Video Abstract

Video Abstract

Close modal
BACKGROUND:

The time providers spend using their electronic health records (EHRs) delivering care and its potential impact on patient care are of concern for the health care system. In studies to date, researchers have focused on providers who primarily care for adults. Scant information exists for pediatricians. Given this gap, it is important to quantify EHR activity for this group.

METHODS:

We studied pediatricians practicing in US-based ambulatory practices using the Cerner Millennium EHR by extracting data from software log files in the Lights On Network for the calendar year 2018 and summarizing the time spent on each of 13 clinically-focused EHR functions according to clinical specialty.

RESULTS:

Our data included >20 million encounters by almost 30 thousand physicians from 417 health systems. Pediatric physicians spent an average of 16 minutes per encounter using their EHR. Chart review (31%), documentation (31%), and ordering (13%) functions accounted for most of the time. The distribution of time spent by providers using their EHR is highly variable within subspecialty but is similar across specialties. Because of data limitations, we were unable to examine geographic or health system–specific variation.

CONCLUSIONS:

Pediatricians, like physicians who care for adults, spend a large portion of their day using their EHR. Additionally, although chart review and documentation accounted for 62% of the activity, as in previously published studies, in our study, we found that chart review accounted for half of that time. Wide variation suggests opportunities to optimize both the processes of entering information and searching for patient data within the EHR.

What’s Known on This Subject:

Although >94% of pediatricians in the United States use an electronic health record (EHR), we know little about how much time they devote to using this vital tool and how time varies by subspecialty.

What This Study Adds:

Pediatricians average 16 minutes per encounter using their EHR, with record review and documentation accounting for 62% of the time. These data provide a baseline from which we can judge the impact of innovations or policy-related EHR changes.

More than 94% of pediatricians across the United States now use electronic health records (EHRs).1,2  There is less literature documenting the time pediatricians spend using the EHR to accomplish their work than there is for some other specialties. A recent study focused on the allocation of time in ambulatory nonpediatric practices revealed that physicians spent 2 of every 3 hours using the EHR or completing desk work.3  More recently, a large-scale study of adult medicine physicians revealed that they devote 16 minutes and 14 seconds per encounter to EHR activities.4  On the basis of a self-administered survey of 1619 pediatricians, pediatricians spend an average of 3.4 hours per day documenting care.1  Tai-Seale et al,5  on the basis of an analysis of EHR log files, found that pediatricians spend 2.48 ± 1.3 hours per day using the EHR, which is similar.

As a response to the length of time physicians spend using their EHR, health systems have devoted substantial resources to optimize EHR workflows to improve physician efficiency.6  The use of medical scribes is one example of an approach by which some practices have demonstrated a degree of success in reducing documentation time.7,8  Optimizing EHR system configuration has also yielded benefits.9 

We undertook a descriptive, multi-institutional, national analysis of ambulatory pediatric medical specialty physician EHR time use, measured by using EHR log files, to better document the time ambulatory physicians in pediatric specialties devote to performing various tasks using their EHR.

We used data on EHR activities in the ambulatory setting from January 1, 2018, to December 31, 2018, by all physicians identified in their EHR system as pediatricians or adolescent medicine specialists in any of 2191 health care organizations across the US tracked in Cerner’s Lights On database.

We extracted deidentified software log data from Cerner’s Lights On database. The Lights On Network consists of a collection of systems that monitor hardware and software system activity, clinical application functionality, and other user activities. These data represent nearly the entire population of sites and physicians who use the Cerner EHR. For this analysis, we were provided with summary statistics (means, medians, SDs) and total counts for most dichotomous data. This analysis data set consists of 29 971 rows of data (one record for every pediatric physician who used Cerner Millennium in 2019). We summarized time data for each clinical activity undertaken by that physician (eg, ordering). Because the research did not involve any interaction or intervention with the subjects and the subjects were not individually identifiable, the research did not qualify as human subjects research under 45 Code of Federal Regulations part 46.10 

We performed a descriptive analysis using summarized EHR log entries reflecting individual software modules and services executed while the provider was using the system. Cerner analysts mapped activities and time data to specific clinical tasks, such as writing documentation, placing orders, reviewing historical notes, or reviewing clinical decision support alerts (Table 1). The algorithm attributed all EHR time to precisely one of these activity categories.

TABLE 1

All EHR Time Was Allocated to One of These Mutually Exclusive, Clinically Focused EHR Function Categories

Clinically Focused EHR Function CategoryClinically Focused EHR Function Description
Chart review Discovering and reviewing clinical results, observations, and notes in the EHR 
Documentation Recording documentation and creating notes 
Message center Reviewing, responding, and acting on messages 
Orders Writing orders 
Patient discovery Performing tasks related to searching for patients, appointment scheduling, and similar activities 
Medication recommendation Reconciling a patient’s medication data and activities 
Allergies Reviewing and updating a patient’s list of allergies 
Problems and diagnoses Reviewing and updating a patient’s problem list 
Alerts Reviewing and responding to alerts 
Health maintenance Reviewing and responding to health maintenance or preventive care clinical decision support prompts 
Patient education Selecting patient education and materials, completing departure information, and printing materials 
History (family, social, surgery) Reviewing and updating family, social, and past surgical history 
Othera EHR times that have not yet been categorized into another specific activity or which represent categories of activities that are too small to uniquely represent 
Clinically Focused EHR Function CategoryClinically Focused EHR Function Description
Chart review Discovering and reviewing clinical results, observations, and notes in the EHR 
Documentation Recording documentation and creating notes 
Message center Reviewing, responding, and acting on messages 
Orders Writing orders 
Patient discovery Performing tasks related to searching for patients, appointment scheduling, and similar activities 
Medication recommendation Reconciling a patient’s medication data and activities 
Allergies Reviewing and updating a patient’s list of allergies 
Problems and diagnoses Reviewing and updating a patient’s problem list 
Alerts Reviewing and responding to alerts 
Health maintenance Reviewing and responding to health maintenance or preventive care clinical decision support prompts 
Patient education Selecting patient education and materials, completing departure information, and printing materials 
History (family, social, surgery) Reviewing and updating family, social, and past surgical history 
Othera EHR times that have not yet been categorized into another specific activity or which represent categories of activities that are too small to uniquely represent 
a

Several categories that individually accounted for small proportions of the total time were aggregated into 1 category entitled “other” for reporting purposes.

One challenge inherent in a secondary analysis of log file data is distinguishing between time spent actively engaged in EHR-related activities and time spent logged into an EHR but not engaged in these activities (such as taking a phone call). To address this challenge, the Lights On Network provided active time values, calculated by using metadata captured in the log files on the basis of a 2-tiered categorization. If a user is logged into the system with activities recorded <45 seconds apart, the user is considered an active user. After 45 seconds, active time continues as long there are ≥3 mouse clicks per minute, ≥15 keystrokes per minute, or mouse movement of ≥1700 pixels per minute.4 

To compute time per encounter, we accumulated all active EHR time in each category during a specific period and then divided by the number of ambulatory encounters completed that period. We considered an encounter completed at the date and time when the note was signed. Physicians could have entered notes directly into the EHR, written and scanned them, or dictated and transcribed them. We computed times using 1 year and 1 month as the period. We chose data aggregated by months for our primary analysis because the more extended period allowed for the fact that providers frequently complete work related to an encounter at later dates. Aggregating by months allowed us to include all the work for an encounter, even when it extended to subsequent days. We report the count of encounters from the 1-year interval because this count most directly reflects the total number of encounters.

We defined after-hours EHR use as any time spent using an EHR between 6 pm and 6 am local time on weekdays and anytime on weekends.

We used Vertica Structured Query Language to compute descriptive statistics for active time per patient encounter by specialty for each clinically focused EHR function for each period across all health systems. Finally, we used Excel (version 1808; Microsoft Corporation, Redmond, WA) and R (version 3.5.1; R Foundation for Statistical Computing, Vienna, Austria) to summarize the data for presentation. We used Excel to perform 1-way analysis of variance and Tukey Honest Significant Difference post hoc testing for differences in total active time between subspecialties.

There was no external funding for this study.

The analysis included data from >20 million encounters by almost 30 000 pediatricians (Table 2). These physicians practiced at integrated delivery networks (34%), regional hospitals (30%), independent physician groups (22%), and academic medical centers (11%). They practiced throughout the United States.

TABLE 2

Number of Physicians, Comparison With 2018 American Board of Pediatrics Count of Pediatricians by Subspecialty, Patient Encounters, and Active Time by Using the EHR in the Ambulatory Setting During 2018 Included in the Analysis by Specialty

Pediatric SubspecialtyStudy Physicians, CountABP Physicians, Count (%)Patient Encounters, CountActive EHR Time, min, Mean ± SD
Surgery 632 — 215 062 6.82 ± 9.85 
Cardiology 1421 2587 (55) 455 782 12.41 ± 9.08 
Critical care 1174 2603 (45) 45 054 13.01 ± 11.54 
General 23 473 65 146 (36) 17 646 576 13.53 ± 8.50 
Gastroenterology 792 1725 (46) 496 532 17.52 ± 9.53 
Pulmonology 474 1181 (40) 253 312 17.16 ± 13.22 
Hematology-oncology 1088 2699 (40) 377 786 17.18 ± 11.37 
Nephrology 336 711 (47) 161 101 17.74 ± 10.66 
Endocrinology 666 1589 (42) 561 755 19.72 ± 12.13 
Infectious diseases 376 1488 (25) 54 674 20.82 ± 10.33 
Rheumatology 171 423 (40) 105 368 26.41 ± 5.56 
Total 31 551 70 748 (44) 20 674 306 — 
Pediatric SubspecialtyStudy Physicians, CountABP Physicians, Count (%)Patient Encounters, CountActive EHR Time, min, Mean ± SD
Surgery 632 — 215 062 6.82 ± 9.85 
Cardiology 1421 2587 (55) 455 782 12.41 ± 9.08 
Critical care 1174 2603 (45) 45 054 13.01 ± 11.54 
General 23 473 65 146 (36) 17 646 576 13.53 ± 8.50 
Gastroenterology 792 1725 (46) 496 532 17.52 ± 9.53 
Pulmonology 474 1181 (40) 253 312 17.16 ± 13.22 
Hematology-oncology 1088 2699 (40) 377 786 17.18 ± 11.37 
Nephrology 336 711 (47) 161 101 17.74 ± 10.66 
Endocrinology 666 1589 (42) 561 755 19.72 ± 12.13 
Infectious diseases 376 1488 (25) 54 674 20.82 ± 10.33 
Rheumatology 171 423 (40) 105 368 26.41 ± 5.56 
Total 31 551 70 748 (44) 20 674 306 — 

ABP, American Board of Pediatrics; —, not applicable.

The average total active time per encounter that pediatric physicians spent in the EHR was 16 minutes, with 12% of this time spent after hours. The mean active time for all clinically focused EHR functions combined varied widely within specialty (P < .001), with clinically significant differences between several groups, as summarized in Table 3.

TABLE 3

Multiple Comparisons Analysis of Mean Active Times Among Pediatric Subspecialties

GroupGroup Mean, minDifference, minLower Bound, minUpper Bound, min
Cardiology versus 12.41 — — — 
 General* 13.53 1.12 0.31 1.93 
 Endocrinology** 19.72 7.31 5.92 8.70 
 Gastroenterology** 17.52 5.11 3.80 6.42 
 Hematology-oncology** 17.18 4.77 3.58 5.96 
 Infectious diseases** 20.82 8.41 6.70 10.12 
 Nephrology** 17.74 5.33 3.54 7.12 
 Pulmonology** 17.16 4.75 3.18 6.32 
 Rheumatology** 26.41 14 11.61 16.39 
 Surgery** 6.82 −5.59 −4.18 −7.00 
Critical care versus 13.01 — — — 
 Endocrinology** 19.72 6.71 5.28 8.14 
 Gastroenterology** 17.52 4.51 3.15 5.87 
 Hematology-oncology** 17.18 4.17 2.93 5.41 
 Infectious diseases** 20.82 7.81 6.06 9.56 
 Nephrology** 17.74 4.73 2.90 6.56 
 Pulmonology** 17.16 4.15 2.54 5.76 
 Rheumatology** 26.41 13.4 10.98 15.82 
 Surgery** 6.82 −6.19 −4.73 −7.65 
Endocrinology versus 19.72 — — — 
 General** 13.53 6.19 5.03 7.35 
 Hematology-oncology** 17.18 2.54 1.09 3.99 
 Nephrology* 17.74 1.98 0.00 3.96 
 Pulmonology** 17.16 2.56 0.79 4.33 
 Rheumatology** 26.41 6.69 4.16 9.22 
 Surgery** 6.82 −12.9 −11.26 −14.54 
Gastroenterology versus 17.52 — — — 
 General** 13.53 −3.99 −2.92 −5.06 
 Infectious diseases** 20.82 3.3 1.45 5.15 
 Rheumatology** 26.41 8.89 6.40 11.38 
 Surgery** 6.82 −10.7 −9.13 −12.28 
General versus 13.53 — — — 
 Hematology-oncology** 17.18 3.65 2.73 4.57 
 Infectious diseases** 20.82 7.29 5.75 8.83 
 Nephrology** 17.74 4.21 2.59 5.83 
 Pulmonology** 17.16 3.63 2.26 5.00 
 Rheumatology** 26.41 12.88 10.61 15.15 
 Surgery** 6.82 −6.71 −5.52 −7.90 
Hematology-oncology versus 17.18 — — — 
 Infectious diseases** 20.82 3.64 1.88 5.40 
 Rheumatology** 26.41 9.23 6.80 11.66 
 Surgery** 6.82 −10.36 −8.88 11.84 
Infectious diseases versus 20.82 — — — 
 Nephrology** 17.74 −3.08 −0.86 −5.30 
 Pulmonology** 17.16 −3.66 −1.62 −5.70 
 Rheumatology** 26.41 5.59 2.87 8.31 
 Surgery** 6.82 −14 −12.08 −15.92 
Nephrology versus 17.74 — — — 
 Rheumatology** 26.41 8.67 5.90 11.44 
 Surgery** 6.82 −10.92 −8.93 −12.91 
Pulmonology versus 17.16 — — — 
 Rheumatology** 26.41 9.25 6.6 11.88 
 Surgery** 6.82 −10.34 −8.55 −12.13 
Rheumatology versus 26.41 — — — 
 Surgery** 6.82 −19.59 −17.04 −22.14 
GroupGroup Mean, minDifference, minLower Bound, minUpper Bound, min
Cardiology versus 12.41 — — — 
 General* 13.53 1.12 0.31 1.93 
 Endocrinology** 19.72 7.31 5.92 8.70 
 Gastroenterology** 17.52 5.11 3.80 6.42 
 Hematology-oncology** 17.18 4.77 3.58 5.96 
 Infectious diseases** 20.82 8.41 6.70 10.12 
 Nephrology** 17.74 5.33 3.54 7.12 
 Pulmonology** 17.16 4.75 3.18 6.32 
 Rheumatology** 26.41 14 11.61 16.39 
 Surgery** 6.82 −5.59 −4.18 −7.00 
Critical care versus 13.01 — — — 
 Endocrinology** 19.72 6.71 5.28 8.14 
 Gastroenterology** 17.52 4.51 3.15 5.87 
 Hematology-oncology** 17.18 4.17 2.93 5.41 
 Infectious diseases** 20.82 7.81 6.06 9.56 
 Nephrology** 17.74 4.73 2.90 6.56 
 Pulmonology** 17.16 4.15 2.54 5.76 
 Rheumatology** 26.41 13.4 10.98 15.82 
 Surgery** 6.82 −6.19 −4.73 −7.65 
Endocrinology versus 19.72 — — — 
 General** 13.53 6.19 5.03 7.35 
 Hematology-oncology** 17.18 2.54 1.09 3.99 
 Nephrology* 17.74 1.98 0.00 3.96 
 Pulmonology** 17.16 2.56 0.79 4.33 
 Rheumatology** 26.41 6.69 4.16 9.22 
 Surgery** 6.82 −12.9 −11.26 −14.54 
Gastroenterology versus 17.52 — — — 
 General** 13.53 −3.99 −2.92 −5.06 
 Infectious diseases** 20.82 3.3 1.45 5.15 
 Rheumatology** 26.41 8.89 6.40 11.38 
 Surgery** 6.82 −10.7 −9.13 −12.28 
General versus 13.53 — — — 
 Hematology-oncology** 17.18 3.65 2.73 4.57 
 Infectious diseases** 20.82 7.29 5.75 8.83 
 Nephrology** 17.74 4.21 2.59 5.83 
 Pulmonology** 17.16 3.63 2.26 5.00 
 Rheumatology** 26.41 12.88 10.61 15.15 
 Surgery** 6.82 −6.71 −5.52 −7.90 
Hematology-oncology versus 17.18 — — — 
 Infectious diseases** 20.82 3.64 1.88 5.40 
 Rheumatology** 26.41 9.23 6.80 11.66 
 Surgery** 6.82 −10.36 −8.88 11.84 
Infectious diseases versus 20.82 — — — 
 Nephrology** 17.74 −3.08 −0.86 −5.30 
 Pulmonology** 17.16 −3.66 −1.62 −5.70 
 Rheumatology** 26.41 5.59 2.87 8.31 
 Surgery** 6.82 −14 −12.08 −15.92 
Nephrology versus 17.74 — — — 
 Rheumatology** 26.41 8.67 5.90 11.44 
 Surgery** 6.82 −10.92 −8.93 −12.91 
Pulmonology versus 17.16 — — — 
 Rheumatology** 26.41 9.25 6.6 11.88 
 Surgery** 6.82 −10.34 −8.55 −12.13 
Rheumatology versus 26.41 — — — 
 Surgery** 6.82 −19.59 −17.04 −22.14 

—, not applicable

*

P < .05; ** P < .001.

In Table 4, the total time spent per encounter is summarized. Chart review accounted for the most time, followed by documentation, and then ordering. Collectively, these 3 activities accounted for 75% of EHR activity. Patient discovery, defined as time spent searching for patients, scheduling appointments, or doing similar activities, accounted for 9% of EHR activity. In Table 5, the breakdown of EHR function time by specialty is shown.

TABLE 4

Time Spent per Encounter on Major Clinically Focused EHR Function by US Pediatricians Caring for Patients in the Ambulatory Setting During 2018

Function CategoryTime per Encounter, s, Mean ± SDTime per Encounter, % of Total
Chart review 281 ± 465 31 
Documentation 282 ± 483 31 
Orders 112 ± 192 13 
Message center 67 ± 121 
Patient discovery 82 ± 130 
Other 37 ± 94 
Problem or diagnosis lists 17 ± 43 
Patient education 11 ± 33 
History 5 ± 25 
Allergy 0.9 ± 5 
Alerts 0.5 ± 2 
Health maintenance 0.3 ± 3 
Function CategoryTime per Encounter, s, Mean ± SDTime per Encounter, % of Total
Chart review 281 ± 465 31 
Documentation 282 ± 483 31 
Orders 112 ± 192 13 
Message center 67 ± 121 
Patient discovery 82 ± 130 
Other 37 ± 94 
Problem or diagnosis lists 17 ± 43 
Patient education 11 ± 33 
History 5 ± 25 
Allergy 0.9 ± 5 
Alerts 0.5 ± 2 
Health maintenance 0.3 ± 3 
TABLE 5

Time Spent per Encounter on Major Clinically Focused EHR Function by US Subspecialty Pediatricians Caring for Patients in the Ambulatory Setting During 2018

SubspecialtyPhysicians, nChart Review, s, Mean ± SDDocumentation, s, Mean ± SDPatient Discovery, s, Mean ± SDAlerts, s, Mean ± SDAllergy, s, Mean ± SDDischarge, s, Mean ± SDHealth Maintenance, s, Mean ± SDHistory, Mean ± SDMessaging, Mean ± SDOrders, s, Mean ± SDProblems and Diagnosis Lists, s, Mean ± SDOther, s, Mean ± SD
Emergency medicine 948 145.8 ± 259.8 302.6 ± 606.1 75.2 ± 138.3 0.3 ± 0.8 0.1 ± 0.6 28.2 ± 61.7 0 ± 0.1 0.1 ± 1.4 18.6 ± 49.6 61 ± 115.4 10.4 ± 23.6 41.6 ± 98.9 
General 23 473 273.9 ± 338.4 271.3 ± 356.5 79.9 ± 94.4 0.5 ± 1.4 0.9 ± 4.2 10.4 ± 29.1 0.4 ± 2.8 5.6 ± 23 64.5 ± 90.4 111.4 ± 145.9 17.5 ± 38.6 36.8 ± 81.7 
Cardiology 1421 283.4 ± 373.4 313.5 ± 454.4 85.2 ± 99 0.2 ± 0.7 0.3 ± 1.4 8.6 ± 32.6 0 ± 0.2 2.7 ± 14.3 50.3 ± 91 66.9 ± 84.9 10.1 ± 22.3 30.7 ± 70.2 
Critical care 1174 292.9 ± 784 383.4 ± 1077 106.9 ± 231.8 1 ± 7.6 0.9 ± 5.1 6.4 ± 64.9 0 ± 0.6 1.1 ± 12.3 33.5 ± 135.8 122 ± 297.9 10.8 ± 52.6 39.2 ± 212.7 
Endocrinology 666 421.3 ± 355 445 ± 436.7 90.2 ± 72 0.4 ± 0.9 1 ± 4.7 14.8 ± 37 0.2 ± 1.1 6.2 ± 24.1 122.4 ± 100.7 133.4 ± 105.5 12.2 ± 18.1 36.9 ± 54.4 
Gastroenterology 792 407.1 ± 510.2 352.6 ± 470.7 94.3 ± 91.5 0.5 ± 1.1 1.7 ± 8.8 15.9 ± 51.8 0.1 ± 0.4 6.5 ± 28.5 145.8 ± 180.4 139.1 ± 162.3 14.5 ± 41.9 42.6 ± 123.6 
Hematology-oncology 1088 413.6 ± 435 362.1 ± 432 126.3 ± 136.3 0.7 ± 1.8 0.8 ± 5.1 5.9 ± 20.9 0.1 ± 1.1 1.4 ± 11.2 70.1 ± 113.6 169.5 ± 192.9 9.3 ± 25.2 43.2 ± 93.9 
Infectious diseases 376 530 ± 789.9 509.2 ± 632.1 143.4 ± 170.3 0.7 ± 2 4.6 ± 50.2 13.7 ± 39.1 0.5 ± 7.2 4.1 ± 19.1 126.5 ± 244.6 160.4 ± 246.3 19.3 ± 36.3 43.7 ± 89.9 
Nephrology 336 425.2 ± 417 417 ± 450.7 111.1 ± 103 0.6 ± 1.2 0.8 ± 3.8 12.1 ± 36.6 0 ± 0.3 3.1 ± 12.6 127.8 ± 148.9 153 ± 150 14.3 ± 20.1 39.9 ± 65.3 
Pulmonology 474 376.6 ± 387 361.4 ± 476.1 109.4 ± 124.6 0.9 ± 2.1 0.7 ± 3.7 14.2 ± 41.2 0 ± 0.4 3.9 ± 18.7 101.5 ± 120.2 145.3 ± 163.3 14.2 ± 27.6 42.7 ± 88.4 
Rheumatology 171 518.1 ± 429.1 639.7 ± 604 135.8 ± 110.5 1 ± 1.7 1.1 ± 4.2 22.5 ± 50 0 ± 0.3 6.4 ± 23.7 248.5 ± 201.4 216.1 ± 172.2 15.1 ± 23.3 43.5 ± 55.4 
Urology 217 143.9 ± 191.4 153 ± 213.3 43.9 ± 52.1 0.1 ± 0.3 0.4 ± 2.8 3.7 ± 9.7 0 ± 0.1 3.4 ± 19.6 35.7 ± 44.8 45.7 ± 59.8 7.6 ± 18 20.5 ± 33.2 
Surgery 632 126.9 ± 169.5 177.9 ± 265.4 48.8 ± 61.9 0.1 ± 0.7 0.4 ± 2.6 3.4 ± 13.5 0 ± 0.1 2.6 ± 33.7 25 ± 43.4 41.5 ± 58.2 5.7 ± 15.6 17.7 ± 35.8 
SubspecialtyPhysicians, nChart Review, s, Mean ± SDDocumentation, s, Mean ± SDPatient Discovery, s, Mean ± SDAlerts, s, Mean ± SDAllergy, s, Mean ± SDDischarge, s, Mean ± SDHealth Maintenance, s, Mean ± SDHistory, Mean ± SDMessaging, Mean ± SDOrders, s, Mean ± SDProblems and Diagnosis Lists, s, Mean ± SDOther, s, Mean ± SD
Emergency medicine 948 145.8 ± 259.8 302.6 ± 606.1 75.2 ± 138.3 0.3 ± 0.8 0.1 ± 0.6 28.2 ± 61.7 0 ± 0.1 0.1 ± 1.4 18.6 ± 49.6 61 ± 115.4 10.4 ± 23.6 41.6 ± 98.9 
General 23 473 273.9 ± 338.4 271.3 ± 356.5 79.9 ± 94.4 0.5 ± 1.4 0.9 ± 4.2 10.4 ± 29.1 0.4 ± 2.8 5.6 ± 23 64.5 ± 90.4 111.4 ± 145.9 17.5 ± 38.6 36.8 ± 81.7 
Cardiology 1421 283.4 ± 373.4 313.5 ± 454.4 85.2 ± 99 0.2 ± 0.7 0.3 ± 1.4 8.6 ± 32.6 0 ± 0.2 2.7 ± 14.3 50.3 ± 91 66.9 ± 84.9 10.1 ± 22.3 30.7 ± 70.2 
Critical care 1174 292.9 ± 784 383.4 ± 1077 106.9 ± 231.8 1 ± 7.6 0.9 ± 5.1 6.4 ± 64.9 0 ± 0.6 1.1 ± 12.3 33.5 ± 135.8 122 ± 297.9 10.8 ± 52.6 39.2 ± 212.7 
Endocrinology 666 421.3 ± 355 445 ± 436.7 90.2 ± 72 0.4 ± 0.9 1 ± 4.7 14.8 ± 37 0.2 ± 1.1 6.2 ± 24.1 122.4 ± 100.7 133.4 ± 105.5 12.2 ± 18.1 36.9 ± 54.4 
Gastroenterology 792 407.1 ± 510.2 352.6 ± 470.7 94.3 ± 91.5 0.5 ± 1.1 1.7 ± 8.8 15.9 ± 51.8 0.1 ± 0.4 6.5 ± 28.5 145.8 ± 180.4 139.1 ± 162.3 14.5 ± 41.9 42.6 ± 123.6 
Hematology-oncology 1088 413.6 ± 435 362.1 ± 432 126.3 ± 136.3 0.7 ± 1.8 0.8 ± 5.1 5.9 ± 20.9 0.1 ± 1.1 1.4 ± 11.2 70.1 ± 113.6 169.5 ± 192.9 9.3 ± 25.2 43.2 ± 93.9 
Infectious diseases 376 530 ± 789.9 509.2 ± 632.1 143.4 ± 170.3 0.7 ± 2 4.6 ± 50.2 13.7 ± 39.1 0.5 ± 7.2 4.1 ± 19.1 126.5 ± 244.6 160.4 ± 246.3 19.3 ± 36.3 43.7 ± 89.9 
Nephrology 336 425.2 ± 417 417 ± 450.7 111.1 ± 103 0.6 ± 1.2 0.8 ± 3.8 12.1 ± 36.6 0 ± 0.3 3.1 ± 12.6 127.8 ± 148.9 153 ± 150 14.3 ± 20.1 39.9 ± 65.3 
Pulmonology 474 376.6 ± 387 361.4 ± 476.1 109.4 ± 124.6 0.9 ± 2.1 0.7 ± 3.7 14.2 ± 41.2 0 ± 0.4 3.9 ± 18.7 101.5 ± 120.2 145.3 ± 163.3 14.2 ± 27.6 42.7 ± 88.4 
Rheumatology 171 518.1 ± 429.1 639.7 ± 604 135.8 ± 110.5 1 ± 1.7 1.1 ± 4.2 22.5 ± 50 0 ± 0.3 6.4 ± 23.7 248.5 ± 201.4 216.1 ± 172.2 15.1 ± 23.3 43.5 ± 55.4 
Urology 217 143.9 ± 191.4 153 ± 213.3 43.9 ± 52.1 0.1 ± 0.3 0.4 ± 2.8 3.7 ± 9.7 0 ± 0.1 3.4 ± 19.6 35.7 ± 44.8 45.7 ± 59.8 7.6 ± 18 20.5 ± 33.2 
Surgery 632 126.9 ± 169.5 177.9 ± 265.4 48.8 ± 61.9 0.1 ± 0.7 0.4 ± 2.6 3.4 ± 13.5 0 ± 0.1 2.6 ± 33.7 25 ± 43.4 41.5 ± 58.2 5.7 ± 15.6 17.7 ± 35.8 

The physician population monitored by the Lights On Network represents a 44% sample of US pediatricians on the basis of comparison with the 2018 American Board of Pediatrics statistics.11  The mean total active time for EHR users per encounter across all specialties was 16 minutes. It ranged from 14.5 minutes (generalists) to 30.8 minutes (rheumatologists). We were not surprised by rheumatologists’ relatively long time, given the nature of their patient population and the volume of chronic illnesses they manage, including chronic pain.12  General pediatricians spent slightly more time than cardiologists but less time than other subspecialists, with rheumatologists using approximately twice as much time as cardiologists or general pediatricians. The proportion of time spent on each activity is remarkably similar across subspecialties (Table 5). Assuming an average pediatrician might care for 25 patients per day,13,14  they would spend 6 hours and 40 minutes using their EHR, a substantial portion of their day. Because these physicians are all using the same EHR, the observed variability must arise from other factors, such as configuration differences, implementation specifics, practice configuration (eg, how the care team divides tasks among themselves), individual provider choices, and similar factors.

Chart review and documentation account for almost 62% of EHR time, consistent with the proportion reported by Sinsky et al3  and by Overhage and McCallie.4  Further classifying this time into chart review and documentation activities provides additional insights, however. Physicians spend one-half of this time (or 31% of their overall EHR time) on chart review. Although not conclusive, studies are beginning to reveal an association between EHR use and improved guideline adherence15  and a decreased incidence of adverse effects due to medical errors.16,17  The time physicians devoted to chart review may be early evidence of providers using the EHR to improve the care they deliver. Emerging technologies, including voice commands to retrieve chart summary information18  and mobile versions of the EHR,19  may help to better balance time spent foraging records with time spent directly interacting with patients.

Documentation accounts for 31% of the time spent in the EHR and is often a target of physicians’ concerns. Documentation may be harder to delegate than some other tasks physicians perform, partly because the physicians would have to convey complex clinical content to another person or system. Adler-Milstein and Huckman20  found, for example, that physicians delegated only 16% of physical examination documentation and only 29% of the history of present illness documentation, whereas they delegated other tasks more often, such as recording vital signs (92%) and medical history (82%). Delegating the latter tasks may be easier because the delegee directly generates and records the data. Delegating recording of data, such as the findings of the physical examination not directly observed by the delegate, is more complex and requires careful verification of the recording of those findings, which may be harder to delegate. The 9% of provider EHR time devoted to patient discovery is surprising and may be amenable to improvement through a redesign.

The proportion of time spent using EHRs on the various patient care tasks performed by different subspecialties appears similar. Given the high variability within specialties, however, we can conclude little about differences or similarities. Acknowledging this statistical limitation, the long mean times spent documenting and short mean times spent coordinating care (messaging) are consistent with our clinical intuition.

The EHR active times by specialty reported in this study can serve as a benchmark for health system leadership, providers, payers, and policy makers. Health system leadership can take advantage of the data to understand the effort required by physicians to complete their work and the value of investing in optimizing the physician EHR workflow. Individual physicians can better understand their performance in the context of other providers by comparing their own EHR use time to the averages for their subspecialty and exploring opportunities to reorganize their practice or obtain further training as needed. Payers can gain insight into the effort required to complete this important part of a physician’s work when using an EHR and consider adjusting their expectations for data capture on the basis of the direct costs incurred. Finally, policy makers can take these data into account when defining their EHR certification processes and data capture expectations.

We recognize several limitations of this baseline study. First, the study is focused on a large but complete population of users from 1 EHR. Although this strength allows for analytic consistency among groups, the result is that the study excludes specific categories of users, such as those in smaller independent practices or those who have selected other large EHR platforms not included. In particular, the sample had a relatively low rate of physicians identifying as specialists in infectious disease and general pediatrics (2 groups likely to be in ambulatory or multispecialty group practices). Despite this deficiency, we were pleased to see results similar to those identified in the EHR time studies performed in adult populations. Second, the scale and scope of this study, which leveraged EHR user activity log files, provided analytical power at the expense of being unable to measure non-EHR time use.

Additionally, providers may have cared for only a portion of their patients using an EHR. This fact is largely controlled for by the per-patient encounter analysis. The study was limited to physicians. In many sites, advanced practice nurses are among the busiest care providers. Additional work should be done to assess EHR use by this group. Outlier records confound active EHR time measurement in Tables 2 and 3. The extreme variances in each of these mean values reflect a nonnormal, multimodal distribution of active EHR time across the time continuum. An analysis of that distribution was not performed. Finally, the Cerner Lights On Network team developed the concept of after-hours use to automate the categorization of work done in settings or times that are nontraditional. Not knowing the providers’ schedule reduces our confidence in the allocation of time to after hours.

Because we treat all weekend EHR time and any evening (after 6:00 pm and before 6:00 am) as after hours, we may be overestimating the amount of after-hours time. In this sense, the amount of time is arbitrary, although reproducible through EHR audit log analysis irrespective of the EHR system under scrutiny. The analysis is also dependent on the local health system correctly classifying provider roles and specialties. In this study, we focused on pediatric and adolescent medicine health care providers. Internationally, it is common for pediatric subspecialists to care for patients in their adult years.21  These adult patients may have unique and unfamiliar problems to address, which may account for increased time spent in chart review or documentation for those specialties. Finally, we were unable to analyze the sources of variation, such as differences in clinical processes or software configuration.

In this study, we estimate the time pediatric and adolescent medicine physicians devote to using an EHR in the ambulatory setting. The significant proportion of their total clinical time that providers spend performing tasks using their EHR is meaningful, and we certainly need to continue to identify and eliminate unnecessary and low-value activities across the entire physician workflow. The wide variability in the time providers within specialties spend using their EHR to care for patients is an important finding and warrants further investigation.

Dr Overhage conceptualized and designed the study, drafted the initial manuscript, collected data, conducted the initial analysis, and reviewed and revised the manuscript; Dr Johnson assisted with the editing of the final manuscript and associated tables, reviewed the analyses, and reviewed and revised the manuscript; and both authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: No external funding.

     
  • EHR

    electronic health record

1
Lehmann
CU
,
Kirkendall
ES
,
Temple
MW
,
Sisk
B
,
Krams
L
,
Schneider
JH
.
Use of electronic health records in pediatric office settings
.
Pediatrics
.
2019
;
144
(
1
Meeting Abstract
):
27
2
Temple
MW
,
Sisk
B
,
Krams
LA
,
Schneider
JH
,
Kirkendall
ES
,
Lehmann
CU
.
Trends in use of electronic health records in pediatric office settings
.
J Pediatr
.
2019
;
206
:
164
171.e2
3
Sinsky
C
,
Colligan
L
,
Li
L
, et al
.
Allocation of physician time in ambulatory practice: a time and motion study in 4 specialties
.
Ann Intern Med
.
2016
;
165
(
11
):
753
760
4
Overhage
JM
,
McCallie
D
 Jr.
.
Physician time spent using the electronic health record during outpatient encounters: a descriptive study
.
Ann Intern Med
.
2020
;
172
(
3
):
169
174
5
Tai-Seale
M
,
Olson
CW
,
Li
J
, et al
.
Electronic health record logs indicate that physicians split time evenly between seeing patients and desktop medicine
.
Health Aff (Millwood)
.
2017
;
36
(
4
):
655
662
6
Washington
V
,
DeSalvo
K
,
Mostashari
F
,
Blumenthal
D
.
The HITECH era and the path forward
.
N Engl J Med
.
2017
;
377
(
10
):
904
906
7
Gidwani
R
,
Nguyen
C
,
Kofoed
A
, et al
.
Impact of scribes on physician satisfaction, patient satisfaction, and charting efficiency: a randomized controlled trial
.
Ann Fam Med
.
2017
;
15
(
5
):
427
433
8
Mishra
P
,
Kiang
JC
,
Grant
RW
.
Association of medical scribes in primary care with physician workflow and patient experience
.
JAMA Intern Med
.
2018
;
178
(
11
):
1467
1472
9
Webber
E
,
Schaffer
J
,
Willey
C
,
Aldrich
J
Targeting pajama time: efforts to reduce physician burnout through electronic medical record (EMR) improvements
. [abstract].
Pediatrics
2018
;
142
:
611
10
US Department of Health and Human Services
.
Human subject regulations decision charts.
Washington, DC: US Department of Health and Human Services. Available at: https://www.hhs.gov/ohrp/regulations-and-policy/decision-charts/index.html#c1. Accessed October 13, 2020
11
American Board of Pediatrics
.
Pediatric workforce trends, spring
2018
. Available at: https://www.abp.org/sites/abp/files/pdf/abp_pediatric_workforce_trends.pdf. Accessed October 13, 2020
12
Tonelli
M
,
Wiebe
N
,
Manns
BJ
, et al
.
Comparison of the complexity of patients seen by different medical subspecialists in a universal health care system
.
[published correction appears in JAMA Netw Open. 2019;2(3):e190147]
.
JAMA Netw Open
.
2018
;
1
(
7
):
e184852
13
Farmer
SA
,
Shalowitz
J
,
George
M
, et al
.
Fully capitated payment breakeven rate for a mid-size pediatric practice
.
Pediatrics
.
2016
;
138
(
2
):
e20154367
14
Medscape. Medscape pediatrics compensation report:
2011
results. Available at: https://www.medscape.com/features/slideshow/compensation/2011/pediatrics. Accessed October 13, 2020
15
Stenner
SP
,
Chen
Q
,
Johnson
KB
.
Impact of generic substitution decision support on electronic prescribing behavior
.
J Am Med Inform Assoc
.
2010
;
17
(
6
):
681
688
16
Furukawa
MF
,
Eldridge
N
,
Wang
Y
,
Metersky
M
.
Electronic health record adoption and rates of in-hospital adverse events
.
J Patient Saf
.
2020
;
16
(
2
):
137
142
17
Appari
A
,
Johnson
ME
,
Anthony
DL
.
Information technology and hospital patient safety: a cross-sectional study of US acute care hospitals
.
Am J Manag Care
.
2014
;
20
(
11
spec No. 17
):
eSP39
-
eSP47
18
Kumah-Crystal
YA
,
Pirtle
CJ
,
Whyte
HM
,
Goode
ES
,
Anders
SH
,
Lehmann
CU
.
Electronic health record interactions through voice: a review
.
Appl Clin Inform
.
2018
;
9
(
3
):
541
552
19
Oscar
R
.
How mobile technology helps meet MU
.
Health Manag Technol
.
2012
;
33
(
9
):
22
23
20
Adler-Milstein
J
,
Huckman
RS
.
The impact of electronic health record use on physician productivity
.
Am J Manag Care
.
2013
;
19
(
11
spec No. 10
):
SP345
-
SP352
21
Lee
KL
,
Lynn
AM
,
Chen
TJ
,
Yang
LY
,
Chiang
SC
.
Visits to pediatric clinics by adult patients: a nationwide survey in Taiwan
.
Int J Environ Res Public Health
.
2018
;
15
(
7
):
1538

Competing Interests

POTENTIAL CONFLICT OF INTEREST: Dr Overhage is a shareholder in Cerner Corporation and was employed by Cerner Corporation when he performed this study. Dr Johnson is a member of the Perception Health Advisory Board and is the director of the Informatics Advisory Committee for the American Board of Pediatrics.

FINANCIAL DISCLOSURE: Dr Overhage is a shareholder in Cerner Corporation; and Dr Johnson has indicated he has no financial relationships relevant to this article to disclose.