xRead - Outside the Box (March 2024) - Full Articles

15314995, 2021, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.29073 by Loyola University Health Sciences Library, Wiley Online Library on [19/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

participating institutions (Table III). There were other small differences noted between institutional means in “ schedule, ” “ inbasket, ” and “ notes ” activities.

TABLE III. Institutional Differences in EHR Clinical Activity Per Day.

Activity Institutional Comparison

Mean Difference in Min, (LCL, UCL)

P adjusted

DISCUSSION Since their development in the 1970s, EHRs are increasingly part of the modern healthcare landscape with over 90% patients having their information in at least one EHR. 8 While the adoption of EHR is fairly uniform, each individual institution has varied implementation and work fl ows creating unique implementations of the same EHR. While the average otolaryngologist across 10 institu tions nationally spends an estimated 70 min with the EHR per clinic day and 7.4 min per patient, there is large vari ance between institutions and between individuals at those institutions. In their triennial membership survey, members of the American Academy of Otolaryngology – Head and Neck Surgery (AAO-HNS) estimates that their weekly time in clinic is 26 – 30 hours per week. 12 At face value, 70 min per day interacting with the EHR may seem low compared to how much work is required in the EHR for each patient visit. It is important to remember that PEP data re fl ect only direct interaction of the EHR with a 5-second time out. Thus, any recorded time excludes the physical exam, potentially radiology image review, any counselling pro vided, teaching, and of fi ce-based procedures. The numerator of recorded time in EHR activities is validated, accurate, and collected automatically in real time. In is important to remember that the metric ’ s fi nal value is also dependent on the denominator. For daily PEP metrics, the calculation of the denominator (clinically active days) is likely an overestimate of the otolaryngolo gist ’ s ambulatory clinics. Firstly and most importantly, every day that the EHR is opened by the attending physi cian counts to the denominator including operative days, administrative days, and even weekends spent catching up on documentation. Further, the denominator does not dis tinguish between half-day clinics and full-day clinics, which are both counted as one clinically active day. This larger denominator results in a series of daily activity met rics which are an underestimation of actual daily EHR time required to see a full day of clinic patients. This cri tique of the daily PEP values is supported by the fewer reported average number of appointments per day, 15.6 compared to nationally reported medians. From the AAO HNS membership survey, median number of visits per year is 2839. 12 Assuming 3 full clinic days, and the median 48 weeks of work, 2839 visits roughly translate into 19.7 patients per full clinic day. Therefore, the per appointment PEP values are likely a better estimate of EHR burden for otolaryngologists. These per appointment values better account for variation in workload, schedules, and EHR uti lization patterns and are more generalizable for inferences and conclusions. For instance, based on the AAO-HNS sur vey estimate of 19.7 patients per day, a typical otolaryngol ogist would spend 146 min (2.4 hours) daily interacting with the EHR. Further, 7.4 min per appointment repre sents nearly half a 15-minute return visit appointment slot. While not all the 7.4 min occurs during clinic

Chart review

NS

In basket 4 vs. 2

− 4.3 ( − 7.6, − 1.1) − 6.4 ( − 10.9, − 2.0) − 6.4 ( − 11.3, − 2.2)

.002

9 v.s 2

<.001 <.001

10 v.s 2

Notes

− 17.7 ( − 31.4, − 4.1) − 14.6 (28.1, − 1.2)

7 v.s 4 9 vs. 4

.002 .020

Orders

− 8.2 ( − 14.8, − 1.5) − 7.4 ( − 15.2, − 0.26) − 9.9 ( − 17.1, − 2.7) − 10.8 ( − 18.2, − 3.4) − 9.6 ( − 16.8, − 2.4) − 8.1 ( − 15.5, − 0.9) − 8.8 ( − 17.0, − 0.6) − 10.3 ( − 19.5, − 1.2) − 10.6 ( − 20.2, − 1.1) − 10.8 ( − 19.7, − 1.8) − 10.2 ( − 19.4, − 1.04) − 10.4 ( − 19.3, − 1.5) − 11.1 ( − 14.7, − 7.5) − 12.6 ( − 18.0, − 7.3) − 12.9 ( − 19.0, − 6.9) − 13.1 ( − 18.0, − 8.1) − 12.5 ( − 17.9, − 7.1) − 12.7 ( − 17.6, − 7.8) − 7.8 ( − 12.8, − 2.7) − 3.5 ( − 6.8, − 0.22) − 3.2 ( − 5.9, − 0.4) − 3.77 ( − 6.7, − 0.8) − 4.9 ( − 7.7, − 2.2) − 3.4 ( − 5.9, − 0.9)

4 vs. 1 6 vs. 1 7 vs. 1 8 vs. 1 9 vs. 1 10 vs. 1 4 vs. 1 5 vs. 1 6 vs. 1 7 vs. 1 8 vs. 1 9 vs. 1 4 vs. 2 5 vs. 2 6 vs. 2 7 vs. 2 8 vs. 2 9 vs. 2 10 vs. 2

.004 .048

<.001 <.001

.001 .014

Other

.026 .013 .016 .006 .016 .008 .028

<.001 <.001 <.001 <.001 <.001 <.001

Schedule 6 vs. 2

.026 .010 .002

7 vs. 2 8 vs. 2 10 vs. 2 10 vs. 4

<.001 <.001

Visit navigator

NS

This data is graphically presented in the boxplots in Figure 3. Using the daily PEP data, the mean time in all EHR activities for each institution were compared using an analysis of variance. Post hoc analysis with adjusted con fi dence intervals is shown in the table below. For ease of inter pretation and visualization only mean differences that were statistically signif icant were included. Activities without statistically signi fi cant differences were left blank and denoted NS for their P adjusted . EHR = electronic health records; NS = not statistically signi fi cant; LCL = lower con fi dence limit of 95%con fi dence interval; PEP = provider ef fi ciency pro fi le; UCL = upper con fi dence limit of 95% con fi dence interval;

in the “ orders ” and “ other ” activities. With respect to average daily time in orders, the mean time for institu tion 1 was greater than six other institutions in this cohort (4, 6 – 10). Both institutions 1 and 2 spent more time in “ other ” activities and compared to the rest of the

Laryngoscope 131: May 2021

Giliberto et al.: National Time Spent in EHR 979