Trust in Mental Health Clinicians Among Patients Who Access Clinical Notes Online

All participants completed an informed consent process, and all study procedures were approved by the local VA institutional review board. This study took place at a single VA medical center (VAMC) providing comprehensive care to 85,000 unique veterans each year, with 18,000 veterans receiving mental health services across a main hospital complex and 11 primary care clinics. Beginning in 2013, clinical notes and all laboratory and imaging results became available through the Blue Button function in MyHealtheVet, the VA’s online patient portal. Veterans must complete a user verification process (“authentication”) to access all functions in MyHealtheVet. When our sample was drawn in May 2014, 51% of patients receiving mental health services at the VAMC were authenticated MyHealtheVet users. In the prior year, 13% of them had downloaded VA clinical notes.

We used VA databases and a purposive sampling strategy to identify a heterogeneous sample of patients who had diverse demographic characteristics and who had utilized a range of mental health services at the VAMC. After obtaining clinicians’ permission to contact patients, we mailed recruitment letters to 402 potential participants; we conducted follow-up screening by phone. We enrolled eligible patients in the order in which they contacted us. We excluded 64 patients who reported not having read their notes at least once. Recruitment was halted when the study team agreed that no new themes were arising in interviews. Twenty-eight patients were enrolled and completed interviews between July and December 2014 (26 interviews in person and two by phone). The average age of participants was 47, ranging from 30 to 69. Fifty-seven percent were women (N=16) and 86% were white, non-Hispanic (N=24). Psychiatric diagnoses of enrolled participants included major depressive disorder (N=19, 68%), posttraumatic stress disorder (N=19, 68%), bipolar disorder (N=5, 18%), and schizophrenia (N=2, 7%). Our sample was highly educated; 18% (N=5) had graduate degrees, and 29% (N=8) had college degrees.

We developed a semistructured interview guide informed by our main research questions, current literature on patient experiences with full health record access (5,10,11), and input from mental health clinicians. The interview guide focused on elucidating patients’ ideas across five domains: note reading practices, including reasons for reading; impressions about what they liked, disliked, and learned from reading notes; experiences discussing notes with clinicians and others; beneficial or concerning impacts of reading notes on their lives and health care; and resources to assist them or other patients with OpenNotes, including advice for mental health clinicians who write notes. [The online supplement to this report includes details on the interview guide.] Each interview lasted approximately 60 minutes. Participants were reimbursed $25.

All interviews were audiorecorded and transcribed. Transcripts were double-coded by two authors using Atlas.ti software (12) and were verified by a third author to ensure consistency. We used a constant comparative approach for our analysis (13), consistent with grounded theory. This approach allows for comparing discrete narratives within common themes to develop a theory of how themes are related to each other and to identify hypotheses for future testing. Using an iterative process to define codes, three authors reviewed transcripts separately and identified initial themes that emerged from the data. Analysts reached consensus on a codebook of 43 codes that were applied to all transcripts. All authors reviewed code reports for theoretical interpretations until agreement was reached on main findings.

Overall, analyses showed that patients receiving mental health care consider strong patient-clinician relationships—and feelings of mutual trust in particular—to be critical to the therapeutic process. Reading clinical notes can strengthen or strain patients’ trust in their clinicians. Patients’ perceptions of the level of transparency and respect in notes contributed to feelings of trust in their clinicians. We elaborate on these themes in the paragraphs that follow. [The online supplement provides relevant quotations from the patients.]

Patients in our sample strongly believed that the patient-clinician relationship is important to their care and that OpenNotes can help foster this relationship: “OpenNotes is a move in the right direction. Like anything when you first start it, it has fits and starts, and people are reluctant and people don’t like change, and psychiatrists like to keep their secrets, you know, and not hurt somebody’s feelings or whatever. It kind of misses the point. The point is to develop this relationship.” Patients felt the patient-clinician relationship develops by establishing good rapport, talking openly, and feeling comfortable with their clinicians. Many respondents considered mutual trust foundational to this relationship: “If you don’t trust your clinician, what value is your conversation?” Patients sometimes worried that clinicians who did not believe them ultimately did not want to help them. Those who had difficulty establishing mutual trust or who had experienced a breach of trust with their mental health clinicians described it as a barrier to therapy, with some patients responding by being more careful about what they said in clinical sessions, reviewing their notes carefully, asking their clinicians questions about their notes, or seeking care from a different clinician within or outside the VA. Patients reported greater trust when their mental health clinicians listened in sessions, focused on patient strengths, and approached the therapeutic process as an equal. Specifically, trust in clinicians—and thus the therapeutic relationship—was strengthened or strained largely on the basis of patient evaluations of the levels of transparency and respect conveyed in notes.

Transparency was evident to patients when notes accurately reflected what happened during clinical sessions. Many appreciated thorough notes such that nothing seemed left out or misrepresented. Notes also clarified for patients whether a clinician was forthcoming in sessions about his or her interpretations and assessments. Patients were more trusting when clinicians reviewed psychiatric diagnoses with them before documenting the diagnoses in their records: “It made that relationship stronger, the trust was there, because they were open to talking to me about [the diagnosis] . . . instead of stuffing it under the carpet.” Although most patients had not spoken with their clinicians about reading notes, clinicians who initiated conversations with patients about what they document garnered trust.

In contrast, patients reported strained trust when they perceived low transparency and when they noticed incongruencies between sessions and notes, including the absence of information discussed in sessions, outdated copied and pasted sections, mistakes (such as incorrect age or gender), and details that misaligned with a patient’s recollection about a session. Patients described these discoveries as surprising, odd, and upsetting. Many were especially surprised to discover diagnoses in their mental health records that had not been discussed with them. Some worried that inaccuracies in their records could negatively affect their treatment, for example, by other clinicians whom they encounter who may perceive or treat them differently. One patient was concerned about how perceived errors in her health record might affect a pending court case. Others who were surprised by their notes wondered if mental health care was worth their time: “I’m . . . giving up a lot of time. . . . I would like you to take it seriously too, not just spit something out on paper and not proofread it.”

Patients felt respected and reported greater trust in clinicians when their notes contained evidence of listening to their patients. Most patients felt that clinicians truly listened if everyday details about their lives, such as grieving the loss of a pet or a patient’s volunteer work, were documented in the clinical notes. Respectful notes also showed evidence of clinician thoughtfulness and consideration; a tone of respect helped patients feel as if they were, as one described, “not just a repeat from the last PTSD person he talked to, but an individual with PTSD.” When reading the notes, patients evaluated whether their clinicians regarded them as whole persons. Patients reported feeling like they were being seen holistically when clinicians documented their strengths and efforts, not just their diagnoses or symptoms. One patient reported, “It’s nice to know when I read the notes that the doctor noticed I was staying on task, because I was really trying to.” Patients appreciated when they perceived that clinicians wrote notes with the patient in mind as a reader.

Conversely, patients’ trust in clinicians was strained when they did not feel respected. “Did [the clinician] actually listen to me? . . . It seems like a very different person written [about] in the notes at times,” expressed one patient, echoing a common complaint. Notes comprising primarily surface-level observations indicated to some patients that clinicians did not understand them as individuals. Also, the tone of notes contributed to some patients’ feeling disrespected. They described such notes as unfriendly, negative, badgering, or angry. Although some accepted that feeling disrespected could simply reflect their personal perceptions, such impressions negatively affected some patients’ trust in certain clinicians.

Like mental health clinicians, patients consider the patient-clinician relationship a powerful tool in the therapeutic process. Patients’ expressions in this study echoed research identifying trust as key to developing or maintaining strong patient-clinician relationships (14). Our findings add to the growing literature demonstrating that sharing clinical notes can facilitate enhanced patient-clinician relationships in general; in a recent study of OpenNotes in primary care, 70% of clinicians reported strengthened relationships with their patients (5). Our study also revealed that trust within patient-clinician relationships can be strained by OpenNotes, especially when patients perceive that the clinician is not being transparent or not respecting the patient.

With these findings, we offer some specific recommendations for mental health clinicians to help strengthen the therapeutic alliance in the context of OpenNotes. First, we suggest that clinicians consider proactive communication with patients about the content of notes and the documentation process. Communicating openly and transparently about observations, interpretations, and diagnoses may help mitigate surprises when patients read their notes. Second, we encourage clinicians to use documentation that highlights the individuality of the patient and that includes patient strengths. Our findings suggest that including unique details from each session could help patients feel heard and understood by their clinicians; also, documentation highlighting patient progress could be therapeutically motivating. However, future research should examine whether—and the extent to which—such strategies could help strengthen the therapeutic relationship in the context of OpenNotes. Further research on mental health clinician perspectives of the therapeutic alliance in the context of OpenNotes may elucidate additional strategies for minimizing opportunities for unintended consequences. Although challenging to conduct, additional research designed to quantify actual impacts and outcomes related to use of OpenNotes may also prove beneficial for health systems considering offering full patient access to mental health records.

These findings and recommendations should be considered in the light of several limitations. This was a small qualitative study. However, we used purposive sampling to recruit a sample with a range of demographic characteristics and psychiatric diagnoses, and rigorous qualitative methods were used to collect and analyze the data. This study was conducted at only one VAMC; findings may be subject to regional or facility norms. In addition, because veterans’ access and participation in their mental health care may be unique, researchers should exercise some caution in generalizing our findings to other health care systems.

In this study, we explored patient perspectives of the therapeutic alliance and how OpenNotes may affect patients’ relationships with their mental health clinicians. We found that reading mental health notes may strengthen as well as strain patient-clinician relationships by enhancing or undermining trust. Patients glean information about clinician transparency and respect from reading mental health notes and by assessing how attentively clinicians listen to and understand them, how accurately clinicians document what happened in sessions, and whether clinicians discuss diagnoses openly. Proactive clinician communication with patients about the content of notes and the note-writing process, as well as documenting strengths and highlighting the individuality of patients, may improve the likelihood of maintaining or developing stronger therapeutic alliances between patients and clinicians in the context of OpenNotes.