Computer Use in the HIV Primary Care Clinic : Patient and Provider Perspectives

Background: Patients infected with HIV face unique psychosocial stressors thus good quality patient-provider relationships are essential. Objective: The objective of this study was to determine whether computers use by providers during outpatient visits for HIV-infected patients negatively affected the encounter. Design/Participants: Two hundred HIV-infected patients and twenty HIV providers were surveyed at an adult HIV clinic in a cross sectional survey study. Main Measures: Main measures used for this study were demographic variables and survey questions based on effects of computer use on interpersonal contact and communication using a Likert Scale. Results: The majority of patients felt that it was appropriate for a provider to use a computer during the visit and were satisfied with the care they received. However, patients who did not own a personal computer, were female, had lower educational status, had detectable viral load, had lower CD4 cell count or had not disclosed their HIV status were more likely to negatively perceive provider computer use. Most providers felt that they missed non-verbal cues when using the computer (70%) and that computer use in the exam room did not improve the relationship with the patient (75%). Conclusion: Providers can be assured that patients do not view the computer as negatively as they do. The study identified specific patient populations which may benefit from providers using the computer less in the exam room.


Introduction
In the United States, in the 1970s, the first Electronic Medical Record (EMR) systems were developed with the goal to track and record patient data for the next visit.With the Health Insurance Portability and Accountability Act (HIPAA) in 1996, there was growing interest to expand the utility of EMR on how it could be used to adhere to laws of privacy and security of health information.In 2009, the Health Information Technology for Economic and Clinic Health (HITECH) Act was passed, which gave twenty billion dollars to promote and expand health information technology in health care.
interacting with the patient [7], there is concern that it will affect the patient provider communication through less eye contact and missed nonverbal communication [8][9][10].
The patient-provider relationship is a significant component of the delivery of healthcare and more so for patients infected with the human immunodeficiency virus (HIV).These individuals face unique psychosocial stressors, including the chronic nature of the disease, need for strict adherence to medication regimens, and often multilayered stigma [11].Research has shown that a patient-centered approach and good quality patient-provider relationships for the care of HIV-infected patients promoted adherence to medication therapy [12][13][14][15][16] and was associated with an undetectable HIV viral load [17].
Research on computer use during the clinical encounter in the outpatient setting has demonstrated a positive to neutral effect on patient satisfaction [18][19][20] although no published studies have addressed computer use and the impact on the patient-provider relationship among HIV-infected patients.We further explore this complex interaction from both provider and patient perspectives by assessing the perceptions of both HIV-infected patients and providers on computer use during an outpatient encounter.These assessments were based on factors known to be associated with patient satisfaction [21] and the effects of computer use on interpersonal contact [22].

Patient participants
All patient participants were recruited from December 2014 to January 2015 at an adult ambulatory HIV clinic at an academic medical center, Yale New Haven Hospital, in New Haven, Connecticut, United States.The EMR system and computers were installed at this site in the year 2000, approximately 14 years prior to the study.The installment of the EMR system into the clinic was mandatory for all ambulatory clinics at Yale New Haven Hospital.
Patients were eligible to participate in the study if they were infected with HIV, had a routine or urgent care appointment with a provider (including attending physicians, resident physicians, physician assistants, or nurse practitioners; excluding nursing visits), were able to provide verbal informed consent, and were able to read and comprehend English.The study was approved by the Yale University Institutional Review Board.
Eligible participants were asked to participate at the time they arrived for their appointment or immediately after their encounter with their provider.After providing verbal informed consent, an anonymous survey was completed by the participants following their encounter with their provider.Surveys were compiled and coded by number prior to data entry.One researcher administered all of the surveys to minimize multiple surveys by any one participant.

Provider participants
All provider participants were recruited from the same adult ambulatory outpatient HIV clinic from February 2015 to August 2015.Providers are required to input data into the EMR for the clinic visits for purpose of billing and to be compensated for the clinic visits.The input of data is fixed, as visits require entry of data such as history, vital signs, physical exam, and concerns addressed during visit, and entry of orders such as blood work and prescription medications.However, the timing of when data is entered into the EMR varies by provider, based on provider preferences and visit with the patient.For example, some providers input data with the patient in the exam room to be efficient and complete work, while others input data after the visit in an effort to maintain the patient-provider relationship.Provider recruitment did not begin until after the patient participant recruitment was completed.Providers were eligible if they were attending physicians, resident physicians, physician assistants, or nurse practitioners; nurses and medical assistants were excluded.
Providers were approached before or after their clinic sessions.After providing verbal informed consent, an anonymous survey was completed by the participant.Surveys were compiled and coded by number prior to data entry.One researcher administered all of the surveys to minimize multiple surveys by any one participant.

Surveys
Distinct surveys were designed for the patient and provider arms of the study.The patient survey was composed of forty-four questions, which included demographics, extent of personal computer use, selfreport of HIV viral load and CD4 cell count, as well as twenty-five items related to computer use during the clinical encounter and the patient-provider relationship.The provider survey comprised a total of thirty-one items, which included questions regarding demographics, computer use, as well as twenty-two items related to computer use during the clinical encounter and patient-provider relationship.A majority of the questions related to computer use during the clinical encounter and patient-provider relationship were the same for both surveys.Responses for both surveys were based on a four-point Likert scale [23] without a neutral option.Neutral option was omitted to avoid the respondents' behavior of survey satisficing and answer choice ambivalence [24].

Statistical analysis
The characteristics of survey participants were summarized using frequency and percentage for categorical variables, and mean (standard deviation) and median (range) for continuous variables as appropriate.The comparison of responses between patients and providers were conducted using Chi-squared test or Fisher's exact test.The four-scale responses were dichotomized as binary outcomes, (i.e.Agree vs. Disagree) for logistic regression analysis.Logistic regression analyses were performed to examine the association of patients' characteristics and their response.Odds ratio and 95% confidence interval were presented.Statistical analysis was performed using SAS 9.4 (SAS Institute, Cary, NC).Significance level was set at P less than 0.05, two-sided.

Patient characteristics
Two hundred patient participants were surveyed with a median age of 52 years (range 20 to 76 years).Thirty-seven percent of patients were female and approximately half of the patients (52%) were African American.The majority of patients had a self-reported undetectable viral load (74%) and CD4 cell count greater than 200 cells/µL (71%).Most patients (86%) had disclosed their HIV status to another person, including a family member, significant other, friend, or colleague.Only a small percentage of patients (5%) were cared for by resident physicians.Other demographic factors are outlined in table 1.
Table 2 summarizes the patient's response to select questions in the patient survey.Patients expressed a general positive attitude toward computer use in the exam room.Among the 200 patient participants, 192 (96%) responded that their providers used the computer during their visit.Among this subgroup, personal computer use, gender, education level, HIV viral load, CD4 cell count, and HIV disclosure

Journal of HIV and AIDS
Open Access Journal status all served as significant predictors of whether a patient agreed with survey questions.
Patients with personal computers voiced less concern for computer use by providers compared to those without personal computers.Patients without personal computers were more likely to agree that, "My provider spends less time listening to me when he/she uses the computer (Odds Ratio (OR) 4.2, 95% Confident Interval (CI) 1.7-10.7,P=0.003)." They were also more likely to agree that "My visit is less personal because my provider uses the computer (OR 2.5, CI 1.3-5.0,P=0.009)." Additionally, patients expressing less frequent computer use were more likely to agree that, "I wish my provider would let me see what he/she is doing on the computer" (Monthly use vs Daily use: OR 1.1, CI 0.3-4.6;Weekly use vs Daily use: OR 5.4, CI 1.7-17.6;P=0.02).Furthermore, monthly users were more likely than daily users to agree that "I wish my doctor would not use computer during my visit" use vs Daily use: OR 8.9, CI 2.0-39.0,P=0.01).
When disaggregating data by gender, female patients were consistently more likely to disagree with several questions related to computer use and screen use during the visit compared to male patients.A higher proportion of women disagreed with the statements, "I understand my lab results better when my provider uses the computer and shows it to me on the computer screen" (OR 0.3, CI 0.1-0.7,P=0.004) and, "I understand my medical problems better when my provider uses the computer and shows it to me on the computer screen" (OR 0.3, CI 0.2-0.8,P=0.008) compared to men.Additionally, female patients were less likely to agree that, "I understand what medications to take and how to take it better when my provider uses the computer" (OR 0.5, CI 0.2-0.9,P=0.02) compared to male patients.In contrast to the differences between women and men, patients with high school degree or above (such a college or master's degree) demonstrated less dissatisfaction with computer use in the exam room compared to those who had not completed either middle school or high school.Those with higher educational degree agreed less with the statement, "My visit is less personal because my provider uses the computer" (High School Degree vs No High School Degree, OR 0.3, CI 0.1-0.7,Bachelor's Degree or Above vs No High School Degree, OR 0.4, CI 0.2-1.0,P=0.02), and were more likely to agree that their provider was skilled at using the computer (High School Degree vs No High School Degree, OR 9.1, CI 1.0-80.7,P=0.04).Additionally,

Age
Years, Mean (SD) Median (min, max) 50 ( 13) 52 (20,76) - Journal of HIV and AIDS Open Access Journal patients with higher educational degrees expressed higher concurrence with the notion that they better understand their lab results when the provider used a computer and showed it to them on a computer screen (Bachelor's Degree or Above vs No High School Degree, OR 6.3, CI 1.9-21.0,P=0.006).
Patients with an undetectable HIV viral load were more likely to agree that their provider is skilled at using the computer (OR 0.1, CI 0.008 -1.02, P=0.022).Compared to patients with a CD4 cell count of less than 200 cells/µL, patients with a higher CD4 cell count (>500 cells/µL) agreed that they could talk easily with their provider when he/she is looking at the computer (OR 7.5, CI 1.6-34.7,P=0.02).Similarly, these patients less often agreed that "My visit is less personal because my provider uses the computer" compared to those who had CD4 cell count less than 200 cells/µL (OR 0.3, CI 0.09-0.8,P=0.04), but agreed at a higher percentage that they understand their medical problems better when their provider uses a computer and shows it to them on a computer screen (OR 1.7, CI 0.8-4.9,P=0.005).
Patients who had not disclosed their HIV diagnosis to others were more likely to answer negatively to the questions regarding provider computer use.They were less likely to agree that, "My provider spends enough time talking to me and examining me while using the computer" (OR 0.3, CI 0.1-0.6,P=0.003) and to the statement "I feel comfortable disclosing private information about myself while my provider uses the computer (OR 0.3, CI 0.1-0.8,P=0.02)."

Provider characteristics
Twenty HIV providers completed the survey (Table 3).majority of the providers were female (70%) and had practiced an average of 14 years (range 0.5 to 50 years).Ninety percent of providers felt comfortable using a computer and 95% (19/20) used the computer in the exam room while seeing a patient.One provider did not use the EMR.This provider was supervising physicians-in-training and was not responsible for direct data entry into the EMR as the provider's primary role was teaching.

Provider survey results
Providers had mixed views regarding computer use during the visit (Table 4).Overall, providers were satisfied with the care they provided, the time spent listening to patients (75%), the medical care provided to patients (90%), and the attention given to patients (80%).

Patient and provider comparison
Patient and provider responses were compared for the questions that were consistent across both surveys.Compared to patients, providers demonstrated more consistently negative responses to questions regarding computer use during the clinical encounter.
Patients expressed more overall satisfaction with the visit and found the encounter more positive compared to providers (Table 5).Additionally, a higher percentage of patients (56%) agreed that, "I can talk easily with my provider while provider uses the computer" compared to providers (15%) who assessed a similar statement, "I can talk easily to my patients while using the computer" (P<0.001).Almost two-thirds of patients (60%) agreed that, "I am satisfied with the attention given to the patient by the provider, " compared to less than one-third of providers (P=0.002).

Discussion
While most providers used the computer in the exam room, they demonstrated more negative perceptions of this topic compared to HIV-infected patients.These results align with systematic reviews that found computer use had neutral to positive effect on the patientprovider relationship in the outpatient setting [18][19][20].
Patients who did not use personal computers showed greater distaste for computer use in the exam room, perceiving this factor as negatively affecting the encounter in some way.For example, patients reporting weekly or monthly computer use were more likely to carry negative perceptions of computer use compared to those who used the computer daily.Research shows that approximately half of HIV- preferred a multi-modal way of learning, which included a combination of visual (learning from charts or flow diagrams), auditory (learning from speech), reading-writing (learning from reading and writing), and/or kinesthetic (learning from touch, smell, sight, and taste), while the majority of women preferred a single learning modality [27].In our study, providers likely provided both auditory and visual approaches when teaching patients using the computer screen during the visit.Female patients may prefer less use of the computer screen for health education, however more research is necessary.
infected patients with personal computers seek health information and knowledge on the Internet [25], using it as a source of coping, empowerment, and support [26].Patients in this study with personal computers validate these findings, and may recognize and value the positive effect of computers on their ability to cope with their illness.
Similarly, women were more likely to respond negatively to questions regarding computer screen use as a mode of communication compared to men.A study that assessed gender differences in learning among physiology students showed that the majority of male students  Patients with higher level of education expressed more positivity regarding computer use in the exam room compared to those with lower education levels.However, one study on tablet computer use in the exam room found the opposite: those with higher education levels (high school degree or higher) perceived use of tablet computers by providers as less secure compared to those who had not completed high school [28].More research can help clarify the impact of education level on patient perception of computer use in the exam room.
Patients with an undetectable viral load more often agreed their provider was skilled at using the computer.Similar to those with higher educational level, patients with higher CD4 cell counts also perceived computer use as a positive factor at a higher rate compared to patients with lower CD4 cell counts.A positive patientprovider relationship has been shown to be associated with improved adherence to antiretroviral therapy [14,16], and interventional trials have shown that improved patient-provider communication results in better disease outcomes [29,30].While different factors of patient-provider communication play a role in the patient-provider relationship, computer use represents a small but critical part of this communication -future research should focus on which aspects of computer use patients find helpful and which lead to improved adherence and better health outcomes for patients, such as undetectable viral load and higher CD4 cell counts.
Patients in this study who disclosed their HIV status expressed more satisfaction with the amount of time the provider spent with them, aligning with research by Holt et al who found that HIV-infected patients use disclosure of their diagnosis as mechanism for coping, increased emotional support, and self-acceptance of having a chronic condition [31].Patients who had not disclosed their HIV status were more likely to carry shame and guilt [32].Reasons for non disclosure of HIV include lack of social support and fear of being stigmatized [32].The findings in this study may be attributed to the fact that those who have not disclosed their HIV status may not have an adequate support system and rely solely on their providers for discussion about HIV, suggesting that these patients may benefit from less computer use in the exam room.

Provider findings
Providers are often concerned that the use of computers in the exam room negatively affect communication with their patient [33].Similarly, despite most providers in this study using computers in the exam room, a majority of them felt that computer use made the visit less personal.The providers in this study likely chose to use the computer while interacting with patients due to the benefits of efficiency outweighing the risk of negatively affecting patient provider communication.
Our study showed a difference in perception between HIVinfected patients and their providers regarding computer use in the exam room, as providers expressed more concern about computer use than patients.Currently, computer training for providers focuses solely on how to use the EMR system interface for specific tasks, such as documentation and billing [34].More formalized training for physicians on how to interact and communicate with patients while using the computer may help ease this tension.

Limitations
An initial limitation to this study was the lack of measurement or control around the degree of computer use by providers.This wide range of usage by providers likely affected how the provider interacted with the patient.Second, values such as HIV viral load and CD4 cell count were self-reported and not validated with external resources.Third, participating patients in the study were invited prior to or just after their encounter with the provider.The patients who were approached prior to their encounter may have been more aware of the provider using the computer in the exam room compared to those who were approached after their encounter, thus influencing the results.However, all patients gave informed consent and completed surveys immediately following their encounter.Lastly, despite having one researcher distribute the surveys, the surveys were completed anonymously, creating a risk of duplicate responses.
In conclusion, most HIV-infected patients were satisfied by the care they received when the provider used the computer in the exam room.However, patients that held a negative perception of computer use in the exam room included those who do not own personal computers, are women, have not achieved high school diploma or higher, have detectable HIV viral load, have low CD4 cell count, or have not disclosed their HIV status.These findings suggest that patients with these characteristics may benefit from less computer use by their provider in the exam room.Since providers viewed computer use as more negatively affecting the relationship compared to patients, the results of this study may provide reassurance that only a small subset of patients share this belief.

Table 1 :
Demographics, Computer Use, and Medical Characteristics of Patient Participants.
PA: Physician Assistant; NP: Nurse Practitioner

Table 2 :
Select Survey Responses: Patient Attitude towards Provider Computer Use in the Exam Room.

Table 3 :
Demographic, Computer Use, and Medical Characteristics of Provider Participants.

Table 4 :
Select Survey Responses: Provider Attitudes Toward Provider Computer Use in the Exam Room.

Table 5 :
Comparison of Patient and Provider Response to Select Survey Questions.