Allowing patients access to their electronic medical record (EMR) is a current trend. Advantages include enhancing doctor–patient communication and transparency, but disadvantages include the risk of patient confusion or anxiety (1, 2). The Medicare and Medicaid Electronic Health Record Incentive Program Stage 2 proposes that at least 10% of a provider’s patients be able to “view, download, or transmit… their health information” (3).

Psychiatry notes frequently contain sensitive information that might cause difficulty if patients were to read them. Previous studies reported psychiatric patient experiences with medical record access. Five small, older studies (1979–1988) found that most patients appreciated reading and discussing their records with staff and were not harmed by the access (4–8). In a 1991 study of 72 psychiatric outpatients, only 28% were upset by what they read, and 51% felt the information was helpful (9). In a 1995 study of 28 psychiatric inpatients, only in one case was it assessed as harmful, and in 22 cases it was “useful or essential” (10). None of these studies were performed after the advent of EMRs.

The use of EMRs can affect how clinicians write notes. A 2010 survey of 56 psychiatric clinicians found that while open therapeutic communications were mostly preserved, they were also less willing to record highly confidential information in the electronic note (11). Knowing that our institution might release psychiatry notes to our patient portal, we were motivated to peer-review our notes. In this study, we focused on psychiatry trainees, attempting to measure their understanding and sensitivity to the content of notes that might be read by patients.

We randomly selected 128 PGY-3, PGY-4, and PGY-5 outpatient notes between July and November 2010, choosing more PGY-3 notes because the core outpatient rotations begin and occur in that year. Ten notes marked “highly confidential” in the initial selection were replaced by regular notes. In our EMR, the “highly confidential” designation indicates intentionally sensitive information, with further safeguards for access. Our Institutional Review Board deemed this study a quality-assurance project and did not require research approval.

Chart reviewers consisted of four board-certified psychiatrists, including one also board-certified in Child and Adolescent Psychiatry (CAP), and two non-psychiatrists (a third-year medical student who had completed his psychiatry rotation and an experienced psychiatry recreation therapist). We chose non-psychiatrists to simulate how patients might respond to reading their notes. One psychiatrist and one non-psychiatrist reviewed each note, and, whereas each non-psychiatrist reviewed half the notes, the number of notes reviewed by each psychiatrist was unequal. The CAP psychiatrist reviewed all child and adolescent notes. The note author was de-identified, but the patient names as used within the note were left intact.

Each reviewer completed a survey after reading the note as if he or she were the patient. The primary outcome was the question “Overall, if a patient read this note, do you think they would be offended, confused, alarmed, injured, upset? Is the note written in a respectful tone, with sensitivity to culture, age, gender, sexual orientation, disability?” with a score of 0–2 (0: No Concern; 1: Some Concern; 2: Major Concern). Secondary outcomes consisted of 1) whether specific content (mention of personality traits/disorders, substance use, psychosis, sexual issues) was present; and 2) if present, whether the patient might be offended, using the previously described 0–2 scale.

Statistical analysis (SAS Version 9.3; SAS Institute Inc.; Cary, NC) consisted of paired t-tests for the percentage of notes and topics of each concern level × reviewer and trainee level. A Generalized Linear Model (GLM) was also used to compare differences among trainee levels. Differences in the concern level of CAP notes were described.

Twenty-two trainees authored 128 notes (8 PGY-3s wrote 69 notes; 9 PGY-4s wrote 34; 5 PGY-5s wrote 25). Overall, 89 notes (70%) were rated of “No Concern” by both reviewers, and ready to be read by patients. Thirty notes (23%) were rated as of “Some Concern” by one or both reviewers; and 9 (7%) were of “Major Concern,” all from psychiatrists. Psychiatrists were more concerned than non-psychiatrists (paired t-test p=0.0001; Table 1) for all concern categories. This finding was also true when analyzed by trainee level (paired t-tests: PGY-3 p=0.0002; PGY-4 p=0.0303; PGY-5 p=0.0042). Among PGY-3, PGY-4, and PGY-5 notes, we found no significant difference on disagreement (GLM model, p=0.7384) between psychiatrists and non-psychiatrists. That is, the trainee level did not make a difference in how psychiatrists and non-psychiatrists rated their notes. Of the 36 CAP notes (14 for ages 3–12, 22 for ages 13–17), the psychiatrist rated 18 as being of “No Concern,” 14 of “Some Concern,” and 4 as of “Major Concern,” as compared with non-psychiatrist ratings of 34, 2, and 0, respectively.

Secondary outcomes described the presence of sensitive content topics and concern level (Table 2). Non-psychiatrists identified these topics twice as frequently as psychiatrists (p<0.05), except for psychosis. However, there were no statistically significant differences in their concern ratings, except for personality traits/disorders.

This study answers the question “What percentage of our trainee notes are ready to be read by patients?” In our sample of 128 outpatient notes, the answer was 70%. These results might surprise psychiatrist-supervisors as too high or too low, depending on each supervisor’s acceptability threshold. Although we would prefer a higher percentage of “No Concern,” we also kept in mind that notes identified as of “Some Concern” might be very reasonable to release to patients. If one accepted “Some Concern” notes as appropriate, then, overall, 93% of our trainee notes would be ready. Patients, not all of whom want to read their notes, might realize there is certain psychiatric jargon and balance of details for the note to be useful. Ideally, patients would not be reading their notes in isolation, but with the opportunity to discuss them with their care-providers.

One of our surprising findings was that psychiatrists raised more sensitivity concerns than non-psychiatrists. We were expecting psychiatrists, being more familiar with the jargon and style of psychiatric notes, to find fewer concerns. There are several possible explanations. Our psychiatrists might be more protective of the patient in not wanting the patient to read the note. Perhaps our non-psychiatrists were more familiar with psychiatry notes than we initially hypothesized, as one has worked for years as a psychiatric recreation therapist and the other was a medical student who had completed his third-year psychiatry clinical rotation and participated in psychiatry research projects. Interestingly, non-psychiatrists identified personality traits/disorders, substance use, and sexual issues content twice as frequently as psychiatrists, but differed only in personality traits/disorders for concern about those topics. Thus, there were definite differences in how psychiatrists versus non-psychiatrists viewed psychiatric content. Perhaps, in this setting, the psychiatrists were not as sensitive to the jargon, and thus did not flag the content. Further research into psychiatrist attitudes about note content and styles could expand these possible explanations.

Another, not unexpected, finding is the higher concern (50%) for CAP notes. When writing these notes, it is sometimes difficult for both trainees and practicing psychiatrists to remember that, at some point, the patient—or perhaps, even more importantly, the parents—might read the note. One of the essential facets of therapeutic work with adolescents frequently involves discussion of materials they do not wish their parents to know, and the manner in which this is disseminated can be very tricky for both the patient and the therapeutic alliance and ongoing work.

We suggest a few methods to potentially improve patient acceptability of psychiatry notes:

• Formal education as part of a professionalism curriculum or quality improvement project, emphasizing sensitivity in the use of language and avoiding pejorative terms such as “unkempt” and “disheveled.”

• Explicit review and discussion of a few notes per week with a clinical supervisor, especially earlier in training.

• More frequent use of a “highly confidential” status for intentionally sensitive information, which would require explicit writer approval before release. This technique might be especially useful for parental request of notes, to allow the provider the option of writing a summary letter to meet the needs of the information release without revealing sensitive specifics an adolescent might not be ready to share.

• Writing or dictating the note in the patient’s presence for feedback and immediate correction of errors. Although our internal-medicine colleagues might be adopting this approach more, it is difficult to imagine for psychiatric patients. For adolescents, this approach might empower them to gain a sense of control of their care, and may even serve as the impetus to discuss difficult subjects with their families.

There are several study limitations. First, we did not standardize criteria with the reviewers for what content constitutes “Some” or “Major” concern. What one reviewer might consider to be of “Some Concern” might be of “No Concern” or “Major Concern” to another. However, that situation reflects reality, as different patients will react differently. Second, we did not record a diagnosis for each note, to rule out any association with the sensitivity of the note. For example, psychotic symptoms might be harder to document in a neutral manner than would depressive symptoms. However, we surveyed for appropriateness of documentation of sensitive topics. Third, a single-site study introduces institutional cultural biases regarding note acceptability, and reviewer bias if the reviewers recognized notes of patients they supervised. Fourth, the non-psychiatrists still had familiarity with psychiatric terms, so their ratings cannot be construed as from a purely patient or layperson perspective. Last, the small number of reviewers and notes per trainee limit generalization of findings.

Further research can help us better understand how patients might react to reading their psychiatric notes. One approach would be to ask patients to read their notes and then survey them directly. Our survey did not include inpatient notes, which might reveal more tension, given the acuteness of psychiatric hospitalization. Future studies can include inpatient and junior resident notes, and even attending psychiatrist and psychiatric allied health practitioner notes.

In summary, trainee EMR outpatient notes are not likely to cause major concerns for patients who read them. Psychiatrists, however, identified more concerns than did non-psychiatrists. Notes on children and adolescents deserve more scrutiny. Further education for trainees on writing more neutral notes could be helpful.

We thank Stephen S. Cha, M.S., Division of Biomedical Statistics and Informatics, Department of Health Science Research, Mayo Clinic, for statistical support.