Patient portal doldrums: does an exam room promotional video during an office visit increase patient portal registrations and portal use?

Abstract

The patient portal is a web service which allows patients to view their electronic health record, communicate online with their care teams, and manage healthcare appointments and medications. Despite advantages of the patient portal, registrations for portal use have often been slow. Using a secure video system on our existing exam room electronic health record displays during regular office visits, the authors showed patients a video which promoted use of the patient portal. The authors compared portal registrations and portal use following the video to providing a paper instruction sheet and to a control (no additional portal promotion). From the 12 050 office appointments examined, portal registrations within 45 days of the appointment were 11.7%, 7.1%, and 2.5% for video, paper instructions, and control respectively (p<0.0001). Within 6 months following the interventions, 3.5% in the video cohort, 1.2% in the paper, and 0.75% of the control patients demonstrated portal use by initiating portal messages to their providers (p<0.0001).

Introduction

The patient portal is an online service that allows patients to access portions of their electronic health record (EHR) and arrange clinic appointments, allows secure communication between patients and medical staff, and facilitates prescription renewals.¹ Patient portals are offered by a number of healthcare providers including the Veterans Administration, Kaiser Permanente, Partners Healthcare, Cleveland Clinic, and Mayo Clinic.^1–4 These examples of patient portals are also described in the current informatics literature as provider-tethered personal health records (PHRs), which are also called a limited form of the integrated PHR.⁵ Although provider-specific patient portals have generally not achieved all the features of an ideal integrated PHR, they still offer significant advantages over standalone PHRs by allowing access to appointment scheduling resources and communication with the healthcare team, as well as giving access to data contained in the EHR.⁶ In addition, patient portals can deliver healthcare decision support such as preventive service reminders.⁷

Although patients benefit from many of the conveniences afforded by the patient portal, many advantages of the portal also accrue to the practice. It is anticipated that organizations fully implementing a patient portal will have a significant financial benefit. Because the portal often standardizes and automates appointment scheduling, medication refills, triage, and patient education, there should be a significant reduction in labor costs of nurses and schedulers who generally do that work.⁸

Despite the potential for cost saving and improved patient care with patient portals, there has been a lack of published research on best practices for encouraging patients to sign up and use the portal. Current literature shows that portal adoption is not rapid, and a Markle Foundation survey found that 90% of the public do not have a PHR.^{9 10} Kaiser Permanente, for example, achieved a registration of only a third of their patients after 2 years with a median time to registration of 409 days.⁴ Recognizing the need to increase portal adoption, Partners Healthcare has employed aggressive marketing, using up to 16 different strategies to increase portal adoption including automated greetings in the practice's telephone system, posters in waiting areas and examination rooms, postcard and letter mailings, staff speaking to the patient in the office or over the telephone, and onsite enrollment with a computer kiosk.¹¹ Recognizing the need for portal adoption research, medical informaticists have suggested a way forward. In their white paper on PHRs, Tang et al advocated the need to leverage a ‘teachable moment’ to get users to realize the value of using the PHR.¹²

In this study, we examine the use of the face-to-face office appointment as a teachable moment to educate patients about the Mayo Clinic patient portal and to promote portal registration. We used software that allowed us to lock patients out of the EHR while we showed them an educational and promotional video on the existing exam room display while they waited for the provider. We compared portal registrations following the video, to registrations following a paper instruction sheet, and to registrations after no portal promotion at all (control).

Methods

The study took place in an adult primary care clinic of the Division of Primary Care Internal Medicine Mayo Clinic, Rochester, Minnesota. The primary-care clinic has 39 staff physicians and 94 resident physicians with a combined patient panel size of 38 181 at the end of 2010. Patients were given access to the portal starting in April 2010. Prior to the study, patients were passively informed about the portal with exam room posters and an accompanying flyer that they could pick up and take with them for further information. The study design was a prospective three-cohort study: video intervention, paper instruction intervention, and a control (no additional intervention). The interventions occurred simultaneously in all cohorts from November 29, 2010 to January 4, 2011. For additional evidence of intervention effect, we also performed a pre–post analysis within each cohort starting from October 1, 2010 to the start of the prospective intervention.

Interventions

For the video cohort, the exam room EHR computer also served as a video-display for the patient. The patients in the video cohort were shown a 2 min video promoting the features of patient online services (see.MOV file, or online link under online services at http://www.mayoclinic.org/patient-visitor-guide). In addition, the video cohort was given an additional paper sheet explaining how to register for the portal. The video explained Mayo Clinic's online services, such as visually demonstrating how patients can access part of their EHR (lab test results, medications, allergies, immunizations) and also showing them the convenience of making online appointments, medication refills, and how they could review already scheduled appointments and create secure messages or initiate e-visits to their provider. Video intervention patients were also asked to complete a survey regarding their satisfaction with the video. Providers were empowered to stop the video if they entered the room before the video ended. Patients in the paper intervention were given the same paper instruction sheet, but did not see the exam room video. The control cohort did not have exposure to either the video or the paper instruction for online services registration. Patients needing translators did not get the video intervention. Except for the interventions, the rooming process for the cohorts was the same.

Measures

The measure for the intervention effect was the proportion of patient online service registrations occurring within 45 days of an eligible appointment. Eligible appointments were determined by excluding those appointments requiring a translator and those who were already signed up. Otherwise, we did not exclude any patient with a provider office visit. In particular, we excluded no one based on internet or computer use or accessibility. Subsequent use of the portal following interventions was determined by portal messaging use within 6 months of the intervention (until June 6, 2011). Portal messaging was patient initiated secure messages to care providers or e-visits with their physician.

To measure the effort in implementation of the video system we used person-hours. Patient satisfaction with the video intervention was measured using a five-point Likert scale: strongly agree, agree, neutral, disagree, and strongly disagree. We used surveys of providers and roomers to measure any disruption of the office visit, potential access problems, and provider satisfaction.

Data collection

The video delivery software¹³ captured the number of times the video was shown to the patients. Patient and provider satisfaction data were collected on paper surveys, and implementation effort for the video cohort was collected with a web-based REDCap database.¹⁴ We used the Mayo Clinic appointment database to capture the number of provider face-to-face visits occurring during the study interval. The Mayo patient portal database captured the date of registration.

Statistics

All provider visits between October 1, 2010 and January 4, 2011 were used in the analysis of registrations except those excluded for foreign language or because the patient had already registered for portal use. Comparisons were made both across interventions and pre- and postintervention (intervention start date was November 29, 2010). Categorical data were compared using the χ² test, and ANOVA was used for continuous variables. We used multivariate logistic regression analysis to determine adjusted ORs. JMP 8.01 (SAS) was the software used for analysis.

This study was approved by the Mayo Clinic Institutional Review Board.

Results

We analyzed portal registrations following 12 050 appointments(table 1). There were 5035 appointments in the three cohort concurrent intervention analysis and 7015 appointments prior to the intervention used for the additional pre–post analysis.

Table 1.

Comparison of video intervention demographics and registration percent to paper instruction intervention and control (no additional patient portal promotion)

	Control (95% CI)	Paper intervention (95% CI)	Video intervention (95% CI)	p Value (Ho: groups are the same)*
Total provider face-to-face appointments	2079	1248	1708	NA
Appointment count excluding foreign language and already signed up for online services	1505	928	1270	NA
Percent of visits eligible (not excluded/total visits)	72.4	74.4	74.4	0.3
Percent female (for eligible visits)	55.3	57.0	55.6	0.69
Percent non-employee (for eligible visits)	89.5	89.1	89.8	0.86
Mean patient age (for eligible visits)	61.6	62.8	62.3	0.31
Registration count within 45 days after appointment	37	66	148	NA
Percent registration during intervention period (95% CI)	2.5 (1.8 to 3.4)	7.1 (5.6 to 8.9)	11.7 (10.0 to 13.5)	<0.0001
Percent registration during preintervention period (95% CI)	2.82 (2.16 to 3.67)	3.33 (2.53 to 4.29)	3.88 (3.01 to 4.88)	0.2
Adjusted OR of registration postintervention to registration preintervention in each cohort (95% CI)†	0.91 (0.59 to 1.4)	2.3 (1.6 to 3.4)	3.39 (2.4 to 4.4)	NA
Portal use (percent of patients who initiated at least one portal message within 6 months of intervention)	0.75 (0.41 to 1.4)	1.2 (0.65 to 2.2)	3.5 (2.5 to 4.7)	<0.0001

^*ANOVA for continuous variable (age), χ² for counts (heterogeneity test).

^†Adjusted for age, employee status, sex.

There were 1038 launches of the video from a total of 1270 face-to-face eligible appointments (82%). We collected 647 patient surveys concerning the 1038 video launches (62%).

Of the 647 patients surveyed following the video, there were only 42 (6.5%) who agreed or strongly agreed to the statement ‘I prefer not to watch a video while waiting.’ All 13 rooming staff involved in the study completed a survey regarding how the patient education video impacted their work flow. With the particular software setup, 12 of 13 (92%) agreed that it was easy to use, and six of 13 (46%) agreed that the ‘video did not take more of my time,’ while an additional four of 13 (31%) were neutral to that statement. None of the rooming staff considered the video made exam room access more difficult. Of the providers involved with the video, 24 of 36 (66%) responded to a survey about their experience. Twenty of 24 provider respondents (83%) agreed that ‘I did not feel the video adversely affected the encounter with the patient.’ Only two of 24 providers (8%) disagreed with ‘My overall satisfaction with the video system is high.’

The video was not shown in 18% of eligible visits. Although we did not collect data on all the reasons why, we do know from troubleshooting logs that software/equipment problems accounted for less than 1% of failures to show the video. Patients had the option to refuse the video, and rooming staff told us that patient refusal was the main reason for not showing it. Our intent was to evaluate the entire video intervention process, so all eligible visits were included in the analysis, even those where the video was not shown.

Discussion

To our knowledge, this is the first study to quantitatively study the use of a promotional video during the face-to-face appointment as a teachable moment to educate patients about the patient portal. We were able to show significantly higher registrations and subsequent portal messaging following the use of a patient portal promotional video. We found no major barriers to the implementation of an exam room video system beyond a modest initial investment of time and resources. Moreover, workflow was not appreciably disrupted for the providers or rooming personnel, and patients did not mind watching the video while waiting in the exam room for the provider.

Although the 11% registration rate following the video is not dramatic, it does represent an absolute increase of 4.5% over the paper intervention and a 9% increase over no additional intervention. The odds for registration after the video were 1.7-fold greater than for paper instructions alone and over five times greater than no registration effort. An increase in registrations also translated into a significant increase in messaging for the video compared to the control and paper interventions. These findings are complementary to and consistent with others. For example, the Partners Healthcare group found that aggressive marketing of their portal (Patient Gateway) resulted in a significant increase in portal adoption (OR 2.92; 95% CI 1.58 to 5.40).¹¹ Our study shows that an additional tool, the exam room video, can be successfully implemented and used in a workflow-friendly way to increase portal registration and subsequent portal message use. The paper intervention also showed a significant increase in registrations compared to the control. Unlike the video, however, the paper intervention cohort did not have a statistically significant increase in messaging at 6 months compared to the control, although there was a positive trend (table 2). This suggests that an increase in registration translated to an increase in use of the portal for the video cohort but perhaps to a lesser degree, if at all, for those in the paper intervention.

Table 2.

ORs comparing cohorts for subsequent portal registration and portal use

	Unadjusted OR	p Value*	Adjusted OR†	p Value*
Portal registration
Video to control	5.2 (3.6 to 7.6)	<0.0001	5.4 (3.8 to 8.0)	<0.0001
Paper to control	3.0 (2.0 to 4.6)	<0.0001	3.1 (2.1 to 4.8)	<0.0001
Video to paper	1.7 (1.3 to 2.3)	0.0004	1.7 (1.3 to 2.3)	0.0004
Portal messaging
Video to control	4.7 (2.4 to 9.6)	<0.0001	4.8 (2.5 to 10.3)	<0.0001
Paper to control	1.6 (0.67 to 3.9)	0.29	1.6 (0.65 to 3.9)	0.31
Video to paper	3.0 (1.5 to 6.0)	0.0015	3.0 (1.5 to 6.3)	0.001

^*H_o: OR=1.

^†Adjusted for age, employee status, sex.

The video fulfills many of the requirements identified by Wald for successful portal promotion in a primary care practice.¹⁵ Wald noted that successful portal promotion required full physician and staff endorsement, and some demonstration of the benefits of the portal. The video demonstrates the major features of the portal, showing patients benefits of the portal while at the same time indicating full endorsement by physicians and staff. However, despite the ability of the video to meet some requirements identified by Wald for successful portal registration, it does not reach outside clinic walls like promotional audio messages during calls to the clinic or email broadcasts. Although the video successfully enhanced portal registrations and portal use when used within clinic walls during a teachable moment, its use should supplement rather than supplant other promotional methods such as telephone and email promotions that have been shown to increase registrations beyond clinic walls.

This study had some limitations. We did not correct our analysis for racial differences which are known to influence registrations. However, similar findings from our pre–post analysis indicate that it is unlikely for race adjustment to change our conclusions, and our greater than 90% white patient population would also minimize an effect size change for unadjusted variation of non-Caucasians between the cohorts. Our patient racial mix itself is a limitation. With our limited non-white population, we cannot generalize our findings to populations where patient registration is known to be more difficult. Also, although message use within 6 months following the video was significantly greater than either the control or paper intervention, we do not know whether this increase in portal use persists for a longer period of time. Another limitation is that the study design did not allow us to isolate solely the effect of the video. The video promotion occurred in the context of the office visit where there was a provider–patient interaction. The video likely prompted some provider–patient discussions which would have been an additional influence above and beyond the video alone. Thus, our findings may not apply to similar videos shown outside the office setting. For example, sending our video to patients via an email with an online link could be an important additional way to boost portal recruitment, but our study is unable to inform readers about the efficacy of that recruitment method.

While exam-room videos may seem to be a costly way to increase portal registrations, practices are increasingly purchasing computers and video-displays for their EHRs. It makes sense to maximally utilize this exam-room technology by having it serve two purposes, as an EHR display and a patient video education device. It is anticipated that this type of video system eventually will be used for other educational purposes. Just as patients were influenced to sign up for the patient portal, the face-to-face teachable moment can be extended to video education on the benefits of screening colonoscopy or prostate-cancer screening, or for education on an advance directive.^16–18 In addition to showing how video can boost sluggish portal registrations, this study shows that portal use also increases following a video promotion. Further research is needed to show whether this promising EHR add-on feature can sustain its value with other uses.