Why Physicians Don’t Ask: Interpersonal and Intrapersonal Barriers to HIV Testing—Making a Case for a Patient-Initiated Campaign

Abstract

In 2006, the US Centers for Disease Control and Prevention recommended HIV testing for all adolescents and adults aged 13 to 64 in health care settings with a HIV prevalence of at least 0.1%. However, 55% of US adults have never been tested and therefore do not know their HIV status. To understand suboptimal HIV testing rates, this study sought to illuminate interpersonal and intrapersonal physician barriers to HIV testing. One hundred and eighty physicians from health centers in Houston completed a survey based on Cabana’s Knowledge, Attitudes and Behaviors model. One-third of the physicians faced at least 1 interpersonal barrier to HIV testing, such as a difference in age or language. Many (41%) physicians faced at least 1 intrapersonal barrier, such as believing their patients would be feeling uncomfortable discussing HIV. Notably, 71% of physicians would prefer their patients ask for the test. A patient-engaging campaign may be an innovative solution to increasing HIV testing and reducing the number of undiagnosed persons.

Introduction

HIV Testing in the United States

Over 1.1 million persons in the United States are infected with HIV, and almost 1 in 6 persons are unaware of their infection.¹ The HIV epidemic in the United States continues, in part, because routine HIV testing remains suboptimal. According to a 2014 US Centers for Disease Control and Prevention (CDC) report, only 36% of US adults aged 18 to 64 have ever been tested for HIV.² There are numerous missed opportunities for HIV testing that have negative consequences.^3–5 A CDC report found that patients diagnosed late with HIV had visited a health care facility a median of 4 times before eventually being tested for and diagnosed with HIV by their health care provider.⁶ Even when patients have clinical conditions suggestive of HIV, 1 study found that 82% of these patients never received a recommendation for HIV testing.⁴ Thus, many patients are diagnosed with HIV late in their disease progression, and nearly a third of people diagnosed with HIV receive a diagnosis of AIDS within 1 year following their HIV diagnosis.⁷ Missed opportunities for HIV testing also increase the probability of ongoing HIV transmission in the community. Studies have found that nearly 50% of new HIV infections in the United States are attributable to individuals who have not been tested and are therefore unaware of their HIV-positive status.⁸

HIV testing offers several benefits. Routine testing for HIV helps find those who are HIV positive so that they may begin antiretroviral therapy to decrease their viral load, the main biological predictor of HIV transmission.⁹ In addition, persons aware of their HIV-positive status engage in fewer risky behaviors leading to a reduction in HIV transmission to others.^9,10 In 2006, the CDC recommended routine HIV testing for all adults in high HIV prevalence areas in order to reduce late diagnoses and prevent the ongoing spread of HIV in the United States.¹¹ However, according to the December 2013 CDC National HIV Prevention Progress Report, although achievements have been made in increasing the percentage of people living with HIV who have been diagnosed as HIV positive, from 80.9% in 2006 to 86.2% in 2010, there remains a substantial group of HIV-positive individuals unaware of their status.¹² The December 2013 CDC Report noted that although the reduction in late-stage HIV diagnoses is on track, greater decreases are necessary.¹² Routine HIV testing and more frequent physician-initiated HIV testing will help to improve HIV testing rates.

Physicians Are Missing Opportunities for HIV Testing

Despite the national recommendations for routine HIV testing and the known benefits of such testing, physicians are still not routinely testing their patients for HIV. In a national survey of over 1500 US adults, 72% of the respondents reported that their physician had never brought up HIV testing in a discussion.¹³ Another study conducted in a high HIV prevalence city found that 89% of physicians had never suggested HIV testing to their patients.¹⁴ In 2013, a study found that 71% of HIV-positive patients had at least 1 health care encounter during the year prior to their HIV diagnosis during which their physicians failed to test them for HIV.³

Physician Barriers to HIV Testing

Physicians have reported a number of barriers to HIV testing, including time,^15–18 competing clinical priorities,^17–19 and concerns about reimbursement.^17,20 Although external barriers to HIV testing exist, physicians may also face interpersonal and intrapersonal barriers. Interpersonal barriers are differing characteristics, such as age, race, or language, which impede effective communication between 2 people. Intrapersonal barriers are an individual’s predisposing, cognitive barriers to communication, such as wrong assumptions or varied perceptions²¹ about patient attitudes toward HIV testing. This type of barrier is a result of personal experiences, values, personality, or education.²¹ To better inform interventions designed to improve rates of physician recommendation of HIV testing for their patients, we first need to identify specific barriers to this practice. The objectives of this study were to (1) determine whether physicians knew and agreed with the 2006 CDC HIV testing recommendations, (2) illuminate the interpersonal and intra-personal barriers impeding physicians from initiating an HIV testing discussion with their patients, and (3) explore potential solutions to overcome physicians’ inter- and intrapersonal barriers to HIV testing.

Methods

Study Location

Houston, the fourth most populous city in the United States,²² is a high HIV prevalence area.^22,23 Approximately 95% of the HIV/AIDS cases in the Houston metropolitan area are in Harris County.²³ In 2011, there were 30 new HIV diagnoses per 100 000 population in Harris County,²³ compared to 15.8 new HIV diagnoses per 100 000 population in the overall United States.⁷ This study took place from January to March 2013 with primary care physicians from 19 publicly funded community health centers in Harris County, Houston, Texas. These community health centers have over 1 million patient visits per year and care for predominantly Hispanic (57.4%) and African American (26.3%) patient populations—the populations most affected by the HIV epidemic in the United States.²⁴

Study Design

An anonymous, Web-based survey for physicians about HIV testing was created based on constructs from the Cabana model aimed at identifying barriers and facilitators of physicians’ adherence to guidelines.²⁵ According to the Cabana model, knowledge, attitudes, and behavioral skills are individual and interacting factors that contribute to physicians’ adherence to guidelines.²⁵ These skills could be classified as inter- or intra-personal factors that affect a physician’s actions. Our study focused on the interpersonal and intrapersonal factors that prevent physicians from initiating HIV testing discussions with their patients. Our study also collected information on physician knowledge, attitudes, and behavior on HIV testing recommendations and self-reported HIV testing behaviors.

Recruitment

To understand the HIV testing knowledge, attitudes, and behaviors of primary care physicians in Harris County’s largest publically funded health care system, purposive sampling was utilized to recruit participants. The names, specialties, and e-mail addresses of all primary care physicians (internal medicine, family practice, obstetrics and gynecology, and internal medicine/pediatrics) at participating community health centers were obtained from leadership at these sites. Physicians were informed of the research study via e-mail, which included a Web address to the consent cover letter and study survey. For some community health centers, leadership sent the e-mail with the study survey Web address to primary care physician list-servs. Participants were also recruited through postcard reminders in their office mailboxes and announcements at medical conferences. To increase the response rate, a reminder e-mail was sent with the survey Web link. Nominal incentives of a US$10 gift card and entry into a US$100 raffle were offered. The Baylor College of Medicine institutional review board approved this study.

Measures

Physicians were asked about their knowledge, attitudes, and behavior about HIV testing adult patients in the primary care setting. To assess knowledge, physicians were first asked “Before this survey, I was Aware/Unaware that the CDC had issued updated recommendations for routine HIV testing in 2006.” Physicians were queried about their attitudes by indicating whether they think HIV testing should be routine for all adult patients aged 18 to 64 and were categorized as agreeing if they responded “Agree” or “Strongly Agree” or were categorized as disagreeing if they responded “Disagree” or “Strongly Disagree.” Additionally, physician behavior was assessed by inquiring about whether physicians test all adult patients aged 18 to 64. Furthermore, physicians also indicated the extent to which they agreed with 5 interpersonal (ie, differences in age, language, and culture, unsure how to initiate the topic of HIV testing, and afraid that the topic of HIV could negatively affect their patient–physician relationship) and 3 intrapersonal barriers (ie, physicians believe their patients would be uncomfortable, offended, or would refuse the HIV test) to HIV testing. With the exception of 3 interpersonal barriers (ie, differences in age, language, and culture) that were rated on a dichotomous scale, where 0 indicated that the difference was not a barrier and 1 indicated that it was a barrier, physicians were considered to have endorsed the existence of a barrier when they answered agree or strongly agree to these survey items and were considered to have not endorsed a barrier when they answered disagree or strongly disagree. Finally, physicians were asked about patient-centric solutions to overcome HIV testing barriers.

Data Analysis

Descriptive statistics were used to characterize study participants as well as physician knowledge, attitudes, and behavior about HIV testing adult patients in the primary care setting. Frequency statistics were used to describe endorsement of each of the 5 interpersonal and 3 intrapersonal barriers to HIV testing.

We then obtained (1) the number and percentage of physicians reporting at least 1 of the 5 interpersonal barriers and (2) the number and percentage of physicians reporting at least 1 of the 3 intrapersonal barriers. Subsequently, chi-square (χ²) tests were performed to examine the associations between whether physicians in this study test all adult patients aged 18 to 64 (yes versus no) and (1) endorse any interpersonal barriers (yes versus no) and/or (2) endorse any intrapersonal barriers (yes versus no). Finally, frequency statistics were reported for patient-centered solutions. All analyses were conducted using SAS Statistical Software version 9.3 (SAS Institute, Cary, North Carolina).

Results

Study Participants

The survey was e-mailed to 561 primary care physicians in community health centers in Harris County. A previous publication reported results from a subpopulation of these physicians.²⁶ A total of 175 physicians completed the survey (response rate: 31.1%). Of the 173 physicians who reported their professional status, 103 (59.5%) were trainees—interns, residents, or fellows—and 69 (39.9%) were faculty. Of the 170 physicians who reported their specialty, 64 (37.7%) were internal medicine, 49 (28.8%) were family practice, 36 (21.2%) were obstetrics and gynecology, and 19 (11.2%) were internal medicine/pediatrics. Of the 171 physicians who reported their gender, 112 (65.5%) were female. The average age of all participants was 34.1 years (standard deviation [SD] = 8.5 years). On average, participants graduated from medical school 8 years ago (SD = 9.0 years, range = 1–45 years). See Table 1 for demographic characteristics.

Table 1.

Demographics and Specialties of Study Participants.^a

Demographic Characteristic	n	%
Gender (n = 171)
Male	59	34.5
Female	112	65.5
Race (n = 165)
Asian	59	35.8
American Indian/Alaskan Native	1	0.6
Black/African American	23	13.9
Native Hawaiian or other Pacific Islander	1	0.6
White	75	45.5
Other	6	3.6
Professional status (n = 173)
Intern	25	14.5
Resident	77	44.5
Fellow	1	0.6
Medical school faculty	69	39.9
Directly employed staff physician	1	0.6
Specialties (n = 165)
Family practice	47	28.5
Internal medicine	64	38.8
Internal medicine/pediatrics	19	11.5
Ob/gyn	33	20.0
Pediatrics	1	0.6
Other	1	0.6

^aN = 175.

Knowledge, Attitudes, and Behavior

A majority (n = 123, 70.3%) of physicians were aware that the 2006 CDC recommendations for HIV testing state that all adult patients in areas of high undiagnosed HIV prevalence should be routinely tested. Furthermore, 153 (87.4%) physicians think that HIV testing should be routine for all adult patients aged 18 to 64. However, only 119 (68.0%) of physicians actually routinely test all adults for HIV.

Interpersonal Barriers

Physicians reported multiple interpersonal barriers that prevent them from offering HIV tests to their patients (see Table 2). Overall, nearly one-third of physicians (n = 52, 29.7%) said at least 1 interpersonal barrier prevented them from offering an HIV test. These barriers include differences between them and their patients in age (n = 14, 8.0%), language (n = 15, 8.6%), and culture (n = 24, 13.7%), as well as being unsure of how to initiate the topic of HIV testing with their patients (n = 14, 8.0%) and being afraid that the topic of HIV could negatively affect their patient–physician relationship (n = 15, 8.6%).

Table 2.

Interpersonal Barriers to HIV Testing.^a

Survey Item	Reporting this Barrier, n	Reporting this Barrier, %
Differences in age between physician and his or her patient	14	8.0
Differences in language between physician and his or her patient	15	8.6
Differences in culture between physician and his or her patient	24	13.7
Unsure how to initiate the topic of HIV testing with his or her patient	14	8.0
Afraid that the topic of HIV could negatively affect their patient–physician relationship	15	8.6

^aNumber (%) reporting at least 1 interpersonal barrier: 52 (29.7%).

Reporting at least 1 of these 5 interpersonal barriers was significantly associated with whether one tested all adult patients (χ²(1) = 5.09, P = .02). Those who reported at least one of these barriers were significantly less likely to report testing all adult patients for HIV relative to those who did not report any interpersonal barriers. Specifically, 29 (55.8%) physicians who reported at least one of these barriers tested all their patients, whereas 90 (73.2%) physicians who reported no interpersonal barriers tested all their patients.

Intrapersonal Barriers

Physician misconceptions about their patients’ attitudes toward HIV testing could also be a barrier to testing (see Table 3). Many (n = 62, 35.4%) physicians believed their patients are uncomfortable discussing HIV, some (n = 29, 16.6%) felt their patients would be offended if offered an HIV test, and a few (n = 12, 6.9%) physicians thought that their patients would refuse an HIV test. Many (n = 72, 41.1%) physicians said at least 1 intrapersonal barrier prevented them from offering an HIV test.

Table 3.

Intrapersonal Barriers to HIV Testing.^a

Survey Item	Reporting this Barrier, n	Reporting this Barrier, %
Think their patients are uncomfortable discussing HIV	62	35.4
Think their patients would be offended if offered an HIV test	29	16.6
Think their patients would refuse the HIV test	12	6.9

^aNumber (%) reporting at least 1 intrapersonal barrier: 72 (41.1%).

Reporting at least 1 of these 3 intrapersonal barriers was not significantly associated with whether one tested all adult patients (χ²(1) = 1.70, P = .19). Those who reported at least one of these barriers were statistically just as likely to report testing all adult patients (n = 45, 62.5%) as those who did not report any barriers (n = 74, 71.8%).

Patient-Centered Solutions

Although 60 (34.3%) physicians reported that information showing patients are receptive to HIV testing would help them offer an HIV test to their patients, more than two-thirds (n = 124, 70.9%) of physicians felt that they would be more likely to test their patients for HIV if their patients asked them for the test.

Discussion

This study, conducted during the routine HIV testing era in the nation’s fifth largest publicly funded health care system,²⁷ found that physicians are generally aware of and agree with the 2006 CDC recommendations for routine HIV testing. Previous studies have found numerous barriers that hinder physicians from recommending HIV testing routinely to all of their patients.^15–20 The most notable barriers are time and competing clinical priorities. Our study has illuminated several interpersonal and intrapersonal barriers and misconceptions that may be impeding physicians from discussing HIV testing with their patients and ultimately recommending HIV testing. Moreover, our study found that at least one-third of physicians face interpersonal or intrapersonal barriers to recommending HIV testing. We found that many physicians fail to offer an HIV test because they believe their patients would be uncomfortable, offended, or refuse the test. However, other research indicates patients may not feel this way. For example, a recent study reported that patients likely would not be offended or uncomfortable if their physician offered the test, given that these patients reported “expecting” and “wanting” HIV testing to be initiated by their physicians.²⁸ Our findings highlight physician cognitive-affective influences (eg, attitudes, perceptions, goals, and motivations) as key factors in the patient–physician relationship. Consistent with the Cabana model, physicians’ perceptions of and assumptions about their patients affect their goals for the encounter and the communicative approach to managing the encounter.²⁵ For example, a physician may feel that offering an HIV test to a middle-aged female married for 20 years might be offensive to her. As another example, a physician may consider offering an HIV test to a homosexual patient but may be uncomfortable talking about it and thus not bring it up.

To overcome physician interpersonal and intrapersonal HIV testing barriers, a pioneering intervention to improve HIV testing in health care settings may be a patient-initiated approach. Our study found that some primary care physicians are unsure of how to initiate the topic of HIV testing with their patients and that 70% of the respondents would be more likely to offer the HIV test if their patients asked them for it. Other studies have also found that physicians want their patients to ask for the HIV test.²⁹ Although there is evidence supporting the fact that patients actually want their physicians to test them for HIV,^30,31 our study found that some physicians are afraid that the topic of HIV would negatively affect their patient–physician relationship. Since physicians are not routinely offering HIV testing and face inter- and intrapersonal barriers, campaigns that will cue patients to engage their physicians in a discussion about HIV testing could be particularly effective.

The quality of physician–patient communication has been associated with (1) greater patient understanding of his or her health and (2) better health outcomes.³² Studies have also found that patients who are more engaged and actively involved in their own health have better communication with their physician and receive more personalized care.³³ Because most patients want to be active participants during the health care encounter,³⁴ they would potentially be receptive to a cue to initiate health discussions with their doctors which in turn should increase HIV testing rates. Notably, one-third of respondents in a national survey reported wanting strategies to help discuss HIV with their physician, with only 46% having ever had a discussion about HIV with their physician.³⁵ Unfortunately, racial and ethnic minority populations face greater HIV disease burdens yet have the most likely to have undiagnosed HIV in the United States.³⁶ Furthermore, racial and ethnic minorities ask their physicians the fewest questions, highlighting the need for targeted campaigns for these patient populations.³⁷

With the 2013 release of the US Preventive Services Task Force Grade A recommendation for routine HIV testing of all persons aged 15 to 65, novel strategies are needed to encourage more widespread HIV testing in health care settings.³⁸ Key objectives of Healthy People 2020 include increasing the proportion of people who have been tested for HIV, improving patient–physician communication, and increasing the amount of patient involvement when making health decisions.³⁹ The “push–pull” capacity model could be the means by which patient activation might work to increase physician recommendations.⁴⁰ This model offers a framework to guide a novel solution to improve HIV testing that is patient initiated, fosters patient–physician communication and thereby helps overcome documented physician HIV testing barriers. This model could be used to create a campaign that prompts, or “pushes,” patients to engage, or “pull,” their physicians into a discussion about the HIV test.

This study is subject to several limitations. First, the results may not be generalizable due to (1) the relatively small sample size and (2) the fact that physicians were all from a single health care system. Second, this study took place in a large, publicly funded health care system in a city with a high HIV prevalence; thus, the results may not be applicable to other health care settings. Third, the study participants were primary care physicians, many of whom were trainees, and may not share the same perspectives as other physicians. Finally, selection bias may have influenced results as participants had to opt-into taking the survey. Despite these limitations, to our knowledge, this is the first study to focus on cognitive-affective influences on the patient–physician relationship among patients and physicians in a large, publically funded health care institution.

Conclusion

Physicians face numerous interpersonal and intrapersonal barriers to offering HIV testing; further research is needed to elucidate innovative solutions to overcoming these barriers. Physicians want their patients to be more proactive in asking for an HIV test. We hypothesize that a patient-oriented campaign could increase patient engagement in their own health, thus reducing both inter- and intrapersonal physician barriers to HIV testing. Improved patient–physician communication and patient-initiated testing can lead to increased routine HIV testing and improved health outcomes.