measure of healthcare utilization (number of office visits during the study period). All statistical analyses were conducted
using SAS software, version 9. 4. All study procedures were approved by the institutional review boards of BIDMC, GHS, and
the University of Washington.
Of the 6,150 patients who responded to the baseline survey and
had at least one visit note available to review during the intervention period, 5,630 (92 percent) accessed at least one doctor note
while 520 ( 8 percent) did not access any notes. Table 1 on pages 30
and 31 presents patient characteristics reported in the baseline
survey for those who did and did not access visit notes. A higher proportion of patients who accessed their notes were older,
white, female, and retired relative to patients who did not view
their notes (all p-values <0.001). Patients with a greater number
of office visits were more likely to access notes (p<0.001). High
proportions of patients (93-97 percent) with common chronic
diseases—including congestive heart failure, asthma, and diabetes—accessed their notes.
Patient-reported ratings of overall satisfaction with their
doctor and of doctor-patient communication were not significantly different between patients who accessed or did not access their notes (data not shown). Internet access was not a significant barrier to viewing notes, as all but 71 of 6,041 patients
(one percent) reported internet use at least twice a month and
85 percent were daily internet users.
Patients’ baseline perceptions of potential facilitators and
barriers to viewing a note were associated with accessing their
clinic note during the study period. Specifically, patients more
concerned about the potential risks were less likely to ultimately view a note (p=0.005). See Table 2 on page 32 for more
After adjusting for patient characteristics that remained sig-
nificant in a multivariate model (age, sex, race), study site, and
the number of office visits during the study, patients had lower
odds of accessing notes if they felt that the visit notes would be
confusing (OR 0.63, 95 percent Confidence Interval (CI) 0.43-
0.94, p-value <0.02), anticipated worrying more as a result of
reading notes (OR 0.72, 95 percent CI 0.53-0.99, p<0.04), or if
they had concerns about privacy (OR 0.77, 95 percent CI 0.61-
0.96, p-value 0.02) (see Figure 1 on page 34). Baseline attitudes
about potential benefits of viewing notes were not statistically
associated with subsequently accessing the notes, though an-
ticipating being better prepared for the visit or understanding
their health better after reading notes attained borderline sta-
Why Most, But Not All, Accessed Their Records
While the majority of patients in the study anticipated many
potential benefits from accessing their clinic notes, a small—
but important—minority of patients did not access their clinic
notes, and these patients reported more baseline concerns
about privacy issues and fear of being confused or worried
by reading their doctors’ notes. While reading their doctors’
clinic notes has the potential to engage them bet ween face-to-face encounters, and thus improving the quality of care and
the patient experience, these patients cannot benefit if they do
not log in to the online portal and read their medical records.
Thus, interventions focused on overcoming these patients’
reservations may prove more productive than simply reiterating known benefits.
Patients’ fear of finding their clinical notes more confusing
than helpful might be appropriately founded. Medical terminology, shorthand acronyms, and large amounts of templated or copied and pasted data are frequently used in clinical
notes. Therefore, improving the clarity, approachability, and
understandability of doctors’ notes is a first step to addressing the worry that the notes will be confusing.
health record vendors could advance technologically feasible
improvements by including integrated hyperlinks in visit
notes, embedded definitions for medical jargon, and provision
of enhanced patient education materials within patient portals. Doctor education on note writing—such as avoiding ac-
Who Reads Their
Table 1: Baseline characteristics of patients who did or did not access at least one
clinical note during the intervention period (Continued from page 30)
Hospitalized at least once in 12 months following OpenNotes accessb
Yes 307 96% 4%
No 5,594 93% 7%
Congestive heart failure/Coronary
334 96% 4%
Asthmab 343 97% 3%
Low back painb 353 96% 4%
Anxiety/Depression 422 93% 7%
Diabetesb 689 97% 3%
a. Some data missing for age, race, education, employment status, and self-reported health.
b. Significant difference between groups with P < 0.05.
c. Healthcare utilization and co-morbidity were obtained through patient medical records at BIDMC and GHS. Information unavailable for Harborview patients.
d. Based upon billed visit ICD- 9 diagnoses during the study period (2010-2011).