Clinical Documentation in Electronic Health Record Systems: Analysis of Patient Record Review During Outpatient Ophthalmology Visits

Abstract

Busy clinicians struggle with productivity and usability in electronic health record systems (EHRs). While previous studies have investigated documentation practices and strategies in the inpatient setting, outpatient documentation and review practices by clinicians using EHRs are relatively unknown. In this study, we look at clinicians’ patterns of note review in the EHR during outpatient follow-up office visits in ophthalmology. Key findings from this study are that the number and percentage of notes reviewed is very low, there is variation between providers, specialties, and users, and staff access more notes than physicians. These findings suggest that the vast majority of content in the EHR is not being used by clinicians; improved EHR designs would better present this data and support the information needs of outpatient clinicians.

Introduction

Electronic health record systems (EHRs) have become an integral part of healthcare within the past decade. As of 2015, 87% of all office-based providers in the United States had an EHR.¹ Adoption within Non-Federal Acute Care Hospitals is almost universal, with over 96% possessing a certified EHR.² While clinicians recognize the benefits of EHR usage^3-5, they have concerns about its impacts on productivity^6-8. Further, increased documentation requirements from new value based purchasing models such as those required by the Medicare Access & CHIP Reauthorization Act of 2015 (MACRA) add to time pressures⁹.

Clinicians face significant usability issues with current EHRs. The flexibility and customizability of EHRs results in systems that are time consuming and difficult to use.^10-12 Recent studies showed that outpatient providers spend 3-6 hours a day using the EHR.^13-15 Furthermore, multiple studies have shown that current physician burnout is greatly affected by EHR systems.^16-18 In response, clinicians often use strategies such as copy forward or copy-paste to save time when writing notes.^19-21 Previous studies have investigated these documentation strategies and practices in inpatient care, and patterns of how notes are reviewed.^22-27 The studies of note review in the inpatient setting have found that clinicians review the assessment and plan sections of a progress note first and are more likely to review notes from the previous 24 hour period.^22,26 However, there is a research gap on documentation review practices using EHRs in an outpatient setting.

The purpose of this study was to address the gap in knowledge with respect to outpatient documentation; specifically, EHR use patterns during outpatient office visits. In this study, we looked at the work done in the EHR for office visits in outpatient ophthalmology at Oregon Health & Science University. Ophthalmology is a specialty that includes surgery and medicine and uses ancillary staff during outpatient visits, similar to many other medical specialties. Using audit logs, we identified the number of prior notes accessed during the course of an office visit by attending physicians, trainees, and ancillary staff.

Methods

This study was approved by the Institutional Review Board at Oregon Health & Science University (OHSU).

Study Environment

OHSU is a large academic medical center in Portland, Oregon. The ophthalmology department includes over 50 faculty providers, who perform over 130,000 annual outpatient examinations. The department provides primary eye care, and serves as a major referral center in Pacific Northwest and nationally. We studied the patterns of EHR use during office visits for 12 attending physicians in 6 different subspecialties of ophthalmology. For all of the physicians, ancillary staff performed an initial exam during the office visit. Trainees were a part of some of the office visits for all of the attending physicians.

Over several years, an institution-wide EHR system (EpicCare; Epic Systems, Madison, WI) was implemented throughout OHSU. This vendor develops software for mid-size and large medical practices, is a market share leader among large hospitals, and has implemented its EHR systems at over 200 hospital systems in the United States. In 2006, all ophthalmologists at OHSU began using this EHR. All ambulatory practice management, clinical documentation, order entry, medication prescribing, and billing tasks are performed using components of the EHR.

Data

EHR data used in this study was obtained from OHSU’s datamart based on principles we have previously described.²⁸ Audit log entries were extracted for a set of office visits in 6 different subspecialties of ophthalmology (general ophthalmology, pediatrics, cornea, retina, neuro-ophthalmology, and oculoplastics), for two ophthalmologists in each subspecialty. The time range for this study were office visits completed between January, 1, 2015 and December, 31, 2017. In order to find a set of “typical office visits” for each provider, we applied the following selection rules for inclusion:

1. Follow-up office visit that was not a post-op visit.

2. Previous office visit was within the past 2 years.

3. One office visit per patient was used—the most recent one.

4. Patient diagnosis was one of three most common for that specialty.

5. Office visit data in the datamart was complete for the office visit.

The data needed for this study included office visit information for the patient (time and date of visit, check in, check out, and close date times, and office visit and patient IDs), a list of all prior notes available for the patient (type of note, date, and department), and audit log entries for the patient office visit. The prior note types were categorized as those for prior office visits and other notes (procedure, diagnostic visit, telephone, etc.). Audit log entries (timestamp, user ID, note ID, and type of entry) were queried from the data mart for each office visit, starting 3 days before the office visit and ending when the encounter was closed. We included the 3 days before the office visit to include any preparation for the visit that may have occurred the business day prior to the appointment time.

Analysis

Audit log entries were analyzed for the number of unique note IDs that were different from the study office visit ID. Based on the user IDs and timestamps associated with the audit log entries, we analyzed which users accessed the notes and when. From the note data, we determined the type and number of notes that were accessed by each user, along with the percentages of notes accessed. ANOVA comparisons with Tukey Honest Significant Differences were performed to determine the significance in the number of notes accessed between the users and note types. All statistics and data manipulations were performed using R.²⁹

Results

The dataset of office visits for our study that met the inclusion criteria for our study included 7,138 office visits for 7,138 unique patients. This was a subset of the total number of office visits (62,358) and patients (18,584) for the 12 physicians in 6 subspecialties during the entire study period from 2015 to 2017 (Table 1). The audit log included entries for trainees in 12.2% of the study office visits, with this percentage varying among the subspecialties from 0.5% of oculoplastics visits to 34.9%of visits in cornea.

Table 1:

Dataset Characteristics. The total number of office visits (62,358) and patients (18,584) for the 6 specialties during the study period (2015-2017), and the number that met our inclusion criteria and were part of the study (n = 7,138). Also, 12.2% of our study visits involved trainees.

Subspecialty	Total for Study Period, 2015-2017		Included in Study
(2 Physicians ea.)	#Office Visits	#Patients	#Office Visits/Patients	#Office Visits With Trainees
General	19712	6085	1813	27(1.5%)
Pediatrics	14773	5227	1547	184(11.9%)
Cornea	18539	4767	954	333(34.9%)
Retina	22926	4028	1760	166(9.4%)
Neuro	8927	4864	460	159(34.6%)
Oculoplastics	11966	4925	604	3(0.5%)
Total	62358	18584	7138	872(12.2%)

Analyzing the audit logs for note IDs, we found that on average each office visit involved reviewing approximately 3 notes, half of which were prior office visit notes (progress notes from an office visit) and the other half were other notes (results, procedures, etc.). (Table 2). This varied by subspecialty, however; neuro-ophthalmology reviewed the most and retina the fewest on average. The number of reviewed prior office visit notes varied on average from about 1 to 2 and the reviewed other notes ranged from under 1 to over 2 per visit.

Table 2:

The number of notes reviewed by specialty. The number of notes reviewed during the course of an office visit by specialty, then further divided by type: office visit or other (results, procedure, etc.). Overall, about 3 notes were reviewed per visit: half prior office visit notes and half other notes.

#Notes Reviewed
Subspecialty	All Notes	Prior Office Visit Notes	Other Notes
General	3.4 ± 2.6	1.7 ± 1.8	0.7 ± 1.2
Pediatrics	2.7 ± 2.5	2.0 ± 1.8	0.7 ± 1.2
Cornea	2.6 ± 2.7	1.8 ± 1.7	0.8 ± 1.4
Retina	2.4 ± 2.7	1.0 ± 2.0	1.4 ± 1.3
Neuro	4.1 ± 3.0	1.7 ± 1.6	2.4 ± 2.0
Oculoplastics	2.6 ± 1.7	1.0 ± 1.2	1.6 ± 1.2
Overall	2.9 ± 2.6	1.5 ± 1.8	1.3 ± 1.5

Breaking down the reviewed notes by type of user, we found that on average, each user reviewed at least one note. Trainees and ancillary staff reviewed closer to 2 notes total during the visit while the attending physician reviewed only one on average (Table 3). Further, the percentage of available notes that were accessed was very low—on average, under 5% of all total available notes and under 12% for available notes from the past year (Table 3). Ancillary staff accessed the highest percentage of notes while attending physicians accessed the fewest; physicians reviewed significantly fewer notes than staff, both in number and percentage (p < 0.0001).

Table 3:

The number of notes reviewed by each type of user. Overall, the attending physician reviews the least number of notes.

	Number & Percentage of Notes Reviewed
	# Notes	% All Notes	% Notes from Past Year
Ancillary Staff	1.8 ± 1.7	4.1 ± 6.4%	11.8 ± 14.2%
Trainee	1.6 ± 3.7	2.8 ± 6.3%	6.5 ± 10.7%
Attending Physician	1.1 ± 1.8*,±	2.6 ± 5.3%*	6.2 ± 10.7%*

^*Number of Notes Physician Reviewed is less than Ancillary Staff reviewed, p<0.0001

^†Number of Notes Physician reviewed is less than Trainee reviewed, p<0.0001

Discussion

This study has 3 key findings: 1) Clinicians review only a very small fraction of the total number of notes available in the EHR during typical follow-up office visits, 2) There is variation among specialties and physicians with respect to this note review in the EHR, and 3) Ancillary staff review more notes than providers during typical follow-up office visits. Taken together, these findings indicate that most data in the EHR is not used, and a fundamental redesign of EHR for note and data review may be warranted.

The first key finding is that very few notes are reviewed during the course of a typical follow-up office visit for all subspecialties. On average only one to two prior office visit notes and non-office visit notes are reviewed, which represents a small fraction of available notes. Even when considering only notes from the past year, the percentages of notes accessed is at most 11%. The low number and percentage of non-office visit notes reviewed is not surprising since many of these notes do not merit review (e.g. phone or messaging encounters). The low number of prior office visit note reviews (Table 2) is more surprising, however. We speculate that there are a few reasons for this: 1) Providers and staff have become used to copying the previous note into the current one so that the progress note becomes a running summary of the patient.^30,31 This saves clicks within the EHR, both for documenting and reviewing. 2) Providers use alternate charting strategies, such as problem list based charting, that relies less on prior office visit notes. For example, the retina subspecialty at OHSU uses this form of charting, which explains their review of fewer prior office visit notes. Again, these charting strategies were developed as a way to save time in the EHR. A previous study in the inpatient setting showed similar results; close to 100% of notes are reviewed within the first 24 hours they were written, but this percentage drops quickly after that. After 3 days, only about 10% of notes are reviewed and after about a year, only about 0.01% of notes are reviewed.²² Taken together, these findings suggest that clinicians tend to access the most recent notes as part of their clinical documentation practices.

The second key finding is that there is variability in the number of note accesses during an office visit depending on the specialty, the type of note and the user. As shown in Table 2, while at least one note was always reviewed during an office visit (usually a prior office visit note), the total number of notes reviewed varies. A further breakdown of note accesses given in Figure 1 shows this variability. Not only did the note accesses vary by specialty and user, they also varied between physicians within the same subspecialty. One subspecialty, neuro-ophthalmology, reviews the most notes while retina reviews the fewest.

Figure 1:

Number of Notes Reviewed by User and Specialty. Average number of notes reviewed by each type of user during an office visit, for each of the 2 physicians in each subspecialty. For most of the specialties, the attending physician reviews the fewest number of notes.

Subspecialties such as neuro-ophthalmology rely on testing, which explains why this subspecialty reviews more non office visit notes. It is more difficult to explain why some subspecialties review multiple prior office visit notes and some only one, as well as why physicians in the same subspecialty review different numbers of notes—this may have more to do with the way the clinicians are using the EHR than with the needs of the subspecialty.

The third key finding is that ancillary staff access significantly more notes than providers during typical follow-up office visits (Table 3). This study was not designed to identify or analyze the underlying reasons for this finding, however. Anecdotal evidence suggests that the reason for this is that the staff and trainees access and copy relevant prior notes into the current progress note so that the physician does not have to search for information during the exam. Another possibility is that staff and trainees may verbally communicate content from these prior notes to the attending physician so that s/he does not need to access the note. Despite these practices, however, physicians still access other notes, especially when with the patient as shown in Figure 2. Further investigation of these workflows is needed to understand the information needs and retrieval strategies of staff, trainees, and providers during office visits.

Figure 2:

When physicians review notes. In all specialties, physicians review the most notes during the office visit compared to the time before the visit and after visit before the encounter closes.

The results from this study suggest that the majority of data in a health record is not being used during follow-up office visits. Current time pressures¹⁵, EHR inefficiencies¹⁰, and bloated notes¹⁹ are contributing to staff and physicians using workflows that review limited information in the shortest amount of time during an office visit. Nevertheless, it is possible that much information, particularly trends, are missed when only a small fraction of notes are reviewed. We feel that this study is evidence that a comprehensive EHR redesign is necessary to better support the information needs of clinicians during an office visit. The power of electronic data storage is not being realized in EHRs; notes are still generated sequentially, and information is still manually searched and retrieved as it was in paper charts. Easily available views and summaries of prior office visits along with trending results are needed to support the information needs of busy providers and staff, while dynamic and collaborative progress notes would reduce note bloat and missed information. Further, these views of prior notes may need customization to better support the different needs of subspecialists. Using careful studies of actual EHR use during clinical care such as this one are key to ensuring redesigns meet the clinical needs of busy providers.

There are several limitations to this study: 1) Audit logs may not contain complete records of note accesses. Notes can be reviewed through patient snapshot views (accessed in less than 5% of our study office visits), which do not generate audit log records with note IDs. In addition, templates can copy forward previous notes, which do not generate audit log entries of note accesses. Nevertheless, we feel that our results accurately represent the limited note review occurring during office visits. 2) Only ophthalmology outpatient office visits were analyzed; no other outpatient specialties were included. We speculate that these results are applicable to other specialties, but further studies are needed to confirm this. 3) Only the most common follow-up office visits for each subspecialty were analyzed in this study. Further work is needed to determine note accesses for other types of office visits, including those for new patients, atypical diagnoses, and pre-operative and post-operative visits.

Conclusions

A major advantage of EHR systems is the availability of clinical data. However, this study suggests that the vast majority of clinical data, even involving patient documentation from prior office visits, is not being used by clinicians. We feel this has important implications for the quality and efficiency of clinical care, and for the design of future EHR systems. Additional collaboration between informaticians, clinicians, and policymakers will be required to create new EHR system designs that can best support the delivery of clinical care.