visit at the location rose from 32 days in 2018 to 58 days in 2019, causing concern for the organization’s leadership. Comparatively, the southern location had a mean patient wait time of 22 days in 2018 and 28 days in 2019.

Using Merritt Hawkins patient wait surveys, the organization determined that the northern location’s service area had the shortest mean patient wait time for a new patient visit (10.8 days) of any midsized market, making it even more pressing for the clinic to make improvements to be more aligned with the market average and avoid patients seeking care elsewhere due to the lengthy wait times.

Monthly patient satisfaction surveys confirmed that the extended wait time negatively impacted patient loyalty, specifically in the patient’s perception of ease of scheduling an appointment. Historically, the clinic scored near or above the national average of 90% for the patient’s ability to get an appointment as soon as they needed one. The southern location continued to score well on this measure; however, in early 2019, the northern location fell below the national average. In November and December 2019, the northern location hit an all-time low satisfaction score on this metric of 60%, which was perceived as an impact of increased demand and growing inability to see patients within a desirable time frame.

The increased demand at the northern location also created a gap in productivity between the physician assistants (PAs) at the northern and southern locations. Measuring each provider’s salary minus the provider’s revenue collections, southern location PA productivity was significantly lower than those at the northern location: In 2019, the northern location’s PA provided more than $46,000 extra productivity than the southern location’s PA, which signaled that the southern location’s PA might be underutilized.

People living in rural counties are more likely to have statistically significant poorer health outcomes than their urban counterparts,1 and decreases in timely access to care has direct implications on rural patients’ overall health, morbidity, and mortality.2

Decreased access for the clinic’s patients was also problematic for the financial viability of the clinic. In the current fee-for-service (FFS) model, decreased patient access leads to a decrease in financial returns. As the industry shifts to value-based care, the negative health outcomes associated with decreased access can lead to decreased financial performance for the organization. In the current FFS model and the future value-based care model of reimbursement, access to care for patients is of vital importance to the financial viability of the clinic.

Patient loyalty was important to the clinic’s financial sustainability and its organizational mission. Continued utilization and increased patient referrals translate to economic benefits for the practice in the FFS model and the transition to value-based care. Patient loyalty was also extremely important to the clinic’s ability to meet its mission of being the leader in treatment and long-term management of venous disease through making patients a priority, providing excellent patient care, and defining success by having satisfied patients who recommend the organization to family and friends. Patient loyalty was a key feature in the clinic’s strategy to meet its financial goals and organizational mission. Word-of-mouth strongly influences health behaviors of patients, and as online communication grows so does the value and importance of word of mouth in the healthcare industry.3

Literature review

A literature review focused on addition of a telehealth program for a rural vascular clinic found evidence to support the use of this model. A telehealth intervention, if implemented correctly, has the potential to address patient access to healthcare services, while maintaining patient satisfaction and quality of care.4 The literature also provided key areas to consider for the implementation of telehealth to be successful.

Telemedicine emerged as an essential component of healthcare delivery in high-volume situations where it has proven difficult to physically bring patients and provider resources together,5 and that it had the potential to extend providers’ reach to patients outside of the proximity of their clinic without adding more clinic locations.6

The increased use of telehealth for patients located long distances from care was important to note because the clinic serviced multiple rural areas and patients. The average patient traveled more than 50 miles to seek care at the clinic; however, patients often traveled several hours and more than 100 miles. Some of the most rural patients traveled more than 280 miles to receive their lower extremity venous care.

The challenge of engaging leadership meant to meet with key leaders to ensure the program was in line with the organization’s strategic plan and the executives support the program on all levels. The challenge of prioritizing telehealth meant that the program was supported by financial resources and manpower to be successful. The challenge of engaging clinicians and staff meant involving clinicians and frontline staff members in training and initiatives to receive buy-in to the program. The challenge of enrolling and educating patients referred to making patients aware of the telehealth option, educating them on the process, and encouraging their participation in the program. Proper evaluation of the program referred to having appropriate goals for the program and measuring the program’s ability to meet those goals. These challenges were considered, and solutions developed to make the telehealth intervention successful at improving patient access to healthcare services.

Rationale

A cause-and-effect diagram was utilized to identify and dissect the patient access problem at the northern location (see Figure 1). Also known as a fishbone diagram, it is a powerful problem identification tool that clearly defines the problem and then stratifies and identifies possible causes.7 The clinic utilized the categorization scheme for service problems, defined as problems in industries that do not produce tangible goods.

The fishbone diagram identified several potential causes that may contribute the problem of patient access to healthcare services. In the policy category, the causes identified were insurance requirements, long exam times, and privacy concerns. In the category of procedures, the causes identified were cluster scheduling, linked appointments and time spent with patients. The causes identified in the category of people were provider availability, provider location and patient location. In the category of technology, the causes identified were cumbersome EHRs, long check-in processes and the inability to offer telemedicine.

The cause-and-effect diagram analysis identified several possible interventions for addressing patient access at the clinic’s northern location. These interventions included added provider time in the northern location to increase available appointments, decreased appointment times, altered scheduling methods and implementation of a telehealth program. These possible interventions were examined individually.

The first intervention that was identified was adding additional providers. The idea of utilizing the southern location’s PA in the northern office for in-office visits was considered; however, this intervention would prove difficult, given the distance between offices.

The idea of hiring an additional healthcare provider was also considered to increase the availability of appointments at the northern location. The ability to find qualified candidates for this location has historically been challenging, and a budget evaluation indicated that this intervention was not financially feasible.

The idea of reducing appointment times and implementing various scheduling changes was also a potential intervention; however, the leadership of the clinic believed that this intervention was in direct conflict with the clinic’s mission of making patients a priority; providing care in an inviting, comfortable, and relaxed environment; and educating patients about venous disease and current treatment options.

The implementation of a telehealth program was an intervention supported by the cause-and-effect analysis. The literature indicated that a telehealth intervention was an evidence-based solution with the potential to address the problem of patient access faced at the northern location. Based upon the evidence, the clinic implemented a telehealth program intervention.

The conceptual model used in this project is the PDSA cycle — a tool recommended by the Institute for Healthcare Improvement8 as the model to document and test an improvement project. With the use of the PDSA model the clinic implemented a telehealth intervention, studied the results, and acted upon the results of the study (see Figure 2).

After the aims of the project were established, the organization’s leadership met to determine the project’s goals. Based upon historical performance of the clinic and industry standards, the leadership established three goals:

1. Decrease patient wait times at the northern location by 10%
2. Improve patient loyalty scores at the northern location by 10%
3. Increase the productivity of the southern location’s PA by 5%.

Methods

Context

The clinic that was the subject of this study is a small specialty vascular practice with locations in the Mountain States region. The clinic is the only dedicated vein care center in the region, which contains a significant number of rurally located patients.

The clinic is one of only three organizations that offered vein care services in the entire region. In 2019, the organization averaged 57 new patients per month, 100 follow-up patients per month, and 118 procedures per month. Of the patients seen at the clinic, 38% are Medicare beneficiaries, 4% are Medicaid beneficiaries, and some form of commercial insurance covers the remaining 58% of patients.

The organization was a sole proprietorship corporation owned and operated by a single physician who works at each clinic. Along with the physician, each location had a PA who conducted most office visits and provided minimally invasive therapies, such as sclerotherapy. The northern location was the only private practice that exclusively provided healthcare services to patients seeking treatment for lower extremity venous disease in the area. Two other private practices previously offered these services; however, those practices closed, resulting in an increased demand for the clinic’s services. Reports generated by the organization’s EHR indicate that the northern location had seen a 23% increase in demand for new patient consults from 2018 to 2019. The increased demand at the northern location led to a decrease in timely patient access, a decrease in patient loyalty, and a gap in provider productivity that existed between the northern and southern locations.

Intervention

The project implemented a telehealth intervention for a small specialty vascular clinic at points in the care continuum where the providers review ultrasound findings and conduct postoperative care. The intervention consisted of real-time audio and visual consultation with one of the clinic’s providers. The telehealth intervention was also offered to new patients for a limited time while the office was closed to in-person appointments during the COVID-19 pandemic.

The care process at the clinic included a new patient evaluation followed by an ultrasound screening, when deemed appropriate during the new patient evaluation. The ultrasound results were discussed, and the findings were reviewed with every patient anywhere from six to 12 weeks after the initial consultation to allow for a conservative management trial should more invasive treatment be indicated. If treatment was indicated, then procedures were scheduled at the location of the patient’s choice. Two postprocedure ultrasounds were scheduled at approximately one week and one month postprocedure. A provider visit was conducted after the second ultrasound follow-up to discuss the findings and determine if further treatment was needed.

Ultrasound review appointments and postprocedure visits, conducted via telehealth, were offered to all patients interested whose insurance covered telehealth services. These visits were conducted by any available provider, regardless of the provider or patient location. The intervention focused on transitioning patients with an established relationship with the clinic to the telehealth platform with the southern location’s PA to increase availability of in-person appointments with the northern location’s PA for new patients.

The primary goal of the intervention was to improve patient access to services by decreasing wait times for new patient appointments by 10% at the northern location. The secondary goal of the intervention was to improve patient loyalty scores regarding availability of appointments by 10% at the northern location and reduce the gap in provider productivity by increasing the southern location’s PA’s productivity by 5%.

Study of the intervention

Secondary data was reviewed from reports generated by the clinic’s EHR, patient satisfaction surveys, and monthly financial reports. Data from a 12-month period preintervention, the seven-month intervention period, and a three-month period postintervention were collected and analyzed.

EHR data included the number of patients seen for a new patient appointment at each location during the study time frame. For each new patient appointment conducted during the study time frame the location of the visit, the date the patient was registered in the EHR, and the date the new patient appointment was conducted were reviewed.

To evaluate patient loyalty, responses from the voluntary patient satisfaction survey, specifically the responses to the question, “Your ability to get an appointment as soon as needed,” were reviewed for the study time frame. This data included the appointment date, appointment location, provider name, and patient response to the survey questions on a 5-point Likert scale was used for the survey, where 0 equaled poor, 1 equaled fair, 2 equaled good, 3 equaled very good, and 4 equaled excellent.

To evaluate provider productivity, data from monthly financial reports were reviewed. The provider productivity data included the location, as well as PA salary and revenue collections, in dollars. This data was converted to a monthly productivity ratio for each location for the study time frame.

Measures

To measure patients’ wait times, the mean number of days waited from new patient contact to the first appointment was utilized. The mean number of days is a valid measure of patient access and wait times utilized in recent research designs evaluating timely patient access to care.9 Data from the clinic’s EHR was collected to measure the mean number of days. This data included the date the patient initially registered in the EHR and the new patient examination date. It was the clinic’s policy to register a patient immediately at the point of inquiry for a new patient appointment. These reports were cross-checked against internal clinic reports to ensure the validity and accuracy of the data.

The patient loyalty data was collected from the patient satisfaction survey vendor. In this survey, the patient responded to the question, “Were you able to get an appointment as soon as needed?” on a 5-point Likert scale. Individual patient responses for the study time frame and monthly average scores were reviewed.

The provider productivity data was collected from the monthly financial performance reports. The provider salary and revenue collections for each location’s PA was converted to a monthly productivity ratio throughout the study time frame. The clinic’s monthly financial reports are balanced, and cross-checked against EHR reports and bank statements ensuring validity and reliability of the data.

Analysis

Quantitative data analysis was used to analyze the telehealth intervention. Data from the clinic’s EHR was collected for the telehealth appointments that were conducted during the intervention and postintervention period (see Table 1). This data was analyzed for appointment type, location and patient demographics associated with the appointments.

Data from the clinic’s EHR was collected for patients who had a new patient consultation during the study time frame. This data was analyzed for location and patient demographics associated with the appointments (see Table 2). The preintervention, during intervention, and postintervention continuous data were compared using t test or equivalent nonparametric tests as appropriate. Chi-squared analysis was utilized to compare nominal patient-level data for the preintervention, during intervention and postintervention patient groups.

Patient wait time data for each appointment during the study period was calculated by finding the difference, in days, between the registration date and the date of the new patient appointment. The results for each patient wait time were analyzed for outliers using z-score calculations, and data points with a z score ± 3 were eliminated from the data set. After, outliers were eliminated, the remaining data was analyzed utilizing the number of days waited from new patient contact to the first appointment for each intervention period utilizing independent t tests. Independent t tests were also utilized to analyze patient wait times by gender.

Mean number of days waited each month was calculated by adding the total number of days waited for each appointment. The sum of days waited was then divided by the number of appointments for that month to determine the mean number of days waited (see Table 3). The mean number of days per month was charted and analyzed for trends and changes.

Analysis of the clinic’s patient satisfaction survey was conducted to measure patient loyalty. In this survey, the patient responded to the question, “Were you able to get an appointment as soon as needed?” on a five-point Likert scale. Individual patient responses preintervention, during intervention and postintervention were collected. Monthly average patient loyalty was collected from the patient satisfaction survey dashboard (see Table 4). The Mann–Whitney U test was used to compare the data preintervention, during intervention, and postintervention and across clinics.

Data from the postprocedure clinic’s monthly financial performance reports was collected and analyzed to measure productivity. The PAs’ collections, outputs, salary, and inputs were collected preintervention, during intervention, and postintervention. A productivity ratio was calculated for each time frame (see Table 5), and a percentage change in productivity was analyzed.

Results

Analysis of the telehealth appointments completed during the research time frame was conducted. Visit counts and types were tabulated by clinic and overall, for each intervention time frame applicable (see Table 1). The total number of telehealth appointments by month was presented graphically (see Figure 3). No trends were identified. Most telehealth appointments (83%) were conducted for ultrasound review appointments, while a small portion were conducted for new patients, 7%, and postsurgery follow-up appointments, 8%. The telehealth appointments were evenly dispersed between locations with 25 of the appointments, 42%, being conducted by providers at the southern location and 34 appointments, 58%, being conducted by providers at the northern location.

Similar analysis was conducted for the new patient appointments for all three intervention periods. New patient appointments by clinic and overall were tabulated (see Table 2). No trends were identified. Throughout the study the northern location saw the majority of new patients, 66%, while the southern office conducted 34% of the new patient visits.

Chi-square analysis was conducted for nominal patient data in the new patient appointment data. There was a statistically significant difference between appointment location and patient gender (p < .001). This analysis shows that 76.1% of the northern location’s patients were female as compared to 65.4% of the southern location’s patients.

Independent t-test analysis indicated that there was a statistically significant difference in wait times between males and females. Male mean wait times were 12.63 days (SE = 2.81) shorter than female mean wait times (p < .001). Independent t-test analysis showed that overall, the southern location’s wait times were 20.74 days (SE = 2.40) shorter than the northern location’s wait times (p < .001). Wait time analysis for all intervention periods indicated there was no significant correlation between patient age and wait times or patient distance from the clinic and wait times.

Overall clinic wait time analysis indicated there were no statistically significant changes in mean wait times during any of the intervention periods. There were no statistically significant changes at either clinic from the during intervention period to the postintervention period when analyzing wait times at individual clinics. However, at both locations, there was statistically significant changes in wait times from the preintervention period to the during intervention period and from the preintervention period to the postintervention period (see Table 3).

For the northern location, the average mean number of days decreased from 62 in the preintervention period to 39 in the postintervention period (see Figure 4). The mean wait time for the northern location was 10.79 days (SE = 4.08) shorter in the during intervention period than it was in the preintervention period (p = .009). The mean wait time decreased by 34.42 days (SE = 5.89) when comparing the preintervention wait times period to the postintervention wait times at the northern location (p < .001). This decrease in mean patient wait time is equivalent to a 37% decrease in mean patient wait time at the northern location.

For the southern location, the average mean number of days increased from 29 in the preintervention period to 51 in the postintervention period. At the southern location, the mean wait time was 13.5 days (SE = 3.83) longer in the during intervention period than it was in the preintervention period (p < .001). The mean wait time increased by 22.45 days (SE = 4.04) when comparing preintervention wait times period to the postintervention wait times at the southern location (p < .001). This increase in patient wait time at the southern office is equivalent to a 62% increase in patient wait time.

There was no significant difference in overall satisfaction scores between periods when analyzing patient satisfaction data for the clinics individually (see Table 4). There was also no significant correlation found between monthly mean satisfaction scores and monthly mean wait time for either clinic individually or the organization overall. However, a Mann–Whitney U test indicated that there was a statistically significant difference in satisfaction scores between the northern and southern location (p = .006) with the southern location scores being significantly higher. A Mann–Whitney U also indicated that there was a statistically significant difference in satisfaction scores between the overall satisfaction scores between the preintervention period and the during intervention period (p = .015) with the during intervention scores being significantly higher (see Figure 5).

There was no statistically significant change in the northern location’s PA’s productivity ratio during any of the intervention periods when analyzing provider productivity data (see Table 5). However, independent t-test analysis showed that the southern location PA’s productivity increased significantly (p = .02) from the preintervention period to the postintervention period from 1.25 to 1.74, a mean increase of .48 (SE = .14). This productivity increase is equivalent to a 40% increase in productivity for the southern location’s PA (see Figure 6).

Discussion

Summary

There were significant changes in all outcome measures identified in the study. The mean patient wait time significantly decreased from the preintervention period to the intervention period and the postintervention period. The greatest decrease was seen from the preintervention period to the postintervention period with a mean decrease in patient wait time of 34 days. This resulted in a 37% decrease in wait time at the northern location, a much larger decrease than the stated project goal of 10%. There was also a significant increase of mean wait time at the southern location from the preintervention period to the intervention period. However, there was not a significant increase in mean wait time in the southern location from the preintervention period to postintervention period.

Patient satisfaction data showed significant improvement overall from the preintervention period to the intervention period. However, this significant change was not present in the postintervention period. The analysis indicated that the intervention failed to meet its goal of increasing the patient loyalty scores by 10% in the northern location.

Provider productivity for the southern location’s PA changed significantly from the preintervention period to the postintervention period. The southern location PA’s productivity increased 40%, far exceeding the stated goal of a 5% increase. Provider productivity for the northern location’s PA did not show significant changes.

Interpretation

The intervention achieved two of its three goals. As indicated by studies examined in evaluation of the evidence, the intervention was successful at improving patient wait times at the northern location. An unforeseen outcome was the significant increase of patient wait times at the southern location. Further analysis into the increase in patient wait times at the southern office is warranted. The increase will need to be examined, and the intervention parameters will need to be adjusted to find an acceptable process to maintain acceptable wait times at each location.

The analysis also indicated that the telehealth intervention was successful at improving productivity of the southern location’s PA. This productivity improvement brought the productivity ratios of the two PAs closer together and helped close the productivity gap between the providers.

The analysis indicated that the telehealth intervention failed to improve patient satisfaction and loyalty at the northern location. Further analysis showed that the mean wait time each month was not significantly correlated with patient satisfaction with their ability to get an appointment when needed. It would be reasonable to conclude that mean wait time is not the primary contributing factor to patient satisfaction in this case.

Overall, the telehealth intervention should be considered successful. The primary goal of improving patient wait times at the northern location was achieved. The clinic was able to improve timely access to care at the northern location without decreasing patient volume. Also, one of the secondary goals of improved provider productivity at the southern location was achieved.

Conclusions

Telehealth interventions can positively affect patient access, patient satisfaction, and provider productivity in the rural specialty care setting. Given the current changes being implemented in telehealth regulations throughout the country, this intervention has potential for long-term sustainability and applicability in the clinic’s care process. Further research into the cause of decreased patient satisfaction scores, and a continuation of the PDSA cycle for telehealth intervention is recommended to maintain and further improve on the outcomes of the study variables.

Notes:

1. Anderson TJ, Saman DM, Lipsky MS, Lutfiyya MN. “A cross-sectional study on health differences between rural and non-rural U.S. counties using the county health rankings.” BMC Health Services Research, 15(1), 441–448. doi.org/10.1186/s12193-015-1053-3.
2. Cyr ME, Etchin AG, Guthrie BJ, Benneyan JC. “Access to specialty healthcare in urban versus rural US populations: A systematic literature review.” BMC Health Services Research, 19(1), 974–990. doi.org/10.1186/s12913-019-4815-5.
3. Martin S. “Word-of-mouth in the health care sector: A literature analysis of the current state of research and future perspectives.” International Review on Public and Non-Profit Marketing, 14(1), 35–56. doi.org/10.1007/s12208-016-0154-y.
4. Welch BM, Harvey J, O’Connell NS, McElligott JT. “Patient preferences for direct-to-consumer telemedicine services: A nationwide survey.” BMC Health Services Research, 17(1), 784–787. doi.org/10.1186/s12913-017-2744-8.
5. Wilson LS, Maeder AJ. “Recent directions in telemedicine: Review of trends in research and practice.” Healthcare Informatics Research, 21(4), 213–222. doi.org/10.4258/hir.2015.21.4.213.
6. Kruse CS, Krowski N, Rodriguez B, Tran L, Vela J, Brooks M. “Telehealth and patient satisfaction: A systematic review and narrative analysis.” BMJ Open, 7(8), Article e016242. doi.org/10.1136/bmjopen-2017-016242.
7. McLaughlin DB, Olson JR. Healthcare operations management, 3rd ed. Healthcare Administration Press, 2017.
8. Institute for Healthcare Improvement. Plan-do-study-act (PDSA) worksheet. Available from: bit.ly/3IYu0Y8.
9. Lewis AK, Taylor NF, Carney PW, Harding KE. “Specific timely appointments for triage to reduce wait times in a medical outpatient clinic: Protocol of a pre–post study with process evaluation.” BMC Health Services Research, 19(1), 831–838. doi.org/10.1186/s12913-019-4660-6.