Psychosocial Factors Are Associated with Electronic Portal Registration

Abstract

Background

Electronic health records often include a portal for secure patient-clinician communication. There is evidence that use of electronic portals increases satisfaction, treatment adherence, safety, and clinical outcomes. We want everyone to enjoy these benefits and we noticed low and uneven portal use. We studied factors that we can address to improve portal use.

Questions/purposes

After controlling for differences in patient characteristics, what psychological and demographic factors are associated with an increased likelihood of registering for an electronic health record portal among people seeking musculoskeletal specialty care?

Methods

We reviewed data on 5672 adult English or Spanish-speaking patients seen in a musculoskeletal specialty office between October 2017 and December 2019. Eighteen percent (996 patients) had missing measures of symptoms of depression and anxiety due to intermittent problems with survey technology, leaving 4676 for analysis, 42% (1970 of 4676) men and 58% (2706 of 4676) women with a mean age of 51±15, 76% (3569 of 4676) of patients were English speaking, 22% (1015 of 4676) were Spanish speaking, and 2% (92 of 4676) spoke another language. Seventy-seven percent (3620 of 4676) of patients were residents of Austin, Texas, USA, 4% (159) were from Pflugerville, Texas, USA, 3% (143) were from Del Valle, Texas, USA, and 16% (754 of 4676) were from other areas of Texas. Ninety nine percent of patients were residents of Texas (4645 of 4676). Twenty-three percent of patients visited the upper extremity team (1077 of 4676), 37% the lower extremity team (1721 of 4676), 21% the back and neck team (1002 of 4676), and 19% the sport medicine team (876 of 4676). Seventy eight percent of patients (3654 of 4676) registered in portal and 22% (1022 of 4676) did not. The omitted population were not different from our study population in terms of age, gender, language, residence, and region of symptoms. We used a two-question measure of symptoms of depression (Patient Health Quality-2 [PHQ-2]) and a two-question measure of symptoms of anxiety (General Anxiety Disorder-2 [GAD-2]). The primary outcome was portal registration. To account for potential confounding, a multivariable logistic regressions model was used to determine the influence of age, spoken language, city and state of residence, care team, number of completed visits and GAD and PHQ scores on portal registration.

Results

After controlling for potentially confounding variables such as state of residence, we found younger age (odds ratio 0.98 [95% CI 0.97 to 0.99]; p < 0.01), speaking English (OR 1.85 [95% CI 1.14 to 3.02]; p = 0.01) rather than Spanish (OR 0.27 [95% CI 0.17 to 0.45]; p < 0.01), seeking care for back or neck symptoms, (OR 3.84 [95% CI 2.60 to 5.66]; p < 0.01) and higher number of completed visits (OR 1.03 [95% CI 1.01 to 1.05]; p < 0.01) were associated with an increased likelihood of portal registration while living in Austin, Texas, USA (OR 0.68 [95% CI 0.53 to 0.87]; p < 0.01) and Del Valle, Texas, USA (OR 0.47 [95% CI 0.30 to 0.74]; p < 0.01) compared with Pflugerville, Texas, USA, or other cities, seeking care for upper extremity (OR 0.74 [95% CI 0.58 to 0.94]; p = 0.01) or lower extremity symptoms (OR 0.68 [95% CI 0.53 to 0.86]; p < 0.01), and greater symptoms of anxiety (GAD score) (OR 0.97 [95% CI 0.95 to 0.99]; p < 0.01) or depression (PHQ score) (OR 0.97 [95% CI 0.95 to 0.98]; p < 0.01) were associated with lower likelihood of registering for the portal. English language, city of residence, and seeking care for back or neck symptoms (due to insurance contracts) were all associated with higher socioeconomic status in our setting.

Conclusions

The association of better mental and social health (financial, employment, housing and food security; connectedness) with registration in a communication portal directs us to be more intentional about efforts to specifically welcome disadvantaged people to participate in the portal and to study the impact and effectiveness of such efforts.

Level of Evidence

Level III, therapeutic study.

Introduction

Electronic health records often include a portal for secure patient communication. These portals are a convenient, private method for patients to communicate with their clinicians, with immediate documentation in the medical record. There is evidence that portals can increase satisfaction [8, 25], adherence [24], safety [25, 29], and clinical outcomes [22, 24, 25, 28]. Additionally, they can help providers with decision-making, patient attendance at important return visits, and timely and appropriate follow-up for patients with abnormal diagnostic test results [12].

Ideally, all patients would have taken advantage of the facilitated interaction with the care team afforded by the portal, but national registration is only 53% [13]. We want everyone to enjoy these benefits in the context of specialty care; however, we noticed low and uneven portal use. Prior studies of people attending diabetes or cancer department specialty care identified age, gender, race, education, income, and clinic type are associated with rate of registration for portal communication [1, 2, 7, 11, 23]. The influence of psychological factors on portal registration is less-well studied: We did not find any accounting for mental health in the primary or specialty care settings. The association of symptoms of depression and anxiety with the content and ratings of in-person patient-clinician communication effectiveness is well documented [5, 20, 26, 27]. We studied factors that we can address to improve portal use.

We asked the following question: After controlling for differences in patient characteristics, what psychological and demographic factors are associated with an increased likelihood that a patient would register for an electronic health record portal among people seeking musculoskeletal specialty care?

Patients and Materials

This cross-sectional study was approved by our institutional review board. We reviewed the records of 5672 consecutive adult English- or Spanish-speaking patients seen in a musculoskeletal specialty office from October 2017 to December 2019. We omitted 18% (996) of records with missing measures of symptoms of depression and anxiety. These questionnaires were incomplete because of periodic computer system issues. The patients in the omitted population were not different from our study population in age, gender, language, residence, and region of symptoms. The records of 4676 adult patients seeing a musculoskeletal specialist remained for analysis.

We recorded patient characteristics including age, gender, language, and city and state of residence. In addition, we documented their integrated practice unit or subspecialty, number of completed and cancelled visits, and portal registration status (yes or no) in athenahealth Inc (Watertown, MA, USA).

Our back and neck specialists do not have a contract to provide care for the county payer that provides care for the poor. All patients seeing the back and neck team have commercial insurance or Medicare. Approximately 40% of patients seeking arm or leg care have commercial insurance, Medicare, or Medicaid. Medicaid is extremely limited in our state; seeing the back or neck care team is associated with a more favorable socioeconomic status.

We performed a descriptive analysis of the demographics, the care team that the patient visited with, registration in the electronic health record portal, completed and cancelled visits, Patient Health Questionnaires 2 and 9 (PHQ-2 and PHQ-9), and General Anxiety Disorder-7 (GAD-7). The study population had a mean age of 51 ± 15 years with 42% men (1970 of 4676) and 58% women (2706 of 4676). Seventy six percent (3569 of 4676) of patients were English speaking and 22% (1015 of 4676) were Spanish speaking. Seventy seven percent (3620 of 4676) were residents of Austin, Texas, USA, 4% (159 of 4676) were from Pflugerville, Texas, USA, 3% (143 of 4676) were from Del Valle, Texas, USA, and 16% (754 of 4676) were from other areas of Texas. Ninety-nine percent of patients were residents of Texas (4645 of 4676). Twenty-three percent (1077 of 4676) of patients visited upper extremity clinic, lower extremity 37% (1721 of 4676), back and neck 21% (1002 of 4676) and sport clinic 19% (876 of 4676). The omitted population were not different from our study population in terms of age, gender, language, residence, and region of symptoms. Seventy eight percent of patients (3654 of 4676) registered in the portal and 22% (1022 of 4676) did not (Table 1).

Table 1.

Demographics of the patients included in this study

People completed the PHQ-2, a validated [16, 18] two-item measure of depression symptoms from the previous 2 weeks. Scores ranged from 0 to 6. A score of 3 or greater triggered administration of the nine-item version (PHQ-9). Scores for the PHQ-9 range from 0 to 27 with higher scores representing greater symptoms of depression. In a tertiary analysis suggested by the reviewers, we analyzed people with a PHQ score of 2 or less compared with people with higher scores. We would have preferred not to categorize continuous variables to avoid reinforcing the social stigma associated with mental health, but the marked floor effect for symptoms of depression and anxiety make this reasonable.

The GAD-7 questionnaire is a validated [17], seven-item measure of anxiety symptoms from the previous 2 weeks. The total score ranges from 0 to 21, with higher scores representing greater anxiety symptoms. In a tertiary analysis suggested by the reviewers, we analyzed people with a GAD score of 2 or less compared with people with higher scores.

To account for potential confounding, a multivariable logistic regressions model was used to determine the influence of age, spoken language, city and state of residence, care team, number of completed visits and GAD and PHQ scores on portal registration (Table 2).

Table 2.

Logistic regression for Athena portal registration (yes or no)

Results

Accounting for potential confounding, registration in the electronic health record portal was associated with fewer symptoms of depression (lower PHQ score (OR 0.97 [95% CI 0.95 to 0.98]; p < 0.01) and anxiety (lower GAD-7 score (OR 0.97 [95% CI 0.95 to 0.99]; p < 0.01), younger age (OR 0.98 [95% CI 0.97 to 0.99]; p < 0.01), and total number of visits (OR 1.03 [95% CI 1.01 to 1.05]; p < 0.01). Portal registration was also more likely for English-speaking patients (OR 1.85 [95% CI 1.14 to 3.02]; p = 0.01) and patients visiting a back or neck care team (OR 3.84 [95% CI 2.60 to 5.66]; p < 0.01); portal registration was less likely for residents of Austin, Texas, USA (OR 0.68 [95% CI 0.53 to 0.87]; p < 0.01), residents of Del Valle, Texas, USA (OR 0.47 [95% CI 0.30 to 0.74]; p < 0.01), patients visiting an upper extremity care team (OR 0.74 [95% CI 0.58 to 0.94]; p < 0.01), and patients visiting a lower extremity care team (OR 0.68 [95% CI 0.53 to 0.86]; p < 0.01) (Table 2).

Discussion

Electronic health records often include a secure communication portal to organize and document communication with patients. Prior studies of factors associated with portal use have mostly focused on primary rather than specialty care and have paid little attention to psychological factors [2, 10, 11, 14, 19, 29]. We studied a large database of patients in a single multispecialty musculoskeletal office that routinely measures symptoms of anxiety and depression to assess psychological factors related to portal participation in specialty care. The identified association between better mental and social health (financial, employment, housing, and food security; connectedness) and higher registration in a communication portal directed us to look for obstacles that prevent a specific population from accessing the benefits of this new technology and to be more intentional about efforts to specifically welcome disadvantaged patients to participate in the portal and to study the impact and effectiveness of such efforts.

Limitations

This study had several limitations. To start, we excluded 18% of patient records because of incomplete information. The program for collecting psychological questionnaires is separate from the medical record, and it was new and subject to periodic technical issues that led to nearly 20% of patients not having mental health data. We found no differences between the omitted patients and the study populations, so their exclusion is unlikely to introduce bias. Another limitation is that we did not track diagnosis, so we do not know the proportion of people with more urgent problems such as fracture, metastasis, laceration, or infection compared with less urgent quality-of-life issues. The type of illness may influence a person’s desire to take advantage the opportunity for ongoing contact presented by the portal. Given the referral patterns in our context, only the upper extremity team sees acute concerns and to a limited degree. Most people seen in our offices have non-urgent concerns; our findings apply best to that context. The degree to which diagnosis influences the perceived potential to benefit from the ongoing relationship provided by portal use merits additional study. A third limitation is that the back and neck unit lacks a safety-net insurance arrangement. As a consequence, their patients are all insured (commercial insurance or Medicare), which creates confounding with socioeconomic status. We did not have 9-digit zip code or any direct measures of socioeconomic status, so we cannot determine the degree to which the presenting problem is confounded with socioeconomic status through insurance type or whether back and neck illnesses themselves influence portal enrollment. Given that our upper and lower extremity and sports units have over 60% of patients under safety net insurance, we interpret associations with the back and neck unit as most likely reflecting socioeconomic differences.

Factors Independently Associated with Portal Registration

The observation that patients with fewer symptoms of depression or anxiety were more likely to register in the portal might reflect their energy and motivation, or perhaps greater hope, and greater engagement and less fear or concern. Studies in specialty cancer and diabetes care identified age, gender, race, education, income as factors associated with portal registration, but did not address mental health. An evaluation of patients recovering from surgery found that healthier patients were more likely to register for a portal than less-healthy patients [21], but again, this did not address mental health. Our findings are consistent with evidence that symptoms of depression are associated with fewer self-care behaviors and participation in educational programs [20]. Based on our findings, the choice to decline portal registration might prompt an inquiry about mood and worry associated with the illness, start a discussion about the potential benefits of portal registration, and an offer of help to ease registration and use of the portal. The evidence that identified the association between lower health literacy and asking fewer questions has led to specific interventions that invite questions from people who might excessively defer to expertise. Similarly, the observation that portal registration and use varies according to personal and social factors highlights the potential value of specific interventions to improve specialist access through alternative communication modes such as a portal. The evidence that greater depression symptoms are associated with lower-rated communication quality in patient-clinician in-person interaction [27] suggests that electronic health portals may help people who may already be disadvantaged by a lack of energy and motivation to have an enhanced interaction with specialists. Future studies might continue to identify barriers to alternative forms of access to specialist expertise, and test interventions to improve use and equity. Such interventions have the potential to relieve symptoms and limitations.

The observation that younger age was independently associated with portal registration is consistent with the results of studies that addressed portal registration in primary care [14] as well as diabetes [6, 23] rheumatology [9, 30], cancer [9], and postoperative care [21]. A study that categorized patients by age found higher registration by younger patients in primary care [2] is in contrast to two other studies that found higher registration by middle-aged patients in speciality care for diabetes [3] and primary care [4, 7]. Older patients might not think of electronic communication as a useful tool; perhaps they feel they are incapable. Given the demonstrated benefits of portal participation, we recommend that clinicians and their teams be intentional about how they address the evidence that older patients make less use of electronic communication, that they take steps to help all people understand the potential benefits, walk them through signing up (perhaps it can become part of routine registration), and practice use of the portal to decrease intimidation. Manufacturers might make portals simpler and user friendly, requiring limited computer literacy, employing simple language in an easy-to-read format for the visually impaired, and perhaps using fingerprint or face identification to sign on to limit the need to remember or keep track of passwords.

The finding that a greater number of visits is associated with a higher likelihood of portal registration is consistent with studies in primary care [19]. More involvement with providers may reflect increased engagement and a better clinician-patient relationship. We interpret portal registration and a greater number of visits as indications that people feel they will benefit from care. Believing that one will benefit from care may involve several factors including the type of symptoms experienced, the underlying disease, agency to seek care, a trusting relationship with clinicians, a sense that care is available and affordable, and an awareness that one is worthy of care. Clinicians and their teams can be intentional about developing strategies for the modifiable factors that can increase the thinking that one will benefit from care. They can help develop a trusting relationship with patients; ensure patients feel that care is accessible and affordable; and establish a care environment that nurtures patients’ sense that they deserve easy access to quality care and that they have control over their own health. Some people might need encouragement to think of specialty care as a longitudinal, interactive process, rather than one associated only with one or a few, brief, in-person visits. Given the importance of trust and relationship to health, these ideas merit additional research.

The finding that English language speakers [10, 11, 19] had higher registration suggests actual or perceived difficulty accessing the portal for non-English speaking patients. Registration may also be confounded by lower socioeconomic status in our context. Our portal is available in Spanish which accounts for a large proportion of our non-English speaking patients. Some interventions that might enhance portal registration and use could be to make people of low socioeconomic status feel welcome and ensure access to a variety of languages. Lower registration in residents of specific cities [1, 15, 21] seems to reflect lower socioeconomic status and, perhaps, proximity to the office, which might decrease the appeal of portal usage. Patients with back and neck pain (the only specialty care team in our context without a safety-net insurance arrangement) were associated with higher portal registration may reflect socioeconomic status. We interpret these findings as evidence that people with less wealth and those who are non-English speakers might benefit from additional encouragement and help using an electronic portal to seek ongoing specialty care for their condition. Future research can measure the impact of such efforts on registration for the portal, experience with care, and improvement in symptoms and limitations.

Conclusions

The association of better mental and social health with communication portal registration suggests that efforts to improve equitable access to speciality expertise must be deliberate and strategic. A portal can provide easier, less-costly communication between clinician and patient that will promote patient health care, which is particularly useful for people with greater symptoms of depression and lower socioeconomic status. Future studies can confirm the association of psychosocial factors on patient portal registration and test efforts to ameliorate the observed inequities.