By Molly Warren and Jennifer Sullivan | April 2015
In the lead-up to the second open enrollment period (OE2), as part of an effort to increase the availability and capacity of in-person enrollment assistance across the county, Enroll America launched the Get Covered Connector. This assistance-scheduling tool was used in 24 states and the District of Columbia in OE2. This brief discusses the benefits of such a tool to outreach and enrollment programs and lessons learned from the first year, including insights into consumer behavior and preferences for in-person assistance.
Among the most important lessons learned during the first open enrollment period (OE1), and reinforced thereafter, was the critical role in-person assistance played in successful enrollment.1 Although not necessary for all consumers, in-person assistance was crucial for certain populations, including many Latinos, African Americans, immigrants, and simply those who understood less about the Affordable Care Act (ACA) or were less computer-savvy going into the enrollment process.2 An examination of the experiences of consumers who interacted with Enroll America (either through the GetCoveredAmerica.org consumer-facing website or through the Get Covered Connector assistance-scheduling tool) during the first month of OE2 underscores the effect in-person assistance has on enrollment: Consumers who received in-person assistance were nearly 60 percent more likely to enroll compared to those who started the enrollment process on their own online.3 (See Figure 1.)
Figure 1: Enrollment Rates Among Consumers Who Started With In-Person Help vs. Online, November – December 2014
The ACA provides for a patchwork of different kinds of in-person assistance, including Navigators, Certified Application Counselors (CACs), In-Person Assisters (Navigator-like entities that received special funding in certain states), and health insurance agents and brokers.
However, the sources and amounts of funding, training and certification requirements, state rules, and degree of coordination vary dramatically from state to state, and even from one community to the next. In many communities, demand for in-person assistance outstripped supply during OE1, especially during the critical final weeks leading up to the deadline.
Given the importance of in-person assistance heading into the second year of enrollment, Enroll America worked to identify the best ways to ensure that there was sufficient capacity among these programs to meet consumer demand and that consumers knew that assistance was available and how to find it.
The resulting multi-pronged approach included a concerted effort to recruit and train CACs, partner with more agents and brokers, and — most significantly — launch a nationwide in-person assistance-scheduling tool, the Get Covered Connector. (See Figure 2.) This report describes lessons learned from the Get Covered Connector, including:
- The value of centralized scheduling tools to assisters.
- Insights into consumer behavior, needs, and preferences.
- Action steps partners — including those using the tool as well as others in the enrollment and assistance community — can take to build on these lessons learned in the future.
Figure 2: Get Covered Connector Tool
Connecting Consumers to Assistance: The Get Covered Connector
During OE1, Enroll America launched the “Locator,” a comprehensive nationwide online listing of in-person assistance searchable by ZIP code. This was a widely used tool and served as the foundation for the Get Covered Connector. To build on this success, Enroll America looked to partners in North Carolina who used a statewide assistance-scheduling tool during OE1. Their tool — spearheaded by Legal Aid of North Carolina with the help of the North Carolina Community Health Center Association and input from the 11 organizations that comprised the state’s main Navigator grant — paired a centralized electronic scheduling system with a statewide toll-free phone number. The electronic scheduling system allowed participating organizations to access assisters’ schedules across the state and schedule consumer appointments with any available local assister. The toll-free number connected consumers looking for in-person help with an administrator who could schedule appointments with assisters throughout the state via the tool. Enroll America’s North Carolina staff saw firsthand the benefits of the unified scheduling system during OE1, and Enroll America used this general concept to build a nationwide scheduling tool for OE2.
The resulting tool, the Get Covered Connector, layers a scheduling system for subscribing assisters on top of the existing “Locator” tool. Consumers and outreach partners throughout the country can visit GetCoveredAmerica.org to use the Connector to find local in-person assistance.4 In 24 states and the District of Columbia, consumers can also view available appointments and schedule an appointment directly through the tool.5 (See Figure 3.) In addition to this consumer-facing scheduling component, the backend portion of the Connector allows subscribing assisters to upload their appointment schedules, track their work, and create program reports.
In total during OE2 and the week-long Special Enrollment Period that followed (offered in most states to allow consumers to complete applications that they started before the deadline), about a half million appointments were made available through the Connector. Consumers were scheduled for nearly 62,000 of these appointments. Among appointments for which data are available (excluding no-shows), about half resulted in successful enrollment.
Figure 3: Get Covered Connector Use During OE2, November 15, 2014 – February 22, 2015
Improving Assister Listings: Getting Consumers to the Right Locations
Starting in OE1 with the Locator and continuing as part of the Connector in OE2, Enroll America has prioritized creating and maintaining a comprehensive online listing of organizations that provide consumers with assistance in applying for health coverage. The list includes information from a variety of sources, including data from HealthCare.gov, and information is carefully and routinely curated by state and regional Enroll America staff, who are in regular communication with assisters in most states.
The Connector may be the single most comprehensive national consumer assistance tool available, because it combines the HealthCare.gov listings with so much other local, granular-level information. This ensures, for example, that the addresses that appear in the Connector are actual locations where consumers can obtain assistance, rather than, for example, an organization’s administrative offices. It also allows for regular updates when new assistance locations become available, such as when new organizations begin providing assistance or schedule enrollment events. Likewise, locations that no longer provide assistance are removed. Consumers who prefer assistance in a language other than English can search by their desired language (with a total of 27 languages offered in OE2, and 1,559 locations offering appointments in languages other than English).
Value of the Connector to Assisters
Although the Get Covered Connector allows free public access to the list of assisters and appointments available, assister organizations were required to subscribe to the tool in order to enter their schedules and take advantage of the tool’s other backend features. One subscription granted an organization 250 logins that they were free to distribute across their organization and to any other coalition partners they chose. In Enroll America’s 11 field states, all staff also had backend access and used the tool to schedule interested consumers with local assisters.6 Experiences with the tool during OE2 offer important insights into the role tools like the Connector can play not only in connecting consumers to in-person help but also in helping grow and strengthen coalitions, improve reporting, and match assister supply and demand.
Value of the Connector to Assisters
→ Increased Visibility for In-Person Help
→ Appointment Reminders
→ Coordination and Planning
→ Strengthening Coalitions
→ Data Reporting and Monitoring
Increased Visibility for In-Person Help
Despite the availability of free in-person enrollment help in every state, consumers reported limited awareness that such help existed. After OE1, less than half of enrollees (44 percent) and less than a quarter of the remaining uninsured at that time (23 percent) knew about free in-person help.7
One clear utility of a central scheduling tool is that it allows assisters and the broader outreach community to refer all consumers to a single place (GetCoveredAmerica.org or the participating organization’s website through an embeddable customized widget) to find and schedule an appointment at a time and place convenient for them. The tool also allows consumers to find help in their preferred language. Some states also coupled the Connector with telephone hotlines (including 211 and other toll-free numbers), so that consumers could call a single number to schedule an appointment anywhere in the state or region.
In addition to making it simpler for consumers to view all their assistance options and make appointments, centralized scheduling also strengthens the connection between assisters and other organizations conducting outreach.
For example, when Enroll America staff or volunteers identified an uninsured consumer interested in getting coverage, they could immediately schedule the person with a specific appointment, rather than simply referring them to an assister organization in their community. The tool was useful in states without Enroll America staff on the ground, as well. For example, in South Carolina, the coalition made sure that schools knew about the tool, and when school nurses encountered a student who was uninsured, they could schedule the parents with an appointment.
Broad adoption across assisters in a community also allowed assisters to accommodate consumers who either could not complete the process in a single appointment, or who needed to schedule an appointment at a different time than the individual assister or organization could provide. This direct scheduling reduced opportunities for consumers to fall through the cracks in the hand-off process, and ensured that those who wanted help immediately knew where and when it was available in their community.
However, it is important to note that assistance-scheduling tools must be paired with a broader marketing and consumer outreach strategy in order to be effective. Simply posting a tool online will not significantly increase assister visibility. The organizations that made most effective use of the tool during OE2 made a concerted effort to drive consumers to it, featuring it as a primary tool on their website, ensuring that earned media pitches mentioned the tool, incorporating links to the tool in emails to consumers and on social media, and directing consumers to the tool in paid advertising. Another strategy that was used in several states, including North Carolina and Tennessee, was to pair the tool with a toll-free number that consumers could call to schedule an appointment, with staff and volunteers using the Connector to look up and schedule appointments for the consumers. Organizations conducting outreach could then advertise assistance and drive consumers to either the online tool or to the phone number, reaching consumers who may not be computer-savvy or have internet access.
OE1 provided strong evidence of the value of following up with consumers multiple times. Consumers were more likely to enroll after each successive contact with an Enroll America staff member or volunteer, and the effect was strongest among African Americans, Latinos, and young adults.8 This knowledge, combined with conventional wisdom about the value of reminder calls for appointments of any kind, made the automated reminder function of the Connector a particularly important one. The tool automatically sent text and email reminders to consumers who scheduled appointments. These reminders included information about the appointment as well as directions to the location. Reminder emails sent by Enroll America also included other information, such as what to bring to the appointment.9 Organizations could also make reminder phone calls to consumers, and Enroll America staff and volunteers helped make these calls in many states.
Coordination and Planning
While free in-person assistance is available in every state, availability of appointments is not always matched with consumer demand. During OE1, this led to enrollment events (particularly those before key deadlines) with more consumers than there were assisters available to provide help. At other times, assisters had capacity to help many more consumers than they did, but there simply was not the demand.
Centralized scheduling allows program administrators to better plan for these ebbs and flows. They can plan to add capacity for an event that fills up quickly, cancel slots that are repeatedly empty, and take no-show rates into account when setting up assisters’ schedules. Managing schedules through the Connector ensures that assisters are busy helping consumers at the times consumers need help and otherwise toggling to other responsibilities like outreach and follow-up when demand is low. This ability to efficiently match assister supply with consumer demand will be increasingly important as more consumers get covered and the uninsured population shrinks. Finding the uninsured will require greater effort, and resources to support in-person assistance programs may decline. Identifying simple ways to increase efficiency and ensure that in-person assisters are maximizing their time will help ensure these programs are increasingly effective and efficient going forward.
The assister community is a patchwork of different organizations and individuals with different funding streams, rules, goals, and incentives. It is therefore unsurprising that a relatively high proportion of uninsured consumers did not know about free in-person help at the outset of OE2.
One promising development that unfolded as the Connector was rolled out to assister organizations and enrollment coalitions during the summer of 2014 was the galvanizing effect the tool had on enrollment coalitions. Fully half of the organizations using the Connector reported that the tool was helpful in building their coalition.10
- Pitching the tool to potential partners gave Enroll America staff and organizations that were early adopters of the tool a reason to reach out and connect, or reconnect, with important allies in the enrollment community.
- Implementing the tool took time and training, which gave groups a reason to be in regular communication about very specific aspects of their work — scheduling, staffing, data collection, and evaluation.
Groups that never had a structured reason to work together, and might even have perceived one other as competitors, had a reason to collaborate on outreach strategy and plans and to troubleshoot challenges together.
Data Reporting and Monitoring
The Connector was built with users’ reporting needs in mind, but meeting the needs a wide range of different kinds of users proved challenging. Ultimately, only about a quarter (26 percent) of organizations that used the Connector used it for regular grant-reporting.11 Organizations’ reporting needs varied depending on whether they were a federally funded Navigator, a health care provider, state agency staff, a private agent or broker, or an outreach-focused organization. Some coalitions (with a single backend Connector subscription) included many different kinds of users, who needed to isolate only their own organization’s outcomes for reporting purposes (rather than mixing the outcomes of multiple coalition members). Some organizations are also already required to use other reporting platforms, so they do not enter any data into the Connector (beyond appointment basics).
Changes will be made to the tool before OE3 to improve functionality for reporting purposes in the future, but given the complexity and diversity of funding sources for in-person assister programs, different users will find different utility in the reporting functions. It is worth noting that the Enroll America Data & Analytics team regularly runs reports and assists field staff and partners in analyzing Connector data so that they can use this information to inform ongoing decision-making.
Three Partners, Three Approaches to Using the Connector
The ways organizations used the Connector varied considerably depending on their needs and capacity. Here are a few examples from our partners on how they fit the Connector into their outreach and enrollment program in OE2:
- Cover Oregon, Oregon’s supported state-based marketplace, used the Connector as the official online “find help” tool for its network of enrollment assisters and insurance agents. Cover Oregon and partners sent consumers seeking information and help getting coverage to its site, which led many consumers to the Connector; in fact, Oregon had more unique visitors into the Connector than any other state.
- The Ohio Association of Food Banks, the state’s largest Navigator grantee, led a statewide coalition of 28 organizations that together used one Connector subscription and integrated it into their outreach and enrollment efforts. In total, they offered assistance at more than 300 locations across Ohio.
- Tenet Healthcare, a broad health care services company that includes 80 hospitals, was a multi-state partner that ran an outreach campaign with on-the-ground organizers and a digital and mailed advertising effort focused on getting consumers to make appointments through the Connector with Tenet’s CACs.
Insights Into Consumers’ Behavior, Needs, and Preferences
In addition to providing program-specific data to maximize assisters’ reach and efficiency, the Connector also yielded new findings that are relevant to the enrollment community as a whole.12
- Seven in 10 consumers completed their appointments.
- Nearly half of completed appointments led to enrollments.
- Consumers were more likely to show up for their appointment if they were in contact with a person before the appointment.
Assisters using the Connector entered an outcome status for nearly 60 percent of their appointments. Of this subset of appointments with outcomes, about seven in 10 consumers completed their appointments as scheduled, 15 percent did not show up, and another 15 percent of consumers cancelled or rescheduled their appointment. Compared with other industries’ standards for no-shows, this represents high turnout, suggesting that consumers value and prioritize getting in-person help and that the automated reminders worked well for this population.13 Of all completed appointments, about half led to enrollment or renewal with the other half not completing the enrollment process (because they were not eligible, chose not to enroll, or simply did not complete the process in the allotted time).
Completion rates differ depending on a number of variables. Appointment completion rates were moderately higher when consumers were in contact with a person before the appointment, either by signing up for an appointment over the phone or receiving a reminder call after otherwise being scheduled for an appointment. Consumers with personal contact prior to the appointment were 22 percent more likely to complete their appointments, compared those who were not in contact prior to their appointment.
Distances to In-Person Assistance
- Consumers traveled an average of 4.5 miles to an appointment.
- Forty-three percent of uninsured 18- to 64-year-olds lived within 10 miles of a Connector appointment.
- Seventy-seven percent of uninsured 18- to 64- year-olds lived within 10 miles of any Connector assistance location.
On average, Connector consumers traveled 4.5 miles to their appointment.14 This varied considerably by state; in Pennsylvania consumers traveled an average of 2.9 miles, while in Georgia they traveled 8.6 miles.
Additional distance analysis looked at the accessibility of in-person assistance through the Connector by seeing how well Connector locations aligned with where uninsured individuals live across the country.15 Nationwide, 31 percent of uninsured Americans 18- to 64-year-olds lived within 4.5 miles of an appointment available through the Connector (even though appointments were only offered in 24 states and DC). When expanding this measure to 10 miles, the proportion increases to 43 percent of uninsured 18- to 64-year-olds. For all locations listed in the Connector (including locations with appointments and those that only listed contact information), 60 percent of uninsured Americans live within 4.5 miles of in-person assistance, and 77 percent lived within 10 miles.16 For a breakdown of distance by state for Enroll America field states, see Figure 4.
Figure 4: Proportion of Uninsured Population Within 4.5 Miles and 10 Miles of Get Covered Connector Locations and Appointments, November 15, 2014 – February 22, 2015
These measures show that in-person assistance is indeed accessible to many uninsured Americans and that the Connector appointment tool had significant reach in OE2, its initial year. However there is still potential for improvement at making assistance more accessible to more of the uninsured.
Popular Appointment Times and Days
- Saturday at 10:00 a.m. was the single most popular appointment slot.
- Weekend appointments were almost twice as likely to be filled as weekday appointments.
- Appointment completion rates were similar for weekends and weekdays, but enrollment rates were 20 percent higher on weekends.
The vast majority (84 percent) of appointments offered in the Connector were during the week. However, weekend appointments were almost twice as likely to be filled, with 19 percent of weekend appointments filled compared to 11 percent for weekdays. In particular, consumers made more appointments on Saturdays (12,000 appointments or 20 percent of all appointments made), and Saturday appointments filled up at higher rates than any other day (20 percent). Of completed appointments the most popular was Saturday at 10:00 a.m. Sunday appointments filled at a rate nearly matching Saturdays (18 percent) although there were fewer appointments offered on Sundays.
This finding reinforces the common understanding that assisters should make their services available at times most convenient for consumers — which, for many working families who are eligible for the marketplaces, is the weekend. Assisters should consider offering more weekend appointments. Since there were many fewer Sunday appointments offered in OE2, adding more Sunday appointments in OE3 could be an opportunity to increase weekend offerings if Saturday capacity cannot be expanded.
Based on the available appointment data, consumers showed up for their appointments at similar rates on weekends as weekdays. However, enrollment rates diverged; consumers with weekend appointments were 20 percent more likely to enroll (52 percent of appointments led to enrollment on weekends compared with 43 percent on weekdays). There are a number of factors that could be driving this difference. Perhaps weekday and weekend appointment attract different kinds of individuals, with weekend consumers recognizing that the only enrollment assistance that fits their schedule for the next week is that day, so they may be more motivated to complete their applications that same day. Or perhaps enrollment events, which are often held on Saturdays, are particularly efficient at converting interested consumers into enrollees.
Behavior Changes Surrounding Deadlines
- Consumers made more appointments before deadlines, with a sharp peak in demand at the end of open enrollment.
- Consumers were more likely to complete appointments and enroll during appointments prior to deadlines.
As expected based on OE1 enrollment patterns, more consumers scheduled appointments, completed appointments, and enrolled in the weeks prior to OE2 deadlines.
Compared with the first two weeks of OE2, twice as many consumers completed appointments in the two weeks leading up to the December 15 deadline and four times as many consumers completed appointments before the February 15 deadline. (See Figure 5.)
Consumers were also somewhat more likely to show up for their appointments and more likely to enroll before major deadlines. Appointment completion rates and enrollment rates were modestly higher (4 percent) during the weeks leading up to coverage deadlines on December 15, January 15, and February 15.
When planning for OE3, assisters should take into account these fluctuations in demand for application assistance over the open enrollment time period. Making sure as many assisters as possible are available for direct application assistance in the weeks before deadlines, particularly the final end-of-open-enrollment deadline, is critical to ensuring as many consumers as possible get the help they need to enroll. Conversely, assisters should plan more outreach and education work at times with lower demand for application assistance in order to use the time more efficiently and effectively.
Figure 5. Consumer Appointments and Enrollments During OE2, November 15, 2014 – February 20, 2015
Ten Tips for Using Centralized Scheduling Tools for In-Person Assistance
1. Commit to using a tool early, well in advance of the desired start date (e.g. the start of the annual open enrollment period), and allow time for onboarding staff, working out initial kinks, training, recruiting partners, etc.
2. Designate a point-person in your organization or coalition whose responsibility it is to understand how the tool works and regularly monitor its use.
3. Post assisters’ schedules publicly as soon as possible. Training and outreach should follow closely after, but getting the schedule posted is a critical first step.
4. Make more appointments available during the most popular times, including weekend mornings, especially those immediately preceding important enrollment deadlines.
5. Train the appropriate staff on how to use the tool, but expect to provide ongoing technical support throughout the open enrollment period.
6. Develop a robust outreach strategy to support the tool’s implementation.
- Reach out to other organizations that work on outreach and enrollment in your community. Ensure they know how to schedule appointments and, if appropriate, invite them to use the tool as well.
- Consider coupling the tool with a statewide or regional telephone hotline, so consumers can make appointments over the phone.
- Develop a digital marketing strategy and encourage relevant organizations to embed the tool on their website.
- Work with local media to push the tool out when they cover enrollment-related stories.
7. Remind consumers about upcoming appointments via email, text, and reminder calls. (Note, the Get Covered Connector automatically sends consumers email and text reminders before an appointment.)
8. Commit to regular data entry (at whatever level of the organization can be relied upon to regularly and consistently record information).
9. Review the data and use it to make adjustments to your program to better meet consumers’ needs and achieve desired outcomes.
10. Remember: A tool is just a tool! It is only one part of an effective outreach and enrollment effort. It makes much of the work more effective and efficient, but it does not replace the need for regular outreach, consumer-tested messaging, planning, or community-based partnerships.
In-person assisters will continue to play a vital role in connecting consumers to coverage, especially those with more complicated situations. Increased coordination among assisters and between the assisters and the broader outreach community will ensure that enrollment efforts are highly effective and highly efficient. Centralized scheduling, as evidenced through implementation of the Get Covered Connector, offers one promising approach for enhancing coordination and making it easier for consumers to find and connect to local assistance.
This piece was written by Molly Warren, Senior Policy Analyst, Best Practices Institute, and Jennifer Sullivan, Director, Best Practices Institute.
Assistance was provided by Anna Gilbert, National Reporting Director, Enroll America Data & Analytics.
The authors wish to thank the following individuals from Enroll America:
- Ernie Anderson, John Gilbert, Chris Mendoza, Lizzy Salinas, Adam Stalker, and Sophie Stern for their input and guidance.
- William Tomasko and Cynthia Youngblood for their editorial and design support.
2 Enroll America, In-Person Assistance Maximizes Enrollment Success, March 2014. Available online at: https://www.enrollamerica.org/in-person-assistance-maximizes-enrollment-success/; PerryUndem Research & Communication and Enroll America, The Affordable Care Act’s First Open Enrollment Period: Why Did Some People Enroll and Not Others?, May 2014. Available online at: https://www.enrollamerica.org/resources/public-education/why-did-some-people-enroll-and-not-others/; Action for Health Justice, Improving the Road to ACA Coverage: Lessons Learned on Outreach, Education, and Enrollment for Asian American, Native Hawaiian, and Pacific Islander Communities, September 2014. Available online at: http://www.apiahf.org/resources/resources-database/improving-road-aca-coverage-lessons-learned-outreach-education-and-enro.↩
3Enroll America analyzed the enrollment rates of two populations during the first month of the second open enrollment period: consumers who scheduled and attended an appointment through the Get Covered Connector (51 percent enrollment rate) and consumers who began the enrollment process online by visiting GetCoveredAmerica.org (32 percent enrollment rate based upon a self-reported data from a telephone survey).↩
4Enroll America listed assister locations for free but charged assisters a fee to use the scheduling component of the tool.↩
5 Assisters in the District of Columbia and the following states made appointments available through the Get Covered Connector during the second open enrollment period: AK, AZ, CA, FL, GA, IL, IN, KS, LA, MI, MO, NC, ND, NJ, NY, OH, OR, PA, SC, SD, TN, TX, VA, and WA.↩
6 During OE2, Enroll America had field staff in AZ, FL, GA, IL, MI, NJ, NC, OH, PA, TN, and TX.↩
7 PerryUndem Research & Communication and Enroll America, The Affordable Care Act’s First Open Enrollment Period: Why Did Some People Enroll and Not Others?, op. cit.↩
8 Enroll America, State of Enrollment: Lessons Learned From Connecting America to Coverage, op. cit.↩
9 Enroll America also tested other email reminder content with a small subset of appointments (approximately 7,000), including the steps to creating a HealthCare.gov account prior to the appointment (in applicable states), and important considerations when selecting a health plan, op. cit.↩
10 Get Covered Connector user survey, March 2015.↩
11 Get Covered Connector user survey, March 2015.↩
12 This section looks at the appointments offered and scheduled through the Connector between November 15, 2014, and February 22, 2015, which includes the 2015 open enrollment period as well as the week-long Special Enrollment Period offered in most states to allow consumers to complete applications that they started before the deadline. Data on appointments offered and scheduled, outcome information was not entered for about 40 percent of all appointments scheduled (25,640 of the 61,847 scheduled appointments). These cases are not necessarily no-shows; they are simply incomplete data entry. Our analysis is based on the best information we had available (e.g. for outcomes, all analysis is based on the 36,207 appointments for which outcome data are available).↩
13 For example, no-show rates for doctors’ appointments (without reminders) are typically between 1 in 4 and 1 in 5 of every appointment scheduled. See: http://www.amjmed.com/article/S0002-9343(10)00108-7/abstract.↩
14 This was determined by calculating the distance between the searched ZIP code or consumer’s address (if they provided it, although not many did) and the address of the appointment.↩
15 In 2013 and 2014, Enroll America and Civis Analytics created a model predicting the likelihood that a U.S. adult is uninsured based on demographic and geographic information from thousands of phone interviews with consumers combined with other publicly available information. This model was applied to a database of 200 million U.S. adults to prioritize outreach efforts in locations and to certain populations who are most likely to be uninsured. For more information: https://www.enrollamerica.org/ blog/2014/10/updating-our-data-model-to-find-the-uninsured/.↩
16 4.5 miles is used as a proxy for accessibility since consumers using the Connector traveled an average 4.5 miles to an appointment.↩